Ecole doctorale « Economie, Organisations, Société » Ecole doctorale, département de la recherche

Thèse, par articles, présentée pour l’obtention du titre de Docteur ès Sciences de GestionDocteur ès Sciences de Gestion

Dissertation proposal for the degree of

Doctor of Philosophy (Ph. D.) in FinanceDoctor of Philosophy (Ph. D.) in Finance

Soutenance publique prévue le 07 juin 2011 par CharlesCharles -- Henri J. Henri J. REUTERREUTER

«« Culture, fCulture, f inanceinance , et institutions n, et institutions n ationalesationales »»

Jury

Rapporteurs Gérard CHARREAUX – Professeur des Universités Université de Bourgogne

Philippe RAIMBOURG – Professeur des Universités Université de Paris 1 Panthéon-Sorbonne

Directeur de thèse Franck BANCEL – Professeur de Finances, HDR. ESCP Europe

Président A définir

Suffragants Dominique JACQUET – Professeur des Universités Université de Paris Ouest Nanterre-La Défense

D’autres suffragants éventuels ont été contactés

2

3

Culture, finance, and national institutions

 Thèse,  par  articles,  ès  sciences  de  gestion  

Ph.  D.  Dissertation  in  finance  Three  essays  in  corporate  finance  

   

by  Charles-­‐Henri  J.  REUTER                                        

Laboratoire  du  CEROS  200  avenue  de  la  République  

90001  Nanterre          

 

E.S.C.P.  EUROPE  Département  de  la  recherche  

Laboratoire  de  finance  79  av.  de  la  République  

75011  PARIS              

4

             

5

Ni  l’Université  de Paris Ouest Nanterre-La Défense, ni ESCP Europe,  n’entendent  donner  

une  approbation,  ou  improbation,  aux  opinions  émises  dans  les  thèses  ;  ces  opinions  

doivent  être  considérées  comme  propres  à  leurs  auteurs.  

                                                       

6

                       

7

                     

A  mes   instituteurs,  mes   enseignants,  mes  maîtres   de  conférence   et   Professeurs.  Merci   de  m’avoir   appris   à  apprendre,  de  m’avoir  appris  à  aimer   le  savoir,  et  de  m’avoir,  chacun  à  votre  tour,  épaulé,  d’étape  en  étape.      A  Claude  Berthaud,  Claude  Simon,  Patrick  Messerlin,  Charles  Goodhart,  Sidney  Winter  &  Mauro  Guillen    Entre  tous,  à  Paul  DiMaggio,  et  à  Franck  Bancel  

         

8

           

9

                   

Lorsque  le  désir  résiste  aux  premières  atteintes  du  bon  sens,  on  lui  cherche  des  raisons.  Et  on  en  trouve  qui  ne  valent  rien.  La   vérité,   c’est   qu’on   ne   sait   comment   nommer   ce   qui   vous  pousse.  Quelque  chose  en  vous  grandit  et  détache  les  amarres,  jusqu’au  jour  où,  pas  trop  sûr  de  soi,  on  s’en  va  pour  de  bon.      Un  voyage  se  passe  de  motifs.  Il  ne  tarde  pas  à  prouver  qu’il  se  suffit  à  lui-­même.  On  croit  qu’on  va  faire  un  voyage,  mais  bientôt  c’est  le  voyage  qui  vous  fait,  ou  vous  défait.  

 Nicolas  Bouvier,  L’Usage  du  Monde  (1963)  

                   

10

               

11

Prologue Culturel aux remerciements

C’est un voyage entamé il y a bien longtemps, sinueux, un voyage peuplé de rencontres sérieuses et légères, graves ou bariolées, un voyage nourri de ces rencontres, et de l’histoire de chacun de mes compagnons de route. Au fond, ce voyage – comme tous les voyages – raconte avant tout ces histoires-là : celles de mes compagnons de route. Et je ne peux pas, ici, résister à l’envie d’abuser du privilège qui m’est offert. J’ai passé une large partie de ma vie dans les livres, que j’ai souvent dévoré sans pudeur, ni modération, notamment lors de mes voyages, très concrets, nombreux et aventureux. Je voudrais pouvoir remercier tous les auteurs qui m’ont influencé, vivants et morts, en citant l’un d’entre eux ; car quel hommage plus beau, et plus sobre, qu’une citation ?

Les rapports scientifiques – c’est-à-dire basés sur l’observation des autres – sont faux et factices : pour connaître une population, il faut à la fois la « vivre » et la « regarder ». C’est pourquoi ceux qui vivent doivent apprendre à regarder, ou ceux qui regardent doivent apprendre à vivre – au choix.

Germaine Tillon, Vivre pour Comprendre in Fragments de Vie (2009)

Et par cette pirouette, j’entends donc remercier tous mes compagnons de route, au sens propre et au sens figuré. Car ce sont eux qui m’ont apporté le matériel pour cette thèse, et l’expérience pour mes louvoiements multiples, entre, et contre les courants. * * *

12

13

Remerciements Ma première pensée va à mon directeur de thèse, le Professeur Franck Bancel. Culture et Finance est le fruit de notre rencontre et de son opiniâtre volonté à ce que nous nous en tenions à cette feuille de route. Il a su reconnaître, à sa valeur, la matière abondante, féconde et informe que j’avais accumulée et il a su anticiper l’incroyable richesse de cette opposition profonde, entre deux champs, que l’on peine à rapprocher formellement. Sa détermination, sa disponibilité, son intuition, ses encouragements sont les fondements indéniables de cette thèse et de chacun de ses articles. Ils sont, j’en suis convaincu, les fondements d’une recherche qui a le potentiel d’être portée fort loin. Merci à vous, donc, Franck, pour ces premières et précieuses indications, dans ce dédale, et pour toutes ces cartes qui pourront ponctuer ma route future. J’ai eu la chance d’errer un moment dans les limbes académiques, à travers les champs disciplinaires, les pays et sur plus d’une décennie. Il me faut remercier mes directeurs de mémoire successifs, leurs influences ne sont pas minimes et elles furent, chacune, galvanisantes : Mauro Guillen à Philadelphie, Charles Goodhart à Londres, Jean-Claude Casanova à Paris. Merci, bien sûr, à Paul DiMaggio, qui m’a accueilli à Princeton, formellement pendant un semestre, informellement pendant une année et qui n’a pas ménagé sa peine, puisque je lui ai transmis un matériel très abondant (bien au-delà des attentes formelles), matériel qu’il a très patiemment annoté à l’encre rouge, de sa main. Merci à toutes les équipes enseignantes, dans le cadre de mes différentes maitrises de recherche, à Wharton (2005, management), à la L.S.E. (1997, economics) et à l’I.E.P. (1996, économie appliquée et relations internationales), avec une pensée particulière pour Patrick Messerlin et une autre pensée particulière pour Sidney Winter. Il serait fastidieux de nommer tout le monde. L’équipe de Wharton, qui m’a accueilli pendant deux années et demie et a financé le début d’une thèse américaine en gestion, centré sur l’usage du concept de culture en gestion internationale (International Business), m’a fourni un bagage fort adéquat pour le type de voyage que j’avais entrepris et allais continuer. C’est un méandre, qui aurait mérité une homélie, tant la profondeur des flots et la force du courant forcent le respect. En passant, merci aux équipes administratives et de recherche à Wharton : Beth McCarthy, Mallory Hiatt, et Robin Woods: comment aurais-je pu vous oublier, ici ? Bien sûr, toute ma gratitude va, en la matière, aux équipes enseignantes de l’ESCP (1992-1995), puis de ESCP Europe (depuis 1997). Merci à Anne Gazengel, qui m’a introduit à la finance, en 1993, d’une façon si vivante et avec un tel élan, qu’il me porte encore aujourd’hui. Merci mille fois aux enseignants du département finance et du doctorat, qui m’ont, à l’occasion, autorisé à être passager clandestin. Merci à toute l’équipe du département finance. Je ne peux pas citer tout le monde, malheureusement, des Professeures aux assistantes, mais merci à tous et pour tout, pour les conseils de recherche, les conseils pédagogiques et tous les autres conseils, de ceux que l’on peut recevoir de ses collègues et qui facilitent tant le quotidien. Merci à l’équipe qui organise les Finance Research Seminars, à l’issue desquels nous avons, en maintes occasions, discuté de mes idées de recherche. Michèle, un grand merci à toi, pour m’avoir ouvert ta porte dans les moments difficiles. Merci pour ta discrétion, ton empathie, ta diligence.

14

Merci à tout le département recherche de l’ESCP. Merci aux directeurs à la recherche et aux responsables du programme Ph.D. Merci pour le soutien financier de la C.C.I.P., sans lequel mon voyage se serait achevé dans une impasse ou sur une corniche. Merci à Christine, à Marie-Andrée et à Sophie. Votre tâche n’est pas facile ; et vous vous en acquittez avec beaucoup de panache, de grâce et de bonhomie. J’ai tout particulièrement apprécié la possibilité qui m’a été donnée d’enseigner, à la fois dans le programme Masters Grandes Ecoles et dans les Part-time Executive Masters. C’est une grande chance, qu’ont peu de doctorants, une marque de confiance. C’est un moyen unique de se rattacher à un horizon concret et à un ancrage professionnel et social, bien déterminé. C’est une porte ouverte aux interactions avec mes ainés, les docteurs, du département finance et d’ailleurs. C’est une distraction importante, utile et salvatrice. Merci en particulier à Cécile Kharoubi et Michaël Tröge, dans le cadre de Finance II et merci à Philippe Raimbourg (et Sylvina Hildevert) dans le cadre du tutorat. N’est-il pas essentiel, que l’enseignement puisse occuper la place qui lui échoie si naturellement aux cotées de l’érudition, depuis la nuit des temps ? Ce fut pour moi un des sommets de mon expérience renouvelée à l’ESCP (Europe), heureux de voir cette si belle association de la recherche et de la pédagogie, ici comme ailleurs (bien sûr !), jalonner les avenues. Merci à toi, Claude Simon, qui m’a appris l’éthique des affaires sans jamais en parler. Dommage que tu aies pris ta retraite, il me semble que la recette pourrait faire des curieux. D’ailleurs, sans toi, je n’aurais sans doute pas repris ma thèse et elle serait restée une escapade exotique dans l’état le plus au nord du deep south. J’ai eu l’insigne honneur d’effectuer un double programme de doctorat et de Ph.D., avec une co-inscription à Paris Ouest (Nanterre) et ESCP Europe. Aussi, il me faut remercier très chaleureusement les équipes pédagogiques et de recherche, de l’université de Paris Ouest, pour m’avoir accueilli, encadré et conseillé. Merci en particulier à Philippe Dessertine, le directeur du laboratoire du CEROS, qui m’a permis de présenter le fruit de mes recherches à plusieurs reprises. Merci à tous les participants réguliers du CEROS et à Jean-Christophe Scilien, merci pour vos questions, lorsque j’ai présenté. Merci pour vos éclairages, divers et variés. Merci de faire vivre le CEROS. J’ai souvent participé à ces séminaires, c’était la règle du jeu. C’est la règle du voyage : quand on choisit un chemin, il faut savoir s’y tenir. Rétrospectivement, c’est l’un des passages vraiment mémorables. Enfin, pouvoir rédiger ma thèse en anglais a été un confort, une facilité, fort appréciés. Merci aux instances de l’université pour l’avoir permis. Merci beaucoup Catherine pour m’avoir facilité la vie et guidé dans les méandres administratifs. Je dois beaucoup à Hervé Laroche, qui a notamment facilité ma participation aux colloques académiques et de recherche, plus qu’il ne lui en incombait. Il est difficile de restituer ici l’importance qu’a revêtue pour moi la participation à ces colloques, nombreux. J’ai beaucoup travaillé dans les Marches de ma discipline. Sans ces colloques, j’aurais sans doute travaillé dans les limbes et peut-être erré, sans jamais trouver le chemin des Champs-Elysées. De fil en aiguille, une sente amenant un chemin, je me dois d’accorder un hommage appuyé aux chercheurs rencontrés dans ces colloques et différents laboratoires, en une vingtaine d’occasions. M’ont remis sur le droit chemin : mes contradicteurs, les présidents de cession, l’audience, les relecteurs anonymes ayant ouvert les bonnes voies (via ferrata, comment progresser autrement en haute altitude ?)…. j’oublie les conversations informelles pendant les pauses café, toujours bien trop courtes pour servir honorablement leur office. D’ailleurs, il s’agit là de la plupart des personnes que je vais pouvoir remercier nominalement.

15

Merci donc à tous ces gens rencontrés à l’A.F.F.I. (Brest, 2009), à l’A.I.B. (Rio, 2010), à l’E.F.M.A. (Aahrus, 2010), au F.M.A. (Turin, 2009), au Groupe de Recherche Monnaie Banque Finance du CNRS (Orléans, 2009 ; Bordeaux, 2010) et au colloque INFINITI (Dublin, 2009, 2010). Je tiens à remercier particulièrement les personnes suivantes, qui m’ont gratifié de leurs analyses, une pause méritée dans un voyage haletant, une pause où l’on regarde en avant, en arrière, pour prendre le pouls des choses, s’équiper selon la météo, faire des projets, reconsidérer sa carte (et ses cartes) : Raj Aggarwal, Thierry Baudasse, Philippe Desbrières, Georges Gallais-Hamonno, John Goodell, Colm Kearney, Eva Liljeblom, Brian Lucey, Laurent Weill et Rohan Williamson. Merci aux membres des laboratoires de recherche du FARGO, de l’INSEAD, de la L.S.F. et du S.S.F.A. Merci également aux éditeurs et relecteurs anonymes des revues Finance, du Journal of Banking & Finance et du Journal of International Business Studies. Je tiens à réserver un hommage appuyé à Gérard Charreaux, qui m’a ouvert les portes du FARGO, à la croisée des chemins, et à Denis Gromb, au début, puis à la fin d’une côte particulièrement difficile. Merci aussi à vous, enfin, Jean-Philippe Bouilloud, pour votre disponibilité, vos relectures et vos conseils, dans une géographie si lointaine de vos bases arrières. Philippe Raimbourg, vos conseils de recherche, au milieu du gué, m’ont permis d’opter pour un chemin qui s’est avéré payant et je vous en suis fort reconnaissant. Charles Goodhart, retrouvé au fil de l’eau, m’a mis en contact avec David Westbrook, dont les préoccupations confluaient. Merci pour votre confiance, votre soutien et vos recherches à tous deux. Et quand, sur le fil du rasoir, on entrevoit à l’issue d’un chemin une route, il peut rester un pont à franchir et, qui sait, un détour à entreprendre. Merci aux équipes du STATEC et de la L.S.F. pour m’avoir accueilli et financé pendant ma dernière ligne courbe. Il serait vain de vouloir restituer l’exacte nature de mes rapports avec mon collègue, docteur en économie de son état, Charles-Henri Dimaria. De quoi pourrais-je ne pas te remercier ? Conseils académiques, administratifs, professionnels, confraternité, relectures. Tout y est passé ? Merci à vous deux, surtout, mes copains de cordées académiques, Anaïs Hamelin et David Lebris, vous qui m’avez tout juste précédé sur la voie lactée, merci pour votre écoute, vos questions, vos conseils, vos relectures, vos pistes à suivre. Puissions nous chercher ensemble, dans une géographie prochaine. Merci pour toujours à mes copains de bordées conceptuelles, sous d’autres latitudes, d’autres vents et l’autre continent : à l’I.E.P. et la L.S.E. (économie), à Wharton (management), à Princeton (sociologie) : Serge Bonzom, Gino Cattani, Rafael Corredoira, Alice Goffman, Julien Hanoteau, Jeawon Kim, Charles Kirschbaum, Xiaohui Lu, David Mes, Lakshmi Ramarajan, William Schnepper et Frans Tieleman, distingués parmi tant, et tant, et tant d’autres. Merci à toi, Thi-Minh Ngo, pour ces innombrables débats loin du café du commerce, depuis octobre 1992, sur les bancs du 79 de l’avenue de la République. Antoine Gailliot, je ne t’ai pas oublié mon alter ego. Bientôt ton tour (pas de point d’interrogation, tu noteras). Et je n’oublie pas tous les autres qui m’ont accompagné pour des bouts de chemin, plus courts ou moins abstraits. Nicolas Fraissé, ton optimisme bien trempé seyait fort bien à ma thèse. Rattrapant ceux qui m’ont tout juste précédé, et précédant ceux qui, j’espère, emboiteront mes pas, merci à mes pairs du programme doctoral d’ESCP Europe, un grand, grand, grand merci à tous, nous qui avons su construire une ambiance d’entraide, de respect mutuel, d’émulation intellectuelle. C’est vraiment peu commun, presque une anomalie typographique qui deviendrait obsédante, si nous étions en Géographie. Bon, nous sommes dans la Gestion, n’est ce pas ? Je ne saurais assez dire, combien votre accompagnement intellectuel et amical m’a aidé à porter le regard suffisamment loin, pour trouver le chemin. Merci à vous tous, que je ne

16

peux nommer par peur d’oublier l’un ou l’autre. Merci tout particulièrement à vous cinq, Magali Ayache, Samer Iskandar, Vanessa Joly, Sébastien Picard et Véronique Steyer, pour ces heures que vous n’avez jamais, ni comptées, ni décomptées. Et tout à coup, au détour d’un virage, on s’aperçoit qu’on est sur l’autre rivage. La boucle est bouclée ? Ma première pensée allait à mon directeur de thèse. Mon avant-dernière pensée va à mon jury de thèse, Messieurs les professeurs Gérard Charreaux, Dominique Jacquet et Philippe Raimbourg. Etre admis à vous présenter les fruits de ce long travail et qui sait, à rejoindre votre communauté, c’est là, pour moi, le plus bel achèvement pour cette étape, étape d’un voyage qui s’est fait, et qui s’est défait, d’un voyage qui m’a fait… et qui m’a construit. Il est difficile de dire, sobrement, la joie que cela me procure, c’est une joie sans tache, à la mesure du respect que vous m’inspirez…

« Last, but not least… »

…je n’avais jamais, jusqu’à aujourd’hui, sondé la vertigineuse profondeur de cette formule éculée. Et comment le dire autrement ? Merci à mes proches, à mon épouse, à mes parents, à mon frère, à toute ma famille, au sens large et des deux côtés de l’onde. Merci pour votre foi, votre patience, votre soutien. Une petite pensée pour mes grand-pères m’assaille ; si Charles avait su que l’on peut faire un commerce d’érudition ? Et si Jean avait su que l’on peut enseigner le commerce ? Jean, toi qui as eu l’insondable bonheur de rencontrer ton arrière petit-fils Jean, au détour d’une escapade, bien sagement écourtée ? Tu n’as pas pu voir ton petit-fils Charles-Henri en habit ? Mais ton épouse, qui est là avec nous et encore pour un bon petit bout de route, te racontera tout cela, je suppose, au détour d’un chemin. Il me reste l’essentiel, la quintessence. Merci à mes deux petits bouts. Votre univers a été un papa à la maison, certes, mais un papa à son bureau, et plus que de raison. Qu’entend-on à la raison, quand on a 2 et 4 ans ? Vous avez appris, dans la persévérance, dans la gentillesse, et bien sûr, dans les larmes, à respecter le travail de papa, à jouer très patiemment à ses côtés, mezza voce. Vous m’avez apporté la sérénité dont j’avais besoin pour parachever ce rapprochement inouï, de la culture et de la finance. Et je vois bien, maintenant que nous sommes à l’orée d’un nouveau bois, que votre persévérance a été bien plus grande que la mienne. * * *

17

Sommaire général abrégé

Notre thèse est présentée à la fois pour le doctorat ès sciences de gestion et pour un Ph. D. in corporate finance. Elle compte une synthèse, une introduction générale, trois articles et une annexe. La synthèse permet de reprendre l’ensemble des éléments de la thèse et de les mettre en perspective. La thèse est, de façon générale, rédigée en anglais. Plus spécifiquement, le prologue culturel aux remerciements, les remerciements, le sommaire général abrégé et la synthèse, sont rédigés en français. Le corps technique de la thèse, comprenant un résumé bref, l’ensemble des chapitres, de un à quatre, et l’annexe, sont rédigés en anglais. Y sont ajoutés, toujours en anglais, une bibliographie générale et les tables (table des figures et des tableaux, index des noms d’auteurs, table des matières détaillée). La thèse comprend enfin les résumés protocolaires de deux cents mots environ, en français et en anglais, en fin du document. * * *

18

19

Dédicace   7  

Epigraphe   9  

Prologue  culturel  aux  remerciements   11  

Remerciements   13  

Sommaire  général  abrégé   17  

Synthèse  générale   21  «  A  Survey  of  Culture  and  Finance  »   23  Les  relations  d’agence  conditionnent-­‐elles  l’interaction  entre  la  politique  financière  

des  entreprises  et  les  cycles  d’affaires  ?   27  La  relation  entre  cycles  d’affaires  et  structure  du  capital,  est-­‐elle  conditionnée  par  

des  facteurs  structurels,  institutionnels  ou  bien  culturels  ?   33  Articulations  entre  les  articles,  positionnement  et  conclusions   41  Appendice  :  citations  proposées  en  exergue  aux  différents  chapitres   45  Bibliographie  pour  le  chapitre   46  

Dédicace  et  épigraphe  (anglais)   51  

Résumé  bref  (anglais)   53  

Chapitre  premier  (anglais)  :  Introduction   57     Table  des  matières  spécifique  et  détaillée   57  1.   Genèse   59  2.   Le  «  Survey  »  et  ses  conséquences   62  3.   Les  deux  essais  empiriques  sur  la  structure  du  capital   65  4.   Le  complément  conceptuel   70  5.   Conclusion  :  positionnement,  essai  d’interprétation  et  directions  futures   71  6.   Bibliographie  pour  le  chapitre   74  

Chapitre  deux  (anglais)  :  «  A  Survey  of  Culture  and  Finance  »   77     Table  des  matières  spécifique  et  détaillée   77     Résumé  de  l’article   79  1.   Introduction   80  2.   Contexte,  enjeux  et  méthodes   85  3.   Les  approches  et  les  succès  dimensionalistes   96  4.   Les  autres  approches  culturelles  en  finance   111  5.   Synthèse  des  conceptualisations  de  la  culture,  émergeant  en  finance   128  6.   Conclusion  :  la  recherche  en  finance  doit-­‐elle  approfondir  ses  approches  

culturelles  ?     139  7.   Premier  appendice  au  chapitre  :  étendue  intuitive  des  définitions  de  la  culture   143  8.   Second  appendice  au  chapitre  :  les  principaux  «  indices  de  culture  nationale  »   148  9.   Troisième  appendice  au  chapitre  :  les  variables  de  contrôle  

«  institutionnelles  »  ,  présentation  de  leur  importance  dans  la  littérature  dimensionaliste   150  

10.   Bibliographie  pour  le  chapitre   152  

20

Chapitre  trois  (anglais)  :  les  relations  d’agence  conditionnent-­elles  l’interaction  entre  la  politique  financière  des  entreprises  et  les  cycles  d’affaires  ?   159     Table  des  matières  spécifique  et  détaillée   159  1.   Introduction   161  2.     Perspectives  institutionnelles  sur  les  politiques  financiers,  l’actionnariat  et  les  

cycles  d’affaires   165  3.     Échantillonnage,  données  et  méthodes   175  4.   Résultats   188  5.     Conclusion   200  6..   Premier  appendice  au  chapitre  :  concentration  d’actionnariat  en  Europe   204  7.   Second  appendice  au  chapitre  :  tests  pour  comparaisons  appariées   206  8.   Troisième  appendice  au  chapitre  :  les  mesures  européennes  de  sentiment  

économique  «  E.S.I.  »   207  9.   Quatrième  appendice  au  chapitre  :  résultats  statistiques  détaillés  et  complets   210  10.   Bibliographie  pour  le  chapitre   212  

Chapitre  quatre  (anglais)  :  la  relation  entre  cycles  d’affaires  et  structure  du  capital  est-­elle  conditionnée  par  des  facteurs  structurels,  institutionnels  ou  bien  culturels  ?   217     Table  des  matières  spécifique  et  détaillée   217  1.   Introduction   219  2.     Revue  de  littérature  et  hypothèses   222  3.     Échantillonnage,  données  et  méthodes   233  4.   Résultats   240  5.     Conclusion   256  6..   Premier  appendice  au  chapitre  :  constitution  de  la  base  de  données   258  7.   Second  appendice  au  chapitre  :  tests  pour  comparaisons  appariées   260  8.   Troisième  appendice  au  chapitre  :  résultats  statistiques  détaillés,  robustesse  

des  variables  de  contexte   261  9.   Quatrième  appendice  au  chapitre  :  résultats  statistiques  détaillés  et  complets   262  10.   Cinquième  appendice  au  chapitre  :  résultats  statistiques  détaillés  par  pays   262  11.   Bibliographie  pour  le  chapitre   265  

Annexe  (anglais)  :  «  Embedding  Culture  and  Finance  :  Divorce  or  a  Fresh  Start  ?  »   269  

Bibliographie  générale  et  tables  (anglais)   307  Bibliographie  générale   309  Tables  des  figures  et  des  tableaux   325  Table  des  références   327  Table  des  matières   333  

Résumés  protocolaires  (bilingues)   338   * * *

Synthèse générale

21

Culture, finance et institutions nationales1

“If one tries to transgress the borderlines of national cultures, one piece of intellectual luggage, that has to be left at home, is the division of labor among social sciences as it has been developed in Europe and North America in the past hundred years [...] Psychology, social psychology, sociology, anthropology, economics and political sciences, all study only facets of the same social reality. Academic inbreeding and atomization in the West have led to extensive production of irrelevant speculations. The system has become self destructive in that it punishes rather than encourages borrowing from related disciplines. Cross-cultural social sciences, therefore, can not be, but cross-disciplinary”.

Hofstede (1994: ix)

“We would expect countries, per se, to matter little in finance. However, this is not the case: countries [are] very important”. “A firm’s country of incorporation is a more important determinant of its financial policies than its industry [This generates] paradoxes that are explored in many papers. [We need] a new theory of international finance that recognizes that countries are relevant” Stulz (2005: 1601 & 1596)

Synthèse générale

Les deux citations proposées en exergue permettent de souligner une tension forte, qui existe

dans la recherche en finance et en sciences de gestion, autour de la notion de « culture

nationale ». En effet, Hofstede parle de « spéculations présentant fort peu de pertinence »2,

alors que Stulz suggère la notion « d’énigme pays », pour caractériser cette différenciation qui

est observée, au niveau national, pour un grand nombre de phénomènes financiers. Nous

avons retenu ces épigraphes car ils se complètent : ils restituent une tension qui, traversant les

sciences humaines, transparait clairement dans la recherche en finance.

1 Nous proposons, en appendice, les titres de nos trois articles, ainsi que les citations proposées en tête de chacun des chapitres. Il y est fait référence dans le corps de cette synthèse. 2 Les traductions, proposées par l’auteur, visent à fluidifier la lecture et ne sont pas nécessairement littérales. Par ailleurs un certain nombre de mots sont sujets à débats en ce qui concerne leurs traductions spécifiques en français il n’existe pas de traduction consacrée. Nous pensons par exemple aux notions de « grounded theory », « d’embededdness » en sociologie, de « market timing » en finance, etc. Un problème similaire se pose pour la traduction du mot « survey », tel qu’il est utilisé dans le titre de notre premier article. Le titre de ce premier article n’est d’ailleurs pas traduit. Enfin, il est à noter que certains concepts, en particulier les notions « d’institutions » et de « culture » recoupent des acceptions extrêmement variables, d’une époque à l’autre, d’une géographie à l’autre, et d’une discipline académique à l’autre (voir par exemple le Survey et l’article en annexe). Cuche (2004) illustre cette dimension pour la culture ; Lévesque, Bourque et Forgues (2001) pour les notions d’institutions ou « d’encastrement » (qui est la traduction que nous retiendrons pour « embededdness »).

Synthèse générale

22

Notre premier article, A Survey of Culture and Finance, correspond à ce que l’on pourrait

qualifier d’une ethnographie culturelle de la production des chercheurs de finance (voir ci-

après). Il permet d’établir scientifiquement la pertinence de l’analyse suggérée par les

épigraphes : l’intérêt pour les approches « culturelles » en finance est fort, mais il est marqué

par un manque de cohérence d’ensemble et par des divergences de fonds, quant aux approches

conceptuelles à retenir. Il permet cependant d’identifier un point focal dans toutes ces

approches : les chercheurs qui s’intéressent à « l’énigme pays », et ceux qui s’appuient sur

l’approche par les « indices de culture nationale » (approches que nous qualifierons de

dimensionalistes , dérivées notamment des recherches d’Hofstede) ont des objectifs

communs : enquêter plus en profondeur sur les déterminants de « l’énigme pays », en finance,

et au-delà, afin d’identifier quelles dimensions, culturelles, religieuses, institutionnelles,

socio-économiques, politiques, etc. génèrent les différenciations qui s’opèrent dans le

fonctionnement des nations et plus généralement dans le fonctionnement des sociétés

humaines.

C’est donc autour de cette tension problématique que nous construisons notre thèse, intitulée

Culture, Finance et Institutions Nationales. Le premier article s’intéresse plus

particulièrement à la première de ces relations, celle entre culture et finance. Cet article est

une contribution empirique et conceptuelle, établie sur une méthodologie qualitative.

Les deux autres articles traitent d’un thème particulier – et central – de la finance

d’entreprise : les déterminants de la structure du capital. Ils procèdent directement de notre

approche conceptuelle préliminaire, sur laquelle ils apportent, chacun, un éclairage

complémentaire. On peut notamment noter qu’ils abordent la question de la culture, telle

qu’elle émerge au travers du Survey, sous quatre angles principaux. Premièrement, les

décisions financières relèvent d’une technique, mais également de normes comportementales

et d’un contexte qu’il s’agit d’expliciter plus avant. Deuxièmement, l’influence nationale se

fait sentir selon différents canaux, convergents, le principal étant sans doute celui des

Synthèse générale

23

institutions, d’ailleurs en étroite interaction avec « la culture nationale ». Troisièmement,

l’attribution d’une étiquette « culturelle » est souvent faite à tort, à des phénomènes d’une

autre nature (institutionnelle, sociopolitique, économiques… Klep, 2001) et il convient, pour

étudier les processus culturels, de bien les distinguer des processus matériels ou structurels

(DiMaggio, 1994). Enfin, nous mettons en œuvre un processus d’échantillonnage conforme

aux recommandations de nombreux chercheurs intéressés par la culture et, ou les institutions.

Nous reprenons ci-dessous, dans un premier temps, la présentation successive de chacun de

nos trois articles. Nous la complétons par une partie complémentaire, qui revient notamment

sur le lien entre les trois articles, et reprend des éléments de l’introduction et de l’annexe ainsi

qu’une interprétation de nos résultats et une conclusion.

1- « A Survey of Culture and Finance »

Notre premier article, A Survey of Culture and Finance, est une contribution empirique,

établie sur une méthodologie qualitative. Nous y analysons de manière systématique, selon

un protocole précis, tous les articles publiés dans les revues à comité de lecture, en finance,

qui font référence à la notion de culture. La transparence et la neutralité de notre protocole

d’échantillonnage systématique (« screening ») est un point important (section II.2.2). Nous

nous inspirons d’études passées, aussi bien en gestion (Wiltbank, et al., 2006), qu’en finance

(Charreaux et Schatt, 2006). Il s’agit de faire émerger les définitions en usage, à partir d’une

analyse du terrain académique. Cette analyse a pour but d’établir quelles sont la nature et la

portée de l’usage du concept de culture dans la recherche financière.

C’est pourquoi nous qualifions cette approche d’ethnographie culturelle de la production des

chercheurs en finance. En ce qui concerne la définition du « terrain » d’étude, il s’agit des

articles revus par les pairs et publiés. Nous retenons la liste des revues qui sont suivies pour

l’établissement des « facteurs d’impact » et disponibles dans la base « ISI-Web of Science ».

Synthèse générale

24

Nous complétons cette liste par des listes « régionales », c’est-à-dire plus spécifiquement les

listes qui ont été établies par le CNRS (la « section 37 ») et des institutions comparables en

Grande-Bretagne et en Allemagne.

Nos questions de recherche sont les suivantes. Premièrement, peut-on caractériser un ou

plusieurs schéma(s) culturel(s) théorique(s), en usage dans la recherche en finance

aujourd’hui ? Et deuxièmement, peut-on considérer culture et finance comme un champs

d’investigation établi ? Les réponses que nous apportons sont, respectivement, négative et

positive. C’est-à-dire que nous ne parvenons pas à identifier un ou plusieurs cadre(s)

conceptuel(s) établi(s), tout en indiquant que le thème semble être suffisamment porteur, pour

être considéré comme un domaine d’investigation en tant que tel. Ce contraste, entre un

intérêt nourri et l’absence de référence établie, nous fournit notre problématique. Notre article

n’est pas simplement une revue de la littérature, ou bien une enquête. Il s’agit d’une analyse

très méthodique des usages, basée sur un protocole d’identification précis et suivie d’une

synthèse structurée ; l’article est un « survey ».

Au-delà des réponses apportées à ces deux questions de recherche, nos résultats

complémentaires sont les suivants. Premièrement, l’usage du concept de « culture » est

quasi-systématiquement lié, dans les travaux identifiés, à l’idée de « culture nationale ».

Deuxièmement, nous mettons en lumière le fait que le concept est désormais d’usage fréquent

dans la recherche en finance. Il concerne un éventail très large de sujets, allant de l’étude des

flux financiers internationaux, à celle de la politique financière des entreprises, en passant par

l’analyse du développement, de l’efficience et de la robustesse des marchés financiers ainsi

que des « structures macro-financières ».

Troisièmement, nous notons une polarisation marquée de la recherche, avec un certain

nombre d’études retenant une approche dimensionaliste (définitions en section II.3.1), tandis

que la plupart des autres travaux ne proposent pas de définition précise. Cela nous conduit à

étendre notre protocole de recherche et nous effectuons une recherche systématique des

Synthèse générale

25

travaux qui peuvent être qualifiés de dimensionaliste en finance (section II.2.2.3), mais qui

n’ont pas, ou pas encore, été publiés, ou bien pas été publiés dans des revues de finance.

Cette enquête complémentaire nous permet de contraster les travaux dimensionalistes publiés

et non publiés, en finance et dans d’autres champs et nous fournit un quatrième résultat. Il

semble que les travaux dimensionalistes identifiés dans les revues à comité de lecture en

finance aient une approche plus parcimonieuse et plus syncrétique que les autres. Par ailleurs,

en dépit de ses forces, il ne semble pas que l’approche dimensionaliste ait connu, en

sociologie ou bien en économie, le succès qu’elle a connue dans certaines branches de la

recherche en gestion (Baskerville, 2003, voir les sections II.5.1, II.5.II et l’annexe). Au final,

nous concluons qu’il ne semble guère que l’approche dimensionaliste puisse aujourd’hui

prétendre à une visée intégrative au sein de la recherche en finance.

Notre cinquième et dernier résultat de recherche est troublant et nous permet de poser un pont

vers différents courants de recherche en sociologie ou vers des courants hétérodoxes en

économie. Le concept de culture est en effet souvent utilisé en rapport avec des techniques et

des professions (Abbott, 1988). Concernant la finance, il caractérise notamment des processus

bancaires, d’attribution des crédits (section II.4.1.2) et dans la littérature financière il est

fréquemment utilisé – il est vrai de façon apparemment anecdotique – dans cette acception et

des acception liées (section II.4.3). Il en résulte que les notions de « culture financière »

(MacKenzie et Millo, 2003) ou de « gestion financière de la firme » (Aoki, 2001: chapter 11

& 13 ; Fligstein, 2001: chapter 7; Aglietta, 2004) sont des notions qui doivent être examinées

avec soin. Cette dernière remarque nourrit notamment une partie de notre recherche empirique

examinée ci-après.

Contexte général de recherche. Il est enfin intéressant de noter qu’il transparait, au travers

de notre protocole d’identification, que les travaux publiés sont surtout le fait de chercheurs

ou d’équipes de recherche établis. On pensera à des chercheurs internationalement reconnus

comme Stulz, Tabellini, ou encore Zingales. Il en résulte qu’il s’agit d’un champ

Synthèse générale

26

d’investigation particulièrement périlleux pour de jeunes chercheurs. Cette constatation n’est

pas surprenante, tant il est vrai que la culture semble être un concept fondateur, un « méta-

concept », pour la recherche en sciences sociales (Cuche, 2004). Par ailleurs, notre synthèse

sur « culture et finance » nous conduit à des considérations méthodologiques nombreuses, qui

évoquent notamment des travaux réalisés sur les positionnement épistémologiques respectifs

de l’économie et de la sociologie économique (Smelser et Swedberg, 1994 ; Lévesque,

Bourque et Forgues (2001). Il n’est pas anodin que la culture soit un concept fondateur en

anthropologie ou en sociologie, un concept dont la compréhension fine peut être à la source

de scissions disciplinaires (par exemple l’anthropologie culturaliste vs. structuraliste, voir

Cuche, 2004 sur ce point).

En ce qui nous concerne, nous attirons l’attention sur l’intérêt que peut présenter la

polarisation des échantillons (« sample-pooling ») pour permettre de faire émerger

l’importance du contexte, autrement que par une approche marginaliste. Cette technique, tout

à fait conventionnelle en statistique, notamment en statistique médicale, est cependant fort

peu utilisée en finance. Nous la mettons systématiquement en œuvre dans le cadre de nos

deux articles empiriques sur la structure du capital des sociétés européennes (les articles deux

et trois présentés ci-après).

Au final, notre article est une première contribution pour référencer en finance les approches

culturelles existant dans les autres sciences humaines. Nous suggérons également l’adoption

de « définitions à rebours » (« backward definitions »), qui ont le mérite d’attirer l’attention

sur les mécanismes liant la culture à un phénomène financier donné, et qui permettent

d’éviter, pour partie, les controverses et enjeux épistémologiques. En phase avec les résultats

de notre analyse, nous recommandons par ailleurs la fertilisation croisée, en utilisant, en

particulier, les connaissances accumulées afférant au dimensionalisme, à la confiance ou

encore aux religions dans différentes disciplines en sciences humaines, gestion, économie,

sociologie.

Synthèse générale

27

2- Les relations d’agence conditionnent-elles l’interaction entre la politique financière des

entreprises et les cycles d’affaires ?

Notre question de recherche est directement explicitée dans le titre de cet article. Etant

donné la centralité du sujet en finance d’entreprise, le nombre restreint d’antécédents de

recherche peut surprendre. Nous pensons en particulier à l’étude de la structure du capital sur

un échantillon paneuropéen, à l’étude de l’interaction entre cycles d’affaires et structure du

capital, ou bien encore à l’étude de la relation entre dispersion de l’actionnariat et structure du

capital (voir sections III.1 et III.2.1).

Échantillonnage et antécédents. En particulier, peu d’études passées se sont exclusivement

intéressées aux déterminants des politiques financières, exclusivement pour les sociétés

européennes cotées. Les études sont le plus souvent centrées sur la situation étasunienne ou

bien elles s’intéressent à d’autres pays en particulier, parfois dans une approche comparative

(deux pays). Les études internationales sont le plus souvent orientées vers des échantillons

très larges incluant des pays développés et émergents, et un certain nombre d’études a été

réalisé sur les principaux « grands pays », mêlant donc des économies appartenant à des aires

géographiques, culturelles et socio-économiques différentes (Ronen et Shenkar, 1985;

Hofstede, 2003 ; Amable, 2005). L’un des intérêts de notre échantillonnage, est qu’il permet

de circonscrire les effets du développement économique, ainsi que la variété qualitative des

institutions légales, politiques, économiques, etc. En revanche il nous permet d’observer les

effets de structures d’actionnariat contrastées, puisque qu’il existe en Europe des proportions

significatives de sociétés à actionnariat concentré et de sociétés à actionnariat dispersé (Faccio

et Lang, 2002, voir l’appendice 1 du chapitre III). Cela nous permet donc de porter notre

attention sur les effets potentiels que cette diversité de l’actionnariat peut avoir sur

l’interaction entre le contexte économique (les cycles d’affaires) et la politique financière des

Synthèse générale

28

entreprises. Ce protocole d’échantillonnage suit d’ailleurs très directement des conclusions et

recommandations de notre premier article, ainsi que des protocoles en vigueur dans les

approches méthodologiques qualitatives en sciences de gestion.

Notre étude s’inscrit dans la foulée d’un certain nombre de contributions récentes sur l’étude

de la structure du capital des sociétés, dans un contexte international (De Jong, et al., 2008;

Fan, et al., 2008; King et Santor, 2008; Margaritis et Psillaki, 2009) ou bien en relation avec

les structures macro-financières (financement principal par les banques vs. le marché;

Antoniou, et al., 2008 ; Gonzales et Gonzales, 2008). Utilisant ces contributions, nous tentons

de démontrer comment une variation dans la nature des relations d’agence au sein de la firme,

telle que résultant d’un changement substantiel dans le degré de dispersion de l’actionnariat,

peut conduire à une dichotomie des politiques financières, en relation aux cycles d’affaires.

Notre hypothèse par défaut est que l’interaction entre cycles d’affaires et structure du

capital est limitée, ou bien secondaire par rapport à d’autres dimensions, notamment les

relations d’agence entre dirigeant et actionnaires, les questions d’optimisation fiscale, le

« market timing » etc. Cette hypothèse par défaut est bien en ligne avec ce que De Jong, et al.

(2008) ont caractérisé comme la « perspective financière traditionnelle » fondée sur le « Role

Model » étasunien d’actionnariat dispersé. Nous en déduisons une proposition de pro-cyclicité

des politiques financières, où l’investissement, l’endettement, le taux de dividende, et la

gestion des liquidités suivent les cycles d’affaires anticipés à la hausse, et à la baisse. Nous

notons d’ailleurs que, dans ce cadre, avec un lien direct entre endettement et investissement,

et un taux d’endettement autour de 50%, le levier comptable des entreprises ne devrait pas

s’en trouver significativement affecté. Ces propositions correspondent d’ailleurs aux

constatations faites sur des échantillons étasuniens (Frank et Goyal, 2009), ainsi que sur notre

base, concernant le sous-échantillon représentant les sociétés à actionnariat dispersé.

Synthèse générale

29

Face à cette hypothèse de pro-cyclicité, nous développons des arguments théoriques qui se

fondent sur la différence de concentration de l’actionnariat : elle entraine notamment une

relation d’agence particulière, un rôle disciplinaire spécifique pour l’endettement, un rapport

de signaux altéré, un gain d’opportunité représenté par le « market-timing » réduit.

Un argument important procède de la constatation par Ruiz-Mallorqui et Santana-Martin

(2010) que l’horizon spécifique des actionnaires de référence en Europe tend à être plus long

qu’aux Etats-Unis, avec des conséquences en termes de liquidité, de diversification, et donc

d’attentes pour ces actionnaires. En conséquence l’alignement des intérêts entre actionnaires

de référence et dirigeants est plus fort3, le besoin de suivi et d’encadrement, aussi bien ex-ante

qu’ex-post, est plus faible, car le contrôle est mieux établi4.

Il en découle que des objectifs de modulation des risques peuvent être privilégiés sur ceux

d’optimisation de court terme : il apparaît que ce type de société gère sa capacité de

financement de façon inter-temporelle, en la desserrant quand les perspectives économiques

remontent et en se protégeant par un endettement anticipé quand elle semble devoir se

détériorer.

En termes d’investissement et de gestion des liquidités, l’existence d’un contrôle mieux établi

(par les actionnaires de référence, la famille ou le dirigeant) et de moyens de pression ainsi

que des recours plus nombreux changent les équilibres dans les relations d’agence. Ce

changement permet de desserrer la contrainte de discipline par la dette, de mieux tirer

avantage des bénéfices potentiels de la latitude managériale, d’avoir une dépendance

différente, en termes d’investissement, face à la conjoncture. Par ailleurs l’alignement plus

3 C’est d’ailleurs ce qui conduit La Porta, et al. (1999) à noter que la relation d’agence qui prédomine et pose problème dans ce type de cas est celle entre actionnaires minoritaires et autres parties prenantes. 4 Le concept traditionnel de « monitoring » généralement traduit par « contrôle » est difficile à bien restituer. Notre argument ici est que, justement, en actionnariat familial ou concentré, le contrôle est plus fort, donc le besoin d’encadrement, de « monitoring » stricto sensu est proportionnellement plus faible. Il y a un équilibre qui se créé entre contrôle effectif et encadrement.

Synthèse générale

30

fort des intérêts entre actionnaires de référence et dirigeant favorise une politique de

dividendes lissés.

Ces différents arguments nous conduisent à proposer une série d’hypothèses spécifiques

concernant l’investissement, l’endettement, le taux de dividende, et la gestion des liquidités,

pour les firmes à actionnariat très concentré. Ces hypothèses découlent d’une proposition

générale de contra-cyclicité des politiques financières dans ce type de firme.

Nos résultats empiriques sont conformes à ces hypothèses, que nous ne pouvons pas, pour la

plupart, rejeter. A minima, nos résultats empiriques illustrent l’inversion des politiques

financières, dans leur relation à la conjoncture anticipée, en fonction de la dispersion de

l’actionnariat. Le message principal que nous retenons donc, est que la modulation des risques

semble prendre le pas sur la question de l’encadrement et du contrôle des dirigeants

(« monitoring »), lorsque l’on passe d’une entreprise à actionnariat très dispersé (moins d’un

tiers du capital est fermé), à une société à actionnariat très concentré (plus de deux tiers).

Un certain nombre de résultats complémentaires émerge en outre de notre analyse.

Premièrement, on détecte une forme de substitution entre la dette obligataire (« long-term

public debt ») et les autres dettes financières pour les sociétés à actionnariat dispersé. Dans

ces sociétés, la gestion de la dette publique de long terme est en effet contra-cyclique, alors

que les autres dettes financières sont pro-cycliques, avec un coefficient plus important que

pour l’effet contra-cyclique. Nous en déduisons que la gestion du risque, liée aux cycles

d’affaires et donc au coût d’opportunité des faillites, est également un aspect important dans

ce type de sociétés, mais que d’autres considérations, émanant notamment de la relation au

marché prédominent. Cela pourrait inclure notamment les aspects suivants : problème de la

gestion de la relation d’agence entre le dirigeant et les actionnaires, « market timing »,

engagement sur une note financière, donc sur un niveau d’endettement implicite.

Synthèse générale

31

Deuxièmement, on note l’absence de relation, pour les sociétés à actionnariat très concentré,

entre politique obligataire et anticipation des cycles d’affaires5. Cela suggère selon nous des

préférences en termes de schémas de financement pour les différents types de société, une

hypothèse qui est reconsidérée dans le troisième article.

Troisièmement, on détecte une relation pro-cyclique plus marquée entre les liquidités (« cash

holdings ») et les cycles d’affaires anticipés dans les sociétés à actionnariat dispersé. Cela

peut être, selon nous, lié à deux pistes complémentaires qui sont le principe de précaution et

les problèmes d’enracinement6.

Quatrièmement, nous identifions de façon consistante que des instruments statistiques

différents, pour les cycles d’affaires, jouent sur la détermination des politiques financières

variées. Il semble en effet que, quelque soit le profil de société, les questions de politique

d’endettement et d’investissement soient plutôt liées à des effets de demande, alors que la

gestion des liquidités et la politique de dividendes soient plutôt liées à des effets d’offre.

Les contributions directes de cet article, au regard de notre thèse, portent notamment sur les

points identifiés en introduction. En effet cet article s’intéresse tout particulièrement à

distinguer, dans les choix de politiques financières observables, quels sont les déterminants

purement techniques et ce qui procède des normes comportementales au sens large. En

particulier nous souhaitons souligner l’ambivalence potentielle de certaines normes de

comportement pour les dirigeants et les membres du comité de direction, face au risque, à

l’investissement, à l’enracinement ou à la latitude managériale.

5 Nous « contrôlons » pour la taille des entreprises, les secteurs industriels, la profitabilité… 6 Plus spécifiquement, le principe de précaution correspondrait au raisonnement suivant : en période favorable, on accroît le cash dans ces sociétés, tout en maintenant un endettement cible, pour se préparer pour une conjoncture dégradée. Les problèmes d’enracinement seraient sensibles pour une part plus significative des sociétés à actionnariat dispersé et donc identifiables en agrégé.

Synthèse générale

32

La littérature existante avait souligné que ces comportements sont étroitement dépendants

d’un cadre économique et institutionnel déterminé (Charreaux, 1996, 2004). Dans ce cadre, la

relation de confiance, qui peut s’instaurer entre le dirigeant et d’autres parties prenantes, est

une question cruciale (Charreaux, 1998) et nous cherchons à démontrer, conformément aux

approches culturelles, que cette question est également imbriquée dans le contexte matériel,

ici la structure générale de l’actionnariat. Il se trouve que la littérature financière avait

essentiellement considéré la question de la relation entre cycles d’affaires et structure de

capital dans le cadre traditionnellement scruté en finance, celui de l’actionnariat dispersé

(De Jong, et al., 2008; King et Santor, 2008 ; Holderness, 2010). Ce cadre masquait des

relations antagoniques au niveau micro-économique, qui se compensent en agrégé.

Cela nous amène à un second commentaire, toujours en regard de la contribution de cet article

à la thèse. En scrutant un cadre géographique et culturel homogène (Ronen et Shenkar, 1985;

Hofstede, 2003), nous cherchons à circonscrire des aspects institutionnels au sens large, afin

d’étudier plus en profondeur une relation donnée, dans sa spécificité. Ce faisant, nous nous

efforçons, appliquant un conseil de DiMaggio (1994)7, d’identifier au sein de cette relation,

quelles dimensions peuvent être déterminées par des éléments matériels plutôt que culturels.

Ce processus nous amène à identifier qu’une part des comportements financiers, des

dirigeants et des membres des comités de direction, sont imbriqués dans un système de

relations institutionnelles diverses, « alignées » (au sens de Williamson, 2000). Ceci nous

conduit directement à notre troisième contribution à la thèse.

7 Ce conseil a trait à la mise en lumière des processus culturels. Il est repris et développé dans les parties II (Survey) et V (annexe) de la thèse.

Synthèse générale

33

3- La relation entre cycles d’affaires et structure du capital est-elle conditionnée par des facteurs structurels, institutionnels ou bien culturels ?

Notre question de recherche est également directement explicitée par le titre de l’article. Alors

que le chapitre précédent s’attachait à démontrer l’importance de déterminants principalement

structurels, c’est-à-dire la structure d’actionnariat, sur les comportements des dirigeants et par

conséquence sur les politiques financières observées en agrégé, nous nous attachons ici à

réexaminer ce lien.

L’une des conclusions de notre premier article est que les approches culturelles requièrent une

vision complexe de la causalité, reposant sur deux idées-forces. D’une part la culture se

présente souvent comme la cause indirecte et nécessaire, plutôt que la cause directe et

suffisante, pour un phénomène donné. D’autre part les phénomènes culturels sont des

contributeurs, in fine, d’une convergence plus générale des dimensions culturelles,

institutionnelles, socio-économiques.

Ce qui nous conduit aux deux citations proposées en exergue (page 46 en appendice). La

citation de Scott souligne cette idée, très présente également chez North (1990), Greif (2006)

ou Aoki(2001), que la culture est une notion fluide, difficile à mesurer et même à identifier,

mais qu’elle est fondamentalement liée à l’existence des institutions, leur pérennité et leur

trajectoire dynamique. Cette conception a pour conséquence qu’il est difficile d’étudier la

culture au niveau national et qu’il est difficile de l’étudier sans étudier en même temps les

institutions nationales. La seconde citation vient souligner la complémentarité fondamentale

entre les différents processus, politiques, économiques, sociaux, au niveau des nations, pour la

détermination des institutions « d’équilibre » (la notion d’équilibre est centrale chez Aoki,

voir par exemple notre contribution en annexe ou bien l’appendice 1 du chapitre II.)

Synthèse générale

34

Concernant la détermination des structures de capital pour les firmes européennes, nous

reprenons ainsi une conception générale, qui traverse différentes sciences humaines,

l’économie et la gestion en particulier, qui est que des observations faites sur un phénomène

donné, dans un cadre national, sont généralement dépendantes d’un ensemble de relais

institutionnels. Ainsi en ce qui concerne la gouvernance, Charreaux (2004) propose une

revue des différentes écoles de pensée pour lesquelles le contexte national, au-delà des

comportements spécifiques des actionnaires et des dirigeants, importe pour la mise en place

d’un cadre de gouvernement de l’entreprise spécifique. Dans une approche liée certains

auteurs anglo-saxons suggèrent que la gouvernance des entreprises résulte d’une « chaîne

institutionnelle » (« governance chain »), dont les chaînons variés se conditionnent l’un

l’autre, incluant par exemple la nécessité d’une presse d’affaires et de sociétés d’audit, libres

et puissantes (Dyck, 2001).

Dans ce contexte, nous cherchons à dépasser notre contribution précédente pour démontrer

comment les interactions structurelles et comportementales, qui lient cycles d’affaires,

structure d’actionnariat et politiques financières sont en fait inscrites sur un substrat légal,

structurel, institutionnel et culturel, qui transcende les différences nationales, pour ce qui

concerne évidemment notre échantillon ouest-européen8.

A ce stade, il est important de mentionner que l’analyse des déterminants des structures de

capitaux, au niveau national, révèle de très grandes disparités d’un pays à l’autre (appendice

5, chapitre IV). L’analyse de ces différences nationales mériterait un examen approfondi et

constituera certainement une des pierres angulaires de nos prochaines étapes de recherche.

Pour le moment nous nous attachons à appliquer le conseil de DiMaggio (1994 : voir

chapitres I et VI) : nous cherchons à identifier quels éléments sous-jacents, structurels,

8 L’appendice 1 du chapitre IV reprend notamment des éléments de l’échantillonnage. Il est notable que nous sommes partis d’un échantillon contenant toutes les bourses sur le territoire européen, de Dublin à Kiev. Cependant, ayant fixé un seuil d’au moins 25 sociétés cotées en moyenne par année, pour chaque bourse retenue, seule la Pologne figure au final dans notre échantillon en compagnie des pays ouest-européens.

Synthèse générale

35

institutionnels ou bien culturels peuvent expliquer, au-delà des fortes disparités nationales

constatées in concreto, une part de convergence dans les structures de capitaux à travers le

continent.

Au final, nous cherchons à illustrer sur ce cas précis l’idée selon laquelle les normes de

comportements, essentiellement rationnelles et calculatoires, des normes financières, sont

imbriquées dans un faisceau de dimensions convergentes. Lorsqu’elles divergent, cela

correspond souvent au réalignement de cadres institutionnels variés qui les polarisent. Ainsi,

la diversité des structures s’en trouve relativement plus réduite en pratique qu’on pourrait le

penser à priori, car les comportements s’alignent de fait pour converger vers un certain

nombre d’équilibres économiques et sociaux (Ragin, 2000).

Cette approche particulière nous conduit à décliner systématiquement l’usage de la

polarisation d’échantillon. Notre objectif est de démontrer que des variables purement

exogènes aux politiques financières de l’entreprise, contextuelles, s’associent de façon

contradictoire aux équilibres observés dans les structures de capitaux. Ainsi nous restaurons la

dichotomie observée dans l’article précédent, dans la relation entre politiques financières et

cycles d’affaires, en fonction des structures d’actionnariat.

Il en résulte que cet article s’articule autour d’un grand nombre d’hypothèses et se concentre

sur la mise en lumière de corrélations, plutôt que de relations de causalité au sens habituel

pour la recherche en finance. Son objectif premier est de mettre en lumière l’alignement d’un

nombre de dimensions contextuelles. De ce point de vue il cherche à appliquer une approche

empirique différente de celles mises en œuvre dans le premier (approche qualitative et

conceptuelle) et le second article (approche plus traditionnelle pour la finance d’entreprise).

Son objectif principal est de restituer, dans un format conditionné par l’exercice académique

par article, le besoin de faisceau convergent de preuves, habituellement recherché dans les

approches néo-institutionnelles.

Synthèse générale

36

Notre première série d’hypothèses procède du constatât de l’article précédent, que la

polarisation de la relation entre cycles d’affaires et politiques financières s’associe à une

politique de dividendes lissés, dans le cas d’une relation contra-cyclique, caractérisant les

sociétés à actionnariat concentré. Dans l’autre cas, c’est plutôt la variable dérivée de Baker et

Wurgler (2002), soulignant l’effet cumulé du « market timing », qui est significative plutôt

que celle de la politique de dividendes.

Nous dérivons de ces observations l’idée que le « market timing » caractérise principalement

les sociétés à actionnariat dispersé ou bien les pays dont les droits des actionnaires sont

relativement plus développés que ceux des créditeurs. L’idée sous-jacente repose sur le lien

d’une correspondance institutionnelle tacite entre structures d’actionnariat et structures

macro-financières : banques vs. marché (Shleifer et Vishny, 1997 ; Black, 1998 ; Aggarwal et

Goodell, 2009). Dans un cadre institutionnel où le financement bancaire est relativement plus

développé, on observe fréquemment un faisceau convergent de pratiques qui incluent

notamment une concentration de l’actionnariat, une association des créditeurs au capital sous

différentes formes (e.g. banque-industrie ou keiretsu), un renforcement relatif des droits des

créditeurs, un rapport de clientèle plus développé dans la politique de dividendes, un horizon

d’investissement plus long pour les investisseurs institutionnels (Ruiz-Mallorqui et Santana-

Martin, 2010), un engagement direct des foyers sur les marchés plus limités (Guiso, et al.,

2001).

Dans ce contexte nous suggérons trois mécanismes. Premièrement, un mimétisme entre petits

actionnaires et actionnaires de référence et la présence d’actionnaires de référence

relativement bien informés, peuvent conduire à limité les fluctuations des cours de marché

autour d’une valeur fondamentale de long terme. Deuxièmement, l’intérêt marginal du

« market-timing » est réduit, la proportion des minoritaires alimentant l’intérêt de la pratique

étant faible et le nombre d’outils « de spoliation» alternatifs étant élevé. Troisièmement, la

Synthèse générale

37

présence forte et l’association d’institutions de crédit à la gestion des entreprises conduit à

remettre plus en phase gestion de la dette et gestion des capitaux propres, l’émission de

capitaux pouvant conduire à un accroissement parallèle du crédit.

Notre seconde série d’hypothèses découle directement de l’article du chapitre précédent.

D’une part, nous montrons que les résultats portant sur la polarisation des politiques

financières, observées sur des variables différenciées, se retrouvent sur des variables non-

différenciées : les leviers comptables et le levier de marché des entreprises.

D’autre part, cette relation vérifiée, nous suggérons que les facteurs sous-jacents qui

conduisent à cette inversion des politiques financières sont également des modérateurs de la

relation. Les facteurs sous-jacents retenus étaient centrés sur la latitude des dirigeants,

l’importance de l’information économique de ces dirigeants et l’impact des coûts d’agence

dirigeant-actionnaires. Nous suggérons que la portée potentielle de ces facteurs est modérée

par la transparence et la renommée des entreprises. Lorsque transparence, renomée et suivi de

l’entreprise par les analystes s’accroissent, la portée de ces facteurs, spécifiques aux

dirigeants, diminue. Ainsi nous suggérons que la polarisation pro- vs. contra-cyclique est

modérée par la taille, la complexité, l’âge et l’ancrage de marché des entreprises.

Troisièmement, nous suggérons que la relation d’agence spécifique, avec suivi et encadrement

des dirigeants, est secondaire à d’autres facteurs pour les sociétés dont les résultats sont très

volatiles. Ces sociétés manquent de transparence. D’autres mécanismes sont souvent en place,

notamment l’association des différentes parties prenantes au capital (c’est d’ailleurs le cœur

de cible des activités de capital-risque). Nous suggérons donc que la relation pro-cyclique,

caractéristique du besoin de suivi et d’encadrement (« monitoring ») est modérée par ces

contextes.

Synthèse générale

38

Notre troisième série d’hypothèses procède de considérations culturelles, avec trois niveaux

concentriques de raisonnement. Nous testons premièrement un lien direct entre culture et

structure du capital ; nous testons deuxièmement un lien indirect ; et troisièmement nous

procédons par association d’idées en retenant une approche similaire à celle de Stulz et

Williamson (2003), fondée sur les systèmes de valeur (« worldviews »).

Chui, et al. (2002) identifient des antécédents culturels directs pour la structure du capital des

entreprises, en se fondant sur les indices proposés par Schwartz (sections II.3 et IV.2.3). Pour

des raisons de disponibilité de cet indice, nous procédons à une réplication de leurs

prédictions en utilisant les indices de Hofstede comme instruments.

De Jong, et al. (2008) notent qu’il est important de bien distinguer les effets directs et

indirects de la culture sur les structures de capital. Il se trouve que la littérature financière a

établi une correspondance forte entre un indice culturel de Hofstede, l’évitement de

l’incertitude (« uncertainty avoidance ») et la structure macro-financière, banques vs. marchés

(Kwok et Tadesse, 2006; Aggarwal et Goodell, 2009). Nous avons insisté sur l’importance

fondamentale que revêt pour nous la structure macro-financière. Cependant, comme pour la

culture, il est difficile de bien différencier effets directs et indirects ; Gonzales et Gonzales

(2008) montrent par exemple que les effets de la concentration bancaire sont primordiaux

pour la structure du capital, indépendamment de la structure macro-financière.

En conséquence, nous testons l’hypothèse que l’évitement de l’incertitude polarise les

politiques financières, c’est-à-dire la relation ente la structure du capital et les cycles

d’affaires anticipés. Cette hypothèse est dérivée d’un double raisonnement, direct : les nations

qui ont une forte tendance à l’évitement de l’incertitude favoriseront les politiques contra-

cycliques (guidées par l’anticipation des risques), alors que les autres favoriseront les

politiques pro-cycliques (guidées par l’optimisation et structurées par le rapport d’agence

Synthèse générale

39

dirigeant-actionaires). L’hypothèse découle également d’un raisonnement indirect, où l’indice

culturel sert d’instrument pour l’identification de la structure macro-financière sous-jacente.

Enfin, dans un troisième temps, nous suggérons que les politiques pro-cycliques, conformes à

la théorie financière orthodoxe et dérivées de l’étude du « Role Model » étasunien, seront

plutôt associées à des contextes culturels correspondant au positionnement étasunien sur

différents indices de culture nationale.

Les résultats que nous obtenons sont, dans l’ensemble, conformes aux hypothèses proposées.

Une fois encore, il est intéressant de constater que la polarisation de l’échantillon a un effet

systématiquement médiateur sur la relation entre le contexte économique et financier et la

structure du capital des entreprises. Cette dichotomie des politiques financières semble en

effet se retrouver la plupart des dimensions structurelles, institutionnelles et culturelles

proposées.

Ainsi un premier résultat de recherche de cet article, découlant directement des approches

culturelles et institutionnelles, est que l’analyse statistique doit s’appuyer non seulement sur

des « variables de contrôle » robustes, mais également sur une relation robuste à la

« polarisation des échantillonnages ». On retrouve dans notre analyse des structures de

capitaux en Europe, c’est-à-dire dans un système moins paradigmatique et plus mélangé que

le système étasunien, un thème ancien du passage de la micro- à la macro-économie. Ainsi

l’étude de l’investissement des entreprises en agrégé a-t-il longtemps achoppé sur la question

des différents motifs qui guident l’investissement au niveau micro-économique et qui se

compensent dans les analyses agrégées.

Deuxièmement, on retiendra que l’analyse de l’interaction entre le contexte économique,

financier et les structures de capitaux est peut-être victime de ce biais particulier. Ainsi ce

thème de recherche est-il contraint d’une part par l’endogénéité de nombre de variables

Synthèse générale

40

spécifiques à la firme, et d’autre part, par l’ambigüité du rapport entre le contexte et les

politiques financières. Cette ambigüité ne peut être que la conséquence naturelle d’une vision

plus comportementaliste de la gouvernance des entreprises (Charreaux, 2004, 2005).

Ce sont ces considérations qui nous conduisent à réinterpréter la citation de Myers proposée

en exergue au chapitre III : « There is no universal theory of capital structure, and no reason

to expect one. There are useful conditional theories, however […] each factor could be

dominant for some firms or in some circumstances, yet unimportant elsewhere ». Il est

possible que l’importance du contexte institutionnel au sens large ait été relativement négligée

par le passé , la plupart des études se concentrant sur le cas étasunien, ou bien sur des

échantillons présentant une diversité de structures politiques, économiques et institutionnelles

trop importante pour être bien analysée comparativement. Auquel cas, les circonstances

changeantes dont parle Myers résulterait de facteurs structurels, institutionnels ou culturels à

identifier, ici une forte disparité dans les structures d’actionnariat, elle-même apparemment

alignée sur nombre de facteurs sous-jacents. On peut donc envisager que derrière la

conditionnalité apparente des relations, il existe des régularités de comportement dont les

déterminants restent à identifier.

Troisièmement, en réponse à la question de recherche de notre titre, notre conclusion peut être

partiellement frustrante. Il ne nous semble en effet pas possible de bien dichotomiser les

différents déterminants structurels, institutionnels et culturels de la structure de capitaux, car

leurs interrelations sont multiples, soumises à une maturation commune, à une dépendance de

sentier, etc. L’analyse doit procéder d’un faisceau d’études et de recherches interdisciplinaires

et favoriser à la fois des analyses diachroniques et synchroniques : Aoki (2001 : chapitre 1)

souligne l’importance, pour l’analyse des institutions, d’aborder leur étude à la fois dans une

approche comparative et dans une approche historique focalisée sur le changement pour le

cadre institutionnel donné.

Synthèse générale

41

4- Articulations entre les articles, positionnement et conclusions

Nous avons déjà amplement souligné comment chacun des articles, deux et trois, se

complétaient, en prolongeant le premier article, le « Survey ». Il est intéressant de noter que

nous considérons le thème de la structure du capital comme potentiellement l’un des thèmes

féconds en finance, pour approfondir les recherches institutionnelles et, ou culturelles (voir

l’article complémentaire, en annexe). En effet, en termes culturels, les dirigeants des

entreprises représentent une population relativement hétérogène, probablement plus proche

des normes abstraites représentées par les « populations nationales » que celle, par exemple,

des opérateurs de marché (les « traders », etc.). Il s’agit donc d’étudier les normes de

comportement d’opérateurs dont les références cognitives et comportementales, d’ordre

national, sont plus directement opérantes.

Dans ce cadre-là, nous nous intéressons très spécifiquement à distinguer dans le pilotage des

entreprises, des aspects qui relèvent purement d’une technique financière, d’aspects qui

seraient plus proprement comportementaux, quels qu’en soient les sources normatives

(contrats, incitations, normes institutionnelles ou culturelles, rapports de pouvoirs, biais

comportementaux, etc.). Nous appliquons là très directement deux conclusions importantes du

« Survey », reprises dans l’annexe, tout en appliquant ci-dessus un conseil de DiMaggio, que

nous avons mentionné dans la synthèse pour notre second article. Il nous semble en effet

intéressant de traiter la finance comme un « système culturel », en cherchant à identifier des

déterminants spécifiques pour les comportements des agents financiers. Dans cette entreprise,

comme dans d’autres recherches de nature culturelle, il sera essentiel, pour identifier une

influence culturelle, de démontrer que les causes matérielles ou structurelles sont insuffisantes

Synthèse générale

42

à restituer le mécanisme considéré9. C’est d’ailleurs sur ce type d’approche que Guiso, et al.

(2009), pour la confiance, ou bien encore Shiller (1997), avec une ambition d’une portée plus

générale10, fondent leur démarche méthodologique pour une approche culturelle des

phénomènes financiers.

Ainsi notre premier article sur les structures du capital (le second article de la thèse) s’attache

à démontrer qu’une modification structurelle des rapports d’agence, entre les principales

parties prenantes à la firme, peut changer significativement les normes qui structurent les

comportements. Il semble acquis que la question de la confiance, très directement soulevée

par Charreaux (2005) en rapport aux variations nationales dans les structures de gouvernance

des entreprises, est un préambule indispensable pour comprendre la question du

« monitoring ». Ceci explique d’ailleurs que nous ayons proposé de traduire ce mot par l’idée

« d’encadrement et de suivi de l’activité du dirigeant », plutôt que par l’idée de « contrôle ».

Dans les structures d’actionnariat très concentré, il y a une forme de contrôle effectif,

complétée par un alignement fort entre les intérêts des principaux acteurs. Dans l’actionnariat

dispersé, l’idée de contrôle prend une acception fort différente : c’est justement une forme

d’absence de contrôle et la divergence palpable des intérêts qui implique la nécessité de

contrats précis ex ante, pour aligner les intérêts, de normes de gestion rigoureuses, pour

encadrer le dirigent et la mise en place d’un comité fort et indépendant, pour permettre le

suivi. Dans ces deux cas abstraits, la notion de confiance se traduira par des acceptions et des

comportements probablement fort différents.

Dans notre second article sur les structures du capital (le troisième article de la thèse) nous

nous attachons à redémontrer que ces comportements financiers, structurellement divergents,

9 Pour reprendre une part du conseil de DiMaggio: “one must demonstrate that such differences [resulting from a cultural effect] do more than mediate structural or material influences” (1994: p. 28) 10 “One difficulty that these researchers [working on the cultural determinants of behaviors] have encountered with experimental work is that of disentangling the “rational” reasons for the imitation of others with the purely psychological” (1997 : 26)

Synthèse générale

43

sont alignés sur un grand nombre de relais institutionnels au sens large. Nous appliquons ici

directement la conception qui est celle de Williamson (2000) sur le nécessaire alignement des

schémas sociétaux ou bien celle des principaux partisans des « systèmes nationaux ».

Pour conclure, nous souhaitons positionner notre travail. C’est une démarche assez rare dans

les thèses en finance, mais elle est fréquente pour certaines thèses en gestion et c’est souvent

la norme pour les thèses en sociologie.

La thèse présentée ici peut s’analyser comme faisant partie d’un ensemble assez large de

recherches qualifiées de néo-institutionnalistes, communes à plusieurs disciplines (Immergut,

1998). On peut également la rapprocher d’un ensemble de recherches en gestion, en théorie

des organisations, en économie ou en sociologie qui considèrent et analysent les différences

entre nations en termes de « systèmes nationaux », alternativement des systèmes sociaux

d’innovation et de production (Amable, 2005), des systèmes sociaux de production (Streeck,

1992; Hollingsworth et Boyer, 1997), des systèmes d’affaires (Hamilton et Biggart, 1988 ;

Guillen, 1994 ; Whitley, 2000), des variétés de capitalisme (Hall et Soskice, 2001), de

l’analyse institutionnelle historique et comparative (North, 1990 ; Aoki, 2001 ; Greif, 2006)11.

Dans le prolongement de ce type d’approche, nous nous intéressons tout particulièrement à

l’importance relative de la finance, du « monde de la finance » et de son influence structurante

sur l’économie, sur les économies nationales, notamment en référence à l’analyse des

structures macro-financières (Mayer, 1990). D’autres références sont mentionnées à cet effet

dans le corps de la thèse, concernant notamment la financiarisation de l’économie et de la

gestion des entreprises. A ce titre, nous proposons à la fin de notre premier chapitre une

interprétation de nos résultats sur les structures des capitaux : cette inversion des politiques

financières pourrait correspondre à ce qu’on pourrait nommer, schématiquement, une

11 Charreaux (2004) reprend d’ailleurs la présentation de certaines de ces écoles, dans le cadre de son article de synthèse sur la gouvernance des entreprises.

Synthèse générale

44

approche financière à la gestion de la firme (la gestion pro-cyclique) d’un côté, et une

approche industrielle et commerciale à la gestion de la firme (la gestion contra-cyclique) de

l’autre. Cette analyse est cependant réalisée sans jugement normatif sur les performances

relatives de ces deux schémas théoriques. Enfin, cette analyse n’est pas sans rappeler des

arguments spécifiques à la transformation, ces dernières décennies, des approches

managériales aux Etats-Unis (e.g. Fligstein, 2001: chapter 7) et dans le monde (Aoki, 2001:

chapter 11 & 13).

Aussi, nous souhaitons refermer cette synthèse en reprenant la conclusion de l’un des articles

identifiés dans le cadre du « Survey » : il semblait alors fort improbable, il y a 35 ans, à un

groupe de chercheurs éminents, que l’approche de la gestion de l’entreprise, par la valeur

actionnariale, soit vouée à la pérennité qu’on lui connait:

Looking ahead, it seems unlikely that financial executives outside the United States would adopt the stockholder wealth maximization model even if they were completely familiar with the academic justification of it. National, cultural and institutional factors are unlikely to allow such an occurrence.

Stonehill, et al., in conclusion (1975: 32)

* * *

Synthèse générale

45

5- Appendice

Nous reprenons ici les citations proposées en exergue aux différents chapitres

Chapitre II “La notion de culture est inhérente à la réflexion des sciences sociales. Elle leur est nécessaire, en quelque sorte, pour penser l'unité de l'humanité dans la diversité, autrement qu'en termes biologiques.” Cuche (2004: 1)

Chapitre III “There is no universal theory of capital structure, and no reason to expect one. There are useful conditional theories, however […] each factor could be dominant for some firms or in some circumstances, yet unimportant elsewhere.” Myers (2003: 216-217)

Chapitre IV “Institutions … are embedded in various types of repositories or “carriers” […there are…] three types of carriers:… cultures, social structures, and routines.” Scott (1995: 52) “[To] investigate the institutional diversity and the complexity of economies [you need to look] into the nature of the interdependencies of institutions across economic, political, organizational and social domains, as well as that of those institutions linking those domains.” Aoki (2001 3)

* * *

Synthèse générale

46

6- Bibliographie

Abbott, A., 1988. The system of professions: an essay on the division of expert labor. Chicago: The University of Chicago Press.

Aggarwal, R., Goodell, J.W., 2009. Financial markets versus institutions in European countries: influence of culture and other national characteristics. International Business Review,

Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: national characteristics as determinants. Journal of Banking & Finance, 33(10): 1770-1780.

Aglietta, M., 2004. Dérives du capitalisme financier. Paris: Albin Michel. 394 pages.

Amable, B., 2005. Les cinq capitalismes. Diversité des systèmes économiques et sociaux dans la mondialisation. Paris: Seuil. 374 pages.

Antoniou, A., Guney, Y., Paudyal, K.N., 2008. The determinants of capital structure: capital market oriented versus bank oriented institutions. Journal of Financial and Quantitative Analysis, 43: 59-92.

Aoki, M., 2001. Toward comparative institutional analysis. Cambridge: MIT Press. 467 pages.

Baker, M., Wurgler, J., 2002. Market timing and capital structure. Journal of Finance, 57(1): 1-32.

Baskerville, R.F., 2003. Hofstede never studied culture. Accounting, Organizations and Society, 28(1): 1-14.

Black, B.S., Gilson, R.J., 1998. Venture capital and the structure of capital markets: Banks versus stock markets. Journal of Financial Economics, 47(3): 243.

Charreaux, G., 1996. Pour une véritable théorie de la latitude managériale et du gouvernement des entreprises. Revue Française de Gestion, 111(special issue): 50-64.

Charreaux, G., 1998. Le role de la confiance dans le systeme de gouvernance des entreprises. (Part Played by Trust in the Corporate Governance. With English summary.). Economies et Societes, 32(8-9): 47-65.

Charreaux, G., 2004. Les théories de la gouvernance: de la gouvernance des entreprises à la gouvernance des systèmes nationaux. Cahier du FARGO n° 1040101.

Charreaux, G., 2005. Pour une gouvernance d'entreprise 'comportementale'. Une réflexion exploratoire. Revue Française de Gestion, 4(157): 215-238.

Charreaux, G., Schatt, A., 2006. La recherche française en finance: une perspective à travers les travaux des enseignants-chercheurs en gestion sur la période 1994-2003. Revue Finance Contrôle Stratégie, 9(1): 239-271.

Synthèse générale

47

Chui, A., Lloyd, A., Kwok, C., 2002. The determination of capital structure: is national culture a missing piece to the puzzle? Journal of International Business Studies, 33(1): 99-127.

Cuche, D., 2004. La notion de culture dans les sciences sociales. Paris: La Découverte - Repères. 124 pages.

De Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance, 32(9): 1954-1969.

DiMaggio, P.J., 1994. Culture and economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology. Princeton: Princeton University Press. 835 pages: p. 27-57.

Dyck, A., 2001. Privatization and Corporate Governance: Principles, Evidence, and Future Challenges. World Bank Research Observer, 16(1): 59-84.

Faccio, M., Lang, L.H.P., 2002. The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3): 365-395.

Fan, P.H.J.P., Titman, S., Twite, G.J., 2008. An international comparison of capital structure and debt maturity choices. SSRN eLibrary.

Fligstein, N., 2001. The architecture of markets: an economic sociology of twenty-first century capitalist societies. Princeton: Princeton University Press. 288 pages.

Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: which factors are reliably important? Financial Management, 38(1): 1-37.

Gonzalez, V., Gonzalez, F., 2008. Influence of bank concentration and institutions on capital structure: new international evidence. Journal of Corporate Finance, 14(4): 363-375.

Greif, A., 2006. Institutions and the path to the modern economy: lessons from medieval trade. Cambridge University Press. 503 pages.

Guillen, M.F., 1994. Models of management. Work, authority and organization in a comparative perspective. Chicago: University of Chicago Press. 424 pages.

Guiso, L., Haliassos, M., Jappelli, T., 2001. Household portfolios. Cambridge: MIT press. 508 pages.

Guiso, L., Sapienza, P., Zingales, L., 2009. Culture biases in economic exchange? Quarterly Journal of Economics, 124(3): 1095-1131.

Hall, D.P.A., Soskice, D.W., 2001. Varieties of capitalism: the institutional foundations of comparative advantage. Oxford: Oxford University Press. 560 pages.

Hamilton, G.G., Biggart, N.W., 1988. Market, Culture, and Authority: A Comparative Analysis of Management and Organization in the Far East. The American Journal of Sociology, 94(Supplement): S52-S94.

Hofstede, G., 1994. Preface. In Kim, U., et al. (Ed.), Individualism and collectivism. Theory, methods and applications. New-York: Sage publications. 348 pages: p. ix-xiii.

Hofstede, G., 2003. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

Synthèse générale

48

Holderness, C.G., 2010. The myth of diffuse ownership in the United Staes. Review of Financial Studies, 22(4): 1377-1408.

Hollingsworth, J.R., Boyer, R., 1997. Contemporary capitalism: the embedness of institutions. Cambridge: Cambridge University Press.

Immergut, E.M., 1998. The theoretical core of the new institutionalism. Politics & Society, 26(1): 5 - 34.

King, M.R., Santor, E., 2008. Family values: ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, (32): 2423-2432.

Klep, P.M., 2001. 'Reculturalisation' in economic and social history? Deel, 64: 6-27.

Kwok, C., Tadesse, S., 2006. National culture and financial systems. Journal of International Business Studies, 37(2): 227-247.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance, 54(2): 471-517.

Lévesque, B., Bourque, G.L., Forgues, E., 2001. La nouvelle sociologie économique. Desclée de Brouwer. 268 pages.

MacKenzie, D., Millo, Y., 2003. Constructing a market, performing theory: the historical sociology of a financial derivatives exchange. American Journal of Sociology, 109(1): 107-145.

Margaritis, D., Psillaki, M., 2009. Capital structure, equity ownership and firm performance. Journal of Banking & Finance, 34: 621–632.

Mayer, C., 1990. Financial systems, corporate finance, and economic development. Hubbard, R. (Ed.), Asymmetric Information, Corporate Finance and Investment. Chicago: Chicago University Press

Myers, S.C., 2003. Financing of corporations. In Constantinides, G., Harris, M., and Stulz, R. (Ed.), Handbook of the Economics of Finance: Corporate Finance. Elsevier North Holland

North, D., 1990. Institutions, institutional change and economic performance. Cambridge: Cambridge University Press. 159 pages.

Ragin, C.C., 2000. Fuzzy-set social science. Chicago: Chicago University. 370 pages.

Ronen, S., Shenkar, O., 1985. Clustering countries on attitudinal dimensions: A review and synthesis. The Academy of Management Review, 10(3): 435.

Ruiz-Mallorqui, M.V., Santana-Martin, D.J., 2010. Dominant institutional owners and firm value. Journal of Banking and Finance, in press

Scott, W.R., 1995. Institutions and Organizations. Thousand Oaks, CA: Sage.

Shiller, R.J., 1997. Human behavior and the efficiency of the financial system. Handbook of macroeconomics

Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance, 52(2): 737-783.

Synthèse générale

49

Smelser, N.J., Swedberg, R., 1994. The sociological perspective on the economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology.

Stonehill, A., et al., 1975. Financial goals and debt ratio determinants: a survey of practice in five counries. Financial Management, 4(3): 27-41.

Streeck, W., 1992. Social institutions and economic performance : studies on industrial relations in advanced european capitalist countries. London: Sage.

Stulz, R.M., 2005. The limits of financial globalization. Journal of Finance, 60(4): 1595-1638.

Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): 313-349.

Whitley, R., 2000. Divergent capitalisms: the social structuring and change of business systems. Oxford University Press.

Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature, 38(3): 595-613.

Wiltbank, R., Dew, N., Read, S., Sarasvathy, S.D., 2006. What to do next? The case for non-predictive strategy. Strategic Management Journal, 27(10): 981-998.

* * *

Synthèse générale

50

51

Culture, finance, and national institutions

 Three  essays  in  corporate  finance  

   

by  Charles-­‐Henri  J.  REUTER    

Ph. D. Dissertation

       Thank  you  to  my  teachers  and  professors  Thank  you  for  teaching  me  to  learn,  teaching  me  to  love  to  learn  and  supporting  me  step  after  step.  To   Claude   Berthaud,   Claude   Simon,   Patrick   Messerlin,  Charles  Goodhart,  Sidney  Winter  &  Mauro  Guillen  

 First  and   foremost,   thank  you  so  much   to  Paul  DiMaggio,  and  to  Franck  Bancel    

       

52

     

 When  desire  evades  the  early  blows  of  common  sense,  you  are   looking   for   reasons.   And   you   find   some,   with   no  bearing.   The   truth   is   that   you   cannot   name   what   drives  you.   Something   grows   inside   of   you,   until   the   day,   you  unconfidently   leave   for   good,   seeing   the   docks   behind.   A  journey   needs   no  motive.   Soon   it   turns   out   as   being   self-­sufficient.  And  when  you  are  settling  on  your  trip  away,  it  is  your  trip  really  that  settles  you,  or  unsettles  you.  

Nicolas  Bouvier,  L’Usage  du  Monde  (1963)1    

     

1 Translation proposed by the author of this dissertation for the benefits of non-French-speakers.

53

Abstract

This thesis is composed of three essays. The first essay adopts a qualitative approach, making

a largely conceptual contribution. The two other essays adopt an empirical approach, and

focus on the determinants of financial policies and capital structures for European listed firms.

Architecture of the thesis: the first essay, entitled ‘a Survey of culture and finance’

emphasizes that special attention should be paid to the varying sources that are causing

national differentiation. At the same time, it highlights that these sources - within the cultural,

institutional and socio-economic spheres - converge to create differentiation. This essay forms

a foundational structure for the empirical research, as well as defining the relationship of this

essay to the other two essays and sections of the thesis. The second essay focuses on an

underlying factor of financial research: agency relations, which, it turns out, cause

differentiation in financial policies across European firms. The third essay shows that this

specific financial mechanism largely overlaps with a large number of underlying contextual

dimensions.

The First essay: ‘the Survey’, details the screening process that has been performed on peer-

reviewed journals in finance, in order to investigate the recent rise of interest in “cultural

approaches”. The aim has been to let definitions emerge, in order to build a field-based

analysis about culture in finance, and about its operationalization.

Abstract

54

The results of that screening process, include the following: first, the concept of “culture” is

mainly connected to the notion of “national cultures”, and, hence, the “country puzzle”.

Secondly, the Survey demonstrates that “culture” is used in a very large range of financial

topics. Thirdly, a marked polarization in approaches is noted. One approach, e.g.

dimensionalism, with its roots in management-research, stands out; while other research areas

provide, on the whole, only fuzzy and ill-defined definitions of “culture”, with no consistency

in approach. As a consequence, the review has been extended to cover all dimensionalist work

that could be identified within the field of finance. This provides the fourth result: the only

framework with current generality, i.e. dimensionalism, is not likely to be adopted in a

unifying capacity within the foreseeable future. Fifthly, the Survey highlights the potential

fruitfulness of considering organized financial know-how as a type of “cultural-system”.

As a result, the research focuses on a subject where the country puzzle is looming: capital

structures, and their underlying financial policies. The sample is conducted in a way that is

consistent with qualitative and institutional approaches: focusing on developed economies,

with some qualitative convergence in judicial, political, institutional, legal and socio-

economic terms. While the sampling of European listed firms has surprisingly limited

precedence (for the study of capital structures), it enables the effects of the variance in firms’

governance and ownership structures to be magnified.

The second essay: ‘do agency relations mediate the interaction between firms’ financial

policies and business cycles?’ investigates the interactions between firms’ financial policies

and expected business cycles. In traditional financial theory, business cycles are of limited

relevance for the understanding of the mechanisms of capital structures. A number of other

dimensions are considered to be of greater importance, such as: agency costs, market timing

and signaling; in past empirical studies, business cycle-proxies are not robust.

Abstract

55

In this essay, two conflicting mechanisms are evidenced, which coexist in aggregate: firms

with dispersed ownership lead pro-cyclical policies, while firms with strongly concentrated

ownership, or under control, lead contra-cyclical policies. The theoretical considerations in

this research unfold from the idea that ownership dispersion implies a different mix in agency

relations within each firm. It entails specificities in agency costs and opportunity benefits of

managerial discretion, and it fosters differing needs for disciplining through debt and

signaling, and, potentially, different market timing behaviors at management and incumbent

shareholder levels.

Overall, this thesis posits the argument that contra-cyclical effects are generally

underestimated in the literature, because they are of limited relevance in the paradigmatic

dispersed-ownership role model, which is characterized by broad, liquid stock markets,

institutionally-coordinated transparency, and robust shareholder rights.

The third essay: ‘Is the relation between business cycles and leverage mediated by

structural, institutional or cultural factors?’ demonstrates that both the pro- and contra-

cyclical mechanisms are mediated by structural, cultural and institutional factors.

Specifically, it is shown that for contextual dimensions, corresponding to the U.S context,

some pro-cyclicality is observed, while there is limited, or no, contra-cyclicality (so the first,

pro-cyclical mechanism dominates in aggregate). This observation is reversed when the

contextual dimensions are of an opposite form to that of the U.S (the second and opposite

mechanism dominates). Furthermore, it encompasses varying dimensions of capital structures

(cyclicality of leverage, market timing, dividend policies) and it extends to the following

dimensions of context: ownership dispersion, institutional anchoring, transparency and risk,

firms’ structural variables, the quality of the contracting environment, or still, measures of

national cultures, particularly those related to bank-based macro-financial systems, known to

proxy for the distinction between bank- vs. market based macro-financing structures.

Abstract

56

Additionally, the third essay complements existing institutional approaches, evidencing

relatedness in underlying contextual dimensions, by reference to the Survey; at the same time

complementing the second essay, by showing that for a primarily financial mechanism, the

driving relation between financial policies and business cycles is in fact related to the broader

institutional environment; hence, the financial mechanism seems to be supported by an

institutional alignment. Interpreting findings in these three essays, an annex argues that the

financial management of a firm could be interpreted through a cultural lens, and assessed as if

it were, in fact, a “cultural-system”.

* * *

Chapter I - Introduction

57

FIRST CHAPTER: INTRODUCTION

DETAILED TABLE OF CONTENTS 1.   Genesis 59  2.   The Survey and implications 62  

2.1.   The Survey 62  2.2.   Dimensionalism and further dissertation-essays 63  2.3.   National cultures, national institutions and national structures 63  

3.   Two empirical essays on capital structure 65  3.1.   European listed firms and pooling: methodological stakes 65  3.2.   Do agency relations mediate the interaction between firms’ financial policies and

business cycles? 67  3.3.   Is the relation between business cycles, and leverage, mediated by structural,

institutional or cultural factors? 68  3.3.1.   Abstract   68  3.3.2.   Concluding  commentary   69  

4.   A conceptual addition 70  5.   Conclusion: positioning, tentative interpretation and directions 71  

5.1.   Positioning 71  5.2.   Tentative interpretation 71  5.3.   Conclusion and directions 73  

6.   Bibliography 74  

 

 

* * *

 

Chapter I - Introduction

58

 

   

Chapter I - Introduction

59

   

 

Culture, finance, and national institutions

 

“If one tries to transgress the borderlines of national cultures, one piece of intellectual luggage, that has to be left at home, is the division of labor among social sciences as it has been developed in Europe and North America in the past hundred years [...] Psychology, social psychology, sociology, anthropology, economics and political sciences, all study only facets of the same social reality. Academic inbreeding and atomization in the West have led to extensive production of irrelevant speculations. The system has become self destructive in that it punishes rather than encourages borrowing from related disciplines. Cross-cultural social sciences, therefore, can not be, but cross-disciplinary”.

Hofstede (1994: ix)

“We would expect countries, per se, to matter little in finance. However, this is not the case: countries [are] very important”. “A firm’s country of incorporation is a more important determinant of its financial policies than its industry [This generates] paradoxes that are explored in many papers. [We need] a new theory of international finance that recognizes that countries are relevant” Stulz (2005: 1601 & 1596)

1. Genesis    

The two quotes reproduced above, underline the tensions generated by the idea of national

cultures in financial academia and cognate research disciplines. Hofstede refers to the

extensive production of irrelevant speculations, surrounding the concept of national cultures;

and Stulz underlines the existence, in financial research, of puzzles and paradoxes connected

to the practical prevalence of countries’ boundaries. Further related discussions will be

presented later in the thesis, which are inclusive of, but not limited to, the central place of the

“U.S. Role Model” in financial research (De Jong et al., 2008: 1956; see chapter IV), and to

the paradigmatic “diffuse ownership” model (Holderness, 2010). Furthermore, considering

the high research-profile reached by the law and finance program, contributions by Stulz and

Chapter I - Introduction

60

Williamson (2003), and by Licht et al. (2005), cast doubt on the practical application of the

law and finance findings; to paraphrase Hofstede, laws, cultures, religions and world-views

may be only facets of the same financial and economic realities within nations; so the precise

causality, from law to finance - a building stone in the law and finance research-program - is

obscured by other institutional and/or cultural influences. Furthermore, this convergence of

dimensions within nations, is likely to be intertwined with socio-political histories (Rajan and

Zingales, 2003). This is, indeed, one of the key facts derived from the first essay (chapter II),

A Survey of Culture and Finance. As a result, from the preliminary focus on Culture and

Finance, the dissertation has been extended to be more encompassing, to include: Culture,

Finance, and National Institutions.

Thus, the dissertation is resultant of a number of complex influences, including a specific

interest in the broader European situation, the candidate’s specific professional, academic and

social experiences, and eventually, the initial PhD ‘roadmap’, as agreed between the candidate

and his director, Franck Bancel. To this initial roadmap, the candidate has brought an

extensive cross-disciplinary research education (economics, finance, management, sociology),

as well as his strong international background (academically, professionally, personally). This

diversity in perspectives, showing in many places in the essays to follow, has been difficult to

harness in order to follow a specific line of inquiry. The first essay of this PhD thesis, the

Survey [chapter II], considers, as the starting point, the following research question:

Is it possible to characterize one or more accepted theoretical background(s) for culture-research, in finance?

The question is answered negatively, which implies that the dissertation cannot follow the

usual routes, anchored around the choices of an appropriate theoretical framework, the

derivation of testable hypotheses, and the ultimate testing with suitable data. This bears a

number of consequences for the architecture of the dissertation.

Chapter I - Introduction

61

In the first instance, the Survey provides more than a simple review of the literature. Firstly,

the chapter sets up a specific research protocol, inspired from Wiltbank et al. (2006) in

strategic management, and Charreaux and Schatt (2006) in finance. Secondly, the Survey

evidences that: in most cases, prominent international researchers, such as Stulz, Tabellini and

Zingales, succeeded in publishing on the subject of culture in finance. Furthermore, it

suggests that this field of inquiry is fraught with perils, and is unsuited for junior researchers;

in part because the cultural stance is foundational to most social sciences (Cuche, 2004), and

because it raises a number of important epistemological issues.

The second and third PhD essays are focused on a specific empirical topic: the study of the

determinants of leverage in Europe. They illustrate the conclusions reached by the Survey,

each applying a different methodological stance to the subject. Both essays directly address

some of the key conclusions of the Survey, including, in particular, considerations about the

intertwining of cultural and institutional factors, as well as methodological considerations.

In parallel, the thesis follows the direction set by the Survey, developing a conceptual

contribution, presented in annex (section V). This addition is original when assessed by the

standards of financial research. Yet, it echoes methodological proceedings in management

(Payne et al., 2009). Furthermore, it presents a truly inter-disciplinary stance, while

addressing some epistemological considerations. These two research activities,

interdisciplinary scholarship and epistemology, generally characterize more senior research

enterprises, involving maturity and perspective. Because the additional essay (VI-annex)

provides insights into the interrelation between the three essays (chapter II to V), and because

it provides some architecture for future research, it is added to the main essays of the

dissertation, and presented in the annex (section V).

The next section explains how conclusions, derived from Survey’s results, lead to research

choices made in the two other essays (chapter III and IV); an outline is made of the purpose

and details of the conceptual addition, and concludes with a tentative interpretation of the

findings.

Chapter I - Introduction

62

2. The  Survey  and  implications  

2.1. The  Survey    

As emphasized, the Survey of Culture and Finance (chapter II) is more than a review of the

literature. It provides evidence that there is still no consensual theoretical framework in

finance, despite a strong rise in interest for the concept. At the same time, this rise in interest

is large and specific enough to establish what could be characterized as a new research-area

within finance. This contrast, between a consistent interest, and the distinct lack of

framework, provides us with a problematic. Overcoming it, with circumstantiated evidence,

provides us with our research focus.

In the Survey, peer-reviewed journals are screened to investigate the recent rising interest in

cultural approaches. The purpose is to let definitions emerge from the screening process,

building a field-based analysis of culture in finance and associated operationalization.

Importantly, the discussion focuses on the scientific production of financial researchers,

searching for theoretical and conceptual references. The transparency, and the neutrality, of

the research-protocol proposed, is a central tenet in this approach (section II.2.2). It could be

said that a culturally centered ethnographic approach has been undertaken in order to analyze

the production of the finance-research community. This enables an assessment to be made of

the nature and reach of the association between culture and finance.

The specific results of this research are delineated in the Survey-abstract (at the beginning of

chapter II). This section includes evidence of the rise in interest previously discussed, as well

as a large range of investigated subjects, and the polarization of research. The resultant

implications are multifold. In particular, an emphasis should be made of the strong

methodological tension between cultural approaches, and the traditional research methods

used in finance. Secondly, ideological, conceptual and political stakes are identified. These

stakes reflect, in part, different disciplinary orientations in economics, management and

sociology. Thirdly, the nature of debates about culture, in finance, is intrinsically connected to

the national puzzle, as exemplified by René Stulz’s quote reproduced at the beginning of the

thesis.

Chapter I - Introduction

63

2.2. Dimensionalism  and  further  dissertation-­essays  

However, the results reached in the Survey exemplify how the use of national culture indices

(i.e. the dimensionalist framework, see section II.3.1) emerges as the only reference, with

some commonality across a number of research articles published in financial journals.

However, the main screening, which has been complemented by the double review process,

(details in section II.2.2.3) exemplifies how dimensionalist research articles appear to face

special challenges when it comes to publication in financial journals (stricto sensu). Besides,

those that reach the financial publication seem to adopt a syncretic approach.

So, the Survey’s results echo earlier findings by Baskerville (2003), who notes that

economists and sociologists have not been keen on adopting dimensionalism. It is noteworthy

that debates and controversies are surrounding the emergence, or the continuation, of this

framework as a mainstream reference in the proximate field of International Business (see

sections II.5.1, II.5.II and V). This provoked some uneasiness when considering the building

of further PhD-essays directly and exclusively on this framework, notwithstanding the fact

that this research has already addressed a very large range of financial topics (Table 2).

2.3. National  cultures,  national  institutions  and  national  structures  

This section will emphasize three conclusions that can be derived from the first essay of this

thesis, and that essentially structure the novel approach taken in the two empirical articles, and

explain how they relate to one another.

Firstly, it is noteworthy that sociologists strongly recommend differentiating cultural from

structural phenomena (DiMaggio, 1994, p. 28). Often, an observed phenomenon can be the

product of power relations, network effects, or still the result of economic interactions. Hence,

it can be wrongly attributed a cultural label (Klep, 2001). Thus, special attention needs to be

paid to varying sources that can cause national differentiation in financial phenomena [This

proposition is labeled P1, to be addressed below].

Chapter I - Introduction

64

Secondly, cultural approaches often convey a sense of causality that is not linear. Often,

culture is the necessary condition that enables further cause and effect relationships. In this

sense, a cultural approach should focus on moderating and mediating influences, as a priority,

among other issues. [P2].

Thirdly, there is an intimacy between groups, their cultures and institutions, the structures that

are set up as a result, which often support and reinforce one another (Jepperson and Swindler,

1994). Therefore, a cultural stance will often favor a situation, where a number of

institutional, or structural, dimensions are aligned with the underlying cultural dimensions.

This results in limited diversity (Ragin, 2000), so that a given national phenomena can be

attributed to a number of apparently independent factors (further commentaries in

section V.2.2) [P3].

These considerations are instrumental in explaining part of the frequent reticence expressed

by economists and sociologists against the dimensionalist framework. They explain the strong

polarization of the current interest in culture identified in finance, probably, in part, due to the

lack of a theoretical framework with sufficient support across the discipline; and these

considerations are likely to be related to the fact that only very senior researchers, or research-

groups, actually publish on the subject of culture in finance.

Addressing this puzzle, through the three underlying causes suggested [i.e. P1-to-P3], is one

of the primary objectives of the two complementary, empirical essays, and the core objective

of the conceptual essay, added in the annex (section V). In short, an intention is made to open

the black box of national embededdness (see section II.5.3), through a specific focus on one of

the central themes in corporate finance: firm’s capital structures.

Chapter I - Introduction

65

3. Two  empirical  essays  on  capital  structure  

In the empirical essays presented in this thesis, a focus is placed on capital structures for

European listed organizations. Some of the conclusions are reached in the first essay -

conclusions, which are have been further formalized in the conceptual addition (VI-annex).

3.1. European  listed  firms  and  pooling:  methodological  stakes  

It is noteworthy that, in the past, only very few studies focused on comprehensive large-scale

samples of European listed firms. In fact, in this thesis, an identification is only made of,

either, studies based on comprehensive international samples, with dozens of countries, or

studies based on a very limited number of countries, belonging to strikingly different

cultural/institutional backgrounds (see sections III.2.1 and IV.2.1.1).

The developed sampling procedure is a direct reflection of the deeper cultural/institutional

concern at the heart of the research: sampling from European listed firms, in a comprehensive

way, limits variation in many key underlying dimensions, such as economic development,

degrees of social protection, unionization processes, accounting, judicial mean-quality, socio-

economic histories, polities, and so on. In short, the sampling philosophy, with its surprisingly

limited past occurrence, is a key device, and a mainstream method in qualitative studies for

addressing causality and multicollinearity [P2].

More specifically, the two empirical essays presented in this thesis are two complementary

analyses of the relation between business cycles and firms’ capital structure.

It is noted that the past financial literature is mainly based on U.S data, and it identifies a

moderately robust, but consistently positive relationship between business cycles and firms’

capital structure (Frank and Goyal, 2009). On this count, an identification is made of, not one,

but two, underlying relations between business cycles and firms’ capital structures1, so that 1 Chiefly, we identify a positive relation, apparently connected to demand-side drivers and associated with dispersed ownership, while a negative relation, connected to supply-side drivers, is associated to concentrated ownership / firms under control.

Chapter I - Introduction

66

there is a conflicting relation with both positive and negative effects that need to be untangled;

this is the primary focus of the first empirical essay (chapter III), while the second empirical

essay (chapter IV) shows that both these two polar relations are intimately related to a

differentiation in the structural, institutional and cultural contexts [P3].

Furthermore, it is noteworthy that in most of the existing literature on capital structures,

dispersed ownership is the dominant working hypothesis (King and Santor, 2008). Thus, the

past literature has ultimately focused on U.S. data (Frank and Goyal, 2009). When it has not,

it has focused on samples, where potentially multiple underlying relations can be obscured by

the dominant shareholder-value-creation model, intimately connected to multinationals

enterprises, dispersed ownership, legal bonding arguments, and a sampling dominated by

financially-managed, large groups, that are listed in many countries (Bancel, 2007 for a

critical view on the causal interpretation of multiple listing; see further comments in

section I.5.2).

Therefore, one primary purpose of the sampling undertaken, is to limit the bias attributable to

the over-representation of the modern multinational corporation - that are highly profitable

and financially managed (Aoki, 2001; Fligstein, 2001) - so as to allow the underlying

diversity in capital structure determinants to re-emerge.

Both essays make extensive use of pooling. The systematic use of pooling, based on varying

contextual dimensions, enables for the testing of the continuity of relations [P2]. In both

articles, it is noted that relations obtained in aggregate are misleading, because they reflect the

combination of underlying and opposite relations. Furthermore, the pooling reveals that,

beyond national differences in Europe, similar structural drivers seem to be at work [P1].

These structural drivers, starting with the degree of ownership dispersion, tend to overlap with

legal (relatively higher creditors’ rights), macro-financial (preference for bank-financing), and

cultural factors (higher uncertainty avoidance, cultural difference to the U.S) [P3].

Chapter I - Introduction

67

3.2. Do  agency  relations  mediate  the  interaction  between  firms’  financial  policies  

and  business  cycles?    

The initial empirical article retains a micro-level approach, based on flow-variables, or yearly

changes in stock variables, and focuses on financial policies. It aims at building a robust

theoretical argument, derived from mainstream financial theories, where the observed

polarization of the relationship between business cycles and capital structures can be derived

from different sets of agency relations, based on ownership dispersion.

Specifically, the essay investigates the interactions between firms’ financial policies and

expected business cycles, in listed firms in Europe, over a 20-year period. It has shown that

these interactions are mediated by ownership structures. Firms with strongly concentrated

ownership, or that are under control, lead contra-cyclical policies; while firms with dispersed

ownership lead somewhat pro-cyclical policies, supporting the traditional expectation, based

on the dispersed ownership role model, that business cycles are of little direct relevance for

financial policies.

The theoretical considerations unfold from the idea that ownership dispersion implies a

different mix in agency relations within the firm. It entails specificities in agency costs,

opportunity benefits of managerial discretion, and it fosters differing needs for disciplining

through debt, for signaling, and potentially different market timing behaviors by managers

and incumbent shareholders.

Our contribution extends, potentially, to a number adjacent research questions, including,

among others, the analysis of performance effects of ownership concentration, the relative

assessment of governance mechanisms, a focus on specific behaviors as part of capital

structures research, or the mediating effect of agency relations for analyzing the effects of

business cycles - and financial shocks - on firms’ financial policies.

Chapter I - Introduction

68

3.3. Is  the  relation  between  business  cycles,  and  leverage,  mediated  by  structural,  

institutional  or  cultural  factors?    

The second empirical article retains a macro-level approach, based on the identification of the

determinants of corporate book- and market-leverage. It revisits the relationship evidenced in

the first empirical article, to show that it can be equivalently obtained from varying

perspectives, such as: legal, institutional, cultural and structural.

3.3.1. Abstract

In traditional financial theories, business cycles are of limited relevance for the understanding

of capital structure mechanisms. A number of other dimensions are more important (agency

costs, market timing, signaling, and so on), and business cycle proxies are not robust in

empirical studies (Frank and Goyal, 2009). In chapter III, it is shown that two conflicting

mechanisms coexist: a pro-cyclical one, related, among other things, to the primary

importance of the disciplinary role of debt. The other, contra-cyclical, mechanism is to be

observed among firms with concentrated ownership. It is derived from risk-related

considerations, comparative benefits and costs of entrenchment or discretion, and it is fostered

by incentive-alignments. This thesis argues that the contra-cyclical effect is underestimated in

the literature, because it is of limited relevance in the paradigmatic U.S. role model with

broad and liquid stock markets, strong shareholders’ rights, and dispersed ownership.

It is shown that, for contextual dimensions corresponding to the U.S. context, some pro-

cyclicality is observed, while there is limited or no contra-cyclicality (the first mechanism

dominates in aggregate). This observation is reversed when contextual dimensions are

opposite to that of the U.S. (the second and opposite mechanism dominates). Further, it

encompasses varying dimensions of capital structures (cyclicality of leverage, market timing,

dividend policies) and it extends to the following dimensions of context: ownership

Chapter I - Introduction

69

dispersion, institutional anchoring, transparency and risk, firms’ structural variables, the

quality of the contracting environment, or still measures of national cultures, particularly the

ones related to bank-based, macro-financial systems.

3.3.2. Concluding commentary

Hence, the relatedness is emphasized in varying contextual, national dimensions [P3], while

trying to untangle the cultural/institutional from the structural factors [P1]. The conclusions

obtained in the first empirical essay are revisited to show that the financial reasoning based on

ownership dispersion and agency relations, overlaps with considerations pertaining to the

relative prevalence of signaling, market timing, the relative strength of stock- vs. credit-

holders, institutional anchoring of firms, or still to the larger cultural context [P3]. In any case,

there are underlying conflicting relations that need to be considered carefully [P2].

Institutional approaches put emphasis on sampling and the development of converging bodies

of evidence to circumvent multicollinearity, and to circumscribe what can be described as a

limited diversity in human designed institutions and policies (Ragin, 2000; and see sections

V.1.1 and V.3.1). Eventually, the second empirical essay does not enable a clear

disentanglement of the varying influences at national levels [P1]. This is no coincidence, and

this may be due to the pervasive influence that national boundaries have, and which Stulz

labels the “country puzzle in finance” (see further section I.5.1).

Chapter I - Introduction

70

4. A  conceptual  addition  

In the additional essay, presented in the annex (section V), the propositions to P1-to-P3,

presented above, are extended to a conceptual contribution.

It is noteworthy that the candidate is indebted to the teaching of Paul DiMaggio on the subject

of culture-processes. The study of culture is a significant pillar of Paul DiMaggio’s research

(DiMaggio, 1994, 1997), however, it is not always recognized as such, particularly in the

international setting, possibly because his institutional theories have been so widely diffused

(DiMaggio and Powell, 1983). Paul DiMaggio suggests that the interrelation between cultures

and institutions is a dimension of primary relevance; it is intricate and it is complex.

Furthermore, he notes that many cultural sources influence individuals and communities,

rather than nations. Similarly, at the national level, many sources, other than cultural, will

cause differentiation, and seemingly cultural phenomena may be triggered by material,

structural or institutional influences (DiMaggio, 1994).

Those relatively consensual sociological findings provide an additional challenge for this

dissertation. Indeed, the cross-national focus of culture-related work in finance is the only

common thread across the different articles identified through the screening-process (section

2.2.3). Exploring and reducing this challenge is a key theme to the conceptual contribution.

The research departs from the advice provided by DiMaggio, to the culture-scholar, and a

conceptual dichotomy (a two-by-two matrix) is proposed, to segregate financial subjects into

more homogenous categories, in terms culture types (professional culture, national culture,

and so on), and relevant methodological approaches. The key message is that one of the

primary contributions that the culture framework adds to financial research is a much-needed

debate on the potential additions that are currently lacking from the traditional methodological

toolkit.

Chapter I - Introduction

71

5. Conclusion:  positioning,  tentative  interpretation  and  directions  

5.1. Positioning  

It is customary in some academic disciplines, to discuss the intellectual positioning of a PhD

dissertation. We suggest that our dissertation could be categorized, either by the range of

research perspectives, belonging to new institutionalism, across disciplines (Immergut, 1998),

or by the research perspectives featuring national differences as a result of systems,

considered alternatively as business systems (Biggart, Hamilton, Whitley), varieties of

capitalism (Hall, Soskice), institutional variations (Aoki, Fligstein, Greif, Scott), social

systems of production (Amable, Boyer, Hollingsworth, Streeck), or systems resulting from

macro-financing structures (Carlin, Mayer), etc.2

5.2. Tentative  interpretation  

To make an illustration, it is proposed, here, that, in closing this chapter, a tentative and

controversial interpretation to the findings is made. Importantly, these following comments

should not be understood as the conclusion to the dissertation-essays. What is provided, here,

is a tentative interpretation, which can serve the purpose of further empirical exploration, or

an opportunity for scientific falsification.

The empirical findings of this research, revisited with the specific lens provided by the Survey

and by the conceptual addition, echo historical work undertaken on the successive dominant

managerial approaches, and the financial view of the firm, in the U.S (e.g. Fligstein, 2001:

chapter 7), and beyond (Aoki, 2001: chapter 11 & 13). Overall the idea of finance, as a

culture, should be taken seriously and explored further (section II.4.3.3).

To simplify, given our time frame and use of mostly Western European data, the suggestion

could be made, that there is an opposition between two styles, two cultures, for governing the

2 An extended and inter-disciplinary review, in the context of governance systems, is provided by Charreaux (2004). Only names of seminal researchers are tentatively mentioned here. More specific references are provided below in the body of the dissertation.

Chapter I - Introduction

72

firm; and the suggestion is put forward that they be labeled, for the sake of argument, the

financial vs. the industrial governance styles. The agency relations of managers-with-

shareholders, the monitoring and signaling, all structure the financial management style; the

company needs to be lean (section III.2.2), to enable proper monitoring and to limit the

opportunity cost of expropriation.

At the opposite end of the polarity, the industrial management style has a longer time-

horizon; it favors risk mitigation over short-term optimization, and it does not respond as

sensitively to changes in economic outlook. Because of this slack, industrial management

styles need alternative institutional pillars, and require financing by banks or families, because

there is a need for investors with longer-time horizons (as in Ruiz-Mallorqui and Santana-

Martin, 2011); investors with stronger monitoring, or controlling, abilities; and investors who

have a stronger incentive-alignment with managers (Shleifer and Vishny, 1997), than in the

other abstract model. Of course, this would drive a need for differentiated financial policies,

in particular relating to dividends (dividend clienteles and dividend catering, as in Baker and

Wurgler, 2004, but based on institutional, rather than behavioral criteria).

In this speculation, culture takes a technical, or professional, meaning: finance is a technically

organized system of meanings and interpretations. Furthermore, it is part of an organized

professional system, with reach beyond borders, in an economic (e.g. technical efficiency),

political (e.g. lobbying) or cultural (e.g. appeal of this interpretative frame) sense. To re-

emphasize, in closing this prospective statement, the financial management of the firm, in the

historical perspective, can be geographically and institutionally traced to the U.S. As a matter

of fact, 35 years ago, to a large team of prominent scholars, it appeared very unlikely to

spread:

Looking ahead, it seems unlikely that financial executives outside the United

States would adopt the stockholder wealth maximization model even if they were

completely familiar with the academic justification of it. National, cultural and

institutional factors are unlikely to allow such an occurrence.

Stonehill, et al., in conclusion (1975)

Chapter I - Introduction

73

5.3. Conclusion  and  directions  

Further exploration of the functionality of these systems, based on specific financial themes,

will be the focus of some of the research, likely to be pursued in the future.

Firstly, the empirical findings, relating to the analysis of European-specific financial policies,

open challenging debates, and will be pursued further. A number of specific propositions can

be derived from the tentative interpretation of the results (section I.5.2). They will be

addressed scientifically. This will involve questioning the potential interrelation between

financial policies, ownership structures (in particular long-term shareholders, or families),

performance, and the general socio-economic and institutional context.

Secondly, the conceptual framework (section V) will be developed further, using

epistemological approaches derived from management research.

Thirdly, a primary conclusion, derived from the Survey, the conceptual addition, and the

effect of pooling in the two empirical essays, is that unconventional methods (for financial

research) can be fruitfully used in financial research. A specific review of unconventional

approaches, and their further use, including, in particular, qualitative methods, are on the list

of the interesting possible directions for future research.

Lastly, an exploration of the idea of financial cultures (section II.4.3.3) will be made in

greater depth, keeping in mind their recent success and foundation is assured, at least in part,

due to their practical efficiency, deductive rationality and relative independence of softer

dimensions, often prevalent in national cultures.

Chapter I - Introduction

74

6. Bibliography  

Aoki, M., 2001. Toward comparative institutional analysis. Cambridge: MIT Press. 467 pages.

Baker, M., Wurgler, J., 2004. A catering theory of dividends. Journal of Finance, 59(3): 1125-1165.

Bancel, F., 2007. La cotation des titres Europeens aux Etats-Unis: une approche critique (The listing of European companies in the US: a critical approach). Revue d'Economie Financiere, 89: 143-162.

Baskerville, R.F., 2003. Hofstede never studied culture. Accounting, Organizations and Society, 28(1): 1-14.

Charreaux, G., 2004. Les théories de la gouvernance: de la gouvernance des entreprises à la gouvernance des systèmes nationaux. Cahier du FARGO n° 1040101.

Charreaux, G., Schatt, A., 2006. La recherche française en finance: une perspective à travers les travaux des enseignants-chercheurs en gestion sur la période 1994-2003. Revue Finance Contrôle Stratégie, 9(1): 239-271.

Cuche, D., 2004. La notion de culture dans les sciences sociales. Paris: La Découverte - Repères. 124 pages.

De Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance, 32(9): 1954-1969.

DiMaggio, P.J., 1994. Culture and economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology. Princeton: Princeton University Press. 835 pages: p. 27-57.

DiMaggio, P.J., 1997. Culture and cognition. Annual Review of Sociology, 23: 263-287.

DiMaggio, P.J., Powell, W.W., 1983. The iron cage revisited: institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2): 147-160.

Fligstein, N., 2001. The architecture of markets: an economic sociology of twenty-first century capitalist societies. Princeton: Princeton University Press. 288 pages.

Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: which factors are reliably important? Financial Management, 38(1): 1-37.

Hofstede, G., 1994. Preface. In Kim, U., et al. (Ed.), Individualism and collectivism. Theory, methods and applications. New-York: Sage publications. 348 pages: p. ix-xiii.

Holderness, C.G., 2010. The myth of diffuse ownership in the United States. Review of Financial Studies, 22(4): 1377-1408.

Immergut, E.M., 1998. The theoretical core of the new institutionalism. Politics & Society, 26(1): 5 - 34.

Chapter I - Introduction

75

Jepperson, R.L., Swidler, A., 1994. What properties of culture should we measure? Poetics, 22(4): 359-371.

King, M.R., Santor, E., 2008. Family values: ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, (32): 2423-2432.

Klep, P.M., 2001. 'Reculturalisation' in economic and social history? Deel, 64: 6-27.

Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2005. Culture, law, and corporate governance. International Review of Law & Economics, 25(2): 229-255.

Payne, G.T., Moore, C.B., Griffis, S., Autry, C., 2009. Social capital research in management: a review of conceptualizations, measurements, and implications. Invited for revision for special annual review issue of Journal of Management. (Invitation Received: 3- 13-09).

Ragin, C.C., 2000. Fuzzy-set social science. Chicago: Chicago University. 370 pages.

Rajan, R.G., Zingales, L., 2003. The great reversals: the politics of financial development in the twentieth century. Journal of Financial Economics, 69(1): 5.

Ruiz-Mallorqui, M.V., Santana-Martin, D.J., 2011. Dominant institutional owners and firm value. Journal of Banking and Finance, 35(1): 118-129.

Shleifer, A., Vishny, R.W., 1997. A Survey of Corporate Governance. Journal of Finance, 52(2): 737-783.

Stonehill, A., et al., 1975. Financial goals and debt ratio determinants: a survey of practice in five counries. Financial Management, 4(3): 27-41.

Stulz, R.M., 2005. The limits of financial globalization. Journal of Finance, 60(4): 1595-1638.

Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): 313-349.

Wiltbank, R., Dew, N., Read, S., Sarasvathy, S.D., 2006. What to do next? The case for non-predictive strategy. Strategic Management Journal, 27(10): 981-998.

* * *

Chapter I - Introduction

76

Chapter II: A survey of ‘culture and finance’

77

SECOND CHAPTER: A SURVEY OF ‘CULTURE AND FINANCE’

DETAILED TABLE OF CONTENTS

1.   Introduction 80  1.1.   Research question 80  1.2.   Organization of the Survey 83  

2.   Background, stakes and methodology 85  2.1.   Defining culture? 85  2.1.1.   Culture-­definitions:  definitional  tactics   85  2.1.2.   Recent  definitions  of  culture  in  economics:  illustrating  epistemological  stakes   86  

2.2.   Screening and identification process 88  2.2.1.   Choosing  journals   88  2.2.2.   Time-­line  and  identification  (main  screenings)   90  2.2.3.   The  double  review  process   92  

2.3.   Classification, trends and first results 93  3.   Dimensionalist approaches and dimensionalist successes 96  

3.1.   Some background on the dimensionalist approach 96  3.1.1.   The  label   96  3.1.2.   Main  sources  and  indices   97  3.1.3.   Cultural  distance   98  3.1.4.   Antecedents  and  disciplinary  path   98  

3.2.   Empirical successes for dimensionalism in finance 101  3.2.1.   Financial  flows  across  nations  and  international  investment   101  3.2.2.   Financial  macro-­structures   103  3.2.3.   Stock-­market  functioning   106  3.2.4.   Dimensionalist  antecedents  to  financial  laws   109  3.2.5.   Corporate  finance  policies   110  

4.   Further culture-research in finance 111  4.1.   Gravity- and other models of trade, as applied to financial flows 111  4.1.1.   “Familiarity  breeds  investment”     111  4.1.2.   Corporate  culture,  “social  capital”  and  merger  flows   115  4.1.3.   Culture,  trust  and  financial  flows   117  

4.2.   Three specific and successful path for culture-research in finance 120  4.2.1.   National  cultures  as  antecedents  of  investors’  rights   120  4.2.2.   Concluding  on  international  flows  and  the  antecedents  of  investors’  rights   121  4.2.3.   Culture  vs.  language  and  distance  in  portfolio  holdings   122  4.2.4.   Corporate  culture,  religiosity  and  investment   123  

4.3.   Anecdotic and less anecdotic references to culture, uncovered in the screenings 124  4.3.1.   The  “cultural  affinity  hypothesis”   124  4.3.2.   The  “country  puzzle”,  as  an  underlying  theme,  in  many  articles   125  4.3.3.   The  “culture  of  investing”   126  

Chapter II: A survey of ‘culture and finance’

78

5.   Emerging culture-views in finance: synthesis 128  5.1.   Dimensionalism in management research, critiques and trajectories 128  5.1.1.   Critiques   128  5.1.2.   Recent  trajectory  in  International  Business   129  

5.2.   Dimensionalism as an emerging framework for financial research? 131  5.3.   National embededdness in finance 132  5.3.1.   Opening  the  black  box  (opening  up  venues  for  culture  in  finance)…   132  5.3.2.   Using  all  the  tools?   134  

5.4.   Culture as layers and considering financial cultures 136  6.   Conclusion: shouldn’t financial research explore culture further? 139  7.   Appendix 1 (chapter II). Intuitive breadth of culture definitions 143  

7.1.   Synthesis 143  7.2.   Definitions in economics 143  7.3.   Definitions in finance 144  7.4.   Definitions in cross-cultural psychology 145  7.5.   Definitions in international management 146  7.6.   Definitions in political sciences 146  7.7.   Definitions in sociology 147  

8.   Appendix 2 (chapter II). Definitions for national culture indices 148  8.1.   Hofstede’s five dimensions 148  8.2.   Schwartz’s four dimensions 149  

9.   Appendix 3 (chapter II). Institutional control variables in the dimensionalist literature presented in our database 150  10.   Bibliography 152  

* * *

Chapter II: A survey of ‘culture and finance’

79

A SURVEY OF ‘CULTURE AND FINANCE’1

La notion de culture est inhérente à la réflexion des sciences sociales. Elle leur est nécessaire, en quelque sorte, pour penser l'unité de l'humanité dans la diversité, autrement qu'en termes biologiques.

Cuche (2004: 1)

In this research chapter, peer-reviewed academic journals in the area of finance have been

screened to investigate the recent rise in interest for “cultural approaches”. The aim has been

to let the definitions emerge from the screening process; thus, building a field-based analysis

about “culture” in finance, and its operationalization. The results of that screening are as

follows: firstly, the concept of “culture” is mainly connected to the notion of “national

cultures”. Secondly, the survey shows that “culture” is now used in a very large range of

financial research areas: from the study of international financial flows, to corporate financial

management, through to the analysis of the development, efficiency and robustness of stock

markets and financial macro-structures. Thirdly, a marked polarization has been noted: some

approaches rely on “national cultural indices” (dimensionalism), while others generally

provide no definition of “culture”. Taken together, these other approaches provide a fuzzy

view, and no consistent framework emerges to compete with dimensionalism. As a

consequence, the review, presented here, is extended to all dimensionalist work that can be

identified in the field of finance, which provides a fourth result: the only framework with

current generality in finance is not likely to be adapted as a unifying model anytime soon.

Overall, the research contributes to a preliminary roadmap for cross-referencing with other

disciplines; it suggests the adoption of backward definitions, and the building upon of

complementarities in existing approaches, including, in particular, dimensionalism, trust,

religions, or, still, the idea of “culture”, as layers, derived, in particular, from recent advances

in cognitive sciences.

1 This essay has been accepted by the review Finance, where it is forthcoming (June 2011). It has been presented at the following academic conferences: A.F.F.I. (spring 2009), E.F.M.A. (2010), the Groupe de Recherche of the C.N.R.S. “Monnaie Banque Finance” (2009), and INFINITI (2009), where it has benefited fro commentaries by reviewers and participants. It has been presented in the following research-laboratories: CEROS, ESCP Finance department and FARGO.

Chapter II: A survey of ‘culture and finance’

80

Methodological note. Our article rests on a research protocol based on the systematic analysis of the financial literature, and cross-referencing from adjacent disciplines. So we propose an indexation of all the relevant material2.

1. Introduction  

1.1. Research  question  

Despite the seemingly recent appeal of cultural concepts in financial research, it is gaining a

large audience. The studies by Stulz and Williamson (2003)*** and by Grinblatt and

Keloharju (2001)***, relating culture to investors’ protection, and to stockholding,

respectively, are now largely acknowledged by the financial community3. Citing classical

works by Greif, Lal, Landes, North, and even Max Weber - Stulz and Williamson (2003: 314,

references therein)*** emphasize their general belief that culture is highly important for the

effective study of financial phenomena. Ramirez and Tadesse (2007: 8)*++ note the growing

importance of references to “national culture indices”, and Hofstede, in the financial and

economic literatures. Some authors go even further. De Jong and Semenov (2002: 2)* suggest

that a specific “cultural view” should be developed, in part to compete with the law and

finance approach, which has blossomed since La Porta, et al. (1997, 1998). Breuer and

Quinten (2009)* suggest the establishment of “Cultural Finance” as an autonomous

discipline. Simultaneously, “cultural biases” have been posited by academics of Behavioral

Finance as early as 1999 (Shiller, 1999: 1), and those ideas are currently being more closely

considered (Statman, 2007, 2008). Yet, in finance, neither any review, nor any consistent

cultural framework, has been proposed to date.

2 We are indebted to an anonymous reviewer of the Revue Finance for this advice, which simplifies reading. Details on the method relating to the main screenings (of financial research-literature stricto sensu), the double review process (extensive review of dimensionalist advances in finance), and the identification of relevant material, are provided in section 2.2. *** Identified through our main screenings (eighteen articles) * Identified through the double review process (24 additions). ++ is a marking for articles, where at least one of the co-

authors has an established track record, in any of the journals screened as part of the main screenings (the J.C.R. list). *++ It results that *++ stands for articles from the financial community lato sensu

° Obtained through the initial screenings and analyzed separately due to the more limited relevance ∆∆∆ References to ‘foundational’ articles for dimensionalism ∆∆ References to seminal critiques for dimensionalism 3 Google-scholar’s citation index is at 408 citations for Stulz and Williamson and 381 for Grinblatt and Keloharju. The citation index on the EBSCO database stands respectively at 56 [102 on Science Direct] and 77 (2010, March 16th)

Chapter II: A survey of ‘culture and finance’

81

In this article, cultural research is surveyed within the field of finance, through a research-

protocol based upon the screening of peer-reviewed journals in finance (section 2.2 for details

and precedents). The survey utilizes the list of journals included in the Journal of Citation

Report [J.C.R.], for the computation of “impact factors”, now extensively used in the

academic profession. The time frame is a long-term horizon. It is dependent upon the database

used: ISI-Web-of Knowledge provides access to the list of the journals followed by the J.C.R.

and goes back to 1945 for some journals. The primary purpose of the survey has been to

answer the following two questions:

Q1. Can one or more accepted theoretical background(s) be characterized for cultural

research in finance, in order to derive testable hypotheses for empirical research?

Q2. Can ‘culture and finance’ be delimitated as a research-area within finance?

Through this process, an important commonality among articles can be characterized: the

articles mostly focus on cross-national comparisons. Investigating, collectively, a broad range

of financial topics, the articles assign observed differences in financial phenomena to

“national cultures”. However, a strong polarization is exemplified in the articles obtained

through the screening: a few approaches rely on “national-culture-indices” (dimensionalism,

definitions and references in section 3.1 and Appendices 1 & 2), while the other approaches

do not, and, taken together, they provide indefinite and inconsistent views on “national

cultures” and “culture” more generally.

Dimensionalist approaches can be traced back to Hofstede (1980)∆∆∆. They are based on

value-statements collected from individuals, in a large-scale survey. These value-statements

are averaged by country, and a factorial analysis provides four principal components, which

characterize four quantitative, and time-invariant, cultural characteristics, for each country.

Chapter II: A survey of ‘culture and finance’

82

Other approaches, which are identified through the screening process, generally provide no

definition of culture at all, but do only in a very few cases. These studies, that do, generally

use the concept of “culture” in combination with a number of institutional characteristics, by

country, to provide an explanation for the observed polarization of the financial data at

national levels. In a few cases, these non-dimensionalist articles empirically, or conceptually,

assimilate culture with other concepts, such as: institutions, religions, “societal trust” (“social

capital”), language and “worldviews”.

Overall, this polarization observed between the dimensionalist approach and other

approaches, seem to reflect varying research paths in adjacent disciplines - in particular, in

management or economics. It is shown below that the only framework, currently identifiable

in finance, is dimensionalist. Yet, the emergence of the dimensionalist framework in finance

is happening while, in adjacent disciplines, considerable debate is taking place about its

prospects. Therefore, it is argued that the current mobilization of such a framework, in

financial research, must be qualified.

So, the answers to the two questions, i.e. Q1 and Q2 presented above, appear negative: firstly,

there is no ready-to-use research-framework for empirical cultural research in finance.

Secondly, culture and finance cannot be considered as a research-area, yet. However, it is

expected that cultural research will soon be constituted into a research-area in finance, for at

least three reasons: interest within the discipline is identifiably strong; the number of financial

topics concerned is very large; and the normative stakes are high.

A primary purpose of this article is to facilitate this emergence, by providing a preliminary

cultural map to financial scholars. This map is established on the varying conceptualizations

of culture that emerge directly from the research-field, which are traced and rooted into

existing knowledge that is derived from adjacent disciplines.

Chapter II: A survey of ‘culture and finance’

83

1.2. Organization  of  the  Survey  

Although this section should start with a definition of culture, it does not. The reasons why,

are explained in section 2, which presents the methodology, database obtained (sixty hits,

eighteen articles for the main screenings) and some key facts that guide the Survey.

Sections 3 and 4 are dedicated to the review of the literature per se, as have been built up

from the main screenings, as well as from the double review process. The review is split into

two parts, to reflect the polarization in cultural approaches. Each part is structured by topics of

interest, focusing on the results obtained by each article, and their relation to culture.

We deal first with the dimensionalist framework (section 3), and we begin by briefly

introducing thirty years of management scholarship in and around that approach (section 3.1).

To provide a synthetic, and comprehensive view, on the progress of dimensionalism in

finance, we complete the main screenings by an extended review. This extension increases

our database from five to twenty-four articles, and preserves the nature our key points, and

further refines them (sections 3.2.)

Section 4 is concerned with non-dimensionalist articles. In a first sub-part, seven articles are

reviewed, in which culture emerges as a indefinite and inconsistent concept, while much

confusion is observed about the distinction between institutions and culture (section 4.1).

Then, three research-articles with original approaches to culture are considered, including in

particular approaches to: trust, religion, world-views and ethnicity (section 4.2). In section

4.3, we reflect on the initial sixty hits obtained in the survey, and we comment on three

additional articles that either reflect some of the hits, or follow cultural traditions that are

more common in other social sciences.

Chapter II: A survey of ‘culture and finance’

84

In section 5 we provide a synthesis, moving beyond the literature review obtained through the

screening. We propose an interpretation of the facts emerging from this grounded

investigation of the use of “culture” in financial research, and answer the two research

questions, Q1 and Q2, proposed above.

In section 5.1 we shed light on debates, and/or controversies surrounding dimensionalism in

management. Provided this background, we reflect on the current emergence and the potential

positioning of dimensionalism for financial research (section 5.2). Next, we delve into the

association, and/or the confusion, between cultural and institutional mechanisms. This sets

some of stakes for cultural research in finance (section 5.3). In section 5.4, we pursue

conceptions of culture not based on “national cultures”, mentioning how they are central, and

growing, in many disciplines. Incidentally, we delve into the rapidly expanding area of

economic sociology, concerned with “financial culture(s)”. Last we shed further light upon

conceptions of culture as “cognitive layers”, because they restore richness and variety in

cultural definitions, as has emerged from the survey, and is further detailed in appendix 1.

In section 6 we conclude by advocating a pragmatic and syncretic approach for cultural

research in finance. Our proposition is based on the idea of backward definitions and on a

careful and considered cross-disciplinary scholarship (as with Hofstede’s quote in exergue to

the dissertation).

* * *

Chapter II: A survey of ‘culture and finance’

85

2. Background,  stakes  and  methodology  

2.1. Defining  culture?  

2.1.1. Culture-definitions: definitional tactics

Schematically four main categories of meaning can be retained for culture, only one of them

being of interest here: the sociological meaning4. It broadly refers to values, norms or beliefs,

shared by some human group, reflecting its specificity. Precise definitions abound, and it is

customary to point to the seminal book by Kroeber and Kluckhohn (1952), listing 251

definitions and statements about culture. In appendix 1, an illustration is provided about the

variety in definitions, as derived from definitions obtained from the authors cited in this

Survey.

The additional essay provided in annex, emphasizes how the understanding of “culture” is

dependent upon a defined geographical or historical area. It is suggested that “definitional

tactics” are an underlying issue when defining culture: the definitions of culture retained by

specific actors, in specific communities, at specific times, reflect who actors are, how

communities interface, and the epistemological, political or social posture of these actors,

rather than providing definitional content, per se. As a preliminary illustration, the next

sections focuses on how the cultural definitions recently proposed in economics5, are resting

upon fundamental intellectual and epistemological postures.

4 The three “residual” categories are culture as in “agriculture” or “cell culture”, “culture as general knowledge” or “education” and “culture as art’. 5 We choose economics –as a meaningful illustration about the variance in definitions- because of the close historical association between finance and economics (Brennan, 1995).

Chapter II: A survey of ‘culture and finance’

86

2.1.2. Recent definitions of culture in economics: illustrating epistemological stakes

Three recent contributions about “culture and economics”, including two books, attest of the

timeliness (Throsby, 2001; Guiso, et al., 2007; De Jong, 2009∆∆∆) of the interest for this

concept in economics. The culture-conceptualizations retained by these scholars reflect the

difficulty in agreeing on a definition and underlying intellectual orientations.

Guiso, et al., focus on the idea of temporal stability to identify cultural elements (see their

definition in appendix 1). Doing so, they echo a conception expressed earlier by Williamson

(2000), who defines culture as one of the elements of the more general “institutions of

embededdness”, elements that are resulting in lasting constraints on the economy: they

operate “on the order of centuries or millennia” (Williamson, 2000: p. 5966).

Relating to a second definition of “culture in economics”, Throsby characterizes culture as a

process, delineated by the specificity of some human group vis-à-vis other groups (appendix

1). So, he retains a conceptualization akin to those in the North-American Sociology (Cuche,

2004). With his conceptualization, he allows for fluidity and change, elements that are, by

definition, excluded with Guiso, et al., or Williamson. Process-wise, instead of focusing on

durability to identify cultural elements, Throsby focuses on cross-sectional homogeneity.

Behavior is important here, as well as meaning. Both can be observed, and their relative

homogeneity (within a group) and heterogeneity (across groups) lead to the identification of a

“culture”.

Now, concerning the third definition, De Jong (2009)∆∆∆ retains a motivational aspect,

through the concept of “values”. He follows the path chosen in cross-cultural psychology (as

in the dimensionalist approach, see section 3.1)7. In this sense, cultures are enduring, as in the

first definition.

6 Williamson suggests three references to explore further what these “institutions of embededdness” comprise: DiMaggio (1994, “culture”), Granovetter (2006, “networks”) and Smelser and Swedberg (1994, “embededdness”). Interestingly relating to this latter concept, Smelser and Swedberg remark that “the concept of embededdness remains in need of greater theoretical specification” and they suggest that instead of providing an overall definition to start with, empirical investigation should focus on four “specific and separate sources of embededdness” directly: “cognitive, cultural, structural and political sources”. 7 Still another classic distinction contrasts conscious (symbols, discourses) vs. unconscious aspects of culture. It is an important distinction in the social sciences literature, as the popular and pervasive “iceberg conception” of culture attests.

Chapter II: A survey of ‘culture and finance’

87

Yet another important distinction emerges, between “constraining” and “enabling” views.

DiMaggio (1994, 1997) reminds us how the cognitive aspects of culture provide ground for

“enabling” views on culture: culture not only constrains, but structures and fosters behaviors,

communication, decision-making, etc. DiMaggio emphasizes how this distinction is central in

the epistemological structuring of scientific disciplines and in the choice of methodologies8.

In fact, he concludes that economics itself can be seen as a “cultural system” (1994: p. 46 and

47), a point that has been recently transposed to finance with considerable force (MacKenzie

and Millo, 2003, published a seminal article, espousing what is becoming known as the Social

Studies of Finance).

Larry Summers (1985: p. 1)° can be quoted to illustrate the point: “the fields of economics

and finance are allied. Indeed finance is sometimes defined as a subfield of economics [… ]

yet we are increasingly witnessing the development of two cultures [emphasis added]”. Larry

Summers’ statement may seem anecdotic, because it is not clear whether it assimilates finance

into something that could be described as a “cultural system”. Yet, it provides ground for

more than just familiarity, or knowledge of the financial field and theory. The distinction with

“economic culture” suggests some form of relative coherence that distinguishes the two sub-

fields in theory and in practice. This coherence excludes anecdotic views. In fact, this

statement may seem anecdotic precisely because of epistemological postures. The idea of a

coherent whole, guiding and structuring behaviors, is an “enabling” view of culture, so that

the idea of a financial culture may seem straightforward to a sociologist, while it is less clear

and apparently anecdotic for a “constraining” view of culture, as in economics. These issues

are further explored in sections 4.2, 4.3 and 5.4. 9

8 DiMaggio calls attention to the dichotomy between enabling and constraining views of culture. He notes a “rough analytic distinction between forms of culture that are characteristically constitutive (categories, scripts, conceptions of agency, notions of techniques) and forms that are predominantly regulatory (norms, values, routines)”. Then he underlines that most economists are in the second category (emphasizing regulatory aspects), treating “economic behavior as analytically distinct from culture, and stress[ing] the ways in which norms and conventions constrain the individual’s […] pursuit of self-interest”, while most anthropologists are in the first category (enabling views). This distinction is emphasized in conclusion to the book’s introductory chapter, on defining culture (De Jong, 2009∆∆∆: introduction). We illustrate further this distinction enabling / constraining with an ad-hoc example with “cultural affinity hypothesis” in section 4.3. 9 In their account of the reticence of the economic field, Guiso, et al., emphasize “in the last fifty years, most economists have been reluctant to rely on culture as a possible explanatory variable” (introduction). They attend to this lack of past interest, putting some of the blame on the “inevitable grittiness” of cultural variables, hardly reconcilable with “the parsimony of deductive models”, their “formal elegance”, the priority for “logical consistency over empirical relevance” and the

Chapter II: A survey of ‘culture and finance’

88

So, provided the complexity and stakes involved in a definition of “culture”, the primary

purpose of this article is to identify which views of culture are currently emerging in finance.

So, rather than providing an upfront definition, it is hoped that it emerges from the field, and

it is expected that this process will, in itself, be informative in many ways. So, a systematic

and scientific process has been established to do so; and it is this process that is described

next.

2.2. Screening  and  identification  process  

2.2.1. Choosing journals

As the purpose has been to favor a grounded emergence of cultural definitions from the

discipline, the social boundaries have been identified and used, in turn, to identify the

financial discipline: peer-reviewed journals of a certain prominence are rested upon, i.e. those

followed by the Journal of Citation Report (JCR). The JCR is a Thomson-Reuter product,

part of the ISI-Web of Science database, that produces and publicizes the “impact factors”,

now largely used in evaluating scholars and institutions. The database provides a category

labeled “Finance and Accounting” containing 52 journals, which need to be untangled.

Discriminating between the accounting and financial disciplines is essential. In accounting,

the interest in cultural approaches is strong, with an established history. It is characterized by

a plurality of accepted cultural frameworks, and debates and controversies among varying

streams of research. Besides, the sheer volume of references that the screening would uncover

in accounting-reviews forbids the mere possibility of reviewing the literature in this article.

Last, provided the strong differentiation between the two disciplines, the suggested process

aiming at letting definitions emerge from the field would become ambivalent.

methodological needs for “designing testable hypothesis” characterizing the economic field. Similar commentaries can be found in De Jong∆∆∆ (chapter 1 and 2). Together with DiMaggio’s Culture and the Economy, they set the epistemological context to whether or not culture variables should be considered or not in finance.

Chapter II: A survey of ‘culture and finance’

89

Chapter II: A survey of ‘culture and finance’

90

To resolve this issue, the JCR’s “Finance and Accounting” category is crosschecked against

available national rankings, that then provide the required distinction. Those categories

available for Germany (by VHB), France (by CNRS) and the U.K. (by ABS) are used

pragmatically, because they all provide this distinction between accounting and finance10.

Overall, 23 journals were identified for the financial research community stricto sensu (table

1). As a robustness check, and to account for potential “regional biases” (Charreaux and

Schatt, 2006), at a later stage, the screening was extended to financial journals that are

mentioned in all these three national rankings simultaneously, while they were not included in

the J.C.R. list. This extension provided only one additional article, and attests to the

robustness of the screening-process.

2.2.2. Time-line and identification (main screenings)

No specific timeline is set, because the survey mainly utilizes the EBSCO-database to

manually screen selected journals; thus, the timeline depends on the number of years included

in EBSCO, for each journal. The Journal of Finance is included since 1946, so 1946 is the de

facto starting point for the screening of the Journal of Finance. This is incidentally the longest

time-period for all financial journals in the list surveyed. Table 1 provides the number of

articles screened for each journal (see, third column).

The process follows a procedure used elsewhere; for instance, in strategic management

(Wiltbank, et al., 2006 for a recent example), or in finance (Charreaux and Schatt, 2006). The

process rests on the screening of abstracts, or titles, of articles, based on some key words.

Words such as ‘culture’, ‘cultures’, ‘cultural’ (shortened “cultur*”) have been searched for,

specifically. This screening results in the identification of a number of other related words that

could be used for further screening, such as ‘religion’, ‘trust’ or ‘language’.

10 Conveniently, Pr. Harzing provides a centralized source, including these three rankings just mentioned, as well as many others. See http://www.harzing.com/jql.htm.

Chapter II: A survey of ‘culture and finance’

91

Chapter II: A survey of ‘culture and finance’

92

However, we restrict ourselves to “cultur*” here for two reasons; first, because additional

words should be subjected to preliminary discussions, and, second, because we precisely wish

to let these connections emerge from the process11.

This process results in the identification of sixty hits, of which 11 are not research-articles, 13

articles can be considered as having an anecdotic use of “cultur*”, 3 articles refer to the

“cultural affinity hypothesis” (section 4.3.1), and 15 articles use specific expressions that are

not related to the core meaning of culture understood in this research, such as pop-culture,

corporate culture, technical cultures, and in particular “financial cultures” (sections 4.3.- and

5.4). Table 1 provides details on this process. The result is eighteen articles (including the

specific result from the regional screening), which are presented in table 2, with some details.

2.2.3. The double review process

As the dimensionalist framework emerges as the only consistent framework spanning several

articles in the database, a “double review process” is undertaken, completing the systematic

screening with an ad-hoc survey of dimensionalism in finance, which is based on the

systematic exploitation of cross-citations (received, made) within the 5 dimensionalist articles

initially obtained, and in any article subsequently identified. Eventually, the base of

dimensionalist articles is extended from 5 to 29 references12. They are detailed in table 3, and

presented with the review of dimensionalism in finance just before section 3.2.

11 An indicative subsequent screening was carried out on related words, as a robustness test. We note that there seem to be few additional informative hits. Note that screening on trust is impeded by ambiguity (investment trust, trust fund, etc.). 12 Overall we have here in the extended dimensionalist survey 29 articles, among those 5 were obtained from the initial screening***. Importantly, for each of the 4 financial areas mentioned above, but “corporate finance policies”, we have one, or two articles obtained from initial systematic screening. This means that extending the review to all dimensionalist articles that we can find in finance deepens the scope without modifying the key messages. It can be noted that among the additional 22 articles*, 9 were published in related disciplines so that we have 13 working papers. Besides we conducted a screening on the background of the authors featured in these additions. They mostly belong to financial faculties. In 13 cases (in 22), at least one co-author can be connected to some other publication in the JCR-financial journals list, so that we would consider the author as part of to the social community of financial researchers. Last we note that a screening, within the JCR journal list, based on “Hofstede” or any of his 4 main indices, did not yield further matches.

Chapter II: A survey of ‘culture and finance’

93

With this “double review process”, the intention is to propose a complete picture of the

dimensionalist framework in finance - its coverage, its results and its relative successes. This

should be helpful in two ways. For researchers specifically interested in dimensionalism, this

provides a more comprehensive picture of the empirical coverage and strengths. For

researchers who are more critical about the approach, this provides a contrast between articles

that made their way into peer-reviewed journals in finance, and other articles, either published

in related disciplines, or not (yet) published.

2.3. Classification,  trends  and  first  results  

Below, a figure is provided, detailing the number of total hits and the number of articles

retained through the screening process. This confirms the sharp increase in interest for

cultural research among financial scholars. Yet, if specific and detailed investigation is recent,

it is noted that anecdotic references to culture have been common for quite some time.

Figure 1. Identified articles: trends Peer-reviewed finance articles with a reference to “culture” in title or abstract (grey bars) and peer-reviewed articles with a non-anecdotic reference to culture (black bars). The scale represents the number of articles: respective 60 hits and 18 articles included in the database.

0  

2  

4  

6  

8  

10  

12  

14  

16  

1948-­‐1950  

1951-­‐1953  

1954-­‐1956  

1957-­‐1959  

1960-­‐1962  

1963-­‐1965  

1966-­‐1968  

1969-­‐1971  

1972-­‐1974  

1975-­‐1977  

1978-­‐1980  

1981-­‐1983  

1984-­‐1986  

1987-­‐1989  

1990-­‐1992  

1993-­‐1995  

1996-­‐1998  

1999-­‐2001  

2002-­‐2004  

2005-­‐2007  

2008-­‐2010  

Chapter II: A survey of ‘culture and finance’

94

The base of eighteen articles is categorized in a number of ways.

Firstly, a check is made as to whether the article is empirical, descriptive, conceptual or

theoretical. All articles eventually included, here, are empirical.

Secondly, a check is made as to whether the focus is cross-national. It is, in about 90% of the

cases. Specifically, it is related to cross-national comparisons in three articles 13. Only two

articles are not related to this dominant “national approach” framework, and these two articles

investigate “investment behaviors”; one article has a specifically local focus (on local U.S

communities: counties), while the other investigates alternative “cultural views” (section 3.3).

Thirdly, broad topic of interest for each article is identified, and they are classified into 4

categories14. The following labels are proposed: “corporate finance policies” (financing,

investment, debt maturity…), “financial flows across nations” (debt, equity, mergers and

acquisitions [M&As], foreign direct investments [FDIs], the “home bias”…), “financial

macro-structures” (governance, stock market vs. banking infrastructures, legal or accounting

infrastructure, development of pensions or life insurance…), “stock market functioning”

(institutional investment, momentum on markets, stock-market volatility...). The range of

topics covered here is extremely large, and it could be questioned, just how financial research

could spare the concept of culture, provided that such a large scope of interests is emerging

from this process.

13 Including: a comparison of investors from Finnish vs. Swedish origin in Finland, a comparison of the nationality of foreign banks in Pakistan as understood as Asian vs. non-Asian, and a comparison of regional civic cultures in Italy. 14 This classification aims at facilitating our review and some of the articles either span across two categories or could be analyzed from different angles so they can be considered as ambiguously classified.

Chapter II: A survey of ‘culture and finance’

95

Fourthly, a precise definition of culture is searched for, and the question is asked as to

whether a definition is provided at all. Occurrences of “cultur*” are counted, references to the

dimensionalist literature are identified, and the context in which “cultur*” appears and is used,

and how it is operationalized (what proxies), are noted. A check is made as to whether it is

part of typical expressions. Figure 2 provides an overview on the relation between cultural

approaches (dimensionalist or not) and the provision of cultural definitions. It underlines a

stark contrast and accounts for the structure of the review.

Figure 2. Identified articles: theoretical frameworks and culture definitions.

Only one non-dimensionalist article has a definition- while only one dimensionalist article has no definition. This database includes the 15 articles covered in the body of our reviews, in sections 3.2, 4.1, 4.2 and 4.3 The scale represents the absolute number of articles * * *

no  def.  

with  def.  

0  

2  

4  

6  

8  

10  

12  

1975-­‐1994   2000-­‐2004   2004-­‐2009  

Dimensionalist works

Non dimensionalist works

with def.

no def.

Chapter II: A survey of ‘culture and finance’

96

3. Dimensionalist  approaches  and  dimensionalist  successes  

This section starts with an outline of the dimensionalist approach, which will be as short as

possible, provided the breadth of research to cover (section 3.1). Then, a review is made of

the dimensionalist results (section 3.2), considering, in succession, the four broad areas of

research mentioned earlier: “financial flows across nations” (7 articles), “financial macro-

structures” (8), “stock market functioning” (7), “corporate finance policies” (7).

3.1. Some  background  on  the  dimensionalist  approach  

3.1.1. The label

The label dimensionalism is proposed in a book edited by Vinken, et al. (2004∆∆∆: chapter 1).

This book compiles contributions, at the occasion of a gathering of “well-known

dimensionalist researchers”, in Holland, in 2004 (including in particular: Hofstede, Inglehart,

Schwartz, or still Triandis).

Dimensionalism is defined as aiming at “finding the ultimate, most frugal, and yet most

meaningful basic set of axes, with which to explain the broad range of attitudes, beliefs, life

styles and the diversity of practices among large populations and/or organizations across

societies” (Vinken, et al., 2004∆∆∆: p.8).

The main focus is on cross-national comparisons and the main tool is the development of

numerical scales, to distinguish nations’ cultures, based on synthetic and abstract

“dimensions”, the national cultural indices.

Chapter II: A survey of ‘culture and finance’

97

3.1.2. Main sources and indices

The dimensionalist large-scale surveys that were mainly identified in the screening, are those

of Hofstede (2001∆∆∆: the revised edition includes references to meta-studies and critiques),

and Schwartz (1994)∆∆∆15: within our extended database of 29 dimensionalist articles,

Hofstede’s indices are used in 24 cases (twice in combination with Schwartz’s); Schwartz’s

indices are used in 6 cases (three times in combination, twice with Hofstede and once with

GLOBE); two articles use another reference (GLOBE, in combination, Smith, et al.).

The tentative measurement of national culture, with indices, can be traced back to Hofstede

(1980)∆∆∆. He used the results of a large-scale survey of IBM employees, in the 70’s, to

identify differences between countries. He collected a large number of statements on the

value-orientation of IBM employees. He averaged them by country, conducting a factorial

analysis on the aggregate scores, still by country. He identified 4 dimensions on which cross-

national differences were clustered. He labeled them “individualism / collectivism” (strong vs.

loose ties between individuals in a society), “power distance” (acceptance of inequality

generated by differences in power), “uncertainty avoidance” (a normative assessment of

uncertain situations), and “masculinity / feminity” (gender roles and relative importance of

self-assertiveness and success vs. harmony in society). More precise definitions are provided

in appendix 2.

Building on Hofstede, Schwartz proposed a complementary approach, theoretically-

grounded16 and oriented by three polarizations: “mastery-harmony”, “egalitarianism-

hierarchy” and “conservatism-autonomy” (appendix 2 for a definition).

15 Other sources that follow comparable objectives include the World-Value Survey (coordinated by Inglehart∆∆∆. see http://www.worldvaluessurvey.org/. for sources and publications), Smith, et al. (1996) ∆∆(∆), Smith, et al. (2002)∆∆(∆) the GLOBE project (coordinated by House, et al., 1999). Relating to the latter, Globe project, there was a prolonged controversy between Hofstede and the main Globe proponents (section 5.1), so that, if GLOBE’s epistemological posture, methods chosen, and results, show similarities with dimensionalism, the inclusion in dimensionalism is strongly debated. Two of these measures (Globe / Smith, et al.) are used, once each, in our database of 29 articles. 16 He proposed value-frameworks based on social theories rather than obtaining them from a factorial analysis ad-hoc. In particular, his frameworks were extensively tested ex ante, accounting for the role of languages and differentiating between value sets based on individual or collective behaviors (table 7.1 p. 89 in Schwartz, 1994)∆∆∆: to illustrate: an individual can’t be both prone toward humility and strong power differentiation, in the values he endorses, for his own personal choices and

Chapter II: A survey of ‘culture and finance’

98

A comparison of Schwartz’s indices with those of Hofstede and Inglehart is available in

Schwartz (2004)∆∆∆, and an empirical assessment of pair wise correlations between

Hofstede’s and Schwartz’s indices is available in Schwartz (1994∆∆∆: p. 109). De Jong

(2009∆∆∆: appendix 1: p. 197-206) provides a more general overview of most national culture

and/or dimensionalist indices, including those by Hofstede, Schwartz, the GLOBE

consortium, Trompenaars and the World Value Survey.

3.1.3. Cultural distance

It is important to mention a specific operationalization, called “cultural distance”. This

approach aggregates the four original national culture indices, from Hofstede, into one value.

Specifically it sums up squared differences across all four cultural indices, which results in

one score for each pair of countries17. The approach was initially proposed by Kogut and

Singh (1988)∆∆∆. This cultural distance measure has been extensively used in management,

and Siegel, et al., (2008) provide, for instance, a detailed review on the use of this measure in

relation to the determinants of FDIs. More recent operationalizations sum up squared

differences, using Schwartz’s, or, still, other indices. The measure is implemented in five

articles in this database; based four times on Hofstede, and once on Schwartz.

3.1.4. Antecedents and disciplinary path

Triandis (2004)∆∆∆ provides an account on how Hofstede’s intellectual heritage influenced

psychologists at the turn of the 90’s, and gave ground to the rise of a new discipline, known

today as: “cross-cultural psychology”.

life. However, it is possible at the societal level. Due to role differentiation, a given society can foster both the pursuit of humility while encouraging strong power differentiation (it is in fact common in Asia). For details see Smith and Schwartz (1997)∆∆∆ among other articles. 17 Aggarwal, et al. (2009)* propose recently a different operationalization retaining the absolute value of the difference for each of Hofstede’s 4 measures separately, plus the usual cultural distance measure.

Chapter II: A survey of ‘culture and finance’

99

Scholarly antecedents are ancient in human sciences. Hofstede (2001)∆∆∆ provides an

extensive outline of these, in his first chapter, tracing them back to post-war influences by

Kluckhohn, et al., or Rokeach (references therein). It is important to note that the key

underlying mechanisms are values, through its motivational grip on individual and collective

behaviors (Hofstede, 2001∆∆∆: p. 11; Oyserman, 2002). De Jong, 2009∆∆∆ re-emphasizes how

other cultural dimensions, such as rituals, heroes and symbols, all rest essentially on the same

core dimension: that of values. Hence the emphasis on measuring values as expressed by

individuals, to aggregate them “within the boundaries of defined social groups”, which has

generally meant countries. Breuer and Quinten (2009: table 1, p.5 )* provide an outline on the

varying cultural models that preceded, and followed, that of Hofstede.

No further elaboration is made here, but multiple alternatives to values are available, in order to

investigate cultures18.

The relative success of Hofstede’s research has been uneven across disciplines. Baskerville

(2003)∆∆ shows how it was much more successful in management related disciplines

(particularly in general management, International Business, organization studies, marketing

or communication) than in economics, sociology or anthropology. So the relative success

must be tackled, discipline-by-discipline. Recently Kirkman, et al. (2006)∆∆∆ provide a meta-

study of all research incorporating Hofstede’s dimensions in management. Ramirez and

Tadesse (2007: appendix)*++ provide a review and a table including cross-references to

Hofstede in management, operational research and economics/finance; they confirm

Baskerville∆∆’s suggestions, above, and highlight that cross-referencing into

economics/finance is very recent, if peaking at all.

18 See in particular Vinken, et al., 2004∆∆∆: chapter 1 on EMIC vs. ETIC approaches and on “particularist” and “post-modernist” views of culture; Earley and Singh, 1995∆∆ or d’Iribarne, 1997∆∆ on anthropological and hybrid approaches; Earley, 2006∆∆ for a focus on meaning and behavior rather than values. This should be reminiscent about the debates ongoing in economics and the variance found in definitions as illustrated in section 2.1

Chapter II: A survey of ‘culture and finance’

100

D

ate

Sour

ceH

ofst

ede

Schw

artz

Cul

tura

l dis

tanc

eO

ther

app

roac

hM

ascu

linity

/ F

Indi

vidu

alis

mU

nc. a

void

ance

Pow

er d

ista

nce

Dis

tanc

eR

elig

ion

Lang

uage

Oth

erM

ain

indi

cato

rsM

acro

-eco

n. e

nv.

Ope

nnes

sTa

xatio

nLe

gal (

gene

ral)

Lega

l (fin

anci

al)

Cor

pora

te g

ov.

Cou

ntry

risk

Polit

ical

env

.So

cial

stab

ility

& …

Aggarwal & Goodell*++

2008

JMFM x x Stock market

functioning Equity premium x x x x x x

Aggarwal & Goodell*** 2009

JBF x x Financial macro-

structuresNational financial macro-structure

x x x x x x x

Aggarwal, Kearney & Lucey*++

2009

WP x x x x x x Financial flows across

nationsFlows of debt, equity in aggregate

x x x x x x x x

Anderson & alii* 2007

WP x x x x Financial flows across

nationsStock-holding [home bias + diversification]

x x x x x x

Beugelsdijk & Frijns*++

2009

WP x x Financial flows across

nationsStock-holding [home bias]

x x x x x x x x

Breuer & Salzmann*++

2008

WP x Financial macro-

structures Governance systems x x x x x x

Desender, Castro & Escamilla* 2011

AJE

S

x x x x Corporate finance policies

Earnings Management [in agg.]

x x x x

Chang & Nourbakhsh* 2007

WP x x x x Corporate finance

policies Cash Holding [in agg.] x x x

Chui & Kwok*++

2008

JIB

S

x x x x x Financial macro-structures

Life-insurance holdings [in agg.]

x x x x x x

Chui, Lloyd & Kwok*++

2002

JIB

S

x Corporate finance policies

Capital Structure [in agg.]

Chui, Titman & Wei*** 2010 JF x x Stock market

functioningMomentum profits [in agg.]

x x x x

Clement, Rees & Swanson* 2003

JAA

F

x x Stock market functioning Market analysis x

Conn & alii *** 2005

JBFA x x Financial flows across

nations M&As

De Jong* 2002

WP x x x x Financial macro-

structuresInflation and independence

x x x x

De Jong & Semenov* 2002

WP x x x x x Stock market

functioningStock-Market Development

x x x

De Jong & Semenov* 2006

IJB

GE

x x x x x Financial macro-structures Ownership patterns x x

Fidrmuc & Jacob*++

2008

WP x x x x Corporate finance

policiesDividend policies [in agg.]

x xGleason, Mathur & Mathur*++

2000

JBR x 1 Corporate finance

policies Capital Structure

Huang* 2009

JDE x x Financial macro-

structuresGrowth in opaque industries

x x xKimbro* 20

02

JAA

F

x x x Stock market functioning

Accounting standards & corruption

x x x x x

Kwok & Tadesse*++

2006

JIB

S

x x Financial macro-structures Financial architecture x x x x x x x x x

Licht, Goldsmidt & Schwartz*++ 20

05

IRLE x x x x Stock market

functioning Investors' protection x x x x

Outreville*** 2008

JRI 2 Financial flows across

nations FDIs x x x

Pirouz* 2004

WP x x x Stock market

functioning Stock market volatility x

Ramirez & Tadesse*++

2007

WP x x Corporate finance

policies Cash holdings x x x x x

Schmeling** 2009

JEF x x Stock market

functioningCross sectional stock returns & sentiment

Shao, Kwok & Guedhami*++

2008

WP x 3 Corporate finance

policiesDividend policies [in agg.]

x x x

Siegel, Licht & Schwartz*++

2006

WP x x Financial flows across

nationsFlows of debt, equity, M&As

x x x

Weber, Shenkar & Raveh* 1996

MS x x x x x Financial flows across

nations M&As

+

Markings: *** for the 4 articles resulting from the main screenings and overlapping with the main database. ** for the regional article. * for the articles obtained through the doublereview process (dimensionalist articles). Details on both processes in section II.2.2. ++ is a marking for articles, where at least one of the co-authors has an established track record, inany of the journals screened as part of the main screenings (the J.C.R. list). It results that *** and ** stands for articles from the financial community stricto sensu, while *++ stands forarticles from the financial community lato sensu. Eventually * stands either for aticles from financial researchers, with no track record in the J.C.R.-financial-journal-list (yet), or forestablished researchers in proximate fields (management mainly). Any unpublished article is notes "WP" for Working Paper in the column for the source. Other journals' names areabridged by their initials. Further notes: 1) clustering of countries based on Hofstede. 2) cultural distance based on Smith, Peterson & Schwartz. 3) Globe measures

Table 3: dimensionalist database resulting from the main screenings and then the double review

Political env.

Details Hofstede's indices

Classification TopicsAuthors

Dim. Approach

Other cultural

Economic env. Law

Chapter II: A survey of ‘culture and finance’

101

3.2.    Empirical  successes  for  dimensionalism  in  finance  

3.2.1. Financial flows across nations and international investment

In this section, we focus on financial flows across borders including debt (bonds, sovereign

bonds) and equity (portfolio management, M&As, FDIs), and we include the study of the

“home bias”. Five articles, in seven, use Hofstede’s indices, and five articles use the cultural

distance approach, corresponding to all our articles using the cultural distance approach, but one.

The general reasoning is straightforward, cultural distance between two nations hinders

financial flows between them. Two articles apply this reasoning to flows of debt and then to

flows of equity: the higher the distance, the lower the flow (Aggarwal, et al., 2009*++ using

all Hofstede’s indices separately and summed up; Siegel, et al., 2006*++ using egalitarian

distance, a measure derived from Schwartz). Relatedly, Beugelsdijk and Frijns (2009)*++ note

that the home bias is higher in foreign countries, with a higher cultural distance relative to the

home country. The reasoning is applied to flows of M&As in two articles. Distance, as

measured by egalitarian distance, has a negative effect on volumes of cross-border M&As

(Siegel, et al., 2006*++, but the effect is not robust for some specifications). Distance has not

only a negative effect on volumes for M&As, but on profitability as well: cumulative

abnormal returns are lower for U.K. companies, which are acquiring other companies abroad,

when the acquisition is made in countries with higher cultural distance (Conn, et al., 2005***

using all Hofstede’s indices aggregated). Lastly, it applies to flows of investment in Outreville

(2008)***: insurance groups choose to locate in countries where the cultural distance is

lower.

A marked difference between studies that overlap with economics is noted, and those that

overlap with management, are noted as differentiated by specific subjects. On the one hand,

Chapter II: A survey of ‘culture and finance’

102

following a tradition in trade economics, the introduction of cultural distance is very recent

(Aggarwal, et al., 2009*++). It is investigated in the section 4.1 of this article, because apart

from Aggarwal, et al., there is no relation to dimensionalism. Additionally, it is noted that this

study is mainly dealing with working papers, the conclusion is made that, on the economic

side, and the use of cultural distance has not yet made its way into the peer-reviewed literature

in finance.

On the other hand, an overlap with management is identified when the focus is on either

M&As or FDIs - two topics with a voluminous research track record in management journals

(as identified through the J.C.R. list). There, the use of cultural distance is fairly ancient

(Siegel, et al., 2008, for a review). Furthermore, it has been controversial; so, for instance,

Morosini, et al. (1998), more than a decade ago, question what was then a decade of

consensual results, and show that cultural distance enhances, rather than reduces, cross-border

acquisition performance. This thread, however, was not followed any further, provided by the

use of dimensionalism, in connection to M&As or FDIs, outside finance journals, stricto sensu,

because the sheer volume of literature forbade it.

However, an inclusion in the database has been made of Weber, et al. (1996)*19. They show

that corporate culture interferes with national culture, in its effects on cross-border

acquisitions. They highlight that measures of cultural distance need further theorization,

before anything can be concluded about the relation between cultural distance and cross-

border M&As; so, for instance, cultural distance can foster rather than hinder cross-border

acquisitions, because partners complement one-another through their differences (i.e. through

their cultural distance).

19 This article is mentioned in Breuer and Quinten (2009), a key reference for dimensionalism in finance. Besides, it provides a useful illustration about some of the difficulties encountered by this line of research. Last, it provides evidence that the notion of “mingling culture layers” has precedence (see section 5, and in particular section 5.4).

Chapter II: A survey of ‘culture and finance’

103

Another aspect, worth emphasizing here, is the large number of institutional control-variables,

in many articles, in the database. Table 3 provides a rough overview, and appendix 3 lists

categories for these control variables, as identified in the extended dimensionalist database. A

further analysis of these references, in relation to national cultures, and the varying aspects of

the institutional context, was initiated by the author, and will be the object of a specific article.

Appendix 3 conveys a sense of the extremely wide range of context dimensions, which were

considered simultaneously with culture, in most of the dimensionalist research-papers.

Lastly, Anderson, et al. (2007)* provide an innovative approach to culture, in relation to the

home bias. They also provide the only article in this section not related to cultural distance.

Interpreting the definitions provided by Hofstede, Anderson, et al., suggests that the relative

magnitude of the home biases should be correlated with national cultural indices, namely the

absolute value of the uncertainty avoidance and the masculinity/feminity indices. The

reasoning is the following: investors from countries where uncertainty avoidance is higher

(less acceptance of uncertainty) will diversify less, because diversification involves investing

in more uncertain stocks. Further, investors from masculine countries will have higher

potential overconfidence and self-attribution biases, which will drive higher exposures

abroad, a lower home bias, ceteris paribus.

3.2.2. Financial macro-structures

This broad subject provides us with a variety of sub-fields, including the determinants of

financial macro-structures (the relative importance of banks and markets: 2 articles), the

development of stock markets (1), the determinants of corporate governance structures

(governance indices; ownership concentration: 2), an article is concerned with growth in

opaque industries (1), another with the consumption of life-insurance (1) and the last article

investigates the connection between inflation, central bank independence and national

cultures (1). All studies refer to Hofstede’s indices, but one, which uses Schwartz’s indices.

Chapter II: A survey of ‘culture and finance’

104

Three articles, in eight, use only one of the Hofstede’s indices, while five of them implement

indices simultaneously (either Hofstede’s or Schwartz’s indices).

Hofstede’s uncertainty avoidance index features prominently in this section, because it is

associated, on a stand-alone basis, with three articles, and it is associated with other indices in

three other articles. Moreover, it is used in the article that results from the main screenings:

Aggarwal and Goodell (2009)***, extending a previous article by Kwok and Tadesse

(2006)*++, establish that countries characterized by greater uncertainty avoidance, favor bank-

based systems20. Importantly, their focus, unlike that of Kwok and Tadesse, is not specifically

dimensionalist, and they show more generally that “national preferences for market financing

increase with political stability, societal openness, economic inequality, equity market

concentration, and they decrease with regulatory quality and ambiguity aversion [i.e.

uncertainty avoidance]” (ibid: p. 1770). They provide specific definitions, neither of culture,

nor of national cultures. This is important because it provides contextual evidence about the

way the financial research community receives, and uses, dimensionalist frameworks: Kwok

and Tadesse published their research in the Journal of International Business Studies

(excluded from the J.C.R. list of financial journals), while Aggarwal and Goodell published in

the Journal of Banking and Finance (further commentaries in section 4.3).

So, uncertainty avoidance is negatively associated with stock market financing. Furthermore,

it is negatively associated with growth in “opaque” industries (i.e. industries characterized by

informational opacity: Huang, 2009)*. It is correlated with stock market development

20 The main difference between both articles relates to their respective theoretical underpinnings. Kwok and Tadesse*++ proceed from a general theoretical structure about national cultures, derived from Hofstede (Kwok and Tadesse: figure 2) while Aggarwal and Goodell*** propose an exploratory process where the chief focus is on characterizing elements, in the general institutional context, to explain observed differences in financial macro-structures. Kwok and Tadesse*++ intend to establish that “national culture plays a significant role [in why] countries differ in the configuration of their financial systems”, while Aggarwal and Goodell*** investigate “how national preferences for financial intermediation (markets versus institutions) are determined by cultural, legal, and other national characteristics”. Both articles include extensive references to additional institutional variables, considered either as complementary exogenous variables (e.g. in Aggarwal and Goodell), or as control variables (Kwok and Tadesse). Kwok and Tadesse include, in particular, complementary cultural variables: religion and an ethnic variable (fractionalization).

Chapter II: A survey of ‘culture and finance’

105

(negatively), with investor’s protection (negatively), with ownership concentration

(positively), with the tandem that associates central bank independence and limited inflation

(positively), with the development of pension schemes (positively). However, Chui and Kwok

(2008)*++ fail to identify a significant relation with life insurance consumption (potentially

because it is mediated by the development of pension schemes).

In the three articles, with simultaneous implementation of three or four Hofstede’ indices,

results are mixed regarding the significance of the indices. Chui and Kwok (2008)*++ identify

a negative relation between life insurance consumption and the three other indices,

collectivism, masculinity and power distance (the higher the three indices, the lower the life

insurance consumption). De Jong and Semenov (2002)* combine uncertainty acceptance and

masculinity to abstractly define an “ability to accept competition”; they show that acceptance

of competition is related to the development of stock markets (so stronger market

development, with higher uncertainty acceptance and masculinity). De Jong and Semenov

(2006)* use the same two dimensions to relate them to ownership dispersion. They show that

ownership concentration increases with masculinity and uncertainty avoidance (so the reverse

relationship, as compared with stock market development). Relating to central bank

independence and inflation, it is power distance, which combines successfully with

uncertainty avoidance: power distance is negatively related to both limited inflation and

central bank independence (De Jong, 2002*).

In a fourth article, with simultaneous implementation of several indices, Breuer and Salzmann

(2008)*++ show that Schwartz’s indices systematically correlate with governance indicators,

and with the underlying macro-financial structure. So, countries with a stronger emphasis on

autonomy, hierarchy, or mastery, tend to have market‐based systems.

Chapter II: A survey of ‘culture and finance’

106

The large use of complementary legal, political and cultural variables is, again, very

noticeable (table 3 and appendix 3). All articles, but one, use complementary political and

legal variables. Three articles use complementary cultural variables (religion in three cases,

ethnic fractionalization once). Overall, the need to complement the study of cultural

determinants to financial macro-structures, by a wide range of institutional variables, is, once

again, very strong.

3.2.3. Stock-market functioning

This domain of research belongs here, due to the recent Journal of Finance publication,

relating momentum profits (on stock markets) to Hofstede’s individualism index (Chui, et al.,

2010***). This section also includes an article obtained through the “regional” screening

(Schmeling, 2009**), and related to the one by Chui, et al.

All articles use Hofstede’s indices. Extending the findings from the preceding section, the use

of indices is more parsimonious here21: three articles use only one of the four indices, while

three of them use two indices in combination with one another. The study of financial macro-

structures may be more dependent on the institutional context of nations than that of stock

market functioning. It results that: the “country puzzle” is less prevalent, and that

Williamson’s “institutions of embededdness” are less relevant (further commentaries in

sections II.5. and in the additional essay, in the annex). This results in approaches that have

fewer comprehensive indices and more specificity. In fact, it is noted that the inclusion of

institutional variables is much more limited here, as compared to the previous two sections

(table 3, in relation to appendix 3).

21 This is in contrast to the overarching approach proposed by Hofstede (2001: p.12)∆∆∆ relating “ecological factors”, “societal norms” (including national culture) and general socio-economic outcomes.

Chapter II: A survey of ‘culture and finance’

107

Furthermore, it is noted that, in this section, most articles are published articles; in two cases

in financial journals, and in four cases in journals at the intersection with law, economics and

financial accounting. This relative success may be supported by the parsimony in approach,

and second, by a more specific characterization of the relationship between national cultures

and the financial subjects being investigated. This echoes the commentary in footnote 20, and

these insights are built upon for the recommendations that are made in sections 4.2 and 4.3.

Chui, et al. (2010***) note that momentum profit strategies are strong in US stock markets,

and weak in many Asian countries. They hypothesize that the ability to derive momentum

profits may be, partly, driven by the individual mechanisms of over-confidence and self-

attribution biases, which in turn are both fostered by (and positively associated with) the

degree of individualism in a given country. Schmeling (2009)**, extending this result,

investigates the tendency of national stock markets to be more or less prone to “over-reaction

and herd-like behaviors”. He shows how these behaviors may be related not only to

individualism, but to collectivism (its opposite) as well, and to uncertainty avoidance. While

individualism triggers individual behaviors such as over-confidence and self-attribution,

collectivism triggers a collective behavior: “herd-like behavior”.

This distinction is particularly interesting, culture-wise, for two reasons. Firstly it provides

specifics on the mechanics of enabling views of culture: here the polarity involves not only

constraints, but generates specific behaviors (it is enabling, see section 2.1 and 5.4). Secondly,

culture would be precisely what characterizes collective behaviors, as opposed to individual

behaviors (and related to psychology). Overall, Schmeling shows that the relation between

sentiment and stock market returns are magnified by collectivism, and uncertainty avoidance.

At the same time, they are moderated by market integrity, a further measure derived from

Chui, et al., which is correlated with the two dimensionalist indices. So the relationships are

complex, involving first and second-order effects, combined with contradictory influences.

Chapter II: A survey of ‘culture and finance’

108

Providing consistent results with the ones just mentioned, Clement, et al. (2003)* show that

the individual experience of market analysts has more leverage on the forecast accuracy in

individualistic countries than in collectivistic countries. Similarly, Pirouz (2004)* combines

individualism with power distance to define what he labels a “relationship-orientation”. He

shows that there is a correlation between relationship orientation (high relationship

orientation meaning high collectivism and power distance), stock market volatility and

limited openness.

Kimbro (2002)* investigates the relation between corruption, on the one hand, accounting,

and financial statement quality, on the other. He shows how the previously identified

positive relationship, between corruption and two cultural variables, collectivism and power

distance, turns negative when the level of economic development and accounting standards

are controlled. So, more individualism drives better monitoring and information, which leads,

in turn, to less corruption opportunities (primary and indirect effect). However, more

individualism leads to a stronger willingness to take advantage of corruption opportunities, so

that it drives more corruption (secondary but direct effect of a generally smaller magnitude).

Kimbro concludes that the smaller, direct, effect of individualism was previously overseen,

because it was covered by the collective externality provided by a higher quality in

accounting and financial statements. This line of analysis is reminiscent of the one by

Schmeling (2009)**, because it shows that the general embededdness of institutional and

cultural phenomena drives first and second order-phenomena, which easily lead to false

assumptions on the basis of over-simplified analyses.

Overall, this section features a large diversity in subjects, provides a large number of articles

featured in peer-reviewed financial journals, and distinguishes itself by a flexible

dimensionalist approach. Further a number of syncretic concepts emerge, such as

“relationship orientation”, or “herd-like behavior”. Last, fine-tuned propositions with second

order causality mechanisms appear, such as Kimbro’s analysis.

Chapter II: A survey of ‘culture and finance’

109

3.2.4. Dimensionalist antecedents to financial laws

One article is positioned at the intersection of the two previous sections (stock-market

functioning, financial macro-structures). The article provides a comprehensive approach,

culture-wise, with far-reaching consequences for the practical relevance of legal determinism

in financial research. Licht, et al. (2005)*++ study the antecedents of financial laws (investors’

rights, procedural formalism), which they relate to a large range of cultural indices.

They show how legal variables are robustly correlated to the cultural constructs, from both

Hofstede (uncertainty avoidance and power distance) and Schwartz (all of them), and are

controlling for other cultural and legal variables (namely, the rule of law, religion), and

economic development. In particular, they identify a robust correlation between harmony,

both shareholders’ and creditors’ rights (negative), and procedural formalism (positive).

Furthermore, formalism is also negatively correlated to affective autonomy. In addition, they

identify a general significance of uncertainty avoidance with shareholders’ rights (negative),

and with formalism (positive), echoing other articles in section 3.2.2 (financial

macrostructures). A wider range of Hofstede’s indices are powerful in the formalism

regressions: uncertainty avoidance, collectivism and power distance, all three indices

correlate positively to formalism. In further regressions on shareholders’ rights, where they

exclude Asian and Australasian Common Law countries (Australia, India, Malaysia, Nepal,

Singapore, Thailand), they find that the same three indices (uncertainty avoidance,

collectivism and power distance) are negatively correlated to the importance of shareholders’

rights. Their argument is that, in these countries, the letter of the law does not reflect the

effective legal practice, so that countries have to be treated practically as “outliers”.

Licht, et al., conclude that their study casts doubt on the “alleged supremacy of statutes” and

on the ability to change the letter of the law, to foster economic and/or financial development,

with effective consequences on behaviors and institutions. Further reflections on these

conclusions can be seen in sections 4.2.1 and 5.

Chapter II: A survey of ‘culture and finance’

110

3.2.5. Corporate finance policies

Provided that there was no result from the main screenings, for this topic, this section will be

brief. Besides, five of the seven articles identified, here, through the double review process,

are working papers. Three articles use Schwartz’s indices, while five of them use Hofstede’s

measures (one article overlaps). Cultural indices are sometimes used on a stand-alone basis,

and sometimes used in combination. Generally, few other institutional variables are used here,

in stark difference to the previous sections; and apart from legal origin and investor’s

protection (they both feature prominently: four times each. This needs to be considered in the

light of the studies relating culture to investors’ rights, in sections 3.2.4 and 4.2.1). No other

cultural variable is used.

The topics covered in this section include capital structure, dividend policies and cash

holdings (twice each, generally one with Hofstede’s and the other with Schwartz’s indices). A

seventh paper is related to earnings management (table 3).

* * *

Chapter II: A survey of ‘culture and finance’

111

4. Further  culture-­research  in  finance  

Five articles from the database were carrying a dimensionalist label. This section is concerned

with the other thirteen articles. Seven articles deal with financial flows across nations. They

are included in section 4.1.1 and 4.1.2. They mostly use indefinite references for culture and,

together, they leave us with some puzzlement concerning what culture is, how it relates to

institutions, to geography or to exchange. A research-article, published in an economic journal

by financial scholars, is considered in section 4.1.3, because it provides an opportunity for

synthesizing international trade- and capital-flows. Section 4.2 is concerned with three

financial articles, implementing each one of the specific cultural approaches. Section 4.3 deals

with the remaining three articles, as well as with those “left behind” after the initial screening.

4.1.  Gravity-­  and  other  models  of  trade,  as  applied  to  financial  flows  

4.1.1. “Familiarity breeds investment” 22

Six articles are reviewed in this section and one is left for the next sub-section. These articles

contribute to a research-stream in the adjacent field of trade-economics, which investigates

the determinants of bilateral trade-flows, sometimes using “gravity models”. These models

traditionally investigate trade-flows patterns, and focus on the relative size of trading partners

and on their geographical proximity (hence the ‘gravity’ label). They were recently expanded

to include the idea of “psychological geography”, among other determinants. Portes, et al.

(2001) consider the importance of the volume of information, available on trading-partners, as

a determinant for economic exchange. While distance is often prejudicial to both information-

exchange and physical-trade, Portes, et al., extend the idea of geographical proximity,

22 This is the title of an article by Huberman (2001)

Chapter II: A survey of ‘culture and finance’

112

conceptually, to the idea of “cultural affinity”, because affinities reduce informational

frictions and facilitate information exchange. While the exact nature of the relation between

distance, information and “cultural affinity” remains unexplored in their article, their research

triggered a wave of empirical testing of culture-related proxies in trade-economics.

Overall, in conclusion of this brief excursus on gravity models, it is noted that if the empirical

investigation of culture-related proxies is recent, references to “cultural similarity” or

“psychic distance” are fairly ancient, as they were proposed as early as the 1950s

(Beckerman, 1956).

The six articles included in this section investigate a range of financial flows including

foreign corporate lending by banks (1 article), spillovers on the sovereign debt market (1),

international cross-listings (1), payments on the interbank market (1), and, cross-country

flows of M&As (2).

While Rosati and Secola (2006)*** identify a strong proximity effect in a gravity model of

inter-bank payment flows in the E.U., Gande and Parsley (2005)*** identify neither cultural,

nor any institutional effect, on the “spillover” of a country’s credit rating onto the credit rating

of other countries. They note that credit-rating spillovers are highly influenced by a common

correlation to financial flows to-and-from the U.S. Sarkissian and Schill (2004)***, studying

the determinants of overseas-listing, conclude that “the same proximity constraints that are

believed to lead to home bias … also exert a profound influence on financing decision”23.

Mian (2006)*** shows that cultural distance limits the efficiency of lending to corporations

by foreign banks, in Pakistan. In his study, he dichotomizes cultural distance by a distinction

between Asian- and non-Asian foreign banks, and he finds that Asian banks do better than

non-Asian foreign banks. Focarelli and Pozzolo (2008)*** show that the amount of cross- 23 The provision of international funds (related to the home bias) and the search for international funds (overseas listings) were arbitrarily included in these sections on international financial flows.

Chapter II: A survey of ‘culture and finance’

113

border M&As depends on a range of proxies, which characterize proximity between two

countries: they use a range of culture-related proxies, including, in particular, distance,

language, and thirdly, the similarity in financial and economic macro-structures. Gleason, et

al. (2005)*** have a related approach, and they show that cumulative abnormal returns for

the acquisition of foreign privatized banks, by British banks, is influenced by cultural

proximity.

Those articles, here, focusing on the occurrence and treatment of culture-related concepts,

have a low occurrence of “cultur*” even through the idea of national cultures, and the

underlying idea of “psychic distance”, is central to the argumentation. In general, “cultur*” is

associated with specific expressions, such as “cultural distance”, “geographic, economic,

cultural, and industrial proximity”, “cultural similarity”, “cultural and, or institutional

similarities”, “geographical and cultural proximity”, “cultural and institutional linkage”,

“cultural barriers”, or “cultural integration”. These expressions convey the idea that distance

between countries has many complementary sources, culture being only one of them.

Moreover, as no definition of culture is ever provided, the precise mechanism by which

culture influences, and the interrelation between varying dimensions: cultural, institutional,

etc., are left undefined.

As a result, the overall picture about the respective contribution, to international financial

flows, of cultures, and institutions, is fuzzy, even confusing. Gande and Parsley (2005)***,

or Rosati and Secola (2006)***, combine national cultures and institutions together, and

analyze them in concert, through common proxies. Gleason, et al. (2005)*** define “cultural

similarity” more narrowly, by geographical distance and a commonality in language. So they

clearly distinguish culture from other “institutional” dimensions, and they incidentally define

institutions with legal variables. Sarkissian and Schill (2004)*** explicitly emphasize that a

Chapter II: A survey of ‘culture and finance’

114

distinction between cultural and institutional linkages should be made because “greater

psychological tolerance” comes as a complement to information exchange, but they

emphasize this important distinction, to be made, in conclusion, as a promising research

direction. Eventually some authors treat proximity as purely generic across a wider range of

sources: Focarelli and Pozzolo (2008)*** consider “geographical distance and economic and

cultural integration” in combination, through combined proxies.

The research, reviewed in this section, aims at uncovering the quantitative impact of

proximity on exchange. It largely foregoes to specify the mechanisms linking proximity and

exchange; their inspiration is largely macro-economic (hence, the relation to gravity models).

Language is most consensual as a proxy for culture and/or institutional differentiation. A

general framework, spanning most of this research, is that culture operates through

information exchanges, so that it is mediated by the use of a common language. Yet, this link

is usually implicit, and most authors believe, following Portes, et al., or Sarkissian and Schill,

that the two mechanisms, i.e. cultural affinity and information availability, are separate and

complementary. Yet they depend, inter alia, upon a common language.

Other sources for cultural proximity are often considered in this literature. They include

commonality in religion, in law or in colonial history, common borders, geographic distance,

similarity in financial or socio-economic systems, etc. However there is no consistent use of

these proxies, and a poor specification dominates this research-field, as to whether these

variables are included as variables, per se, or as proxies; or whether these variables are

included in combination, with, or in contrast, to one-an-other, and to cultural variables.

Chapter II: A survey of ‘culture and finance’

115

4.1.2. Corporate culture, “social capital” and merger flows

A seventh article, on inter-regional exchange flows, is by Allessandrini, et al. (2008)***.

They develop a typology of M&As in the Italian banking industry, distinguishing what they

call cleaning strategies from portfolio strategies. A cleaning strategy corresponds to the

situation, where, in the situation of one bank acquiring another, the acquiring bank cleans the

portfolio of the acquired bank of inefficient borrowers. A portfolio strategy corresponds to the

situation, where the acquiring bank completely re-orients lending in the acquired bank,

depending on its specific know-how, its organizational processes and its preferred clientele.

Allessandrini, et al., show that cleaning strategies dominate in what they call in-market

acquisitions, while portfolio strategies dominate in out-market acquisitions.

It is through this distinction between in- and out-market, that the cultural variables enter into

play. Allessandrini, et al., characterize in- from out-market by “geographical and cultural

distance”. They implement cultural distance as the difference in relative “social capital”

between varying Italian regions. They combine two definitions for defining “social capital”:

first, the number of blood-bags donated by citizens in any given Italian region, and, second,

the average regional voter-turnover at a specific and important referendum. While providing

no definition of the “cultural”, they use it very specifically, with diverging meaning-

categories. First, they use it in connection to a region-specific “culture of mutual trust”, which

they label “social capital”, following a stream of literature in economics (see section 4.1.3).

Second, they use it in another way, related to “corporate culture”.

So, their article departs from those in the preceding section (4.1.1) in three significant ways.

First it clearly distinguishes geographical from cultural distance: so a bank can be “culturally

close” to the working environment found in a “distant province”. Second, short of a specific

Chapter II: A survey of ‘culture and finance’

116

definition, they provide an extent of references and definitions for their two categories of

meaning.

Third, they implement two contrasting views of culture, simultaneously. The expression

“corporate culture” (five occurrences) denotes a construct that is separate from regional

“social capital”, related in particular to the construct of “societal trust”. So the two meaning

categories are conceptually quite separate24. The “corporate culture” is in particular related to

a specific know-how, which serves as basis for the portfolio strategy (ibid, p. 700 and 701). In

this sense, the corporate culture of the bank is intimately related to its lending strategy, and it

clearly evokes the expression “credit culture”, that can be found in Gleason, et al. (2005)***.

This conceptual category is, in fact, very common in the material initially collected through

the screening (the initial sixty hits). Credit cultures describe this technical knowhow, from

which banks derive superior performance, in itself, or from acquisitions, when they transfer it

to acquired banks. In the initial screening, this use of culture was classified as technical and

brushed it aside, as anecdotic. However, further analysis is required, and provided in sections

4.3.3, and 5.4.

Eventually, it should be noted that in this research article, a concrete example of two types,

two kinds of cultures that inter-relate and overlap, exist. The first is akin to national cultures,

here, in a regional sense, while the second is to be related to a mix of professional (technical)

and organizational types, blended together into the specific bank’s culture. This is a practical

illustration of what will be labeled as cultural layers, later, in section 5.4.

24Even though they are likely to be connected in practice. The authors mention “corporate culture” in relation to the condition of the workplace environment for the acquired banks. They suggest that this new environment consists of a new culture with two components, the in-the-acquired-bank culture and the around-the-acquired-bank culture (cf. p. 705, expressions are our own derived from their vocabulary). So there is variation in the local geographical “provincial” cultures, but the in-the-acquired-bank culture is likely to be reliable reflection of the local environment.

Chapter II: A survey of ‘culture and finance’

117

4.1.3. Culture, trust and financial flows

The previous paragraph accounts for the emergence of a specific understanding of culture as

civic culture, or societal trust. This specific view, emerging from the screening, has depth and

precedence in economics, and beyond. Given the extended survey of dimensionalism in

finance (resulting in table 3), we wish to provide, here, a brief, and complementary, outline on

culture as trust. Luigi Zingales provides, in particular, 10 research articles on his webpage25,

mostly co-authored by Guiso and Sapienza, all of them on the subjects of “culture and

economics”. These articles explore, further, this aspect of cultures, called, alternatively,

generalized trust, civic capital, social capital, and societal trust. The research-stream

identified here, through this connection with research by Guiso, Sapienza and Zingales, has

antecedents in the political sciences (Banfield, 1958; Putnam, 1993) and in economics

(Arrow, 1972: see Guiso, et al., 2009 for details).

A recent article by these co-authors (Guiso, et al., 2009) is of interest here, because it explores

a topic extensively covered in this research: the antecedents of financial flows, trade-,

portfolio-, investment-flows as well as M&As and FDIs. Besides, this topic is explored with

respect to dimensionalism (section 3.2.1) and includes most of the other identified articles in

the screening (the two preceding sections 4.1.1 and 4.1.2)26. This article is of particular

interest for four reasons. Firstly, it intends to unpack the source of relatedness between

nations, separating out rational, geographical, institutional and cultural sources. Secondly, it

extends the range of topics tackled at once, investigating three endogenous variables

25 http://faculty.chicagobooth.edu/luigi.zingales/research/topics/culture.html 26 This article was not obtained through our screening and the issue of whether it belongs to this review, or not, reflects the complexity of our task. This article was presented in prominent financial conferences. However, it was published in an economic journal, so that it falls, technically, outside of our scope, even though its authors are financial scholars with an established track record in the most prominent peer-reviewed financial journals (as identified by the J.C.R.-list). A further analysis of the 10 articles mentioned by Luigi Zingales, on his website, in connection to culture and economics, shows that only one of them has been published in a financial journal (J.C.R.-list). We suggest that this reflects the epistemological roots of financial research and the relative larger heterodoxy of some recent economic research in comparison to finance (Zingales, 2000; Brennan, 1995).

Chapter II: A survey of ‘culture and finance’

118

simultaneously. Thirdly, it provides many details, as well as sophisticated definitions of the

culture-related variable, trust. Fourthly, it brings forward some new methodological guidance

for the treatment of culture-related concepts in economics and/or finance, in particular with

respect to causality and the use of instrumental variables.

Guiso, et al. (2009), characterize culture, either as the stable aspect of individuals’ priors, and

preferences, or as the stable social constraints in which individuals make judgments and

choices. They operationalize it through “interpersonal trust”. They do not define this concept,

rather, they split it into three categories: “personalized”, “mutual”, or “generalized” trust.

They provide an extended discussion on how trust relates to the “culture” of individuals, as

opposed to “rational expectations”. They suggest a relation to Arrow’s seminal works,

because they view trust as either derived from repeated and successful interpersonal

exchanges, so it is a “rational expectation”, or as related to stable priors and preferences,

hence it is “cultural”. In this sense, an extended conceptual and empirical discussion about

varying aspects, components, of trust, rational and cultural ones, overturns the need for a

definition.

It is very interesting to compare this article by Guiso, et al., to a dimensionalist article

discussed earlier (section 3.2.1), by Siegel, et al. (2006)*++. Siegel, et al., use egalitarianism

to investigate a very similar range of international investment flows. The comparison of the

two articles teaches us that, in their methodological proceedings, both articles instrumentalize

their culture variable, trust for Guiso, et al., egalitarianism for Siegel, et al., with identical

instruments: the number of years countries were at war across centuries, ethnic

fractionalization within countries and commonality in dominant religion.

Chapter II: A survey of ‘culture and finance’

119

This illustrates a point brought forward by this survey: the dimensionalist approach and other

culture-approaches, for instance, trust, need not be opposed. Here, the two articles pursue

identical objectives (explaining three types of quasi-identical, or related, financial flows), they

define culture in seemingly opposing ways (Guiso, et al., refer to the constraining

conceptualization, while Siegel, et al., refer to the enabling one, as emphasized in the

Introduction), but both articles instrumentalize it with the same set of instruments, and with

comparable success. So taken together, it seems that egalitarianism and mutual trust between

European countries (pair by pair) explain the same set of phenomena, while they are caused

by the same underlying factors.

This opens up a number of questions about what Williamson labels the “institutions of

embededdness”. Are trust and egalitarianism spurious in this investigation? Is the one causing

the other? Which one causes the other? Are they both driven by a third underlying variable27?

This illustrates the kind of questions that will be raised in section 4.3.

Additionally, the fact that neither article has found its way into a peer-reviewed journal in

finance is interesting. So Guiso, et al., have published in an economic journal while Siegel, et

al., have cast their article in a new light, and are successfully positioning it in the management

field. We suggest that this reflects disciplinary inbreeding, as discussed in the Introduction,

the Synthesis and Conclusion. Culture is to be understood and accepted differently, by

different academic fields. Relating to a financial angle, as considered through peer-reviewed

journals, the discipline still seems reluctant to accept either type of approach; whether the

dimensionalist one, or the one based on trust, as exemplified by Zingales’ web-site, be

positioned as an economic rather than a financial research-contribution.

27 This potential connection is the object of extended scholarship in other human sciences. For instance Todd, 1990, relates justice and ethics, on a regional or national basis, to factors such as mutual trust, egalitarianism and power structures.

Chapter II: A survey of ‘culture and finance’

120

4.2. Three  specific  and  successful  path  for  culture-­research  in  finance  

Our screening resulted in the identification of two well-known footnote 3 page 44 works on culture,

in finance: the studies by Stulz and Williamson (2003)*** and by Grinblatt and Keloharju

(2001)***. We review them in turn, before considering the article by Hilary and Hui (2009)***.

4.2.1. National cultures as antecedents of investors’ rights

Investors’ rights, along with legal origin, rule of law and enforcement, have been extensively

used, in the law and finance approach, to investigate a number of financial issues, spanning

from the quality of corporate governance, to the development of financial markets, and

beyond to economic and political freedom (La Porta, et al. 2004). This stream of research

considers investor’s rights as exogenous. It provides policy recommendations in

consequence28. In this context, the article by Stulz and Williamson (2003)***, along with the

one by Licht, et al. (2005)*++ considered earlier, show how financial laws can correlate with

enduring and stable underlying socio-cultural characteristics, casting doubt on whether

changes to the letter of the law can have effective consequences in practice.

Stulz and Williamson provide a detailed definition for culture (appendix 1) and they use two

proxies: religion and language. They note that religion has to be considered as a cultural

variable, per se (not as a proxy), and they provide extensive scholarly references drawn from

the fields of economics, economic sociology, law, economic history and religious history.

In contrast, they note that the second variable, language, proxies for “world-views” or “world-

systems” and they contrast an Anglo-American worldview (“Anglo-Saxon”, operationalized

as the English-speaking countries), with Hispanic and other worldviews.

28 This includes for example and among other things, a positive correlation observed between English law and (specific references to be asked from survey-authors)

• The protection of corporate shareholders (La Porta, et al., 1998, 2000) • The valuation of firms (La Porta, et al., 1999) • The efficiency of financial institutions • The regulation of economic activity, including the quality of corporate governance (La Porta, et al., 2000) and the

regulation of entry (firm creation, Djankov, et al., 2002) • Growth and macro-economic performance (La Porta, et al., 1997, Gleaser, et al., 2004, Botero, et al., 2004).

Chapter II: A survey of ‘culture and finance’

121

Importantly, Stulz and Williamson (2003)*** provide a discussion on how Protestantism

relates to decentralization and individualism (conceptually, not in relation to Hofstede’s

individualism index), with a suggested relation to societal trust as well29. Further they

exemplify how the religious spirit and related social norms have interacted with national

institutions, to foster cross-national differentiation across Europe. This is in fact one of the

reasons why they retain the “dominant national religion”, as the variable of choice, precisely

because the dominant religion best reflects this process of path-dependant maturation

processes within nations. Stulz and Williamson further provide references for the relation

between religious faith, and lending, an/or investing activities.

4.2.2. Concluding on international flows and the antecedents of investors’ rights

Overall, we wish to underline three key issues. First, the two articles considered here, like the

two articles about international capital flows (section 4.1.3), the dimensionalist ones, and the

ones on trust, religion, or worldviews, all provide evidence on the mechanisms that relate

culture with financial outcomes. So these comparisons exemplify that culture must be

examined in connection with a specific mechanism relating it to the outcomes considered.

Second, taken together, these articles, emphasize the strong underlying, and implicit, relations

between varying cultural, and institutional dimensions. In the last two sections (4.1.3 and

4.2.1), a dimensionalist approach, and an alternative approach, achieve similar results, on

similar issues, with competing cultural approaches. This illustrates both the disciplinary stakes

that are at play, and the potential for a middle ground, not yet achieved.

Last, in many cases, and in particular in Stulz & Williamson (2003)***’s article, the authors

emphasize potential transversal linkages between views of culture.

29 Note that, on a similar investigation of investors’ rights, but considered from a dimensionalist angle, Licht, et al., identified that power distance and uncertainty avoidance mattered across the board, with individualism being important for some of the specific rights considered. These synthetic national indices may correlate with the degree of centralization and individualism as identified by Stulz and Williamson. At least, they are all correlated to the same set of legal variables.

Chapter II: A survey of ‘culture and finance’

122

4.2.3. Culture vs. language and distance in portfolio holdings

Grinblatt and Keloharju (2001)*** investigate the importance of investor’s native cultures,

understood here as ethnicity, as a determinant of investors’ stockholding patterns. They do not

provide a definition of culture, but rather "classify the culture of the firm based on the name

and native language of the CEO (Finnish/Swedish)”. Then they compare the ethnic origin of

investors with that of the shares they are holding.

Here culture, unlike section 4.1 on financial flows, is conceptualized and operationalized

separately from language and distance: the language of the company is that of the annual

reports, and the culture of the company is the native ethnic culture of the CEO, either Swede

or Finnish. So the authors provide a test of the relative importance of culture as opposed to

language or distance, on the nature of portfolio-holdings. They show that culture matters.

Most importantly, they show that the culture-influence is moderated by distance and by

education (or training): the culture-effect disappears for larger distances30, and it disappears

for savvy and institutional investors.

This underlines an important and specific aspect of “national cultures”: not every citizen, in a

given country, is subject to the same cultural influence, and a national culture’s influence can

be individually circumscribed through education, experience or other circumstances. This

relation between a synthetic index of national culture (as in dimensionalism), and the

individuals considered in any empirical investigation, is problematic. It is, indeed, one of the

key issues in the debates surrounding dimensionalism in management; it is one of the pillars

of the critiques (see sections 5.1 and 5.2 for further developments and references).

30 For companies that are nationally renown, or that are registered on the capital city’s stock-market

Chapter II: A survey of ‘culture and finance’

123

4.2.4. Corporate culture, religiosity and investment

The article by Hilary and Hui (2009)*** shows that, in effect, higher religiosity has an

implication for the variance in equity returns (lower), ROA (lower), investment (lower),

growth (lower) and break-even point for investment (higher).

This article stands out, in this survey, in many respects. First, its focus is on comparing

behaviors across local communities within the U.S., not across nations: the comparison is

conducted at the county level (read county, like borough).

Second, it investigates the influence of religiosity, rather than the effect of religious

affiliation, i.e. its focus is procedural rather than substantial (we use this terminology, because

it is standard in the sociology of culture, e.g. DiMaggio, 1994). Third, and consequently, it

shows that the observed firms’ outcomes are related to the relative intensity of religious

practice (i.e. the religiosity of executives, a procedural aspect), rather than on any specific

religious affiliation, i.e. any specific cultural norm (a substantial aspect). This focus on the

intensity of a given cultural (religious) practice is unique in the dataset obtained from the

screenings. However, it reflects a long-standing scientific proceeding in anthropology or

sociology - as well as in insurance research31.

Fourth, the reason this article has been included in the database is not related to a use of words

based on “religion”, even though it is often associated with culture-research as exemplified

above. This article was distinguished in the screening, because of its use of “corporate

culture”. In fact the study aims at “examining the influence of community religion on

corporate decisions”. The implicit link is that the narrow geographical focus on counties

accounts for a direct link between the “corporate culture”, that of the CEO and that of the

surrounding environment. 31 In the field of insurance research, this focus is not recent; for instance Burnett and Palmer (1984) investigate a number of “behavioral biases” in life insurance holdings across nations; they note that “traditional work ethic, fatalism, socialization preference, religion salience, and assertiveness” are important international determinants. In all these cases, the focus is on the intensity of the identification with a norm rather than on the norm itself.

Chapter II: A survey of ‘culture and finance’

124

Thus, this article underlines the prominence of a number issues acknowledged earlier. If

nations, and related national cultures, provide the strongest common link within the database,

there are other cultural sources to be considered, in particular corporate cultures. Second, the

focus on a specific norm, or a specific definition, does not need to be the most appropriate

way to proceed. As can be seen, in the previous section (4.2.2), an emphasis on mechanisms

can help to specify cultural definitions. Moving one step further, the mechanism itself can be

the key dimension, independent of the content: the procedural supersedes the substantial. By

extension, the dimensionalist literature focuses on values as the building block to

operationalize “national cultures” (section 3.1). Hence, it rests on the motivational aspect of

cultures. But what is the importance of the relative motivational intensities to these values,

and their interactions?

4.3.      Anecdotic   and   less   anecdotic   references   to   culture,   uncovered   in   the  

screenings  

4.3.1. The “cultural affinity hypothesis”

A number of articles explore the “cultural affinity hypothesis” (Calomiris, et al., 1994) in the

U.S. They were excluded from the survey, because of their specificity. What is the

hypothesis? “This hypothesis implies that white loan officers, because of a lack of familiarity

with minority applicants, will rely more heavily on characteristics that can be observed at low

cost (e.g. objective loan application measures) in evaluating the creditworthiness of minority

applicants, relative to white applicants” (Hunter and Walke, 1996: p.57)°.

This exemplifies, how culture, familiarity and trust interrelate, echoing earlier remarks.

Further, this exemplifies the difficult, abstract, but central notion of culture as enabling rather

than constraining: indeed, because of differing values, communication habits, social

Chapter II: A survey of ‘culture and finance’

125

references, and so on - something lacks in the credit relationship between minority

(“colored”) borrowers and “white” loan officers. This missing dimension would alleviate the

constraints of the lending process, of the lending relationship, to facilitate the lending. So

while discrimination in lending is really about additional constraints, for “colored” borrowers,

the “cultural affinity hypothesis” is really about an interpersonal relation that enables the

lending, for “white” borrowers. Colored borrowers are not discriminated against. Rather some

white borrowers are favored. The cultural affinity hypothesis is different from standard

discrimination, precisely because it is not about constraining lending, but because it is about a

failure to facilitate, and to enable it.

4.3.2. The “country puzzle”, as an underlying theme, in many articles

A large number of articles, initially obtained, refer to culture as a key source for country

differentiation. Yet, most articles do not pursue the analysis further: the frequency of use of

“cultur*” is very low, often limited to a single occurrence in the abstract.

This is echoing the idea that cultural concerns societal, or “societal” (O. Williamson, 2000)

analysis, but not economic analysis. These articles, considered in particular together with

those included in section 4.1, provide us with an inconsistent, and confusing view on national

cultures, national institutions, and on the way they interrelate. They reflect what Stulz (2007:

p.1596) labeled the “country puzzle” in finance: countries have a significant empirical effect

on many financial issues, yet no financial theory proposes any framework to date.

A few articles with a low occurrence of “cultur*” in the body of the text were included in the

database, because the occurrence seemed connected to a central argument made in the article.

This included in particular Stonehill, et al. (1975)***, Grinold, et al. (1989), and O’Bar and

Conley (1992)***.

Chapter II: A survey of ‘culture and finance’

126

The two former articles exemplify the “country puzzle”, each in its own way. So Grinold, et

al. (1989: p.1)*** investigate the stakes involved with managing global portfolios. They

emphasize the fact that “cultural differences among countries raise some important questions

regarding investor preferences, market efficiency, and security analysis”. They mention that

the relative importance of countries, vs. industries, must guide investment-management,

because countries’ influence is more prevalent than that of industries. The focus of their

effort, in this article, relates to an empirical split of both sources, national and industrial, in

the global risk-return profile of the portfolio.

Stonehill, et al. (1975)*** present the results of a global survey on the practice of financial

policies in five developed countries. Back in 1975, they conclude, that: “looking ahead, it

seems unlikely that financial executives outside the U.S would adopt the stockholder wealth

maximization model”. Again, this article is somewhat secondary to the main purpose of this

research thsis, and it has a strong historical connotation. Yet, it is important in helping to

replace the financial discipline and financial practice upon specific trajectories. It emphasizes,

further, the relative prevalence of national backgrounds, about four decades ago; while it

exemplifies how something akin to a “financial culture” can be situated historically, and

geographically.

4.3.3. The “culture of investing”

The article by O’Bar and Conley (1992)*** must receive, like the one by Hilary and Hui

(2009)***, a special place in this survey, because it is difficult to classify, specifically. It

provides a financial ethnography, on the motives guiding investment decisions by managers,

in institutional investment-funds, in the U.S. It concludes the ethnographic approach with the

idea that corporate culture and corporate politics dominate more technical and rational

“financial criteria” in managing investment-funds32.

32 In managing funds, “fund executives appear to be motivated more by the kinds of cultural influences that drive less consequential decisions. These include the quirks of institutional history and corporate politics, the desire to displace responsibility, and the demands of maintaining smooth personal relationships”, p. 21

Chapter II: A survey of ‘culture and finance’

127

This article, unlike the two other just reviewed, offers a high frequency of occurrence for

“cultur*”, and it provides a specific definition (appendix 1), which accounts, in particular, for

both “corporate-“ and “national-cultures”. Further, “cultur*” is used in combination with a

number of other expressions, or layers, including “business culture”, “culture of capital”,

“civil service culture” and “public fund culture”. This extension in the way culture is

conceived, and utilized, may seem puzzling. It reflects a reality that emerges from the

screening undertaken, here. In fact, the question of how anecdotic these, and other uses, are

should be considered seriously. It is arguable that they do not just reflect a way of speech, but

a deeper reality, engrained in the notion of “culture”. Besides they convey shades of meaning

in line with ancient academic knowledge throughout a range of more qualitative disciplines

(Cuche, 2004).

It must be noted that there is, in particular, a connection between the developments in section

4.1.2 about a “credit culture”, and a number of expressions emerging from the screening, and

substantiating the idea that financial techniques constitute a specific culture33. Further, they

echo Larry Summers’s (1985)° statement about the emerging specificity of a “financial

culture”, provided in section 2.1. Last, they mirror the expression “equity culture”, used by

Guiso, et al. (2001) to describe a increasing willingness, among the middle and upper classes

in continental Europe, to invest their savings in stock markets.

Overall, this crossing path, does not come out as a simplification of the survey, but it reflects

research in other fields, and, in particular, in sociology, anthropology, and more; and more

frequently, in International Business. These remarks are central to providing justification for

the developments in section 5.4, below.

33 Database available from the author on demand. References include "cross cultural differences in risk perception", “Embracing the risk: the changing culture of risk and responsibility” (a book title), “the equity culture: the story of the global stock market” (book title), “developing a risk-aware culture that involves honest discussions between firms and clients about the nature of risks they are exposed to”. A comment that “many studies are now recognizing the important institutional, developmental and cultural differences that exist between stock markets in industrialized and developing countries". Anecdotic references to “hot money cultures” (on stock markets and in funds), to the “cultural gap between… financial scholarly research” (and professional finance), or "wall street cultural hang-up” after the recent crisis.

Chapter II: A survey of ‘culture and finance’

128

5. Emerging  culture-­views  in  finance:  synthesis  

This review underlines the polarization of cultural research around dimensionalism, and

illustrates how dimensionalism appears to be the only consistent framework, emerging in

finance, through our process (notwithstanding efforts by Zingales, et al., on trust, in

economics). A number of issues must be formulated. First, dimensionalism has triggered

much controversy in management, at large (International Business, Accounting, Marketing,

etc.). Second, it is emerging in finance at a time, when it is questioned in proximate fields, for

instance in International Business, with increasing force and consensus. Third,

dimensionalism is only starting to emerge in finance, and it does so in a very specific manner.

5.1. Dimensionalism  in  management  research,  critiques  and  trajectories  

5.1.1. Critiques

It is beyond the scope of our survey to provide a consistent and detailed review of the

critiques of dimensionalism. They are strong, ancient, and voluminous.

On the one hand, inside dimensionalism alone, a number of debates are ongoing, in particular

with regards to the Globe approach that is resting on comparable bases, but that is not

mentioned as part of the stream by Vinken, et al. (2004∆∆∆, see the Hofstede vs. Javidan-and-

House’s exchanges in the Journal of International Business Studies, and other sources). A

number of researchers also emphasize that Schwartz’s approach overcomes a number of

important weaknesses in Hofstede’s framework (Schwartz and Ros, 1995∆∆(∆); Kagitcibasi,

1997∆∆; Smith, et al., 2002∆∆(∆)). It is noted, here, that Hofstede’s indices still largely

predominate in empirical research, as exemplified in the financial database, presented.

Chapter II: A survey of ‘culture and finance’

129

On the other hand, a number of direct theoretical and conceptual attacks have been carried out

on Hofstede’s framework. References abound34, they can be found in particular in Hofstede’s

exchanges with McSweeney (2002∆∆ [101/450] and subsequent articles), or with Baskerville

(2003∆∆ [50/148] and subsequent articles). Further critiques, with important cite-frequencies,

are found in Adler and Graham (1989∆∆ [20/265]), d’Iribarne (1997∆∆ [8/55]), Harrison and

McKinnon (1999∆∆ [-/115]), Chenhall (2003∆∆ [130/483]).

Similarly, critiques relating to the cultural distance framework, as derived from Hofstede,

abound: Shenkar (2001∆∆) provides a seminal critique, and Harzing (2004∆∆) provides both a

review and a critique.

5.1.2. Recent trajectory in International Business

To provide the reader with a sense of the current trajectory of dimensionalism, a specific

screening has been conducted, over one decade, centered on the main peer-reviewed journal

for International Business: the Journal of International Business Studies.

Overall, it seems that there is a consensus emerging on the need to “re-balance” cultural

research, by shifting the approach away from that of mean-values by country (focusing for

instance on variance as well), to using sources other than Hofstede’s, or still, to favoring

approaches other than the dimensionalist one. In particular the emphasis is on contrasting

varying cultural sources (community, gender, ethnicity, professions, organization, and so on),

and on accounting for the need to better specify the effects of each cultural source on the

phenomena considered.

34 We retain here only some of the most cited ones, so we provide the cite frequency in brackets for both EBSCO and Google scholar, search executed on March 27th. It will be noted that these critiques belong to a variety of disciplines, some of them relatively further apart from management research (sociology), and some other being more central (International Business, Marketing, Accounting…)

Chapter II: A survey of ‘culture and finance’

130

More specifically, Tung (2008) ∆∆(∆) re-emphasizes the need to consider intra-country cultural

variation (see earlier propositions by Au, 1999∆∆). She insists, as well, on the need to consider

cultural change. Smith (2006)∆∆, emphasizes, among other things, the need to verify that

relations at the individual levels (as operationalized in surveys), do hold in aggregate (at the

level of the national culture index). He stresses that a better analysis of the effect of economic

development is needed, and that wealth-related effects must be disentangled from cultural

effects, stricto sensu. Lenartowicz and Roth (1999)∆∆ recall that the validity of a country

grouping must be established beforehand, and that the use of a national culture indices must

be made, depending what is being studied, and how it relates to culture (a focus on

mechanisms). They propose several alternative methods. In a related enterprise, Sivakumar

and Nakata (2001)∆∆(∆) examine how countries can be grouped depending on Hofstede’s

indices. They show how the dimensions retained, and the weighting of these dimensions, can

lead to a large volatility in the way countries are grouped together, compared and opposed.

Leung, et al. (2005)∆∆ as well as Earley (2006)∆∆ emphasize how dimensionalist approaches

should be supplemented by other approaches in the future. In particular Leung, et al. (2005)∆∆

re-emphasize the importance of varying cultural “layers” (Global, national, organizational,

group-culture and then individual behaviors); they re-emphasize the processes of cultural

change, the interaction between cultural layers and the varying ways in which values can

influence behaviors (either causal links, mediating or moderating effects). They conclude:

“much of previous research has adopted what we view as a simplistic view of culture, which

tends to examine the static influence of a few cultural elements in isolation from other cultural

elements and contextual variables…Hofstede dimensions fall into this category…” (p. 374).

Earley underlines what approaches, other than those based on values, are seen to be

promising. He recalls that “meaning” moderates the relation between values and behaviors.

He concludes his review with the idea that “it may well be time that this form of large-scale,

multi-country survey [i.e. Hofstede’s or Globe’s] be set aside” (p. 922).

Chapter II: A survey of ‘culture and finance’

131

5.2. Dimensionalism  as  an  emerging  framework  for  financial  research?  

The previous section emphasizes how dimensionalism is coming under increasing scrutiny in

International Business research. Financial researchers, choosing to adopt a dimensionalist

approach, should acknowledge these issues and critiques, as is now being done in

International Business. Besides, dimensionalism is emerging in finance, at a time when it

seems to be somewhat re-qualified, or even re-considered in proximate fields.

Consequently, it is unclear whether the cultural framework, which is emerging from several

articles simultaneously in this screening, could be adopted, within financial research, as an

overarching framework of reference. Evidence suggests that economists and sociologists have

not been keen on adopting this framework in the past (Baskerville, 2003∆∆), while the

dimensionalist wave seems to be currently abating in management, at least as a “grand

design” (Leung, et al., 2005∆∆; Earley, 2006∆∆).

Furthermore, only the “smoothest” dimensionalist studies were identified in the JCR-

screening: the articles identified through the main screenings, (***) often propose streamlined

approaches, with an emphasis on mechanism, rather than on national cultures per se, and a

syncretic use of varying institutional indices. Two studies focus on the determinants of the

success in firms’ extension abroad: acquisitions in banking for Conn, et al. (2005***), and

FDIs in insurance for Outreville (2008***). Both of them use a cultural distance index as an

important and complementary addition to other institutional variables. Three other studies, all

of them published most recently (fall 2009, February 2010) parsimoniously use the

dimensionalist approaches: two studies use one of Hofstede’s dimensions while the third

study uses only two indices in combination. Aggarwal and Goodell (2009)*** underline how

they intend to investigate a variety of cultural and institutional factors simultaneously, to

Chapter II: A survey of ‘culture and finance’

132

develop a clearer picture of the determinants of financial macrostructures35. They provide no

specific definition of culture, but a rationale as to why that particular index, uncertainty

avoidance, could be important in their context, along with other institutional, financial and

economic determinants. Chui, et al. (2010***) propose a focus on individualism as a mediator

for a number of specific, seemingly related, and well-identified behavioral biases. Schmeling

(2009)** extends this latter study and tries to practically characterize the notion of “herd

behavior” through a reference to two related dimensionalist indices.

Overall, a flexible and pragmatic approach using ad-hoc dimensionalist references, based on a

precise delineating of the mechanisms at play, may be the more promising direction in

finance. This is what can be learned from the analysis of the databases and from the emerging

and comparative use of “cultur*” in the peer-reviewed literature in finance. This may explain

why researchers such as De Jong, Breuer and Quinten, expect the development of

dimensionalism to happen in relative autonomy from the financial discipline (see introduction

for references).

5.3.        National  embededdness  in  finance  

5.3.1. Opening the black box (opening up venues for culture in finance)…

The survey, presented here, outlines the importance of nations, and national cultures, for

many research articles. Relating to the dimensionalist papers, it is noted that the articles do

not come without extensive additional references to institutions. Conversely, in other

approaches, much confusion between culture and institutions has been observed (sections 4.1,

4.2.1 and 4.3). Overall, it seems that the emergence of the culture theme in financial research

35 They investigate the national determinants of macro-financial structures and state their positioning in the following manner: “why do countries … differ in the composition of financial intermediation, some relying on banks more while others relying more on markets? It would seem that such differences among countries may be related to the legal, cultural, and other such national characteristics, but these national characteristics as have been largely ignored in the literature …”.

Chapter II: A survey of ‘culture and finance’

133

corresponds to a renewed interest in trying to untangle the sources of national embededdness:

cultural, political, institutional, structural (networks), etc. It is noted, in passing, that this

effort has started some time ago, in other disciplines. For instance Guillen and Suarez (2005)

highlight the coexistence of five schools of thought, in relation to the exploration of the

broader institutional context, in Organization Theory; dimensionalism being only one of them,

and law and finance another. Further, the conclusions of the essays by Maseland (2009), or by

De Jong (2009∆∆∆: chapter 2), on the recent re-emergence of culture in economics, applies

very well to financial research, but with a potential time lag.

Stulz (2007)’s proposal of “the twin agency” model is an inspiring theoretical step to explore

the “country puzzle”. It opens up venues for exploring the black box of national

embededdness (The expression is original to the author). It echoes Williamson’s call to pay

attention to the much-needed alignment between the “institutions of embededdness” and the

varying levels of economic analysis (levels of societal analysis 1, 2, 3 and 4 in his

terminology). It corresponds to the primary importance given by North (see appendix 1) to

“informal institutions” and “culture”, in economic research.

The contributions by Stulz and Williamson (2003)***, and by Licht, et al. (2005)*++ cast a

new light on the successes of the law and finance research stream. The empirical regularities

observed in law and finance research articles may be robust, but if the laws are related to

enduring and stable cultural characteristics, country by country (understood as world-views,

national religions, egalitarianism and harmony, and still other dimensionalist indices), then

how can the results of law and finance be normatively interpreted? And how can

recommendations be derived? There are many robust empirical regularities, which have been

evidenced in law and finance, but short of clear conceptual mechanism, are there reliable

predictions?

Chapter II: A survey of ‘culture and finance’

134

For these reasons, and provided the formal, and less formal attention, now given to cultures in

financial research, it is arguable that culture and finance is set to become a full-blown

research-area soon. This answers to the second research-question proposed in introduction.

5.3.2. Using all the tools?

Future research may pragmatically draw from existing frameworks, from adjacent disciplines,

in management (in particular with emphasis on dimensionalism), in economics and

development-economics (in particular with special emphasis on trust or religion), or in

economic sociology (see below). It is suggested that this should not be done without detailed

cross-references. In fact, a detailed analysis of the database shows that successful cultural

research in finance generally comes with very specific references. For example Stulz and

Williamson (2003)*** provide extensive scholarly references drawn from the fields of

economics, economic sociology, law, economic and religious history. Guiso, et al. (web-site

references provided earlier) provide a stream of articles on trust, including experimental

investigation in trust-games, exploitation of survey questionnaires, comparison of varying

methods from varying disciplines, and they derive conceptual mechanisms from ancient

scholarship in political economy and economics (references provided above). Furthermore, a

number of authors provide an extensive analysis of risk-taking behaviors by individuals, of

their acceptance of uncertainty and of related concepts. They draw from a range of sources,

from psychology to economics, to investigate the behaviors of market-participants and

investors (see for instance Statman, 2007, 2008; Schmeling, 2009**; Chui, et al., 2010***).

All of these approaches (trust, religion, risk-appetite) intend to unpack specific culture

influences, through specific cultural lenses. They need not be opposed, and in fact, they

present large overlaps, often in relation to national and institutional influences.

Chapter II: A survey of ‘culture and finance’

135

Stulz and Williamson provide a discussion on how Protestantism relates to decentralization

and individualism, with a suggested relation to societal trust. Furthermore, they show how the

religious spirit and the related norms interact with national institutions to foster the

development of cross-national differentiation. In relation to the opposite connection, Guiso, et

al., remind us on how religious people have differential trust-levels (at the individual level, in

a given country, for instance the U.S). In aggregate, and in particular at the level of regions or

nations, they show that there are lasting differences, in the way individuals relate to their

community, and they identify, and measure lasting differences in the way societal trust/ civic

culture structures socio-economic and political interactions. Lastly, Statman relates social

capital (societal trust), individualism, culture and religion, to, and contrasts them with various

socio-economic categories (such as revenue, education, age, gender).

Therefore, is argued that: one, complementarities rather than divergences should be

emphasized across these approaches, and, two, that there is room for a variety of references in

financial research, provided appropriate cross-references are made from adjacent disciplines.

Furthermore, authors within the dimensionalist literature have extensively investigated the

importance of serial correlations of their indices with many country-level measures, across a

range of domains. There is evidence of serial correlation of national culture indices with

varieties of capitalism (Schwartz, 2007∆∆∆), with the rule of law, corruption and democratic

accountability (Licht, et al., 2007∆∆∆), with economic, demographic and political structures

(Schwartz, 2004∆∆∆), as well as with many other cultural variables, mentioned earlier. De Jong

(2009∆∆∆: p. 49.) summarizes similar, but somewhat less extensive, results for Hofstede’s and

the Globe’s indices.

Overall, the issue of national embededdness has been looming large in the social sciences,

including economics, and many venues exist for financial scholars. For instance generic

Chapter II: A survey of ‘culture and finance’

136

efforts to describe nations, going beyond the idea of national culture indices exist. So Hall and

Soskice (2001) propose an index based on a dichotomy between contractual vs. relational

economies. This distinction echoes historical evidence provided by Greif (2006) who studies

the development of trade by Genovese and Maghribi traders comparatively, from the 10th to

13th centuries. He shows how reputation and family ties (among Maghribi traders) can be

substituted for analytical accounting, contracting, and legal enforcement (with the Genovese).

The definitional issue of what culture is, remains, as does how financial scholars should tackle

it. This is what is considered next. In particular, it is highlighted that over-simplified views of

culture are likely to fail, while, in conclusion, it is suggested to backtrack definitions of

culture in approach tentatively labeled backward definition of cultures.

5.4. Culture  as  layers  and  considering  financial  cultures  

Firstly, we wish to stress the emphasis put on globalization, and on the existence of a

spreading “global culture” by a number of sources, which were mentioned earlier (Inglehart,

1997∆∆∆; Leung, et al., 2005∆∆; Shiller, 1999). Shiller insists that the countervailing influences

of “national cultures” and a “global culture” should be considered, when studying the

behavioral specificities of interest to financial scholars. Inglehart∆∆∆ derived from the World

Value Survey a theory called “post-modernism”, where economic development provides

ground for a “post-modern” culture, which is largely orthogonal to national cultures. In any

case, this push for globalization in finance (Stulz, 2007), and beyond (Guillen, 2001), sets a

number of issues that have to be considered, when relying on “national cultural indices”.

Recent research in International Business outlined the importance of cultural layers, national

culture being only one of them. This focus on culture as the combination of varying frames of

reference, within the individual, their potential contradictions, shifts, re-combinations is much

Chapter II: A survey of ‘culture and finance’

137

in line with traditional cultural research, specifically within the American social sciences

(Cuche, 2004). Moreover, recent advances in cognitive research provide ground to believe

that it is compatible with the functioning of the individual mind. DiMaggio (1997) underlines

how recent advances in cognitive sciences, in psychology and in sociology, provide basis for

the development of a relatively consensual research effort, structured around the idea of

cultures as layers and toolkits.

A benefit of these approaches is that they restore some of the anecdotic breadth that our

survey brought about, including aspects of culture related to organizations, techniques,

professions, genders, ethnicity, and (specifically with regards to our screening) with financial

cultures.

At this juncture, we must mention the growing field, within economic sociology, that

considers finance as a culture (MacKenzie and Millo, 2003, building on earlier and

comparable analyses of economic phenomena). This research left outside the scope of this

research until now, and as resulting from our field-based screening, is of interest in two ways.

Firstly, it seems to have gained some fame, including among financial practitioners and

scholars, particularly in the wave of the recent financial crisis. Secondly, it corresponds to an

elusive but well engrained idea, illustrated in the previous paragraph, that the business of

finance requires a culture, to lend successfully in banks, to trade and cover risks on markets

etc. After all, if three significant sources for culture emerge, by now, as being a global culture,

countries and organizations, it is surprising that professions should not feature prominently.

They are, and have always been, a central tenet in society (Durkheim, 1897/1967; Abott,

1988). Further, market finance with its “cognitive complexity and mature mathematical

models” (MacKenzie and Millo: p. 136) has a powerful cognitive base to be globally shared

across countries, ethnicities, social classes, etc., provided it is taught and learned. So, the

Chapter II: A survey of ‘culture and finance’

138

presumption is that financial cultures should supersede national and socio-economic cultures,

quite efficiently, in many circumstances, and in many ways. This would be particularly true

for a financial culture, technical and deductive, as opposed to, say, a national culture,

seemingly more value-driven.

MacKenzie and Millo (2003) follow the creation of the CBOE (Chicago Board Option

Exchange) and its seeding across the U.S, and the globe, to provide a case for a vision of

finance as a “performative culture”. Other recent research delineates varying views on

“financial cultures” (Fligstein, 200136; Westbrook, 2009), views that do not entirely, nor

necessarily overlap with the “performative” view.

These streams of research about finance, currently beyond the social boundaries of the

financial discipline, contribute, at the very least, one key message. The emerging conception

that the business of finance requires some specific professional cultures has theoretical

precedence, and academic depth; it relates to the practice of finance in banks or on markets; it

emerges, in a seemingly anecdotic way, from the screening undertaken, here, in this research.

It follows that the consideration of national cultures will need to account for this powerful -

and in many ways global - source of norms and behaviors: financial cultures.

* * *

36 He underlines the rise to prominence of two financial cultures over the last three decades, first a financial culture and then a shareholder culture

Chapter II: A survey of ‘culture and finance’

139

6. Conclusion:  shouldn’t  financial  research  explore  culture  further?  

This review should provide the reader with more perspectives than answers. The purpose has

been, precisely, to underline the fact that culture-research is not trivial, and does not fit well

into the traditional methodological toolkit used in finance. And culture-views, currently seen

to be emerging in financial research, through the field-based screening process undertaken,

here, can be characterized by an overall lack of consistency. This should not come as a

surprise, as is argued the essay annex, and this will explain why many financial (and

economic) scholars view culture-research as outside the natural range of their investigations.

Hence, the first question is answered negatively: can we characterize one or more accepted

theoretical background(s) about culture-research in finance, to derive testable hypothesis for

empirical research?. Culture is subject to what we have labeled definitional tactics, and it is a

central concept in the social sciences; many speak about a “meta-concept”. Culture-research is

ancient, broad, political and rife with controversies. It is not foreseen that finance will witness

the emergence of an overarching theoretical framework that would support the modeling and

empirical investigation of culture. Dimensionalism is very promising, but it seems unlikely to

be adopted as an overarching and integrative framework within financial research (see,

section 4.2). Yet, we do not believe that this should discourage researchers interested in

culture.

We have exemplified that culture has, at least implicitly, been around in financial research for

a long time. It has crystallized many preoccupations centered on the empirical signification of

countries. The ideas of “country proximity” and of “familiarity”, that present over the last half

a century in the study of trade and financial flows, illustrate this point. In the field of

investment management, the identification of countries as a primary source for diversification,

Chapter II: A survey of ‘culture and finance’

140

beyond industries, has been a key preoccupation for a long time too. Stulz (2007) recently

emphasized how a “country puzzle”, involving nine areas of financial research, has become a

central polarity for research in our field.

So, can we delimitate ‘culture and finance’ as a research-area within finance?

It has been around for a long time. Stakes are high. In particular, we illustrated how the

“country puzzle”, “institutional embededdness” and/or cultural factors are limiting the

normative assessments for a number approaches in finance, in particular in law and finance.

The interest is rising further, and some research is now being integrated: two significant

dimensionalist papers just came out, one in the fall of last year (2010), and another this

February (2011). Two articles with “alternative views” attest to the large audience the topic is

now received over the last couples of years (Stulz and Williamson, Grinblatt and Keloharju).

At the fringe of financial research (economics, management), established financial scholars

are calling for further investigation of culture-related variables (trust, religion, and so on), and

dimensionalist researchers call for the independent establishment of “Cultural Finance”.

International Business Scholars interested in Finance have already pushed this agenda further.

Overall, the time seems to have come to open the ‘black box’ of the “institutions of

embededdness”. So, we expect that culture and finance will grow further, and become a

research area in its own right.

Yet, our screening process has emphasized the ambiguity in the current uses of “culture”. In

particular, ambiguity occurs in relation to specific know-how in financial theories, financial

conventions and financial practices. While this may appear anecdotic, it is not. The alignment

of these occurrences has a long tradition in sociology, and other disciplines, which

Chapter II: A survey of ‘culture and finance’

141

investigates “finance as a culture”, and requires serious consideration. Furthermore, this

profound ambiguity in use may have some depth, and it is aligned with recent advances in the

cognitive sciences, sociology, psychology and international business (culture as layers). It

raises serious questions as to the mingling of national cultures with the business of finance

and its professional cultures.

As a result, trying to shy away from controversies about what culture is (or should be), and in

the belief that culture references are bound to continue to grow in finance, we suggest that

finance researchers should pragmatically implement an approach we tentatively label

“backward definitions”.

Researchers could define culture from the financial phenomena being studied, and they should

relate it to collective behaviors observed and grounded in some specific culture. The culture

will then have to be defined ad-hoc, with appropriate references, and with a focus on the

mechanism at play, backtracking the definition (hence a backward definition). Once you have

delineated the mechanism, linking a group’s behavior to some financial outcome, and once

the specificity of this group is identified through its distinctive collective behavior, a specific

culture with specific content should emerge naturally. Stated differently, and roughly, the

“whos” and the “hows” will clarify the “what” of cultural content. In fact, we note that the

“successful” dimensionalist papers, which are referenced here, provide this focus and

specificity with an emphasis on precise mechanisms. They mostly concentrate on one or two

dimensionalist measures, which are known to relate to the object of their study, in proximate

disciplines (such as, but not limited to, psychology, sociology, law, politics and economics).

Advocating a pragmatic approach based on these backward definitions, we believe that

financial research should restore the diversity in meanings to “culture”, with a primary, but

Chapter II: A survey of ‘culture and finance’

142

not exclusive, emphasis on national cultures. And we see no reason why it should, a priori,

include or exclude dimensionalist approaches. Some of the dimensionalist indices are

grounded in extended psychological research, and provide a very strong background for

financial research (this includes in particular, and for example, decades of investigation on the

relevance of individualism / collectivism). Further analysis of the variety in cultural

approaches (layers), and of the way they relate to the variety of financial subjects, is a

promising proposition for future research.

It will not be easy to encompass culture-research into finance, and it will require changes in

methods, including more conceptual and qualitative approaches, converging bodies of

evidence, and in particular, extensive as well as patient and careful cross-disciplinary

borrowings (Hofstede’s quote in exergue to the dissertation). However, simplified approaches

will be prejudicial to the understanding of culture in finance, as well as to the progress of

financial research along this axis.

So to prosper into a research-area ‘culture and finance’ should avoid controversies, it should

progress pragmatically around the notion of backward definitions, with emphasis on

mechanisms and collectives, and it should borrow from the wealth of culture-knowledge

accumulated in adjacent disciplines.

* * *

Chapter II: A survey of ‘culture and finance’

143

7. Appendix  1  (chapter  II).  Intuitive  breadth  of  culture  definitions  

Intuitive breadth of culture definitions, as collected from cited authors, and organized by discipline. When the concept defined is not ‘culture’, it is provided in parenthesis: Author, date (concept). Disciplinary classification is indicative. Authors are by alphabetical order.

7.1. Synthesis  

Drawing inference from the sample of definitions provided here, culture is associated with the following list of concepts: aesthetics, assumptions, beliefs, belief systems, categorizations, codes of conducts, cognition, concepts, convention, customs, discourse, doctrine, ethos, expressive symbols, facts, identities, ideologies, institutions, knowledge, language, material culture, meanings, models, motivational goals (values), norms, patterns of sampling information from the environment, practices, prejudices, priors, recipes, representations, rituals, role definitions, rules, schemata, self-definitions, shared attitudes, signs, suppositions, standard operating procedures, social institutions, strategies, style , symbols, tools, traditions, unstated assumptions, values, shared values. Only values, norms or beliefs seem to carry generality through their frequency of occurrence.

7.2. Definitions  in  economics  

Aoki, 2001: "the basic nature of social capital in the social exchange domain may remain relatively robust over time, which roughly corresponds to what is normally referred to as a cultural pattern…"

Aoki, 2001 (institutions): "self-sustaining system of shared beliefs… In order for beliefs to be shared by the agents in a self-sustaining manner and regarded by them as relevant to the consequences of their choices, they must have substantive bases"

De Jong, 2009∆∆∆: he reviews different sets of definitions and distinguishes between broad & narrow definitions of culture. The broad definitions "have in common that they refer to the entire social system; the complex 'whole' and the 'totality'. By stressing the entirety, these definitions lack focus and become open-ended.... In my view, a problematic feature of these definitions is that they contain both the possible sources of behaviors (e.g. beliefs), the behaviors itself, and its results". The more narrow and more useful definitions, all "have some common features: i) values are essential, ii) they refer to a group, iii) they refer to a trend or a pattern, and iv) the cultural elements are humanely devised aspects that are transmitted from generation to generation. All definitions refer explicitly or implicitly to values"

Greif, 1994 (institutions and culture): "institutions -the non-technological constraints on human interactions- are composed of two interrelated elements: cultural beliefs (how individuals expect other to act in various contingencies) and organizations (the endogenous human constructs that alter the rules of the game)"

Chapter II: A survey of ‘culture and finance’

144

Guiso, Sapienza & Zingales, 2006, providing a “narrow definition of culture“: “those customary beliefs, values, and social constraints that ethnic, religious and social groups transmit fairly unchanged from generation to generation” [with reference to the Webster dictionary]"

North, 1990 (institutions): institutions have three dimensions: they consist of informal constraints, of formal constraints and of the enforcement of these constraints.

North, 1990: Culture is defined as the antecedent of informal constraints: “Where do informal constraints come from? They come from socially transmitted information and are part of the heritage that we call culture”. Culture provides “customs, traditions and codes of conducts”. Culture is of major importance for economic phenomena “the formal rules make up a small (although very important) part of the sum of constraints that shape choices; a moment’s reflection should suggest to us the pervasiveness of informal constraints”.

Williamson, 2000 (embededdness): there are "four levels of social analysis […] the top level is the social embededdness level. This is where the norms, customs, mores, traditions, etc. are located. Religion plays a large role at this level. Although Level 1 analysis is undertaken by some economic historians and other social scientists... level 1 is taken as given by most institutional economists... The concept of embededdness, both at the level of society and in the context of ongoing network relations, has been advanced to help explicate these issues (Granovetter, 1985). The vast literature on culture (Paul DiMaggio, 1994) is also pertinent. Neil Smelser and Richard Swedberg discuss these and related issues in their introduction to the Handbook of Economic Sociology, where they observe that different kind of embededdness -cognitive, cultural, structural and political- should be distinguished, and conclude that the concept of embededdness remains in need of greater theoretical specification”

7.3. Definitions  in  finance  

Breuer & Quinten, 2009*** (culture as values): "culture may be understood as a complex entity of cognitions, shared by the members of a social group. The focal point of the cognitions is (core) values, which are assumed to steer individual behavior"

Castro, Desender & Escamilla de Leon, 2007°: "culture refers to the complex of meanings, symbols, and assumptions about what is good or bad, legitimate or illegitimate that underlie the prevailing practices and norms in a society (Bourdieu, 1977). Value emphases are the essence of culture seen this way. They are the implicitly or explicitly shared, abstract ideas about what is good, right, and desirable in a society (Williams, 1970). They justify and guide the ways that social institutions (e.g., the family, education, economic, political, religious systems) function, their goals and modes of operation. Social actors (e.g., organizational leaders, policy-makers, firm managers) draw on these cultural value emphases to select actions, evaluate people and events, and explain or justify their actions and evaluations"

O'Bar & Conley, 1992***: “the set of shared beliefs and practices that define a society's (or an organization's) way of life. Culture provides the mental map that guides individual members of the society through their daily live”

Chapter II: A survey of ‘culture and finance’

145

Schiller, 1997: "the concept of culture central to sociology and cultural anthropology ever since the work of Taylor (1871), Durkheim (1893) and Weber (1947) is related to the selective attention that the human mind exhibits. There is a social cognition, reinforced by conversation ritual and symbols that is unique to each interconnected group of people; to each nation, tribe, or social group. The array of facts, suppositions, symbols, categories of thought that represent a culture have subtle and far-reaching affect on human behavior"

Sekely & Collins, 1988***: "while there is a great deal of discussion and debate over exactly what constitutes culture, most definitions include the following elements: social institutions, belief systems, aesthetics, language, and material culture. It is this last element, which includes a society's economic structure and technological capabilities that probably is most closely related to the capital structure tendencies of a society. Specific areas most likely to influence capital structure include the different legal and tax systems, which give rise to differences in property rights across cultures. However, the potential impact on capital structure is not limited to the material portion of culture. There are a number of other' aspects of a society's culture which also could impact on the financial structure of a firm"

Stulz & Williamson, 2001***: “for our purpose, a suitable definition of culture is the one that North (1990) cites from the work of Boyd and Richerson (1985): culture is defined as ‘transmission from one generation to the next, via teaching and imitation, of knowledge, values, and other factors that influence behavior’”

Throsby, 2001: “a set of values, beliefs, traditions, customs, etc. which serve to identify and bind a group together”.

7.4. Definitions  in  cross-­cultural  psychology  

Licht, Goldschmidt & Schwartz, 2007: "definitions of culture abound, but the common denominator of all the definitions is that culture represents shared values and beliefs "

Schwartz, 1999°°° (cultural vs. individual values): "the explicit and implicit value emphases that characterize a culture are imparted to societal members through everyday exposure to customs, laws, norms, scripts and cultural values (Bourdieu, 1972; Markus & Kitayama, 1994)"

Schwartz, 2004°°° (culture as values, at the collective level): "the rich complex of meanings, beliefs, practices, symbols, norms, and values prevalent among people in a society. The prevailing value emphases in a society may be the most central feature of culture… These value emphases express shared conceptions of what is good and desirable in the culture, the culture ideals"

Schwartz, 2004°°° (culture as values vs. other sources): "note that I refer only to value dimensions. In the introduction I argued that these are particularly significant dimensions for comparing cultures because they affect so many different aspects of life. But other dimensions of cultural difference, such as the tightness or looseness of normative systems, holistic vs. analytic styles of thought, degree of emotional expressiveness, and time perspective, are also important"

Triandis, 1996 (cultural syndromes)°°°: "the patterns of shared attitudes, categorizations, self-definitions, norms, role definitions, and values…”

Chapter II: A survey of ‘culture and finance’

146

7.5. Definitions  in  international  management  

Earley, 2006: “culture is not a value or a set of values; culture is the meaning we attach to aspects of the world around us… Many shortcomings of current research on cross-cultural issues can be connected to the obsession we have values as culture rather than meaning as culture. Even the traditional scholars dealing with values (Parsons & Shils, Kluckhohn & Strodtbeck, Mead, Rokeach) did not fall into this trap that values are culture... Second meaning systems are imperfectly shared across individuals and/or segments (sub-populations) within the same society"

Hofstede, 1980°°°: "collective programming of the mind […] distinguishing the members of one human group from another"

Hofstede, 2001°°°: "a social mind"

Hofstede, 2004°°°: "Culture is to society what memory is to individuals. It includes what has worked in the history of the society -tools concepts, ideologies, norms, values, prejudices, standard operating procedures, unstated assumptions, pattern of sampling information from the environment- that most members of the society teach to the next generation. This teaching is done by example or explicitly..."

House & Javidan, 2004: “we define culture as shared motives, values, beliefs, identities and interpretations or meanings of significant events that result from common experiences of members of collectives and are transmitted across age and generation”

House, Quigley & de Luque, 2010: ... following their preceding 2004 definition... "we emphasized the 'sharedness' of the cultural indicators among members of a given collective and noted that the specific criteria used to distinguish among cultures (i.e. dimensions of culture) were likely to depend on the preferences and/or discipline of the investigator and the issues under investigation"

Siegel, Licht & Schwartz, 2006: "social players interact with partners assumed to share the same priors (beliefs) and to be guided by similar sets of motivational goals (values)"

7.6. Definitions  in  political  sciences  

Inglehart, 1997°°°: "a culture is a system of attitudes, values, and knowledge that is widely shared within a society and is transmitted from generation to generation. While human nature is biologically innate and universal, culture is learned and varies from one society to another. […] By culture, we refer to the subjective aspect of a society's institutions: the beliefs, values, knowledge, and skills that have been internalized by the people of a given society, complementing their external systems of coercion and exchange. This is a narrower definition of culture than is generally used in anthropology, because our purpose here is empirical analysis"

Chapter II: A survey of ‘culture and finance’

147

Inglehart, 1997°°° (culture stability): "the more central and early learned aspects of culture are resistant to change, both because it requires a massive effort to change central elements of an adult’s cognitive organization, and because abandoning one's most central beliefs produces uncertainty and anxiety. In the face of enduring shifts in socioeconomic conditions, even central parts of culture may be transformed, but they are more likely to change through intergenerational population replacement than by the conversion of already socialized adults"

7.7. Definitions  in  sociology  

Cuche, 2004: “the concept of culture is inherent to thinking in the social sciences. In a way, it is a necessary concept to fathom singularity, within the diversity of the human race, in a non-biological way (la notion de culture est inhérente à la réflexion des sciences sociales. Elle leur est nécessaire en quelque sorte pour penser l'unité de l'humanité dans la diversité autrement qu'en termes biologiques)”

DiMaggio, 1994 (culture processes): “shared cognition, values, norms and expressive symbols”

DiMaggio, 1994 (culture processes): “representations of culture as a toolkit or repertoire […rather than] the latent-variable view of culture as coherent, integrated, and ambiguous”

Jepperson & Swindler, 1994 (culture's dimensions): culture's dimensions ordered from the least conscious to the more expressive [identical paragraphs in original]:

“Codes, rules, schemata, models Identities, practices, recipes, strategies, norms, values Convention, custom, tradition Symbols, signs, rituals Knowledge, discourse, representations, doctrine, ideology Ethos, style “

Jepperson & Swindler, 1994 (institutionalization of culture): “some cultural elements are more institutionalized than others, i.e.. more linked to other cultural elements, more embodied in standardized routines and formal organizations, more taken for granted as fixtures of the environment [...these forms of culture that are] congealed in forms that require less by way of maintenance, ritual reinforcement and symbolic interaction than the softer realms we usually think of as culture"

Scott, 95 (institutional carriers): “Scott distinguishes among three main institutional carriers, namely cultures, social structures and routines” (in Guillen, 2004)

* * *

Chapter II: A survey of ‘culture and finance’

148

8. Appendix  2  (chapter  II).  Definitions  for  national  culture  indices  

De Jong (2009: appendix 1, p. 197, 206) provides a complete overview of the varying dimensionalist indices.

8.1.  Hofstede’s  five  dimensions  

The following citations are obtained from Hofstede (2001: p. xix & xx) “Power-distance is the extent to which the less powerful members of organizations and

institutions accept and expect that power is distributed unequally. The basic problem involved is the degree of human inequality that underlies the functioning of each particular society”.

“Uncertainty avoidance is the extent to which a culture programs its members to feel either

uncomfortable or comfortable in unstructured situations. Unstructured situations are novel, unknown, surprising, different from usual. The basic problem involved is the degree to which a society tries to control the uncontrollable”.

“Individualism on the one side versus its opposite collectivism is the degree to which

individuals are supposed to look after themselves or remain integrated into groups, usually around the family. Positioning itself between these poles is a very basic problem all societies face”.

“Masculinity versus its opposite femininity refers to the distribution of emotional roles

between the genders, which is another fundamental problem for any society to which a range of solutions are found; it opposes ‘tough’ masculine to ‘tender’ feminine societies”.

“Long-term versus short term orientation refers to the extent to which a culture programs its

members to accept delayed gratification of their material social and emotional needs”

Chapter II: A survey of ‘culture and finance’

149

8.2. Schwartz’s  four  dimensions  

The following short definitions are obtained from Chui, Lloyd & Kwok (2002*++: p. 101-102) Extended definitions can be found in Schwartz (1994: p. 102-105 for group-level value types; the individual-level value types are different and a definition is provided in table 7.1, p. 89)

Conservatism/ autonomy Conservatism includes values that are important in close-knit harmonious relationships, in

which the interests of the individual are not viewed as distinct from those of the group. These values are primarily concerned with security, conformity, and tradition.

Intellectual and Affective Autonomy are values likely to be important in a society that

views the individual as an autonomous entity entitled to pursue his or her own interests. Intellectual autonomy places an emphasis on self-direction and affective autonomy

emphasizes stimulation and hedonism.

Mastery/Harmony Mastery accentuates active mastery of the social environment through self-assertion. Such

values promote the active efforts of people to change their surroundings and get ahead of others.

Harmony lays emphasis on harmony with nature. Egalitarianism/ hierarchy Egalitarian Commitment emphasizes the transcendence of selfish interests. This group of

values concerns voluntary commitment to help improve the welfare of other people. Hierarchy stresses the legitimacy of hierarchical roles and resource allocation. * * *

Chapter II: A survey of ‘culture and finance’

150

9. Appendix   3   (chapter   II).   Institutional   control   variables   in   the  

dimensionalist  literature  presented  in  our  database  

We provide, below, a categorization of varying “institutional” variables used in any of the 29 articles contained in the dimensionalist database. The categorization is our own and contains 85 items.

The scope of this survey has not included a consideration of what institutions encompass. The concept may seem more appealing because it is less ‘fuzzy’. Yet, potential discrepancies within disciplines may be counterbalanced by a larger variance in meaning across disciplines than is the case for culture. As an illustration, let us note that, in finance, institutions may describe the “institutional investors”, or the bodies that contribute to the functioning of the markets (rating agencies, regulators…) in a narrow sense. Whereas, in a broader sense, it is often associated with the legal infrastructure in which financial markets and firms operate. In a broader sense still, it also includes elements of the political or social environment.

Below, we reproduce a list of contextual elements that have been considered as part of the institutional environment by at least one researcher in the database. It does not provide a list of ‘institutions’. Illustrative definitions on institutions and on the relation between institutions and culture are provided in appendix 1. Defining and characterizing ‘institutions’ for financial research should be the object of further research.

We provide more details on our website. In fact, an excel spreadsheet is available that lists a total of about 120 “institutional” indicators, along with original labels, initial sources (one specific source per indicator) and corresponding dimensionalist papers implementing the variable. Corporate governance environment Accounting

• A composite index of overall disclosure quality and intensity

• Accounting quality • Accounting transparency • Transparency and timeliness of financial

information Boards. Board efficiency Market efficiency

• Average transaction cost for investment management

• Protection of minority shareholders (index)

• Take-over market: existence of hostile takeovers

Ownership concentration • Fraction of insider ownership • Fraction of ownership by minorities • Median ownership by top three

shareholders • Fraction of firms with no controlling

shareholder

Legal environment Legal (financial)

• Existence of legal reserve to prevent dissolution of company

• Anti director rights index & shareholder protection & aggregate ‘investor protection’ index

• Creditor rights • Securities law disclosure quality • Securities law private litigation

importance • Outside Investor Protection • Anti Self-Dealing index

Legal (general) • Case Law dummy • Legal Origin dummy • Contract enforcement index • Legal Enforcement index • Legal formalism & legal flexibility • Rule of law • Efficiency of the judiciary

Chapter II: A survey of ‘culture and finance’

151

Culture & geography Culture

• Dominant religion • Existing religions (% or dummies) • Common language • Major languages in countries • Common colonial tie • Ethnic fractionalization • Ethnic, language & religious

fractionalization, separate and aggregate indices

Distance • Distance from the equator • Distance between capital cities for

country pairs Wars. Number of times a country was at war Economic environment The macro-economic environment

• Quality of the banking system • Antitrust: perceived effectiveness of

antitrust Policy • Antitrust enforcement expenditure • Inflation • Dependency ratio (proportion of pop.

pensioned) • Natural resource abundance • Existence of national deposit insurance

scheme • Existence of an exchange rate system • Degree of concentration of countrywide

firms • Infrastructure quality: facility of

transportation • Existence of capital controls

Main indicators • GDP per capita • Development: growth in GDP per cap

growth • Gross domestic product (GDP) • National economic inequality

Openness • Degree of overall globalization • Social and economic openness • Trade effectiveness / trade barriers • Trade openness (Trade over GDP)

Taxation • Tax on investment by non residents • Corporate tax rate • Dividend tax systems

Political environment Country governance

• Government quality or government effectiveness

• Political legitimacy of government • Measure of political constraint on

governmental action • Civil liberties • Corruption, corruption perception • Economic freedom • Quality of checks and balances in the

legislative function • Risk of expropriation, property rights • Principal component: voice and accoun-

tability, political stability and absence of violence, government effectiveness, regulatory burden, rule of law, freedom from graft

• Aggregate index for institutional quality (rule of law, self-dealing, public efficiency, public enforcement, expropriation)

• Property Rights: degree of protection of private property and property rights

Political risk • Country risk index & varying political

risk indices Political system

• Federalism • Lest/right; coloration of governments • Political stability • Former socialist state dummy

Elements of social stability and protection • Newspaper circulation • Principal component of revolution,

assassination and corruption; or same indices separately

• Rigidity of Labor Laws • Labor relation quality: implicit contracts

= employee’s tenure • Social benefits: sickness and health

benefits / unemployment benefits / social security laws

• Tenure: average employee tenure with one employer

• Union power: legally mandated worker participation on boards

• Union density Socio-political environment Human capital

• Human Capital index • Life expectancy • Literacy rate • Secondary school enrollment

Chapter II: A survey of ‘culture and finance’

152

10. Bibliography  

We are indebted to an anonymous reviewer of the Revue Finance for this advice, which simplifies reading. Details on the method relating to the main screenings (of financial research-literature stricto sensu), the double review process (extensive review of dimensionalist advances in finance), and the identification of relevant material, are provided in section 2.2. *** Identified through the main screenings (eighteen articles). * Identified through the double review process (24 additions). ++ A marking for articles, where at least one of the co-authors has an established track

record in any of the journals screened as part of the main screening (the J.C.R. list). *++ It results that *++ stands for articles from the financial community, lato sensu. ° Obtained through the initial screenings and analyzed separately due to the more

limited relevance. ∆∆∆ References to ‘foundational’ articles for dimensionalism. ∆∆ References to seminal critiques of dimensionalism. It should be noted that there is

“extension”, so that there is overlap between the key sources and critiques ∆∆(∆).

Abbott, A., 1988. The system of professions: an essay on the division of expert labor. Chicago: The University of Chicago Press.

∆∆ Adler, N.J., Graham, J.L., 1989. Cross-cultural interaction: the international comparison fallacy? Journal of International Business Studies, 20(3): 515-537.

*** Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: national characteristics as determinants. Journal of Banking & Finance, 33(10): 1770-1780.

*++ Aggarwal, R., Kearney, C., Lucey, B., 2009. Is gravity a cultural artifact? The role of cultural distance in foreign portfolio investment. Infiniti conference proceedings.

*** Alessandrini, P., Calcagnini, G., Zazzaro, A., 2008. Asset restructuring strategies in bank acquisitions: does distance between dealing partners matter? Journal of Banking & Finance, 32(5): 699-713.

* Anderson, C.W., Fedenia, M., Hirschey, M., Rantala, H., 2007. There's no place like home: cultural influences on international diversification by institutional investors. Working paper (available from SSRN)

Arrow, K., 1972. Gifts and exchanges. Philosophy and Public Affairs, 1(343-362)

∆∆ Au, K.Y., 1999. Intra-cultural variation: evidence and implications for international business. Journal of International Business Studies, 30(4): 799-812.

Banfield, E., 1958. The moral basis of a backward society. New York: Free Press.

∆∆ Baskerville, R.F., 2003. Hofstede never studied culture. Accounting, Organizations and Society, 28(1): 1-14.

Beckerman, W., 1956. Distance and the pattern of intra-European trade. Review of Economics and Statistics, 38(1): 31-40.

Chapter II: A survey of ‘culture and finance’

153

*++ Beugelsdijk, S., Frijns, B., 2009. A cultural explanation of the foreign bias in international asset allocation. Australasian Finance and Banking conference (available from SSRN)

Brennan, M.J., 1995. Corporate finance over the past 25 years. Financial Management, 24(2): 9-22.

*++ Breuer, W., Salzmann, A., 2008. Cultural Dimensions of Corporate Governance Systems. Working Paper (available from SSRN)

*∆∆∆ Breuer, W., Quinten, B., 2009. Cultural finance. Working paper (available from SSRN)

Burnett, J.J., Palmer, B.A., 1984. Examining life insurance ownership through demographic and psychographic characteristics. Journal of Risk & Insurance, 51(3): 453-467.

Calomiris, C.W., Kahn, C.M., Longhofer, S.D., 1994. Housing-finance intervention and private incentives: helping minorities and the poor. Journal of Money, Credit & Banking, 26(3): 634-674.

Charreaux, G., Schatt, A., 2006. La recherche française en gouvernance d'entreprise: un panorama. Banque et marchés, 82: 7-23.

∆∆ Chenhall, R.H., 2003. Management control systems design within its organizational context: findings from contingency-based research and directions for the future. Accounting, Organizations & Society, 28(2): 127-168.

*++ Chui, A., Kwok, C., 2008. National culture and life insurance consumption. Journal of International Business Studies, 39(1): 88-101.

*++ Chui, A., Lloyd, A., Kwok, C., 2002. The determination of capital structure: is national culture a missing piece to the puzzle? Journal of International Business Studies, 33(1): 99-127.

*** Chui, A.C.W., Titman, S., Wei, K.C.J., 2010. Individualism and momentum around the world. Journal of Finance, 54(1): 2143-2184.

* Clement, M.B., Rees, L., Swanson, E.P., 2003. The influence of culture and corporate governance on the characteristics that distinguish superior analysts. Journal of Accounting, Auditing & Finance, 18(4): 593-618.

*** Conn, R.L., Cosh, A., Guest, P.M., Hughes, A., 2005. The impact on U.K. acquirers of domestic, cross-border, public and private acquisitions. Journal of Business Finance & Accounting, 32(5/6): 815-870.

Cuche, D., 2004. La notion de culture dans les sciences sociales. Paris: La Découverte - Repères. 124 pages.

∆∆ d'Iribarne, P., 1997. The usefulness of an ethnographic approach to the international comparisons of organizations. International Studies of Management and Organization, 26(4): 30-47.

* De Jong, E., 2002. Why are price stability and statutory independence of central banks negatively correlated? The role of culture. European Journal of Political Economy, 18(4): 675.

Chapter II: A survey of ‘culture and finance’

154

∆∆∆ De Jong, E., 2009. Culture and economics: on values, economics and international business. Routledge.

* De Jong, E., Semenov, R., 2002. Cross-country differences in stock market development: a cultural view. Working paper (available from SSRN)

* De Jong, E., Semenov, R., 2006. Cultural determinants of ownership concentration across countries. International Journal of Business Governance and Ethics, 2(1): 145-165.

DiMaggio, P., 1997. Culture and cognition. Annual Review of Sociology, 23: 263-287.

DiMaggio, P.J., 1994. Culture and economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology. Princeton: Princeton University Press

Durkheim, E., 1897/1967. De la division du travail social. Paris: Les Presses universitaires de France. 416 pages.

∆∆(∆) Earley, P.C., 2006. Leading cultural research in the future: a matter of paradigms and taste. Journal of International Business Studies, 37(6): 922-931.

∆∆(∆) Earley, P.C., Singh, H., 1995. International and intercultural management research: what's next. The Academy of Management Journal, Vol. 38(2): 327 - 340.

Fligstein, N., 2001. The architecture of markets: an economic sociology of twenty-first century capitalist societies. Princeton: Princeton University Press. 288 pages.

*** Focarelli, D., Pozzolo, A.F., 2008. Cross-border M.&A.s in the financial sector: is banking different from insurance? Journal of Banking & Finance, 32(1): 15-29.

*** Gande, A., Parsley, D.C., 2005. News spillovers in the sovereign debt market. Journal of Financial Economics, 75(3): 691-734.

*** Gleason, K., McNulty, J.E., Pennathur, A.K., 2005. Returns to acquirers of privatizing financial services firms: an international examination. Journal of Banking & Finance, 29(8): 2043-2065.

Greif, A., 2006. Institutions and the path to the modern economy: lessons from medieval trade. Cambridge University Press. 503 pages.

*** Grinblatt, M., Keloharju, M., 2001. How distance, language, and culture influence stockholdings and trades. Journal of Finance, 56(3): 1053.

*** Grinold, R., Rudd, A., Stefek, D., 1989. Global factors: fact or fiction? Journal of Portfolio Management, 16(1): 79-88.

Guillen, M.F., 2001. Is globalization civilizing, destructive or feeble? A critique of five key debates in the social science literature. Annual Review of Sociology, Vol. 27: 235 - 260.

Guiso, L., Haliassos, M., Jappelli, T., 2001. Household portfolios. Cambridge: MIT press. 508 pages.

Guiso, L., Sapienza, P., Zingales, L., 2007. Does culture affect economic outcomes? Journal of Economic Perspectives, 20(2): 23-48.

Guiso, L., Sapienza, P., Zingales, L., 2009. Culture biases in economic exchange? Quarterly Journal of Economics, 124(3): 1095-1131.

Chapter II: A survey of ‘culture and finance’

155

Hall, D.P.A., Soskice, D.W., 2001. Varieties of capitalism: the institutional foundations of comparative advantage. Oxford: Oxford University Press. 560 pages.

∆∆ Harrison, G., McKinnon, J., 1999. Cross-cultural research in management control systems design: a review of the current state. Accounting, Organizations & Society, 24(5): 483-506.

∆∆(∆) Harzing, A.W.K., 2004. The role of culture in entry mode studies: from negligence to myopia. Advances in International Management, 15: 75-127.

*** Hilary, G., Hui, K.W., 2009. Does religion matter in corporate decision making in America? Journal of Financial Economics, 93(3): 455-473.

∆∆∆ Hofstede, G., 1980. Culture's consequence. New-York: Sage publications.

∆∆∆ Hofstede, G., 2001. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

House, R.J., et al., 1999. Cultural influences on leadership and organizations: project Globe. In Mobley, W.H. (Ed.), Advances in global leadership. Stamford, CT: JAI Press. 171-233.

* Huang, R., 2009. Tolerance for uncertainty and the growth of informationally opaque industries. Journal of Development Economics, 87(2): 333-353.

Hunter, W.C., Walke, M.B., 1996. The cultural affinity hypothesis and mortgage lending decisions. The Journal of Real Estate Finance and Economics, 13(1): 57-70.

∆∆∆ Inglehart, R., 1997. Modernization and postmodernization: cultural economic and political change. Princeton: Princeton University Press. 440 pages.

∆∆ Kagitcibasi, C., 1997. Individualism and collectivism. In Berry, J., Segall, S., and Kagitçibasi, C. (Ed.), Handbook of cross-cultural psychology. Needham Heights: Allyn and Bacon. 1-49.

* Kimbro, M.B., 2002. A cross-country empirical investigation of corruption and its relationship to economic, cultural, and monitoring institutions: an examination of the role of accounting and financial statements quality. Journal of Accounting, Auditing & Finance, 17(4): 325-349.

∆∆(∆) Kirkman, B.L., Lowe, K.B., Gibson, C.B., 2006. A quarter century of Culture's Consequences: a review of empirical research incorporating Hofstede's cultural values framework. Journal of International Business Studies, 37(3): 285-320.

∆∆∆ Kogut, B., Singh, H., 1988. The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19(3): 411-432.

Kroeber, A.L., Kluckhohn, C., 1952. Culture: a critical review of concepts and definitions. Harvard: Peabody Museum of American Archaeology and Ethnology.

*++ Kwok, C., Tadesse, S., 2006. National culture and financial systems. Journal of International Business Studies, 37(2): 227-247.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 1997. Legal determinants of external finance. Journal of Finance, 52(3): 1131-50.

Chapter II: A survey of ‘culture and finance’

156

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political Economy, 106(6): 1113-55.

La Porta, R., Lopez-de-Silanes, F., Pop-Eleches, C., Schleifer, A., 2004. Judicial check and balances. Journal of Political Economy, 112(2): 445-470.

∆∆(∆) Lenartowicz, T., Roth, K., 1999. A Framework for culture assessment. Journal of International Business Studies, 30(4): 781-798.

∆∆(∆) Leung, K., et al., 2005. Culture and international business: recent advances and their implications for future research. Journal of International Business Studies, 36(4): 357-378.

*++ Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2005. Culture, law, and corporate governance. International Review of Law & Economics, 25(2): 229-255.

∆∆∆ Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2007. Culture rules: the foundations of the rule of law and other norms of governance. Journal of Comparative Economics, 35(4): 659-688.

MacKenzie, D., Millo, Y., 2003. Constructing a market, performing theory: the historical sociology of a financial derivatives exchange. American Journal of Sociology, 109(1): 107-145.

Maseland, R., 2009. A history of thought about culture and economy. In Culture and economics: on values, economics and international business, De Jong (Ed.), Routledge, 260 pages.

∆∆ McSweeney, B., 2002. Hofstede's model of national cultural differences and their conserquences: a triumph of faith - a failure of analysis. Human Relations, 55(1): 89-118.

*** Mian, A., 2006. Distance constraints: the limits of foreign lending in poor economies. Journal of Finance, 61(3): 1465-1505.

Morosini, P., Shan, S., Singh, H., 1998. National cultural distance and cross-border acquisition performance. Journal of International Business Studies, 29(1): 137-158.

*** O'Barr, W.M., Conley, J.M., 1992. Managing relationships: the culture of institutional investing. Financial Analysts Journal, 48(5): 21-27.

*** Outreville, J.F., 2008. Foreign affiliates of the largest insurance groups: location-specific advantages. Journal of Risk and Insurance, 75(2): 463-491.

Oyserman, D., 2002. Values, psychological perspectives. In Smelser and Baltes (Ed.), International Encyclopedia of the social and behavioral sciencese. 16150-16153.

* Pirouz, D.M., 2004. National culture and global stock market volatility. Working paper (available from SSRN)

Portes, R., Rey, H., Oh, Y., 2001. Information and capital flows: the determinants of transactions in financial assets. European Economic Review, 45(4-6): 783-796.

Putnam, R., 1993. Making democracy work. Princeton University Press, Princeton, NJ.

Chapter II: A survey of ‘culture and finance’

157

*++ Ramirez, A., Tadesse, S., 2007. Corporate cash holdings, national culture and multinationality. William Davidson Institute Working Papers

*** Rosati, S., Secola, S., 2006. Explaining cross-border large-value payment flows: evidence from 'TARGET' and 'EURO1' data. Journal of Banking & Finance, 30(6): 1753-1782.

*** Sarkissian, S., Schill, M.J., 2004. The overseas listing decision: new evidence of proximity preference. Review of Financial Studies, 17(3): 769-809.

*** Schmeling, M., 2009. Investor sentiment and stock returns: Some international evidence. Journal of Empirical Finance, 16(3): 394-408.

∆∆∆ Schwartz, S.H., 1994. Beyond individualism/collectivism: new cultural dimensions of values. In Kim, U., et al. (Ed.), Individualism and collectivism: theory, method and application. Newbury Park: Sage

∆∆∆ Schwartz, S.H., 2004. Mapping and interpreting cultural differences around the world. In Vinken, H., Soeters, J., and Ester, P. (Ed.), Comparing cultures: dimensions of culture in a comparative perspective. Leiden: Brill Academic Publishers

∆∆∆ Schwartz, S.H., 2007. Cultural and individual value correlates of capitalism: a comparative analysis. Psychological Inquiry, 18(1): 52-57.

∆∆(∆) Schwartz, S.H., Ros, M., 1995. Values in the West: a theoretical and empirical challenge to the individualism-collectivism dimension. World Psychology, 1(2): 91-122.

∆∆ Shenkar, O., 2001. Cultural distance revisited: towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32(3): 519-535.

Shiller, R.J., 1999. Human behavior and the efficiency of the financial system. In Taylor, J.B. and Woodford, M. (Ed.), Handbook of macroeconomics. Amsterdam: Elsevier. 1305-40.

*++ Siegel, J.I., Licht, A.N., Schwartz, S.H., 2006. Egalitarianism and international investment. Working paper (available from SSRN)

Siegel, J.I., Licht, A.N., Schwartz, S.H., 2008. Egalitarianism, cultural distance, and foreign direct investment: a new approach. Working Paper (available from SSRN)

∆∆ Sivakumar, K., Nakata, C., 2001. The stampede toward Hofstede's framework: avoiding the sample design pit in cross-cultural research. Journal of International Business Studies, 32(3): 555.

∆∆ Smith, P.B., 2006. When elephants fight, the grass gets trampled: the GLOBE and Hofstede projects. Journal of International Business Studies, 37(6): 915-921.

∆∆(∆) Smith, P.B., Schwartz, S.H., 1997. Values. In J.W. Berry, Segall, M.H., and Kagitcibasi, C. (Ed.), Handbook of cross-cultural psychology. Needham Heights, MA: Allyn and Bacon. 79-116.

∆∆(∆) Smith, P.B., Dugan, S., Trompenaars, F., 1996. National culture and the values of organizational employees. Journal of Cross-Cultural Psychology, 27(2): 231-264.

Chapter II: A survey of ‘culture and finance’

158

∆∆(∆) Smith, P.B., Peterson, M.F., Schwartz, S.H., 2002. Cultural values, sources of guidance, and their relevance to managerial behavior. Journal of Cross-Cultural Psychology, 33(2): 188-208.

Statman, M., 2007. Local ethics in a global world. Financial Analysts Journal, 63(3): 32-41.

Statman, M., 2008. Countries and culture in behavioral finance. CFA Institute, 25(3): 38-44.

*** Stonehill, A., et al., 1975. Financial goals and debt ratio determinants: a survey of practice in five counries. Financial Management, 4(3): 27-41.

Stulz, R.M., 2007. The limits of financial globalization. Journal of Applied Corporate Finance, 19(1): 8-15.

*** Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): 313-349.

° Summers, L.H., 1985. On economics and finance. Journal of Finance, 40(3): 633-635.

Throsby, C.D., 2001. Economics and culture. Cambridge University Press. 208 pages.

Todd, E., 1990. L'invention de l'Europe. Paris: Seuil. 537 pages.

∆∆ Tung, R.L., 2008. The cross-cultural research imperative: the need to balance cross-national and intra-national diversity. Journal of International Business Studies, 39(1): 41-46.

∆∆∆ Vinken, H., Soeters, J., Ester, P., 2004. Comparing cultures: dimensions of culture in a comparative perspective. Leiden: Brill Academic Publishers.

*∆∆ Weber, Y., Shenkar, O., Raveh, A., 1996. National and corporate cultural fit in mergers/acquisitions : an exploratory study. Management Science, 42(8): 1215-1227.

Westbrook, D.A., 2009. Out of crisis: rethinking our financial market. Paradigm. 176 pages.

Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature, 38(3): 595-613.

Wiltbank, R., Dew, N., Read, S., Sarasvathy, S.D., 2006. What to do next the case for non predictive strategy. Strategic Management Journal, 27(10): 981-998.

Zingales, L., 2000. In Search of New Foundations. Journal of Finance, 55(4): 1623-1653.

* * *

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

159

THIRD CHAPTER:

DO AGENCY RELATIONS MEDIATE THE

INTERACTION BETWEEN FIRMS’ FINANCIAL

POLICIES AND BUSINESS CYCLES?

DETAILED TABLE OF CONTENTS 1.   Introduction 161  2.   Financial policies, ownership and business cycles in institutional perspective 165  

2.1.   Sampling European listed firms 165  2.2.   Lean financial policies (dispersed ownership firms) 167  2.3.   Financial smoothing (firms with concentrated ownership or under control) 170  2.3.1.   Entrenchment  and  managerial  discretion   170  2.3.2.   Business  cycles,  agency  relations  and  block-­holders   171  2.3.3.   Smoothing  policies   172  2.3.4.   Hypotheses   173  

3.   Sampling, data and methods 175  3.1.   Pooling depending on the concentration of ownership 175  3.2.   Firm-, industry- and country-specific data 176  3.2.1.   Firms’  financials   176  3.2.2.   Business  cycles  and  other  context  variables   177  

3.3.   Variables 183  3.4.   Methodological proceedings 186  3.5.   Empirical models 187  

4.   Results 188  4.1.   Short notes about overall results 190  4.2.   Interactions between financial policies and business cycles (1): debt 191  4.2.1.   Broad  results   191  4.2.2.   Splitting  total  financial  debt  into  long-­term  public  debt  issuance  and  a  residual  category   191  

4.3.   Interactions … (2): dividends, cash holdings and investments 194  4.3.1.   Dividend  policies   194  4.3.2.   Investment  policies   195  4.3.3.   Cash  holdings   195  

4.4.   Interactions …(3): varying proxies for anticipated business cycles 196  4.4.1.   Interest  rates   196  4.4.2.   Debt  and  investment  policies  vs.  dividend  and  cash  policy   197  4.4.3.   Leverage  and  anticipated  business  cycles   199  

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

160

5.   Conclusion. 200  6.   Appendix 1 (chapter III): average ownership concentration in Europe 1990-2007.

204  7.   Appendix 2 (chapter III): pair wise correlations 206  8.   Appendix 3 (chapter III): the E.S.I. measures 207  

8.1.   Presenting the measures 207  8.2.   Analyzing E.S.I. measures 208  

9.   Appendix 4 (chapter III): main results with all coefficients reported 210  10.   Bibliography 212   * * *

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

161

Do agency relations mediate the interaction, between

firms’ financial policies, and business cycles?1

There is no universal theory of capital structure, and no reason to expect one. There are useful conditional theories, however […] each factor could be dominant for some firms or in some circumstances, yet unimportant elsewhere. Myers (2003: 216-217)

1. Introduction  

Recently Hackbarth, et al. (2006) note that “little attention has been paid to the effects of

macro-economic conditions on […] capital structure choices”. They propose a model where

leverage is countercyclical, in line with some evidence obtained earlier by Korajczyk and

Levy (2003), on a sub-sample of financially unconstrained U.S. firms.

This limited theorizing, and empirical investigation, relating business cycles to analyses of

leverage is surprising, because business cycles influence many underlying determinants of

capital structures. So for instance they are known to influence shareholders’ default policies

(Hackbarth, et al., 2006; Chen, 2010), marginal tax benefits (Graham, 1996), investment sets

(Gomes and Schmid, 2010), the cost of debt (Hackbarth, et al., 2006), credit spreads and bond

prices (Chen, 2010), the value of collaterals (Jimenez, et al. 2006; Benmelech and Bergman,

2010), market capitalizations (for instance they should modify optimal portfolio allocation:

Avramov and Chordia, 2006). Hence the theoretical expectation is that business cycles should

influence capital structure policies, because they influence the underlying variables, for

1 This essay is under revision for the Journal of Banking and Finance, but has not yet been re-drafted. It has been elaborated following the presentation, at varying academic conferences, of Reuter (2009), an earlier version of the article presented in the next chapter. It has benefited from key questions, about that article, by Prs. G. Charreaux, C. Moussu, and P. Raimbourg. The present article, however derived from some of these insights, is the sole production of the author.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

162

instance when the perspectives taken by the trade-off, pecking order, signaling or market

timing theories are retained.

We note, in passing, that survey evidence provides some ground to the mere idea that,

expected booms and busts are considered, by financial executives, in the planning and

execution of their financial policies (Graham and Harvey, 2001; Bancel and Mitoo, 2004;

Brounen, et al., 2006).

In this paper we investigate the suggestion that the lack of aggregate evidence, on this issue,

may be partly related to the mediating effect, that ownership dispersion has, on the relation

between business cycles and capital structures. In fact, we show that that empirical evidence,

as derived from a sample based on European listed firms, is consistent with the idea that

ownership dispersion polarizes financial policies, followed by firms, in response to

anticipated business cycles.

Interestingly, the literature on ownership dispersion and capital structure is about as limited,

as the one on business cycles and capital structure. Margaritis and Psillaki (2009: p. 622) note

that “the relationship between ownership structure and capital structure remains largely

unexplored”. They mention three international one-country studies. On the same subject, King

and Santor (2008, p. 2425) cite a number of theoretical approaches, mostly from the 80s, and

some empirical literature, primarily on U.S. data, all of which is focused on the marginal

effects of entrenchment on capital structure, in a context where ownership is presumed to

range from wide dispersion to limited concentration.

Theoretically we expect that ownership dispersion should affect the underlying mix of agency

relations in the firm, so that it should condition the nature of the financial policies followed by

the firm, particularly in response to anticipated business cycles. First, monitoring by external

block-holders strongly gains in efficiency (Shleifer and Vishny, 1986, 1997; Holderness,

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

163

2010), to a point where some form of ex-post control may be exerted, which modifies the role

of debt as a disciplinary device (Berger, et al., 1997). Second, the existence of block holders

has an effect on information asymmetry, adverse selection, signaling or still market timing.

Block-holders have varying ways to collect information, sometimes semi-private or private

information; and they may be associated to the board. Third, some block-holders may target a

long-term shareholding, or may de facto face a liquidity constraint (particularly outside the

U.S., Ruiz-Mallorqui and Santana-Martin, 2010). This should affect their opportunity cost of

bankruptcy. This will exacerbate the opposition between new and old shareholders, since

block-holders will naturally belong to the latter category. Fourth, because of this “tying” of

their investment, and, or, because of their specific time-horizon, block-holders may require

less risky debt-policies because they cannot easily diversify away, and they may have longer

investment horizons. Sixth, the shift to the background, of minority shareholders’ concerns,

will affect optimal debt-equity choices under constraints (irrespective of potential governance

externalities), as related for instance to dilution or expected dividends.

In this paper we focus on an institutional context, European listed firms, where ownership

concentration is medium to high, with a some proportion of direct control (Faccio and Lang,

2002). In this specific institutional setting, and in relation to business cycles, we investigate

how ownership structures could condition the execution of financial policies. We look

specifically at a number of key financial policies in parallel, which account together for the

build up of capital structures: we look at debt changes, including in particular public debt

issuance, we look at investments (capital expenditures), changes in cash holdings, dividends,

and the resulting changes in the book and market leverages.

The financial theory does not delve much into the interaction between financial policies and

business cycles, and indeed we expect, in line with traditional financial theorizing, that this

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

164

interaction is of limited relevance, in the case of dispersed ownership firms (at best a

moderating influence). However, we expect it to matter in the case of concentrated ownership

firms, or firms under control. Our reasoning is that a different mix of agency costs, in these

firms, is conductive to different policy-mix, with more weight for considerations pertaining to

the opportunity cost of bankruptcy, dilution (Levy and Hennessy, 2007), less weight for

signaling, potentially different pecking orders...

Our core result is that there is, indeed, a systematic empirical polarization, in the way

financial policies relate to information about expected business cycles, depending on whether

the firm has primarily dispersed, or primarily concentrated, ownership. This result, derived

from European data, echoes stylized facts about financial preferences in U.S. firms, most

notably depending on size or status, or still on private vs. public firms (Frank and Goyal,

2008). Overall, our contribution is to attract attention onto the potential, for a contrasted

relevance of varying capital structure approaches, depending on the institutional context.

It is noteworthy that we depart from the “abundant literature” (King and Santor, 2008:

p. 2424) on the performance effects of ownership structures. So do we depart from the

modeling of the interaction between capital structure, and business cycles ab abstracto.

Rather we look empirically at specific financial policies that underlie capital structure choices,

and we intend to investigate how strongly ownership structures can influence financial

policies.

In section 2 we provide some more background, and derive our hypotheses. In section 3 we

describe our data and methodology. Section 4 is allocated to results, while we conclude in

section 5.

* * *

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

165

2. Financial   policies,   ownership   and   business   cycles   in   institutional  

perspective  

As just noted, we delve into the potential interaction between corporate finance variables,

which are not traditionally associated: capital structure choices and business cycles, on the

one hand, capital structure choices and ownership dispersion, on the other. However we focus

on one of the most researched topics in corporate finance: capital structure, a topic on which

additional research can still provide some new insights, as emphasized in by Myers’ quote

provided in exergue.

Our specific contribution is to suggest a potential relevance of the institutional context for

understanding this traditional subject. To this effect, we note that past empirical investigation of

capital structures, across internationally diverse institutional backgrounds, is relatively recent.

This is where we start (section 2.1), outlining some gaps in this literature, before we focus on

two key propositions, deriving specific hypotheses for firms with primarily dispersed ownership

(section 2.2) and for firms with concentrated ownership, or under control (section 2.3)

2.1. Sampling  European  listed  firms2  

Past empirical research on capital structure has primarily focused on the U.S., investigating

international samples only recently (De Jong, et al., 2008; Frank and Goyal, 2008; Mahajan

and Tartaroglu, 2008). Further, we note that the bulge of these recent international studies,

either focus on small subsamples of the most developed countries, or they focus on broad

2 We are indebted to an anonymous reviewer (at an important U.S. academic finance conference) for attracting our attention on this issue: citing his conclusion to a very short comment, “how are European findings relevant to other countries? If at all?”. So we provide a motivation for our specific, and relatively novel sampling. We note that the interest for contrasting results, as derived from U.S. data, is peaking up (King and Santor, 2008). Further De Jong, et al. (2008: p. 1956) note that “the conventional theories on capital structure [were] developed using listed firms in the United States as a role model…”, and they call for evidence derived from more diverse sampling, echoing similar remarks in Zingales (2000: p. 1624), or Frank and Goyal (2009).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

166

samples of international firms, with both developing and developed countries (Antoniou,

et al., 2008; Bessler, et al., 2008; Booth, et al., 2001; Demirguc-Kunt and Maksimovic, 1999;

Fan, et al., 2008; Gonzales and Gonzales, 2008; Mahajan and Tartaroglu, 2008; Rajan and

Zingales, 1995; Wanzenried, 2006)3. Consequently cross-country evidence on non-U.S., but

listed firms, is either limited to samples featuring a very small but institutionally very diverse

set of countries, or hampered by issues attached to economic development, or still, in a

handful of cases, based on firms with specific international status (cross-listed indexed firms).

The European context is interesting for our purpose. We follow a standard scientific

procedure in institutional, cultural or qualitative studies, which consists in limiting the

variance on important criteria, to enable the examination of specific dimensions. Cultural

scholars, in particular, have long emphasized the potential fruitfulness of using regional

groupings to derive meaningful conclusions (Ronen and Shenkar, 1985 for an early review;

Schwartz, 1999; Lenartowicz and Roth, 1999; Hofstede, 2003 for more details)

Specifically, our European focus provides us with some converging socio-economic and

political properties, with a relatively favorable environment for business, with adequate rule

of law and contracting environment, with advanced economic development, relatively strong

financing systems, whether through markets or banks, with commonalities in political

systems, featuring strong checks and balances, a somewhat contained corruption, etc. Further

there is now much commonality in regulations, as European regulations are now well

3 Authors mentioned here focus on small subsamples of the most developed countries, for instance on the G-5 (Antoniou, et al., 2008) or on the G-7 (Mahajan and Tartaroglu, 2008; Rajan and Zingales, 1995), or they focus on broad cross-samples of international firms, involving both developing and developed countries. In these large international samples, development effects are complicating analysis (Demirguc-Kunt and Maksimovic, 1999; Fan, et al., 2008, Gonzales and Gonzales, 2008). Booth, et al. (2001), focusing on a sample of 10 developing countries, note that cross-country comparisons are impeded by specific economic-policies and institutional considerations. We note a handful of regional studies. Deesomsak, et al. (2004) focus on the Asian-Pacific region and two studies, with a specific European focus, investigate mostly small, unlisted firms (Giannetti, 2003; Hall, et al., 2004). Wanzenried (2006) investigates listed firms in four European countries comparatively, while Bessler, et al. (2008) use the Dow Jones STOXX 600, so they include firms in 16 countries, but only 425 listed firms overall. Our third essay, in chapter IV of this dissertation, provides some evidence that this may be introducing a bias.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

167

developed, and increasingly common, in particular with respects to financial regulations,

while there is more recent convergence in accounting frameworks (IFRS reporting).

Additionally, this European sampling provides us -in historical perspective- with the

successful and prolonged coexistence of both types of firms: closely held and widely

dispersed. These structures are in line with international realities (La Porta, et al., 1999; Denis

and McConnell, 2003; Holderness, 2010), with a mean level of concentrated ownership

oscillating between one, and two third, by country, and over the period (appendix 1). This

coexistence, of contrasted ownership structures, overlaps with contrasted macro-financial

structures: European countries feature both dominant bank-financing and market-financing

systems (Aggarwal and Goodell, 2009, 2010).

We note, in passing, that the overall levels of market capitalization in Europe, and the Gross

European Product are comparable to that of the U.S., So empirical results derived from pan-

European data may produce contrasting evidence, of particular interest, to researchers and

policy makers alike.

2.2. Lean  financial  policies  (dispersed  ownership  firms)  

Our general proposition, detailed below from existing literature, is that policies, in controlled

or concentrated ownership firms, will favor some financial smoothing4, to mitigate risk, while

deriving more benefits from managerial discretion, and bearing less costs from entrenchment.

In contrast, dispersed ownership firms will have to conduct leaner financial policies to ensure,

as a priority, the mitigation of agency costs between managers and shareholders. At the same

4 We propose this label for convenience. It does not seem to have any established pattern of meaning, and different authors use it with different objectives. In macro-economics, Titelman and Uthoff (United Nations, ECLAC) use it as a synonym for counter-cyclical financing, with a meaning related to the meaning we wish to suggest here.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

168

time, managers in these firms need to cope with these constraints, imposed ex ante, as an

important governance mechanism. This implies, by the same token, that the potential benefits

from managerial discretion are more difficult to reap off. Overall we suggest that the

maximization of capital structure and the execution of financial policies occur under a

different set of constraints, with potentially with differing time horizons.

In dispersed ownership, because of shareholder’s dispersion, managers can derive benefits

from a collective action problem, and extract perks, build an empire or build up their situation

through entrenchment (Shleifer and Vishny, 1989). So Berger, et al. (1997) define

entrenchment as the manager’s ability to evade discipline, specifically through lower debt

levels. Discipline is relying on ex-ante frameworks, rules, processes and contracts, to align

interests between shareholders and the manager. It is relying on ex post monitoring, to ensure

that these processes are effectively in use. Eventually robust legal procedures are primarily

needed as an ex-ante dissuasion mechanism, completed, when needed, with ex-post legal

proceedings; but a trial’s outcome is by nature uncertain and its primary purpose is to render

credible the ex-ante threat (Westbrook, 2009).

So under dispersed ownership ex-post managerial discretion is mostly regarded as

entrenchment because it limits the effectiveness of ex-ante-devices. It results that dispersed

ownership firms should exhibit lean financial policies. Further in the presence of

entrenchment, across a significant number of firms, the breakdown of discipline will entail a

stronger pro-cyclicality than warranted.

Hence we can derive a number of more specific hypotheses, starting with firms with dispersed

ownership; these sets of hypotheses are characterized by the subscript (a).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

169

First, we expect that debt will serve primarily as a disciplining, or a signaling, device. It is

correlated with expected profitability, and hence expected business booms and busts.

H1a. Debt. Net yearly debt changes are positively related to expected business cycles.

The discretion granted to managers needs to be curtailed, to ensure that potential agency

costs, between the manager and shareholder5 are kept under control. So demand-shocks need

to be accommodated appropriately (investment policies, etc.), and cash needs to be disgorged.

H2a. Dividend. Yearly dividends are positively related to expected business cycles.

H3a. Investments are positively related to expected business cycles.

H3bis(a) Investments are more specifically related to demand-driven business cycles.

In both cases, the expected positive relation is moderated, at the firm level, by the quality of

the firm’s governance: the relation observed in aggregate consists of a fully positive effect on

the one hand, and a positive effect, of lower intensity, reflecting the lower governance quality,

in some proportion of the sampled firms.

Relating to cash holdings, a number of conflicting approaches can be retained. If cash

holdings are considered as a precautionary measure, it should proportionally fall with

expected booms. If it is considered as a tool, used by managers, to build up entrenchment, it

should exhibit some form of pro-cyclicality: upon expected upturns, managers build up their

discretion, as embodied by the cash they can dispose of. So the specific sign of the correlation

is indeterminate, and it is left for the empirical analysis.

H4a. No prediction.

5 considered here as an homogenous group, with converging interests. When there are block holders, a key dimension, in agency relations, is the relation between large and minority shareholders (La Porta, et al., 1999)

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

170

Last, we note that upon an expected upturn, the previous hypotheses suggest that managers

will build up their asset-base (investments) while increasing their debt level. As a result book-

leverage (debt over assets at book value) will have both an increasing denominator and

increasing numerator, so the net change in book leverage could be both ways, and it is likely

to be insignificant if the book leverage is around one, and the new debt serves to finance

investments.

H5a. Book leverage is not significantly related to business cycles.

2.3. Financial  smoothing  (firms  with  concentrated  ownership  or  under  control)  

2.3.1. Entrenchment and managerial discretion

Entrenchment need not be associated with ownership concentration, and conversely,

managerial discretion, carries benefits. Hence, Claessens, et al. (2002), as well as King and

Santor (2008), show that, in a context of moderate to higher concentration of ownership,

ownership concentration can lead to better valuations, provided an alignment in interests

between managers and shareholders, which means in their studies, “no pyramid”, or a

congruence between cash flow- and voting rights. More generally, Claessens, et al. (2002)

note that entrenchment and alignment effects must be “disentangled”. They show that

ownership concentration exhibit a U-shape relationship with entrenchment.

Second, some authors note that entrenchment can be favorable to all shareholders. For

instance the purpose of defensive measures, such as poison pills or parachutes, is precisely to

entrench managers for the benefits of all standing shareholders (Sarig and Talmor, 1997).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

171

Third, there is a contradiction between the need to discipline managers, and the need to grant

them some discretion to foster strategic decision-making, exploit strategic opportunities in the

face of competition, and informational uncertainties (Garvey and Swan, 1994). This is the

way managers exploit new opportunities and create value in the literature in Strategy

(Hambrick and Mason, 1984; Charreaux, 2004), and this know-how, form which value is

derived is labeled “managerial discretion”.

As a result we expect that financial policies, in firms with control or concentrated ownership,

need not be lean, because the disciplining motive is more evenly balanced against the

consideration of potential benefits to managerial discretion.

2.3.2. Business cycles, agency relations and block-holders

Financial policies, in concentrated ownership firms, should follow contra-cyclical patterns for

a number of reasons. First the opportunity cost of bankruptcy may be higher, not only for the

management, but for varying shareholders who hold, large share-blocks. This is because risk

mitigation becomes a primary objective, competing with short-term share-value

maximization, when their share block is not liquid, or their investment horizon is longer.

Second there may be a psychological bias that increases the opportunity cost of bankruptcy

for both managers and block-holders. Third, varying stakeholders (banks, business partners,

sovereign funds, government circles…) may gain not only from their shareholding, but from

credit- (e.g. debt, commercial credit, etc.) as well as business relation. Fourth, these

stakeholders may face severe credit constraints, like the firms themselves, in particular in the

downward, or low phase of the business cycle. Fifth, downturns may present strategic

opportunities (acquisitions…) that should be prepared upfront, given contained opportunity

costs of entrenchment. Sixth, provided the lock-up of capital, or the investment horizon,

block-holders are not at will to fully diversify or stock-pick, so that they may expect risk-

mitigation from the management to compensate (precise funding plans, dividend policies…).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

172

2.3.3. Smoothing policies

The relatively more limited relevance of manager-shareholder agency costs, compared to

costs derived from other agency sources, is an important aspect fostering contra-cyclical

policies. In the literature derived from the seminal contribution by Jensen and Meckling

(1976), debt is considered as a central disciplining device, to constraint managers away from

excessive risk-taking or entrenchment. In other theories related to agency, financing can occur

specifically in response to managers’ incentives (Myers, 2003). But with the presence of large

block-holders or in the case of family firms, there is a larger incentive alignment between

managers and block-holders and a key governance aspect is the relation of these actors with

minority shareholders (La Porta, et al., 1999). Irrespective of the financing prospects (on

markets), or a normative assessment of firm’s governance, this should substantially affect the

financial policies. In signaling or market timing theories, as derived from adverse selection

arguments, the informational advantage of managers is very large. It is probably much lower

in “insider systems”, such as those described in countries where the financing by banks

dominate (bank-based financial systems). In trade-offs theories of capital structure, the

opportunity cost of bankruptcy is structuring the choice of leverage. Yet holders of share-

blocks may have a stronger incentive alignment with management than with minority

shareholders (Ruiz-Mallorqui and Santana-Martin, 2010).

In both cases the precautionary motives gain more weight, compared to monitoring issues,

and the balance between he opportunity costs of entrenchment vs. discretion is shifted; this

reduces the need for lean policies, while the comparative advantage of contra-cyclical policies

is increased. Our prediction is that this changing combination of opportunity costs, related to

different agency relations, is fostering contra-cyclical financial policies.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

173

2.3.4. Hypotheses6

We need to emphasize that our contrast between dispersed- and concentrated-ownership firms

will proceed from two thresholds, fixed at more than two third, or less than one third of tightly

held shares in the total shareholding (see data section for further details). Firms falling in the

middle category will serve as a test case, and should exhibit either an intermediate behavior,

or replicate the behavior of either of the two other groups. The rest of this section provides

hypotheses for firms in the upper category, those with more than two-third of tightly held

shares. We provide the (b) subscript for all these specific hypotheses.

Provided the relative importance taken by risk considerations, firms smooth out their financial

constraints inter-temporarily, based on the business cycles effects on firms’ cash flows. So

they decrease their debt to increase their debt capacity upon good prospects and build up debt

upon bad outlooks (for instance before bad economic conditions bring about a credit squeeze).

H1b. Debt. Net yearly debt changes are negatively related to expected business cycles.

Similarly the dividend policy is expected to be contra-cyclical (in proportion, for the dividend

payout ratio), because firms may target a specific absolute value dividend, while they build up

their reserves as a precautionary measure against adverse economic conditions, or for

potential strategic opportunities.

H2b. Dividend. Yearly dividend payout ratios are negatively related to expect business cycles.

Relating to investments, we expect a positive correlation with expected business cycles, as in

dispersed ownership firms, but we expect that this positive relation should be smaller than in

dispersed ownership firms. This would be because opportunity costs of entrenchment are

lower and opportunity benefits of discretion are higher.

6 A summary of hypotheses is provided for convenience after the table containing the key results: table 4

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

174

Consequently H3b and H3bis(b) are the similar to H3a and H3bis(a):

H3b. Investments are positively related to expected business cycles.

H3bis(b). Investments are more specifically related to demand-driven business cycles.

And we add a further “comparative” hypothesis:

H3-comparative. The coefficient for investments is higher (more positive) in dispersed

ownership firms, compared to concentrated ownership firms.

Third, relating to cash holdings, the reasoning is similar to that with investments and

dispersed-ownership firms. So while the prediction for concentrated ownership firms is

ambiguous, as for dispersed ownership firms (H4a), the comparative assessment is that these

firms have a stronger precautionary motive, while the cost-benefit analysis of entrenchment

and discretion, is more favorable. As a result we expect on the one hand a more negative

precautionary effect or a less positive entrenchment effect.

H4-comparative. Cash holdings. Net changes in yearly cash holdings are lower in absolute

value for firms with concentrated ownership or under control.

Last, as we expect a contra-cyclical change in the numerator (debt) and a pro-cyclical change

in the denominator (assets), we should observe a negative change in book leverage, in

concentrated ownership firms.

H5b. Book leverage is negatively related to business cycles.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

175

3. Sampling,  data  and  methods  

3.1. Pooling  depending  on  the  concentration  of  ownership  

Our key theoretical arguments are based on the concentration of ownership. While extensive

research efforts have enabled recent progress, in evidencing quantitatively and qualitatively

the ownership structures across countries (La Porta, et al., 1999; Claessens, et al., 2000;

Faccio and Lang, 2002), there is evidence that the issue is intricate, and the data unreliable,

without complementary case-by-case analysis (Holderness, 2010).

Provided these data limitations, we rely on the Worldscope item labeled “closely held shares”,

from which the “public market capitalization” variable is derived. Hence this “public market

capitalization” represents the proportion of the total value of the shares, which are publicly

and regularly traded, over the total market capitalization. These two items are not

systematically reported, but they exhibit some consistency despite missing years.

We average the “public market capitalization” for each firm, over each firm’s sampling

period: the sampling period is either the length of their listing, or from 1990 to 2007, our

overall sampling period. We conduct a number of tests to insure that the firms’ average

“public market capitalization” is a meaningful variable. In particular we insure that less than a

third of yearly values are reported, and we conduct complementary tests on the variance, and

on the changes in the underlying yearly values7.

We split our sample into three comparable groups, size-wise. Those with more than two third

of closely held shares, on average, are described as the firms with concentrated ownership, or

under control, which is simplified into firms with concentrated ownership from now on.

Those with less than a third of closely held shares are labeled as the “dispersed ownership

firms”, and the rest serves as a middle ground and a test case as emphasized above.

7 We pool our sample on the mean-value, as well as the standard error of the “public market capitalization” , as robustness tests in our main proceedings below

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

176

Holderness reports discrepancies between mean ownership concentration, as derived from

Worldscope, and as derived from his own computations based on data provided by Claessens,

et al. (2000) or Faccio and Lang (2002). However these discrepancies, even though they are

significant for cross-country comparisons, do not seem to be sufficient to lead to a large

number of re-categorization among our three classes of firms. Further any misclassification on

our account would primarily consist in classifying firms with either “concentrated” or

“dispersed” ownership in the middle category.

Furthermore, we note that Faccio and Lang (2002) point out that they could not trace back

28% of the ownership data, because they eventually obtained non-listed companies, with no

reported shareholders8. The data about ownership is by nature difficult to handle; yet we

suggest from the previous considerations that our pooling into three categories, however not

perfect, with a polarization across a middle category, should average out some of the

concerns, while providing ground for some empirical progress.

Our main purpose is to reject the general hypothesis that the two polar categories follow the

same financial policies, hence providing some ground for our specific hypothesizing on the

polarization of financial policies, following substantial changes in agency relations.

3.2. Firm-­,  industry-­  and  country-­specific  data  

3.2.1. Firms’ financials

We restrict ourselves to a brief synthesis of the data gathering process, for the sake of length

and clarity9. We collect firm-specific data primarily from Worldscope (through Thomson-

One-Banker). We collect data, since 1990, for all companies that have been, or are, listed on a

8 they assigned these firms to the “family firm” group, following Claessens, et al. (2000). 9 A more comprehensive description of the data collection effort is provided in Reuter (2009)

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

177

stock exchange on the European territory (64 cities with stock exchanges). First, from an

exhaustive list of 10 633 identified company tickers, dead or alive, we eliminate firms with no

financial/accounting data at all (34%), firms with sic codes 6*** or 9*** (16% of counts),

firms with less than 5million Euros in sales (11% of counts)10. Second, from a total of 4182

companies with exploitable data we clean our database, bringing the count down to 3434

firms (40 885 firm-years): we remove companies from stock exchanges with less than 25

companies listed on average per year; we remove companies with less than 3 years of data;

we remove firms (or years at the end of listing-spans) with missing data on any of the

endogenous variables (i.e. market-cap, assets, debts); we remove some of the years/companies

with missing data that can not be properly reconstituted. Third, we trim the data at 5 standard

deviations removing further 129 firms-years (3 firms entirely). We use this database,

removing further American-registered companies and Swiss-listed firms, for which we do not

have the sentiment data11.

3.2.2. Business cycles and other context variables

We use two primary measures of business cycles, while controlling for stock-market activity,

long-term interest rates, unemployment and inflation. We proceed with extreme care in the

way we build up our regression models, because of pair wise correlations across all these

“context variables” (appendix 2). Moreover, we implement an instrumental variable approach.

Our two measures for business cycles are derived from a survey that has been carried out by

the European Community, for more than 20 years (appendix 3). It is a robust index, broadly

used in economic policy and macroeconomics, and in academic research in finance12.

10 …on average per year, eliminating many firms with significantly missing data for exogenous variables, missing years in data-ranges, etc. See King and Santor, 2008 for a related procedure based on market capitalization 11 For all companies, but one, the place of primary listing is that of official company registration. 12 Geyer, et al. (2004) and Ferreira, et al. (2008) analyze the relations between these indices, output growth and interest rates in Europe

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

178

The first measure we use is a manager-specific measure labeled manager’s economic

sentiment indicator (managers’ E.S.I.); the second measure is a measure derived from a

survey of consumers (consumer’s E.S.I.). Both measures formalize business cycles’

expectations, for managers on the one hand, for consumers on the other, into a numerical

scale. While consumer confidence is now widely used in economic policy making, and in

research, in economics and in finance, the measure that we use for managers is less common.

It is probably due to the fact that it is more specific to Europe, and other governing bodies

have not as consistently computed it.

The distinction between these two measures of expected business cycles is important in two

ways; first, as shown below, both E.S.I. measures capture some specific and independent

pieces of information about business cycles. We contend that the managers’ measures proxy

for economic up- and downturns that are more supply-side related. Being more specific to

business executives, this E.S.I. measure is more asymmetric with relation to the information

used by stock market agents: manager’s E.S.I. captures a greater share of the information

asymmetry, which is driving some of the agency costs and hence potentially some financial

policies.

Second, we expect somewhat differential effects for E.S.I. measures depending on whether

they are derived from managers or from consumers, particularly in dispersed-ownership

firms: manager’s E.S.I. should be specifically related to entrenchment effects that could be

prevalent in some proportion of the dispersed ownership firms, and consequently observed in

aggregate (due to heteroskedasticity depending on firm’s governance quality).

Overall we include six context variables: in addition to the 2 confidence indicators just

mentioned, we include a measure of the long-term interest rates, obtained from Datastream.

We use the yield on 10-years government bonds for the respective national governments in

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

179

the 15 countries that were eventually retained in our sample. This measure is particularly

important because it has been used as a variable containing information about prospective

business cycles. However, because it is, at the same time, directly related to financing and

investing decisions, its overall effects are obscured by the combination of direct and indirect

effects. So we had no specific hypothesizing regarding this variable and the direct effect is

standard in financial theory. Increasing interest rates decrease additional debt, investments,

and increase the opportunity cost of cash.

Further, we obtain from Datastream other control variables, including indicators for inflation

(Consumer Price Indices-CPIs), broad stock market indices (generally the Morgan Stanley

Dean Witter’s broad stock market indices, for the varying national stock markets considered),

and an indicator of unemployment by country. All our context variables are available on a

monthly basis. They are used on a yearly basis in the regressions because firms’ financials are

available on this basis only.

All these context variables are implemented on a flow basis, i.e. as a net change in the index

from one period to the next, normally year-on-year13. In this way time frames are matched to

those of the financial variables obtained form Worldscope. All these six context variables,

collected on a monthly basis over 20 years, were centered-reduced, before taking net yearly

changes. This made particular sense provided that these context variables are indices with

13 Specifically, this means the difference from the end of December of one year to the end of December of the previous year in all regressions conducted with firm financials. As a robustness check, alternative measures of net differences in indices were derived and tested on initial “baseline models”. These included net differences on a 3- or 6-months basis before December, as well as a year-on-year difference with a six-months lag (a year-on-year difference based on end of June of the previous years compared to financial data. Results for the baseline models were much the same with the 6-months window or 6-months lags showing limited difference in significance and signs. Complementary analyses on co-variances in context variables and the instrumental variables approaches were based monthly data (sometimes quarterly data), “filtered back up” to match the yearly data provided by Worldscope. Quarterly financials are very rarely available with European data. Robustness checks included additional analysis with further data from two sources, both available on a quarterly basis. This included data on capacity utilization in industry available from the European Union (same source as for E.S.I.) and hourly labor costs available from Datastream. The analysis on a quarterly data yields similar results as those based on monthly data with a narrower scope.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

180

different base-years. Further many other variables are computed as a percentage (e.g. debt as a

percentage of capital, see below), on the basis of a year-on-year change.

It is very important to mention that a thorough analysis of the interactions between these

variables was conducted, regressing in particular managers’ E.S.I. on the other 5 context

variables. This analysis (appendix 3) was carried out on the indices themselves (on “stocks”)

as well as on “flows”, and all specific analyses were conducted on a monthly basis, with

univariate and multivariate regressions, with OLS for the “flow” variables (net monthly

changes) and with more refined techniques for the “stock” variables. Four conclusions were

derived from this analysis14.

Below, we present the main results of the pooled regression for our six endogenous variables

on firm-and industry-specific variables. We present them with a restricted range of context

variables for the sake of clarity. We present the pooled results with either of the two E.S.I.

measures, and then with a specific instrumental variable (I.V.) approach. In the I.V. approach

that is presented, we regress managers’ E.S.I. on consumers’ E.S.I. and on long-term debt,

and we retain the residual from this regression, consumer’s E.S.I. and long-term debt for the

I.V. approach. These three variables were those that were retained as most meaningful

through the extended analysis of context variable conducted here, and the other regressions

scopes that were tested do not alter significantly our results.

14 First, monthly regressions based on aggregate data (European level data is used) were very different from those based on country-level data. In particular coefficients of esicsumer in the varying univariate, and multivariate, regressions of esimnger were nearing 1 in aggregate data while they were generally around, or just above, 0.5 with country-level data (appendix 3). Second and consequently, when appropriate granularity is retained, esimnger provide specific and separate information from other business cycle proxies, and in particular esicsumer. Third, the overall levels of pair wise correlations, based on country-level data, are lower than one would expect and exhibit interesting cross-industry, cross-country properties. Fourth, the instrumentalization of esimnger differs depending on whether flow or stock variables are retained. Based on stocks, stepwise regressions indicate that esimnger can be instrumentalized on esicsumer and broad stock market levels. On a flow basis, the stepwise procedure indicates that net changes in the LT interest rates are more meaningful than those stock market indices

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

181

Table 4: sources, definitions and variables (financial policies in Europe)Financials Worldscope definitions and specificationsFloat The float is the public market capitalization.

Year-end public market capitalization = Year-end maeket price x (Common Shares Outstanding - Closely Held Shares)

EBITD EBITD is earnings before interest, taxes, amortization and depreciation. "EARNINGS BEFORE INTEREST, TAXES AND DEPRECIATION (EBITDA) represent the earnings of a company before interest expense, income taxes and depreciation. It is calculated by taking the pretax income and adding back interest expense on debt and depreciation, depletion and amortization and subtracting interest capitalized."

DnA Depreciation and AmortizationIncBT Income before TaxesLTD Long-Term DebtMC Year-end Market Capitalization (roughly the closing share price times the

number of ordinary shares as a personal check indicates)MC(pub) The year-end public market capitalization is "Market Price-Year End * (Common

Shares Outstanding - Closely Held Shares)"PayOut Dividend pay-out ratio, "Sum of the trailing twelve months dividends divided by

sum of the trailing twelve months earnings"PPE(net) Net Property Plant & EquipmentRE Retained EarningsSTD Short Term Debt; specifically, "ST DEBT AND CUR PORT LT DEBT represents

that portion of debt payable within one year including current portion of long term debt and sinking fund requirements of preferred stock or debentures"

TA Total AssetsTaxes Income TaxTD Total Debt. Author have checked that TD = STD + LTD did hold for the sampleTS Total SalesTSE Total Shareholder Equity is computed in this way: "Preferred Stock + Common

Stock Net + Capital Surplus + Retained Earnings - Treasury Stock +/- Other Liabilities"

Financials Variables retained and computationsExogenous variablesC_MLev Change in market leverage from Market leverage: ML = TD / (TD + MC) C_BLev Change in book leverage from Book leverage: BL = TD / TAC_TD_TA Change in financial debt over total assetsNLTDI_TA Net long term public debt issuance over total assetsC_ResidD_TA Change in other financial debt over total assetsC_Cash_TA Change in cash holdings over total assetscapex_TA Capital expenditures over total assetsDivpayout Dividend payout ratioFirm-specific variables C_Profits Change in profitability from Profitability : EBITD / TAC_size Change in logarithm of size from Size: ln (TA)C_tobinsq Change in tobin's Q from Tobin’s Q: QMV / TA: Quasi-market value (QMV) =

TA - + MCC_tangible Change in tangibility from Tangibles: PPE(net)C_non_D_TS Change in non-debt tax shield from Non-debt tax shield: DnA / TASize The logrithm of sizeCumul_NEI_TA Cumulated net equity issue over total assets from "neteqissubw": A ratio based on

Baker & Wurgler, computed as the sum, from the first year of listing (or by default 1990) to the current year, of [Change (TSE)-change(RE)]/TA

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

182

Pooling variablepubcapav Firm's average public market capitzlization over listing period: Average of

MC(pub)/MC over the sample period*class_pubMC Splitting pubcapav (just above) into three categories; sortpmcap=0 if

pubcapav<33%, =2 if pubcapav>66% and 1 otherwhise*Industry-specific variablesC_med_BLev Change in median industry book leverage from the median book leverage per

industry based on the E.U./ Nace categorization, "Industry" below.C_med_MLev Change in median industry market leverage from the median market leverage per

industry based on the E.U./ Nace categorization, "Industry" below.BLev_to_ind Firm's book leverage minus median industry book leverageMLev_to_ind Firm's market leverage minus median industry market leverageIndustry Nace classification used by the European Comission, in particular in the context of

the computation of the economic cycle and confidence indices. Aggregated into 5 categories: Manufacturing industries, Services, Construction & Retail trade industries and an 'other' category.

Country variables included in the regressions as presentedYear Dummies for the years 1990-2007Country 15 countries: aut, bel, deu, dnk, esp, fin, fra, gbr, grc, irl, ita, nld, pol, prt, swe*esimnger** 12-moonths change in E.S.I. for managers from the European E.S.I. measures.

Details in appendix 2mgerregres** Residuals of the I.V. approach to the change in E.S.I. manager. Details p. 16 and 22

of the manuscriptesicsumer** 12-moonths change in E.S.I. for consumers from the European E.S.I. measures.

Details in appendix 2dtstrm_LTR2** Change in long-term interest rates from Long-term interest rate (10-years

government bond). Source: Datastreamdtstrm_St_mkt** Change in stock market indices by country from Broad stock market indices.

Source: DatastreamO_CPI Change in consumer price index from Consumer price indices by country. Source:

OECDO_Emplot Change in unemployment from Unemployment index. Source: OECDFurther country variables used in robustness testsInflation Inflation rates by counrty, source: European Union websitedtstrm_CPI Consumer price indices by country. Source: Datastreamdtstrm_STR Short term interest rates (government bonds). Source: Datastreamdtstrm_LTR1 Medium-term interest rates. Source: DatastreamO_LTR_gvt Long-term interest rate (10-years government bond). Source: OECDO_St_mkt Broad stock market indices. Source: OECDO_STR Short term interest rates (government bonds). Source: OECDO_LTR Long-term interest rate (interbank market). Source: OECDQuotes in italic are definitions or specifications from Worldscope. * at the end of a definitioncharacterizes a time-invariant variable. All variables preceded by "C_" are computed as the year-on-yeardifference. All variables preceded by "D_" are dummy variables. ** Some context variables(E.S.I.measures mainly) are implemented with different lags and potentialy based on year-changes orsemester-changes. Further details in footnote 6 in the body of the text. "O_" stands for Oecd source while"dtstrm_" stands for data from Datastream. The first sub-table provides specifics about firm-financials extracted from Worldscope. The second subpartof the table presents the computations we executed on the firms'financials. The third sub-part of the tablepresents details and sources about industry and country-specific variables

Table 4 (continued)

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

183

3.3. Variables  

For the most part, our definitions are standard (table 4), directly related to those of

Worldscope, and they follow recent contributions on international Capital Structures (e.g.

Antoniou, et al., 2008; Fan, et al., 2008; Frank and Goyal, 2009, who provide mostly

converging definitions). The table 4 provides the definitions, for the main variables, including

alternatives variables for the context, which were tested, but not reported because of lack of

either significance. The table provides ratio-definitions as well. Generally the main variables

are presented, but it is primarily their year-on-year differences that are used subsequently. The

table 5 provides the main statistics.

We are interested in understanding the mechanisms driving financial policies, so this is the

reason why we analyze our variables in terms of flows, on a yearly-change basis, rather than

in terms of stocks. We contend that this provides a more accurate picture on the relation

between manager’s information about business cycles, and manager’s behaviors, resulting

financial policies, and the way agency relations mediate these interactions.

Among the endogenous variables we include the yearly change in both measures of leverage

(book, market), the changes in outstanding debt, the changes in cash holdings as well as

yearly investments. Noteworthy are the following points: in both measures of leverage we

include all financial debt, i.e. long-term plus short-term debt; we split ex-post this variable

into two mutually exclusive measures of debt: public long-term debt issuance on the one

hand, and changes in other financial debt outstanding, net of the public long-term debt

issuance. Investment is measured through capex, which is the best proxy, provided our data-

source.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

184

Table 5: statistics (analysis of financial policies in Europe)Variable Mean Std. Dev. Min MaxDescriptive variablesyear 2001 n.s. 1991 2008countryindustryclass_pubMC 1.01 0.76 0.00 2.00Exogenous variablesChange in market leverage 0.008 0.122 -0.965 0.933Change in book leverage 0.005 0.105 -3.464 3.049Change in financial debt over total assets 0.009 0.170 -10.414 2.264Net long term public debt issuance over total assets 0.007 0.103 -6.622 2.764Change in other financial debt over total assets 0.004 0.152 -10.414 1.640Change in cash holdings over total assets 0.000 0.130 -10.421 0.970Capital expenditures over total assets 0.063 0.114 -0.172 4.227Dividend payout ratio 0.279 0.314 0.000 1.000Firm-specific endogenous variablesChange in profitability -0.002 0.190 -4.420 6.100Change in tangibility -0.002 0.067 -1.420 2.029Change in logarithm of size 0.088 0.320 -3.726 4.719Change in tobin's Q -0.084 1.273 -37.763 30.132Change in non-debt tax shield 0.000 0.045 -1.759 1.539Change in median industry book leverage 0.002 0.018 -0.139 0.166Change in median industry market leverage 0.002 0.042 -0.218 0.313Firm's book leverage minus median industry book leverage 0.023 0.180 -0.281 4.988Firm's market leverage minus median industry market leverage 0.050 0.221 -0.446 0.895Logrithm of size 5.303 1.998 -2.814 12.529Cumulated net equity issue over total assets 0.113 0.509 -2.791 2.990Context variables12-moonths change in E.S.I. manager -0.064 1.250 -4.691 3.02912-moonths change in E.S.I. consumer -0.065 1.060 -4.742 2.643Change in stock market indices by country 0.116 0.635 -2.487 1.968Change in long-term interest rates -0.134 0.435 -1.952 1.217Change in unemployment -0.034 0.565 -1.495 2.482Change in consumer price index 0.035 0.736 -1.912 1.798Residual of the I.V. approach to the change in E.S.I. manager -0.032 0.804 -2.558 1.858

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

185

Relating to our exogenous firm-specific variables, they include two sets of variables.

Following Frank and Goyal (2009), we include the four firm-specific “core variables”, which

are found to be robust to range of specifications in their very long time-horizon, U.S.-focused

analysis of capital structures. The four measures are profitability, size, tangibility, and the

market-to-book ratio. A second set of variables includes additional variables, which make

specific sense provided our focus; they include changes in the dividend payout ratio and in the

non-debt tax-shield. In these six baseline variables we mainly deal with variables measured as

proportions, but for size, measured in logarithm. Here all variables are measured in terms of

“flows”, as with the endogenous variables.

To these variables we add the variable introduced by Baker and Wurgler (“EFWAMB”, 2002)

and we add some of the variables just mentioned, but measured as “stocks” in addition to the

flow measure included previously. This concerns in particular the logarithm of size and the

dividend payout ratio (so we check whether the fact of being a large firm, or a high-dividend

paying firm, has an influence on the yearly financial policies).

We include industry-specific variables, based on varying definitions, and in varying forms.

We primarily use the median industry leverage in stock and in flow, based on two-digits SIC

codes, or on the aggregation used in the Nace categorization of the European Union, as

implemented in particular in the branch-specific measure of manager’s E.S.I. We include as

well the difference between firm’s leverage and firm’s median industry leverage (with the

appropriate amends for book- and market leverages).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

186

3.4. Methodological  proceedings  

Our methodological proceedings are structured by a number of specific dimensions. First, we

depart from the bulge of the literature, which has analyzed the overall debt ratios in

equilibrium, or dynamically, rather than based on net changes in financial policies. We note,

in passing, the general lack of consensus on the potential convergence in capital structure

mechanisms, in international samples (see section 2.1). So our approach provides us with less

precedence than usual and limited consensus. Third, our approach is based on year-changes in

ratios, or proportions, or centered-reduced indices, so this limits the potential for auto-

correlation, endogeneity and heteroskedasticity in residuals, an issue that had been prominent

in past studies of capital structure (Antoniou, et al., 2006; Barclay and Smith, 2003). Fourth

our specific focus is not on the firm-specific determinants of financial policies, but on their

interaction with manager’s information and business cycles. So we retain a range of firm-

specific determinants that serve as controls, while they are generally, in most studies, the

focus of the investigation. Fifth, we investigate simultaneously a range of six endogenous

variables, so it is important for the sake of clarity to streamline and systematize our approach.

So we decide to use a stable set of exogenous control variables across the regressions.

As a result, we provide extended robustness checks around a base-line model, based on what

were identified as the “core variables” by Frank and Goyal (2009). In particular every

endogenous variable is added to the model on an univariate basis, as a marginal addition to

the baseline model (equations 1-1 to 1-6 below). Then multiple additional variables are added

on a multivariate basis in a full model (equation 2-1 to 2-6 below). Overall we calibrate a

large number of models, unreported here. Further our focus is on the interaction between

expected business cycles, financial policies and the mediating role of ownership dispersion, so

firms- and industry-specific variables are treated as control-variables in the rest of the study15.

15 27 regressions for each of the 6 endogenous variables were eventually formalized and are available on demand. These robustness checks attest of the stability in the results of the successive coefficient obtained from the range of models. These results are broadly consistent with those reported in the literature. Some further comments in section 3.6 and 4.1

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

187

3.5. Empirical  models  

Our baseline models are: (EQUATIONS 1-1 to 1-6)

ENDOGENOUS i ((t+1) – t) = α+γ1FIS(1)i((t+1)–t)+D(1)t+D(2)k+D(3)j+εit

Our full-size models are:

ENDOGENOUS i ((t+1) – t) = (EQUATIONS 2-1 to 2-6)

α+γ1FIS(1)i((t+1)–t)+γ2FIS(2)i((t+1)–t)+γ3MIL?t+γ4FSit+γ5Fk((t+1)–t)+γ6ESIk((t+1)–t)+D(1)t+D(2)k+D(3)j+ε i t

The models retained for pooling are:

ENDOGENOUS i ((t+1) – t) = (EQUATIONS 3-1 to 3-6)

Pooling by ownership class**

α+γ1FIS(1)i((t+1)–t)+γ2Δ(MIL)j((t+1)-t)+δ1ln(Size)it+δ2B&Wi∑t=1 to t+δ3ESI*k((t+1)–t)+D(1)t+D(2)k+D(3)j+εit

Where i,j,k,t and ? stand respectively for firms, industries, countries, years or any-one of these. ENDOGENOUS stands for the year-difference in any of the 6 endogenous variables. Note that we extend this to 8 variables (and 8 regressions models) below, because we split net financial debt change into two sub-components (net pubic long-term debt issuance and other financial debt changes), and because we include results for market leverage. FIS(1) are firm- and industry specific variables based on year-on-year differences. They include five of the six-core variables mentioned by Frank and Goyal (2009): size, profitability, tangibility, Tobin’s Q, median industry leverage and we add the change in non-debt tax shield, provided our range of endogenous variables. FIS(2) are other firm- and industry specific variables following the literature, based on year-on-year differences. MIL are additional measures related to median industry leverage, including the firm-to-industry difference in leverage (? stands for i) and the absolute value of the median industry leverage (? stands for j). FS includes a range of firm-specific variables used in absolute value (corresponding to firm-specific variables used in FIS(1) and FIS(2)). Note that in the models retained (equations 3-.), we only retain the log of size in absolute value. B&W is the cumulative measure adapted from the EFWAMB variable initially proposed by Baker and Wurgler: (see section 3.3). F stands for varying financial or economic variables, including in particular long-term interest rate, stock market indices, unemployment, inflation, etc (see annex 1 for further variables’ details). E.S.I. are the European Union Economic Sentiment Indicators, either for managers, or consumers, or still for the residual of manager’s ESI regression on consumer’s ESI and long-term interest rates. D1 to 3 stands respectively for time-, country- and industry dummies. ε is the residual that we presume normally distributed and homoscedastic because we took year-on-year differences in all endogenous and firm-specific variables (but for a few flow variables such as investment of public L.T. debt issuance), and because we included dummies for potential sources of heteroskedasticity.* we repeat the regression sets with EITHER managers’ E.S.I., OR consumers’ E.S.I. OR the I.V. approach to managers’ E.S.I. (section 3.2). **Three ownership classes as defined in section 3.1. We primarily use OLS regressions, provided that we work on year-differences. Early

robustness checks on the baseline model contrast OLS approaches with panel-data regression

(under a Maximum Likelihood specification); results were very similar.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

188

4. Results  

Our focus is on the way ownership structures mediate the interaction between business cycles

and financial policies. For this reason we provide only very brief commentaries about the

successive regressions, as identified in equations series 1- & 2-. Similarly commentaries

pertaining to the regressions in equations series 3- are focused on the variables of interest.

So we present our results twice, in the tables 6, next page, and 10, in appendix 4. Table 6 is an

extract of table 10, where the coefficients for the business cycles and interest rates variables

are reported only. The reader will notice that these coefficients are identical in both tables.

Table 10 provides the extended results of the pooled regressions, got the six plus two

exogenous variables.

We start by providing brief commentaries about the overall results (4.1) before we focus on

the specific interactions of interest (section 4.2 to 4.4). There we examine first results

pertaining to the interaction between debt and business cycles (section 4.2) before we

examine other financial variables (investments, dividends, cash holdings, section 4.3). In

these two sections we consider results irrespective of which business cycle proxy is

considered and section 4.4 is allocated to the specific distinction between our respective

proxies. It deals with results pertaining to book-leverage as well.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

189

Table 6. Main results (analysis of financial policies in Europe)This is an extract of table 10 in appendixPooling the regressions (financial policies) on firm-, industry-specific var. & business cycle proxies

Low Float Medium Float High Float Low Float Medium Float High Float Low Float Medium Float High Floati.e. c-o. firms i.e. d-o. firmsi.e. c-o. firms i.e. d-o. firmsi.e. c-o. firms i.e. d-o. firms

1-Regressing change in total debt (over total assets)E.S.I. manager n.s. -0.0031** n.s.Residuals of I.V. approach n.s. n.s. n.s.E.S.I. consumers -0.0046**-0.0041*** n.s. -0.0045**-0.0046*** n.s.L.T. interest rates n.s. n.s. n.s.1a-Regressing public long-term debt issuance (over total assets)E.S.I. manager n.s. n.s. n.s. Residuals of I.V. approach n.s. n.s. n.s. E.S.I. consumers n.s. -0.002* -0.0022** n.s. -0.0023** -0.0023** L.T. interest rates n.s. n.s. n.s. 1b-Regressing change in residual financial debt (residual net of public long-term debt issuance)E.S.I. manager n.s. -0.0021** n.s. Residuals of I.V. approach n.s. n.s. n.s. E.S.I. consumers -0.0037** -0.0023** 0.0035** -0.0034* -0.0023** 0.0038** L.T. interest rates n.s. n.s. n.s. 2-Regressing dividend payout ratioE.S.I. manager -0.0085** n.s. n.s.Residuals of I.V. approach -0.0121*** n.s. 0.0117***E.S.I. consumers n.s. n.s. n.s. n.s. n.s. n.s.L.T. interest rates -0.0264***-0.021*** -0.022***3-Regressing capital expenditure / total assets E.S.I. manager n.s. n.s. n.s. Residuals of I.V. approach n.s. n.s. n.s. E.S.I. consumers n.s. n.s. 0.0013** n.s. 0.0026*** 0.0022***L.T. interest rates -0.0107*** -0.0169*** -0.0051***4-Regressing change in cash / total assetsE.S.I. manager 0.0017* n.s. 0.0031** Residuals of I.V. approach 0.0027** 0.0033** 0.0068***E.S.I. consumers n.s. n.s. n.s. n.s. n.s. n.s. L.T. interest rates -0.0053** -0.0073***-0.0151***5-Regressing change in book-leverageE.S.I. manager -0.002* -0.0018** n.s. Residuals of I.V. approach n.s. n.s. n.s. E.S.I. consumers -0.0043***-0.0028*** n.s. -0.0043***-0.0035*** n.s. L.T. interest rates n.s. 0.0054*** n.s. Addition-Regressing change in market-leverageE.S.I. manager -0.0097*** -0.0079*** -0.0081***Residuals of I.V. approach -0.0078*** -0.0038*** -0.0035***E.S.I. consumers -0.0096*** -0.01*** -0.011*** -0.0108*** -0.0112*** -0.0111***L.T. interest rates 0.011*** 0.0111*** 0.0059**

Summary of hypotheses: H1. Net yearly debt changes are positively (negatively) related to expected business cycles for firms with dispersed(concentrated) ownership. H2. Yearly dividend payout ratios are positively (negatively) related to expect business cycles forfirms with dispersed (concentrated) ownership. H3. Investments are positively related to expected business cycles for alltypes of firms, and H3-specific. Investments are more specifically related to demand-driven business cycles. H3-comparative. The coefficient for investments is higher (more positive) in dispersed as compared to concentrated ownershipfirms H4-comparative. Net changes in yearly cash holdings are lower in absolute value for firms with concentratedownership or under control. H5. Book leverage is not significantly (negatively) related to business cycles for firms withdispersed (concentrated) ownership. Subscripts a / b in the body of the text stand respectively for hypotheses pertaining tofirms with dispersed / concentrated ownership (e.g. H1a for dispersed ownership firms vs. H1b for concentrated ownershipfirms, etc.)

We regress the five endogenous variables on the 'baseline' plus one of the three business cycle proxies: this iscorrespondonging to EQUATIONS 2-1 to 2-3 but the "base-line coefficients are not reported here (they are instead inappendix 3). We use O.L.S. becasue we use mostly year-on-year differences and eight regression results, instead of 5, arepresented because of the split of total debt change into its two components (long term public bonds and other financialdebts) and because of the reporting of results for market leverage as a complement. The 'baseline' regression (implementedin these regressions, but not reported for ease of reading) includes on top of the business cycles proxies that are presentedhere the following list of variables: C_Profits, C_tangible, C_size, C_tobinsq, C_non_D_TS, C_med_MLev, MLev_to_ind,Size, Cumul_NEI_TA.

Adding ESI Managers only to regressionsAdding ESI COnsumers only to regressions Instrumental variable approach

POOLING:

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

190

4.1. Short  notes  about  overall  results  

Our firm- and industry-specific “baseline” results are mostly conventional. Let us note that

the Tobin’s Q provides some variation and appears to be systematically misleading in

univariate regressions (reversion in signs in multivariate regressions). Besides it offers some

discrepancies depending on model scope and on pooling criteria.

Some of the variables are meaningful in both absolute value and in differences, especially size

and the dividend payout ratio, and in those two cases opposite coefficients are obtained for the

variable in absolute value and in difference. This could indicate mean-reversion for dividend

payouts, and a non-continuous relationship with thresholds for size. Further research is

warranted.

Third, the Baker and Wurgler ‘s variable fares well, confirming their specific insight that the

building up of leverage ratios is path-dependent.

Fourth, the variable computed as the difference between firms’ leverage (book or market)

minus the median’s industry leverage is significant and positive, implying that a higher

leverage than the industry induces increases in debt, higher cash holdings, higher capital

expenditures and a higher resulting book-leverage. We suggest this could be due either to

considerations pertaining to firm’s quality or to heteroskedasticity in firms, which deserves

further research too.

Now we focus on the results of the pooled regressions, reported in Table 4 (annex 2), which

reports only the coefficients obtained for the business cycle proxies.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

191

4.2. Interactions  between  financial  policies  and  business  cycles  (1):  debt  

4.2.1. Broad results

We specifically hypothesized that net yearly debt-changes are positively (negatively) related

to expected business cycles in dispersed (concentrated) ownership firms: H1a (b). Our results,

for the first regression, on overall net financial debt changes, disconfirm H1a (dispersed

ownership firms), while they are consistent with H1b: concentrated ownership firms and firms

in the medium range, exhibit a negative relation between debt-changes and expected business

cycles, with a slightly higher coefficient for concentrated ownership firms, as compared to the

middle range.

This is our main result, and this suggests that some financial smoothing may indeed occur in

non-dispersed ownership firms, to smooth out debt capacity in relation to business cycles.

4.2.2. Splitting total financial debt into long-term public debt issuance and a residual

category

Provided this mixed result, we expand our analysis to two kinds of debt. As Worldscope

provides two items labeled “public long-term debt issuance” and “public long-term debt

reduction”, we compute “net public long-term debt issuance” (the difference in the two

items), and we split our regression of financial debt changes into two mutually-exclusive,

completely-exhaustive categories: public long-term debt changes and other financial debt

changes. We repeat our range of regressions for these two additional variables and test further

H1a & b on public long-term debt and on net other financial debt.

This provides us with a more complete and more complex picture.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

192

Relating to public long-term debt, we now see that that the coefficient is negative and

significant, by about half the previous value, but it is only significant for dispersed ownership

firms and for the middle category. This implies that when business prospects deteriorate,

dispersed ownership firms tend to issue long-term public debt.

Relating to the other category, net other financial debt changes, the coefficients are now

significant for all firms, and they are negative for concentrated ownership and middle-range

firms, while they turn out positive for dispersed ownership firms.

So, concluding for concentrated ownership firms, we have some evidence that there is no

relation with public long-term debt, while the issuance of other net private financial debt is

contra-cyclical. We conclude that we cannot reject H1b.

Now, for dispersed ownership firms, the negative, contra-cyclical, coefficient obtained for

public long-term debt (-0.0023) is overcompensated by a positive and higher coefficient in

absolute value (+0.0038) for other financial debt. Hence this suggests two contrary effects that

may explain the absence of significance for the coefficient initially obtained for total debt

changes. Overall, for dispersed ownership firms, the net effect, adding the respective

significant coefficients obtained from the split, would correspond the one previously

identified (0.0038-0.0023= +0.0015, i.e. close to +0.0019). So from this new set of results, we

can confidently reject H1a (pro-cyclicality), as relating to public long-term debt issuance, why

we cannot reject H1a (pro-cyclicality) for other net financial debt. Further from the

aggregation of the specific results, we cannot reject pro-cyclicality overall as the sum of the

two specific coefficients is positive16.

16 For the middle-range firms, the results are consistent with the two other categories, providing relations mostly in line with those of concentrated ownership firms, with lower coefficients.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

193

So, firms with dispersed ownership are substituting long-term public debt with private debt in

relation to the accommodation of business cycles: upon expected business downturn,

dispersed ownership firms may issue public bonds, to secure financing, and mitigate risks

(this may correspond to market timing as well, if interest rates decrease with expected

downturns). These firms would amortize public long-term debt, without re-issuing it with

expected upturns. This pro-cyclical mechanism would enable to mitigate risks, while the

related and overcompensating contrary effects with private debt would ensure financing, and

potentially disciplining through debt (alternatively, this may corresponds a given financial

policy required by a given rating).

So what does this suggests? This suggests that concentrated ownership firms may be indeed

associated with some financial smoothing, because manager-shareholder agency costs are

lower, while there is some inter-temporal management of the firm’s debt capacity. This

corresponds to qualitative evidence as gathered by surveys (Graham and Harvey, 2001;

Bancel and Mitoo, 2004; Brounen, et al., 2006). For dispersed ownership firms, this suggests

that the lean company proposition holds, while there is some room for sir mitigation through

debt-substitution.

Eventually we note that the counter-cyclicality for concentrated ownership firms is derived

from private debt management only. This suggests that financing may not only occur

differently, in response to anticipated business cycles, but through different channels as well.

Furthermore, we note that median-range firms present a mixture of the patterns observed for

both dispersed ownership and concentrated ownership firms17. Last we note that debt changes

are specifically related to consumers’ E.S.I. and we consider this further in section 4.4.

17 We note that in all these regressions we control for firm’s size (appendix 4) and we obtain a positive and significant coefficient for dispersed ownership firms across the boards (i.e. the bigger the size, the more addition of debt of any kind), while coefficients’ signs are mixed for medium-range companies and insignificant for concentrated ownership firms.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

194

4.3. Interactions  …  (2):  dividends,  cash  holdings  and  investments  

4.3.1. Dividend policies

Relating to the dividend policies, we cannot reject either of H2a & b.

In dispersed ownership firms, a positive, pro-cyclical, relation is obtained between the

dividend payout ratio and expected business cycles, while this relation is negative, contra-

cyclical in concentrated ownership firms. It is not significant in the middle range.

This would imply that firms with dispersed ownership disgorge excess cash flows more than

proportionally, paralleling the anticipated business situation and expected future cash flows.

This corresponds to varying theoretical frames including signaling arguments, or

corresponding to effective governance: shareholders get rewarded for the subsumed risk, with

the residual share of excess returns (increase in dividend payout ratio).

For firms with concentrated ownership this implies that, upon a positive (negative) business

outlook, a smaller (larger) proportion of earnings is distributed. It suggests that there may be

an absolute value-target for the dividend policy. This corresponds to evidence collected on

family firms, to arguments about dividend clienteles, or still to our suggestion that the

mitigation of inter-temporal risks bears relatively more weight than manager-shareholders

agency costs. Obviously this may bear consequences in terms of majority large shareholders-

to-minority shareholders agency costs, which warrants specific investigation. A minima, this

shows that a changing mix in agency relations bears consequences on financial policies.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

195

4.3.2. Investment policies

We find some evidence consistent with our hypothesizing as well. More specifically, we can

reject H3b for concentrated ownership firms while we cannot reject H3a for firms with

dispersed ownership. As a consequence, H3-comparative cannot be rejected either (the

coefficient for dispersed ownership firms is more positive).

Overall this provides support for our suggestion that firms with dispersed ownership need to

be lean, as a consequence of the disciplining motive, derived from traditional governance

considerations (in the perspective provided by the dispersed ownership role model, De Jong,

et al., 2008). The fact that we cannot identify any significant relation, between our anticipated

business cycle proxies and investments, in concentrated ownership firms, is somewhat

puzzling. But it is in line with our argumentation. It suggests that time-horizons or investment

policies may follow different patterns, depending on ownership structures, which is for

instance consistent with specific arguments in Jensen (1989) or Zingales (2000).

4.3.3. Cash holdings

We had conflicting prediction derived from theory, and we observe a positive and significant

sign for all classes of firms, which is likely related to the primarily endogenous determination

of cash holdings.

However, we note that the coefficient for dispersed ownership firms is twice that of

concentrated ownership firms, while both are significant. This provides ground for hypothesis

H4-comparative, which we can not reject. Overall, we find that cash holdings are pro-cyclical

in all types of firms and, the more dispersed the ownership class, the larger the coefficients

obtained in absolute value, both for the E.S.I. measure (positive coefficient) and for interest

rates (negative coefficient).

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

196

An interpretation of this result will be somewhat speculative. First, this is a sign that

endogeneity, or entrenchment effects (pro-cyclicality), are dominating precautionary

measures (counter-cyclicality). Second, this is a sign that, overall, pro-cyclicality is more

pronounced in dispersed ownership firms, consistently with H4-comparative and with our

general propositions about lean vs. smoothing policies. Third and complementarily, we note

that, if the coefficient for the E.S.I. measure is twice larger, in dispersed compared to

concentrated ownership firms, the reverse holds for interest rates, reinforcing the idea that

expected business cycles, as compared to interest rates drive cash-holding policies in

dispersed, as compared to concentrated ownership firms.

4.4. Interactions  …(3):  varying  proxies  for  anticipated  business  cycles  

We now turn to the specifics of the three complementary regressions (equations 3-*), adding

either managers’ E.S.I., or consumers’ E.S.I., or still the I.V. approach. We did not

hypothesize on the relative effects of our three main variables, but for general mechanism

relating to investment policies (H3bis).

Relating to that specific prediction, we cannot disconfirm H3bis. We further identify some

noteworthy results.

4.4.1. Interest rates

First, we note that long-term interest rates do not matter for any of the debt regressions. We

conduct additional regressions (not reported here) with the baseline model, adding only long-

term interest rates in addition. None of the coefficients obtained in those regressions, for

interest rates, is significant, in any of the 3 regressions, for any ownership structure. This

confirms the results reported in Table 6 in relation to the I.V. approach (with interests rate,

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

197

consumers’ E.S.I. and the residuals from the regression of managers’ E.S.I. on the two other

variables). This result is surprising because one would expect some debt management in

relation to interest changes, as exemplified by survey evidence (see earlier citations). We

suggest that this lack of relation may be due to a combination of financial sophistication in

banking credit (flexible rates), and, or in banking markets (ability to renegotiate without re-

contracting). More specific evidence, country by country, is needed.

Second, and still relating to interest rates, we note that there are significant and negative

coefficients across the board (in all ownership classes) for the three complementary analyses

of investments, dividends and cash holdings. This is confirmed by complementary analyses

based on the ‘baseline’ plus the interest rates alone, as above, results that are not reported

here. These results are more in line with conventional wisdom because interest rates are

central in the investment process, they represent the opportunity cost of holding cash (from a

managerial perspective) on the one hand, or holding risky shares on the other.

4.4.2. Debt and investment policies vs. dividend and cash policy

We note that there are clear and marked differences concerning which proxy for the expected

business cycles matters. While manager’s specific information (E.S.I. manager) does not

matter for debt or investment policies, consumer confidence (ES.I. consumers) does not

matter for dividend or cash policies.

This polarization of observed correlations between expected business cycle proxies and

specific financial policies is puzzling; first, the specific association between dividend and cash

policies echoes findings by Bates, et al. (2009) who observe that the “change in cash ratios

and net debt is the result of a secular trend, [which] is concentrated among firms that do not

pay dividends”. Second, this association seems to provide consistence to the debate pertaining

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

198

to the association between entrenchment and cash holdings/ dividend policies. So while some

suggests that poorly governed firms hold more cash (for instance Dittmar, et al., 2003), other

suggests that poorly governed firms use more intensively cash, so that they hold less cash

(Harford, et al., 2008). Further, Bates, et al. (2009) suggests that the general increase in cash

holdings, in historical perspective in the U.S., is primarily driven by precautionary arguments.

In any case the polarization in our findings in a context where we suggest both a divergence in

precautionary policies and in the opportunity costs of entrenchment relates to the two side of

the debate. We speculate that entrenchment and precautionary arguments may carry different

weight and interrelate in different ways, depending on ownership dispersion, echoing our

findings on the contra-cyclicality of debt vs. debt substitution in section 4.2.

In any case, the polarization of our results, depending on the variable we use to proxy for

business cycles, suggests marked differences in the way financial policies unfold. Relating to

manager-specific information about business cycles (E.S.I. managers), we observe a marked

pro-cyclicality of dividend and cash-holdings policies in dispersed ownership firms, while

there is evidence, in concentrated ownership firms, of dividend smoothing and more limited

pro-cyclicality in cash holdings.

Relating to consumer confidence, a piece of information more readily observable to all

economic actors, we note no effect on either dividend or cash holdings for any class of firm,

while we note a strong pro-cyclicality of both investment and debt policies in dispersed

ownership firms. In firms with concentrated ownership, there is no cycle-sensitivity of

investment and some financial smoothing in relation to economy-wide growth anticipations

(consumer confidence). So it seems that general growth, and its expected effects on free cash

flows, has an incidence on investment, in dispersed ownership firms, and debt, in both cases,

but with opposite effects.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

199

4.4.3. Leverage and anticipated business cycles

Book leverage is contra-cyclical in concentrated ownership firms (H5b), while it is not related

to business cycles in dispersed ownership firms (H5a), hence providing further support to our

proposition that the management of dispersed ownership companies is lean, which is broadly

consistent with the traditional financial view that firm’s policies should not be directly and

importantly influenced by anticipated business cycles.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

200

5. Conclusion.  

Provided our specific focus and the limited amount of past studies on the subject, we proceed

to a largely explanatory study about the interactions between capital structures, business

cycles and ownership structures. Our study proceeds in an institutional context where mean-

ownership concentrations are moderate, and neither dispersed ownership nor concentrated

ownership firms dominate the sample, in aggregate.

This research design departs from the more traditional literature in many ways, reflecting

some more recent contributions either on international capital structures (De Jong, et al.,

2008; Fan, et al., 2008; King and Santor, 2008; Margaritis and Psillaki, 2009), or on financial

intermediation and financial macro-structures (i.e. market-financed vs. bank-financed macro-

financial structures, Aggarwal and Goodell, 2009), or still on the interaction between the two

domains (Antoniou, et al., 2008; Gonzales and Gonzales, 2008). Building on this blossoming

literature, we try to provide a theoretical and practical sense about how a specific mix of

agency relations, corresponding to varying degree of ownership dispersion, should (in theory)

and does (empirically) lead to a polarization in financial policies.

We focus on a sample of firms corresponding roughly to the universe of the European listed

firms over the last two decades. This provides us with a population of listed and liquid firms,

in a context where there is much converging quality standards in economic, legal, regulatory,

accounting, or political terms, while mean-concentration of ownership is moderate to high,

much in line with worldwide realities (Holderness, 2010).

Our baseline, in line with traditional financial theorizing, is that the interaction between

capital structures and business cycles should be of limited relevance. However, we suggest

that this proposition, that we label the lean financial policy, should correspond to what

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

201

De Jong, et al. (2008) have characterized as the “traditional perspective” in financial theory,

based on the “U.S. role model”, with dispersed ownership. So given our institutional context,

characterized with the prolonged co-existence of firms with primarily dispersed vs. primarily

concentrated ownership, we propose an alternative proposition that we label financial

smoothing. We suggest that financial smoothing policies correspond to firms with primary

concentrated ownership, and they are characterized by a marked contra-cyclicality, where

risk-mitigation receives relatively more attention by the management, and the board, while

signaling, market timing, or disciplining receives relatively less attention. Theoretically we

interpret this as the consequence of the shift in the mix of agency relations, implying different

agency costs, opportunity benefits of managerial discretion, as well as an alignment in risk-

mitigation motives among management and block- or control-holders.

We delineate our proposition into specific hypotheses, pertaining to varying dimensions of the

financial policies being followed: these include, public debt issuance, change in private

financial debt, dividend policies, investment policies, cash holdings and book-leverage.

We note, empirically, a marked polarization of the relation between capital structures and

anticipated business cycles, depending on ownership dispersion: the anticipated business

cycle-proxies exhibit a polarized relation with all of our exogenous variables but one (change

in cash holdings). Three additional conclusions are derived. First, debt management in

dispersed ownership firms exhibits a more complex pattern, with some debt-substitution.

Long-term public debt issuance is contra-cyclical (unlike total financial debt, which is more

pro-cyclical than long-term public debt is contra-cyclical). This could imply either some risk

mitigation or a form of market timing. Second, changes in cash-holdings are more pro-

cyclical in dispersed ownership firms, which suggests either a different precautionary

perspective, or a different importance of entrenchment effects in aggregate. In any case, it

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

202

seems that both mechanisms, the precautionary motive and entrenchment’s effects in

aggregate, play out differently, depending on ownership dispersion. Third, different proxies

matter in relation with anticipated business cycles, with managers’ specific information being

relevant for dividend and cash policies, while economy-wide information (consumer

confidence) matters for debt and investment management.

To streamline and simplify the argument, we reckon that business cycles are of little relevance

in the context of the dispersed ownership firm, where a primary structuring dimension, in the

execution of the financial policies, is the agency relation between the manager and

shareholders. As exemplified, debt and investment management need to be pro-cyclical to

ensure the disciplining of the manager, and so is the dividend policy, while some effects of

entrenchment are observable in aggregate.

In firms with concentrated ownership (their average proportion of floating shares is lower

than a third of the total market capitalization), policies enfold in different ways, and primarily

reflect a concern for the mitigation of risks, as represented by expected business cycles. While

investment is not related to the business cycle, debt management (primarily private financial

debt) is contra-cyclical and the dividend policy likely reflects an alignment in interests among

the management, control- and block-holders.

So the answer of our question in title is yes, agency mediates the interaction between business

cycles and financial policies.

Our potential contribution extends to a number of adjacent research questions. First, we

suggest that business cycles relate differently to debt and investment management, depending

on ownership dispersion. This provides some specifics on how a theory of investment could

be mediated by some underlying institutional dimension (Jensen, 1989; Zingales 2000), here

ownership concentration. Second, because we suggest that risk mitigation, and ripping off the

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

203

benefits of managerial discretion, may be a primary concern in firms with concentrated

ownership, a thorough analysis of performance effects should be carried out, in international

samples, keeping in mind varying investment mechanisms, different time-horizons and

sampling concerns, in particular in relation to censoring (M&As, bankruptcies, etc.). Third, a

relative assessment of governance mechanisms, depending on ownership concentration, could

take into account potentially positive performance effects that minority shareholders may

free-ride. After all, it has long been recognized that small bond-hollers may free-ride effective

monitoring by larger bond-holders (Shleifer and Vishny, 1986). This potential benefit should

be compared to the opportunity costs of expropriation (La Porta, et al., 1999). Fourth, Frank

and Goyal (2008: p. 21) note in their synthesis on past and present capital structure research

“no-one has tried to distinguish among alternative sources of pecking order behaviors”. We

suggest, in this article, that a more refined, and polarized approach to agency, could provide

thicker material to investigate managerial behaviors empirically. In short, we provide some

specific evidence on how an institutional context can mediate financial relations and

managerial or shareholder behaviors. So we provide some specifics on a rising field of interest

in financial research, as exemplified by De Jong, et al. (2008: p. 1955): “financial policies

and manager’s behavior are influenced by the institutional environment and international

operations […] more research is needed”.

* * *

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

204

6. Appendix   1   (chapter   III):   average   ownership   concentration   in  

Europe  1990-­2007.  

A key item in our study is the governance regime that firms belong to. We identify the

governance regime from the proportion of the float in the total stock market capitalization.

The float is a variable provided by Datastream that quantifies the share of stocks that are

traded publicly, provided the number of outstanding shares. Hence it proxies for the existence

of blocks or long-term investors (families, sovereign investors, institutional investors such as

insurers or pension funds…).

We compute the all-years average and the standard deviation of the float for each company.

On both criteria, we separate our sample in three categories. For the average of the float over

the listing period we distinguish companies that have a float lower than 33%, those that have a

higher float than 66% and others (respectively 28%, 28% and 42% of firm-years). For the

standard deviation of the float, we set the thresholds at 5% and 25%. Relating to the latter, we

note that 2/3rd of the low volatility firms are low float firms. Otherwise, there seem to be no

further overlap.

Overall, our sample exhibit contrasted governance structure: there are more than a third of

firms exhibiting both concentrated and dispersed ownership structures in any of the 15

countries investigated here. More specifically and roughly speaking, a third of countries have

a median float around 33%, a third around 50% and a third around 66%. Not surprisingly, the

U.K. and Ireland have highest median float (more dispersed-ownership) while countries

generally featured as bank-based financial systems (with universal banks, a larger proportion

of banking assets as compared to total market capitalization, etc.) have lower median floats.

Interestingly, there seems to be a slight upward trend in the median and average float for most

countries, but the U.K. and Ireland, so that overall median differences have decreased over the

last two decades. Median and averages by industry do not exhibit any noticeable pattern at

first sight, but for the upward general trend just mentioned. The next table provides detail

about this variable.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

205

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

AU

T - A

ustr

ia40

%37

%33%

33%

34%

35%

35%

37%

35%

34%

34%

34%

37%

37%

37%

37%

37%

38%

39%

38%

40%

BE

L - B

elgi

um40

%41

%46%

46%

46%

46%

46%

46%

45%

43%

42%

39%

38%

38%

38%

38%

38%

38%

39%

40%

53%

DE

U -

Ger

man

y30

%36

%41%

40%

40%

39%

39%

39%

36%

36%

37%

35%

35%

34%

34%

35%

35%

35%

35%

35%

43%

DN

K -

Den

mar

k60

%56

%60%

60%

61%

61%

60%

60%

57%

55%

53%

53%

54%

54%

54%

54%

54%

54%

54%

54%

56%

ESP

- Sp

ain

50%

51%

59%

56%

55%

55%

55%

55%

53%

52%

51%

50%

50%

50%

50%

50%

50%

50%

50%

50%

52%

FIN

- Fi

nlan

d60

%57

%66%

65%

63%

63%

63%

62%

59%

58%

56%

57%

55%

55%

55%

55%

55%

55%

55%

55%

58%

FRA

- Fr

ance

30%

36%

45%

45%

44%

44%

43%

43%

39%

38%

36%

36%

34%

34%

34%

34%

34%

34%

34%

35%

36%

GB

R -

Gre

at B

rita

in70

%69

%73%

74%

74%

74%

73%

73%

71%

70%

69%

69%

69%

68%

68%

68%

68%

67%

67%

67%

68%

GR

C -

Gre

ece

20%

23%

27%

28%

31%

32%

33%

33%

30%

29%

28%

25%

21%

20%

20%

20%

19%

20%

20%

20%

34%

IRL

Irel

and

70%

66%

71%

71%

69%

70%

70%

70%

70%

70%

70%

65%

63%

62%

62%

62%

62%

63%

63%

63%

66%

ITA

- It

aly

40%

42%

46%

46%

46%

46%

47%

47%

47%

46%

44%

43%

42%

41%

41%

41%

40%

40%

40%

40%

37%

NL

D -

Net

herl

ands

50%

50%

51%

51%

51%

53%

52%

52%

52%

53%

51%

50%

48%

48%

48%

48%

49%

48%

49%

49%

54%

POL

- Pol

and

30%

35%

30%

45%

46%

47%

47%

45%

46%

42%

41%

37%

34%

33%

30%

30%

30%

21%

PRT

- Por

tuga

l30

%33

%36%

35%

33%

32%

34%

36%

30%

33%

34%

34%

33%

33%

34%

33%

33%

33%

34%

34%

31%

SWE

- Sw

eden

50%

48%

65%

63%

60%

61%

61%

62%

59%

55%

50%

49%

45%

46%

45%

45%

44%

43%

43%

44%

52%

Ret

ail t

rade

50%

52%

63%

63%

62%

61%

59%

60%

56%

54%

51%

48%

47%

47%

46%

47%

48%

47%

48%

47%

56%

Con

stru

ctio

n50

%48

%63%

62%

60%

59%

58%

56%

54%

54%

54%

53%

51%

50%

48%

48%

48%

48%

47%

48%

60%

Man

ufac

turi

ng50

%48

%57%

56%

55%

55%

54%

54%

52%

50%

48%

47%

46%

45%

45%

45%

45%

45%

45%

45%

56%

Serv

ices

50%

47%

61%

62%

61%

62%

61%

59%

55%

54%

50%

47%

45%

45%

46%

46%

46%

46%

46%

47%

55%

Oth

er in

dust

ries

50%

48%

54%

54%

53%

52%

52%

52%

49%

49%

48%

47%

45%

44%

44%

43%

43%

44%

45%

45%

54%

Tota

l50

%48

%57

%57

%56

%55

%55

%55

%52

%51

%49

%48

%46

%45

%45

%45

%45

%45

%45

%46

%55

%

AU

T - A

ustr

ia17

1922

2527

3134

3842

4753

5555

5656

5960

5825

BE

L - B

elgi

um27

2828

2929

2931

4153

6367

6868

7072

7777

757

DE

U -

Ger

man

y155

164

178

193

205

217

249

270

318

398

489

492

486

472

473

479

479

464

148

DN

K -

Den

mar

k42

4549

5154

5566

6975

7681

8384

8686

8683

8361

ESP

- Sp

ain

3746

5051

5252

6267

7075

8082

8687

8990

9290

32FI

N -

Finl

and

3134

3841

4446

5462

6986

9899

9999

99103

105

104

84FR

A -

Fran

ce130

145

155

161

172

179

222

268

320

372

427

447

454

455

459

470

471

433

113

GB

R -

Gre

at B

rita

in333

344

350

364

380

393

449

494

528

551

597

634

663

676

722

794

823

806

512

GR

C -

Gre

ece

1516

3240

5359

8695

105

132

179

196

203

206

212

214

211

210

30IR

L Ir

elan

d17

1820

2121

2222

2525

2628

2626

2626

3031

319

ITA

- It

aly

4747

4949

5259

6675

87103

139

155

160

165

173

182

183

180

29N

LD

- N

ethe

rlan

ds48

5355

5760

6469

7786

9197

9797

9799

101

100

9733

POL

- Pol

and

16

2022

2533

4049

5264

81101

126

128

125

15PR

T - P

ortu

gal

810

1113

1620

2527

2833

3838

3738

3838

3737

12SW

E -

Swed

en30

3645

5156

5969

92109

129

152

165

174

175

182

192

196

192

71R

etai

l tra

de66

6767

7480

8498

116

126

136

157

153

166

169

175

185

187

179

96C

onst

ruct

ion

5760

7279

8484

104

106

105

115

114

119

121

126

128

127

127

124

41M

anuf

actu

ring

533

573

615

641

693

738

828

909

1018

1106

1163

1212

1233

1233

1267

1338

1343

1303

496

Serv

ices

8794

100

107

115

120

165

223

298

417

585

636

654

668

689

737

754

728

280

Oth

er in

dust

ries

194

211

228

246

255

279

331

371

401

448

555

569

582

593

628

654

665

651

268

Tota

l93

710

0510

8211

4712

2713

0515

2617

2519

4822

2225

7426

8927

5627

8928

8730

4130

7629

8511

81

Tabl

e 7.

Pub

lic m

arke

t cap

italiz

atio

n as

a p

erce

ntag

e of

tota

l mar

ket c

apita

lizat

ion

and

sam

ple'

s cha

ract

eris

tics

5853

2294

1041

3

Med

ian

& a

vera

ge p

ublic

mar

ket c

apita

lizat

ion

by c

ount

ry &

indu

stry

Num

ber

of fi

rms

by y

ear

& b

y co

untr

y or

indu

stry

(tot

al n

umbe

r fo

r al

l col

umns

)77

993

9

Aver

age

year

ly p

ublic

mar

ket c

apita

lizat

ion

by c

ount

ry &

indu

stry

Aver

age

float

Med

ian

float

Cou

ntry

6329

1315

1290

1395

450

2000

1478

888

3810

2

504

2175

2381

1893

1824

274

5781

29

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

206

7. Appendix  2  (chapter  III):  pair  wise  correlations  

12

34

56

78

910

1112

1314

1516

1718

1920

2122

2324

251

Cha

nge

in m

arke

t lev

.2

Cha

nge

in b

ook

lev.

62%

Leve

rage

3C

hang

e in

fina

ncia

l deb

t (/T

A)

48%

59%

4N

et lo

ng te

rm p

ublic

deb

t iss

uanc

e (/T

A)

37%

49%

61%

5C

hang

e in

oth

er fi

nanc

ial d

ebt (

/TA

)36

%44

%86

%-9

%6

Cha

nge

in c

ash

hold

ings

(/TA

)-3

%-4

%21

%7%

21%

7C

apita

l exp

endi

ture

s (/T

A)

4%4%

5%10

%3%

-1%

8D

ivid

end

payo

ut ra

tio2%

2%5%

4%4%

1%4%

Oth

er e

xoge

nous

var

iabl

es9

Cha

nge

in p

rofit

abili

ty-1

8%-2

4%-9

%-2

%-9

%14

%-3

%10

Cha

nge

in ta

ngib

ility

13%

14%

9%4%

8%-2

0%11

%1%

-6%

11C

hang

e in

loga

rithm

of s

ize

18%

9%45

%34

%35

%34

%6%

2%15

%-5

%12

Cha

nge

in to

bin'

s Q-1

9%-7

%-6

%-5

%-5

%2%

1%-1

9%13

Cha

nge

in n

on-d

ebt t

ax sh

ield

6%5%

-7%

-6%

-5%

-11%

3%-1

%-1

3%11

%-1

7%-2

%Fi

rm-s

peci

fric

var

iabl

es14

Cha

nge

in m

edia

n in

dust

ry b

ook

lev.

13%

5%4%

5%2%

-3%

-3%

2%-5

%3%

15C

hang

e in

med

ian

indu

stry

mar

ket l

ev.

25%

7%3%

5%1%

-6%

-6%

3%-2

%-1

5%6%

69%

16D

iffer

ence

firm

's le

ss in

dust

ry b

ook

lev.

20%

33%

16%

18%

10%

-2%

4%-8

%-4

%3%

-2%

3%1%

-4%

-2%

17D

iffer

ence

firm

's le

ss in

dust

ry m

arke

t lev

.27

%18

%8%

7%6%

-1%

-12%

-4%

4%-7

%2%

-3%

-4%

77%

Indu

stry

var

iabl

es18

Logr

ithm

of s

ize

3%1%

8%7%

5%6%

-1%

14%

-2%

7%2%

2%15

%15

%19

Cum

ulat

ed n

et e

quity

issu

e (/T

A)

5%-1

%10

%6%

9%14

%4%

15%

-17%

16%

-8%

9%2%

4%-1

2%-9

%24

%N

atio

nal e

cono

mic

"co

ntex

t" v

aria

bles

2012

-moo

nths

cha

nge

in E

.S.I.

man

ager

-20%

-5%

-2%

-3%

5%5%

-4%

5%12

%-4

%-2

7%-5

3%-4

%-3

%21

12-m

oont

hs c

hang

e in

E.S

.I. c

onsu

mer

-19%

-5%

-2%

5%4%

-4%

7%13

%-4

%-2

4%-4

8%-3

%-3

%76

%22

Cha

nge

in st

ock

mar

ket i

ndic

es b

y co

untry

-23%

-4%

2%2%

7%4%

-4%

10%

20%

-7%

-21%

-63%

1%2%

-3%

-4%

52%

59%

23C

hang

e in

long

-term

inte

rest

rate

s1%

-5%

-3%

2%-4

%2%

-1%

-3%

1%1%

32%

18%

8%24

Cha

nge

in u

nem

ploy

men

t-2

%-6

%-4

%-4

%-2

%1%

4%-1

3%4%

1%1%

2%-6

%-2

3%-3

2%-1

9%-3

2%25

Cha

nge

in c

onsu

mer

pric

e in

dex

-1%

-2%

-1%

-1%

-1%

2%-3

%1%

-4%

-2%

-1%

6%-3

%-1

8%9%

26R

esid

ual o

f the

I.V.

app

roac

h-1

0%-3

%-2

%-3

%-1

%3%

-1%

3%-2

%-1

%4%

-2%

-17%

-29%

-2%

-2%

75%

15%

15%

29%

-4%

15%

Tabl

e 8:

pai

rwis

e co

rrel

atio

ns (a

naly

sis

of fi

nanc

ial p

olic

ies

in E

urop

e)

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

207

8. Appendix  3  (chapter  III):  the  E.S.I.  measures  

8.1. Presenting  the  measures  

In this study we use the Economic Sentiment Indicator (ESI) developed by the European

Commission. It is available for most European countries and was developed 3 decades ago,

with robust sampling and measure over the period18. This index comes in different versions:

managers’ surveys, consumers’ surveys and capacity utilization. It is executed on a monthly

basis and aggregated over industries, over countries or for Europe as a whole with varying

perimeters (e.g. 12, 13, 15, 16, 25 or 27 countries). There are 7 survey-questions for

managers’ surveys19 and 12 questions for consumers’ surveys (with some more questions in

quarterly and annual surveys for both). The managers’ surveys deal with company’s business

situation, its demand for inputs, employment, upstream and downstream prices. The consumer

surveys deal with both household-specific issues (wages, employment, prices) and general

assessment of their country’s economic situation.

Generally, the E.S.I. and economic sentiment indices in general are widely used in practice.

Policy makers in the E.U. and in member countries routinely use the E.S.I. The E.S.I. and

other economic sentiment indices (such as the ones produced by the O.E.C.D., by Datastream,

by Dow-Jones, by Michigan University,… ) are commonly used by stock market actors (high

citation frequency of “economic sentiment” on Reuters News, Dow-Jones News, PR

Newswire…). This cite-frequency in specific stock market information sources is a tangible

18 The index is available for download from the European Commission’s website and the survey methodology is explained in the related User Guide (2007 2007). In particular the section 3.1 of this user guide provides details on the sampling methodology. 19 Industries based on the NACE classification were originally sampled from three broad categories: retail, construction and manufacturing. ‘Services’ were added in the early 90s and the related survey included 6 questions instead of 7 questions. Financial services were then added very recently. The 7 survey-questions for industry-managers are: Q1 How has your production developed : + = - Q2 Do you consider your current overall order books to be...? : + = - Q3 Do you consider your current export order books : + = - Q4 Do you consider your current stock of finished products to be...? : + = - Q5 How do you expect your production to develop : + = - Q6 How do you expect your selling prices to change : + = - Q7 How do you expect your firm’s total employment to : + = -

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

208

sign that economic sentiment are integrated into mechanisms of price-discovery on stock

markets. Additionally, such indices are often cited in the business press (Business Week,

Financial Times, New-York Times…), and so they are likely to be used by business decision

makers as well.

The E.S.I. has been used in research across human sciences, in particular in political sciences

and in economics. In economics, it has been shown to associate to accurate economic

forecasting, to stock market activity, to GDP growth or to household spending.

In Financial research, it has been used in two recent studies dealing with the antecedents of

government yield spreads in the EMU. Geyer, et al. (2004) show that the E.S.I. marginally

accounts for government yield spreads; long-term corporate bonds and swaps being primary

drivers20. Ferreira, et al. (2008) show that the ESI can be used instead of past ex-post data on

output and consumption to predict economic growth, hence replacing yield spreads. They

conclude that “this ability of yield spreads to capture economic agents expectations may be

the actual reason for the predictive power of yield spreads about future business cycles”.

8.2. Analyzing  E.S.I.  measures  

We analyze the sentiment measures provided the European Union in three ways. Graphically,

we analyze measures of managerial sentiment by industry and by country. They follow

similar trends generally, but exhibit specific sub-trends that are sufficiently marked. Second

we compute pair-wise correlations for all measures in each country. These pair wise

correlations cover the range of correlations between 0 to 1, so that on the one hand, specific

20 The ESI is indeed based on a short-term assessment, so it should not come as a surprise that “long spread models” explain most of the variation in yield spreads, and that ESI are a complementary source of explaining power.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

209

implementation must be done carefully, but on the other hand the correlations are lower than

one could expect.

Third, we regress managerial sentiment on varying other context variables, based on indices

or on period-to-period differences (i.e. months for variables available with monthly data, or

quarters for other variables). Further we repeat regressions with aggregated European data and

country-based measures. We note striking differences and important losses when data is

aggregated. In that case every context-variable ends to proxy for the general economic

situation, i.e. pan-European growth. .

We attract attention on the fact that the coefficient of regressing centered-reduced measures of

manager’s E.S.I. on consumers’ E.S.I. is about 0.5, so that there is separate explanatory power

in both the managers’ and consumers’ ESI.

Table 9. Key results for the analysis of manager's E.S.I. based on other context variablesRegressing E.S.I. for managers on other variables on a MONTHLY basis

Multivariate Univariate Multivariate UnivariateCoef. P(.) Coef. P(.) R2 Coef. P(.) Coef. P(.) R2

By countries on indices By countries on monthly-differentiated dataE.S.I. consumer 0.703 0% 0.785 0% 65% 0.416 0% 0.532 0% 30%Broad stock market indices 0.273 0% 0.434 0% 19% -0.044 32% -0.291 0% 1%Long-term Debt 0.143 0% -0.180 1% 3% 0.475 0% 0.785 0% 21%Unemployment index 0.089 0% -0.127° 9% 2% -0.040 32% -0.158 0% 1%Inflation (CPI) -0.040 0% -0.273 0% 8% 0.006 85% 0.154 0% 1%

For pan-European data on indices For monthly-differentiated pan-European dataE.S.I. consumer 1.001 0% 0.916 0% 84% 0.488 0% 0.585 0% 44%Broad stock market indices 0.026 73% 0.468 0% 22% 0.093 30% 0.500 0% 10%Long-term Debt 0.003 94% -0.126° 6% 1% 0.193° 7% 0.348 2% 2%Unemployment index 0.053 15% -0.354 0% 12% -0.143 24% -0.432 0% 3%Inflation (CPI) 0.215 0% -0.241 0% 6% -0.057 24% -0.038 51% -3%

Regressing E.S.I. for managers on other variables on a QUARTERLY basis, based on COUNTRY-level dataMultivariate coeficients Univariate coeficients reported in row

E.S.I. consumer 0.624 0% 0.807 0% 65%Broad stock market indices 0.204 1% 0.444 0% 19%Long-term Debt 0.080 22% -0.160 1% 2%Unemployment index 0.132 0% -0.152 3% 2%Inflation (CPI) 0.023 65% -0.271 0% 7%Labor costs 0.052 18% -0.120 7% 1%Capacity Utilization 0.204 0% 0.535 0% 27%These regressions are conducted using the absolute level of centered-reduced indices, and using panel data regressions

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

210

9. Appendix  4  (chapter  III):  main  results  with  all  coefficients  reported  

Table 10. Main results - all coeficients are reported (analysis of financial policies in Europe)

Low Float Medium Float High Float Low Float Medium Float High Float Low Float Medium Float High Floati.e. c-o. firms i.e. d-o. firms i.e. c-o. firms i.e. d-o. firms i.e. c-o. firms i.e. d-o. firms

1-Regressing change in total debt (over total assets)Change in profitability -0.1485*** -0.1321*** -0.123*** -0.149*** -0.1328*** -0.123*** -0.1489*** -0.1349*** -0.123***Change in tangibility 0.1772*** 0.3655*** 0.1553*** 0.1683*** 0.3677*** 0.155*** 0.1682*** 0.3782*** 0.149***Change in logarithm of size 0.2645*** 0.2622*** 0.2492*** 0.2656*** 0.2613*** 0.2486*** 0.2746*** 0.2657*** 0.2484***Change in tobin's Q n.s. 0.0075*** 0.0036*** n.s. 0.0075*** 0.0035*** n.s. 0.0077*** 0.0035***Change in non-debt tax shield n.s. -0.0856*** -0.1747*** n.s. -0.0878*** -0.1836*** n.s. -0.0878*** -0.1861***Change in median industry book lev. 0.1213*** 0.1339*** 0.1876*** 0.0934** 0.1312*** 0.2066*** 0.0951** 0.1141*** 0.197***Difference firm's less industry book lev. 0.0841*** 0.0819*** 0.0783*** 0.083*** 0.082*** 0.0792*** 0.0783*** 0.081*** 0.0787***Logrithm of size 0.0022** n.s. 0.0036*** 0.0021** n.s. 0.0036*** 0.0023** n.s. 0.0036***Cumulated net equity issue (/TA) n.s. 0.0061** n.s. n.s. 0.0055** n.s. n.s. 0.0051** n.s.E.S.I. manager n.s. -0.0031*** n.s.Residuals of I.V.a. (! esimnger cleaned) n.s. n.s. n.s.E.S.I. consumers -0.0046*** -0.0041*** n.s. -0.0045** -0.0046*** n.s.L.T. interest rates n.s. n.s. n.s.1a-Regressing public long-term debt issuance (over total assets)C_Profits -0.0402*** -0.026*** -0.0389*** -0.0405*** -0.0254*** -0.0383*** -0.0395*** -0.0257*** -0.0368***C_tangible n.s. 0.1596*** 0.039** n.s. 0.1593*** 0.0387** n.s. 0.1646*** 0.0302* C_size 0.1153*** 0.1237*** 0.121*** 0.1194*** 0.1229*** 0.1207*** 0.1198*** 0.1245*** 0.1199***C_tobinsq 0.0025* 0.0021** 0.003*** 0.0027** 0.002** 0.003*** 0.0028** 0.0021** 0.0031***C_non_D_TS 0.2767*** -0.1299*** n.s. 0.3585*** -0.13*** n.s. 0.3602*** -0.1302*** n.s. C_med_MLev 0.0746** 0.1041*** 0.1671*** 0.0653* 0.0938*** 0.1584*** n.s. 0.0883*** 0.1533***MLev_to_ind 0.0381*** 0.047*** 0.0474*** 0.0374*** 0.0475*** 0.0489*** 0.0363*** 0.0473*** 0.0489***Size n.s. 0.0021*** 0.0021*** n.s. 0.002*** 0.002*** n.s. 0.0021*** 0.002***Cumul_NEI_TA -0.0088** 0.0043** -0.0068*** -0.0091** 0.0041* -0.007*** -0.0094** n.s. -0.0067***esimger12 n.s. n.s. n.s. mgerregres12 n.s. n.s. n.s. esicsumer12 n.s. -0.002* -0.0022** n.s. -0.0023** -0.0023** d_ltr212 n.s. n.s. n.s. 1b-Regressing change in residual financial debt (residual net of public long-term debt issuance)C_Profits -0.1362*** -0.1103*** -0.0873*** -0.1367*** -0.1115*** -0.0878*** -0.137*** -0.1131*** -0.0891***C_tangible 0.1781*** 0.2391*** 0.1287*** 0.1733*** 0.242*** 0.1286*** 0.1728*** 0.2465*** 0.1297***C_size 0.2252*** 0.1747*** 0.1483*** 0.2257*** 0.1741*** 0.1482*** 0.2343*** 0.1772*** 0.1488***C_tobinsq n.s. 0.0062*** n.s. n.s. 0.0062*** n.s. n.s. 0.0063*** n.s. C_non_D_TS n.s. n.s. -0.1533*** n.s. n.s. -0.1615*** n.s. n.s. -0.1652***C_med_MLev 0.091** n.s. n.s. n.s. 0.0555* n.s. n.s. n.s. n.s. MLev_to_ind 0.0689*** 0.0489*** 0.0387*** 0.068*** 0.0486*** 0.0384*** 0.0632*** 0.0478*** 0.0379***Size n.s. -0.0009* 0.0017*** n.s. n.s. 0.0018*** n.s. n.s. 0.0018***Cumul_NEI_TA n.s. 0.0044** 0.0057** n.s. 0.0038* 0.0056** n.s. 0.0039* 0.0057** esimger12 n.s. -0.0021** n.s. mgerregres12 n.s. n.s. n.s. esicsumer12 -0.0037** -0.0023** 0.0035** -0.0034* -0.0023** 0.0038** d_ltr212 n.s. n.s. n.s. 2-Regressing dividend payout ratioC_Profits n.s. 0.0291** 0.0498*** n.s. 0.0276** 0.0482*** n.s. 0.0293** 0.0479***C_tangible 0.1656*** 0.1198*** n.s. 0.1739*** 0.1198*** n.s. 0.1595*** 0.1155*** n.s. C_size n.s. -0.0253*** -0.0307*** n.s. -0.0236*** -0.028*** n.s. -0.025*** -0.028***C_tobinsq n.s. 0.0059*** 0.0048** n.s. 0.006*** 0.0051** n.s. 0.0061*** 0.0049**C_non_D_TS -0.1224* -0.1184** -0.1515* n.s. -0.1186** -0.1626* n.s. -0.1188** -0.1603*C_med_MLev -0.3176*** n.s. n.s. -0.2117** n.s. -0.1663** -0.2389** n.s. n.s. MLev_to_ind -0.2338*** -0.1587*** -0.1597*** -0.2332*** -0.1622*** -0.1646*** -0.2371*** -0.1645*** -0.1637***Size 0.0235*** 0.0167*** 0.0241*** 0.023*** 0.0168*** 0.0243*** 0.0236*** 0.0172*** 0.0243***Cumul_NEI_TA 0.0803*** 0.084*** 0.0432*** 0.0821*** 0.0836*** 0.0427*** 0.0807*** 0.0833*** 0.0435***esimger12 -0.0085** n.s. n.s. mgerregres12 -0.0121*** n.s. 0.0117***esicsumer12 n.s. n.s. n.s. n.s. n.s. n.s. d_ltr212 -0.0264*** -0.021*** -0.022***

Pooling the regressions of financial policies variables on firm-, industry-specific and business cycle proxies. All regression-coefficients are reported unlike in table 4, which is an extract of this table

Adding ESI Managers only to regressions Adding ESI COnsumers only to regressions Instrumental variable approach

POOLING:

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

211

Table 10 (continued).

Low Float Medium Float High Float Low Float Medium Float High Float Low Float Medium Float High Floati.e. c-o. firms i.e. d-o. firms i.e. c-o. firms i.e. d-o. firms i.e. c-o. firms i.e. d-o. firms

3-Regressing capital expenditure / total assets Change in profitability n.s. -0.0179*** -0.0095*** n.s. -0.0177*** -0.0093*** n.s. -0.0148*** -0.0091***Change in tangibility 0.1429*** 0.2009*** 0.207*** 0.1401*** 0.2009*** 0.2053*** 0.1343*** 0.171*** 0.2025***Change in logarithm of size 0.0179*** 0.0365*** 0.0203*** 0.0179*** 0.0363*** 0.0196*** 0.0178*** 0.0262*** 0.0187***Change in tobin's Q n.s. 0.0016* 0.0013*** n.s. 0.0015* 0.0013*** n.s. 0.0014** 0.0014***Change in non-debt tax shield 0.0766*** 0.0411* 0.034** 0.0859*** 0.0417* 0.0384** 0.0826*** 0.0348** 0.0402***Change in median industry book lev. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 0.0885*** n.s. Difference firm's less industry book lev. n.s. n.s. -0.0053* n.s. n.s. -0.0063** n.s. 0.0132*** n.s. Logrithm of size -0.0039*** -0.0013** 0.0009*** -0.0038*** -0.0014** 0.001*** -0.0022** n.s. 0.0011***Cumulated net equity issue (/TA) 0.0056* 0.0097*** 0.0066*** 0.0053* 0.0099*** 0.0065*** n.s. 0.0069*** 0.0064***E.S.I. manager n.s. n.s. n.s. Residuals of I.V.a. (! esimnger cleaned) n.s. n.s. n.s. E.S.I. consumers n.s. n.s. 0.0013** n.s. 0.0026*** 0.0022***L.T. interest rates -0.0107*** -0.0169*** -0.0051***4-Regressing change in cash / total assetsC_Profits 0.0814*** 0.0559*** 0.083*** 0.0816*** 0.0553*** 0.0827*** 0.0797*** 0.0555*** 0.0816***C_tangible -0.2657*** -0.3516*** -0.3782*** -0.268*** -0.3551*** -0.3823*** -0.2653*** -0.3581*** -0.3845***C_size 0.0876*** 0.1319*** 0.1279*** 0.088*** 0.133*** 0.1284*** 0.0886*** 0.1345*** 0.129***C_tobinsq n.s. 0.0052*** -0.0031*** n.s. 0.0053*** -0.003*** n.s. 0.0053*** -0.0032***C_non_D_TS -0.1415*** -0.0449** -0.2164*** -0.1454*** -0.0433** -0.2201*** -0.1464*** -0.0418** -0.2203***C_med_MLev -0.1038*** -0.0797*** -0.091** -0.1181*** -0.0995*** -0.1287*** -0.1141*** -0.0842*** -0.0773* MLev_to_ind 0.014*** 0.0222*** 0.0138* 0.0145*** 0.0219*** 0.0132* 0.016*** 0.0216*** 0.013* Size n.s. n.s. n.s. n.s. n.s. n.s. 0.0011* n.s. n.s. Cumul_NEI_TA 0.0302*** 0.0319*** 0.0276*** 0.0303*** 0.0316*** 0.0277*** 0.0301*** 0.0322*** 0.0279***esimger12 0.0017* n.s. 0.0031** mgerregres12 0.0027** 0.0033** 0.0068***esicsumer12 n.s. n.s. n.s. n.s. n.s. n.s. d_ltr212 -0.0053** -0.0073*** -0.0151***5-Regressing change in book-leverageC_Profits -0.1688*** -0.1478*** -0.1148*** -0.1694*** -0.1483*** -0.1144*** -0.1699*** -0.149*** -0.1155***C_tangible 0.1964*** 0.2001*** 0.1407*** 0.1891*** 0.2037*** 0.1416*** 0.1905*** 0.2056*** 0.14***C_size 0.0552*** 0.0596*** 0.0531*** 0.0552*** 0.0596*** 0.0535*** 0.0574*** 0.0597*** 0.0537***C_tobinsq 0.0025*** 0.0026*** 0.0033*** 0.0027*** 0.0027*** 0.0033*** 0.0027*** 0.0027*** 0.0033***C_non_D_TS 0.1725*** 0.0381** n.s. 0.1702*** 0.0373** n.s. 0.1806*** 0.0377** n.s. C_med_MLev 0.1473*** 0.0981*** 0.1713*** 0.1302*** 0.0895*** 0.1851*** 0.1246*** 0.072*** 0.1721***MLev_to_ind 0.0866*** 0.0801*** 0.0887*** 0.086*** 0.0808*** 0.0908*** 0.0825*** 0.0801*** 0.0899***Size -0.0021*** -0.0008* n.s. -0.0022*** -0.0008* n.s. -0.0021*** n.s. n.s. Cumul_NEI_TA -0.0195*** -0.0134*** -0.0097*** -0.0195*** -0.0136*** -0.0099*** -0.0198*** -0.014*** -0.0098***esimger12 -0.002* -0.0018** n.s. mgerregres12 n.s. n.s. n.s. esicsumer12 -0.0043*** -0.0028*** n.s. -0.0043*** -0.0035*** n.s. d_ltr212 n.s. 0.0054*** n.s. Addition-Regressing change in market-leverageC_Profits -0.1157*** -0.1112*** -0.1107*** -0.1154*** -0.1105*** -0.1097*** -0.1125*** -0.1102*** -0.1098***C_tangible 0.2243*** 0.2058*** 0.1098*** 0.2174*** 0.2097*** 0.1139*** 0.2192*** 0.2023*** 0.1125***C_size 0.0911*** 0.0929*** 0.0827*** 0.0911*** 0.0933*** 0.0831*** 0.091*** 0.0934*** 0.0826***C_tobinsq -0.015*** -0.0088*** -0.0076*** -0.0146*** -0.0084*** -0.0075*** -0.0142*** -0.0081*** -0.0074***C_non_D_TS 0.1361*** 0.1267*** n.s. 0.1436*** 0.125*** n.s. 0.1555*** 0.126*** n.s. C_med_MLev 0.4874*** 0.5673*** 0.5956*** 0.5083*** 0.5583*** 0.5956*** 0.4521*** 0.5233*** 0.5695***MLev_to_ind 0.163*** 0.1576*** 0.1796*** 0.1614*** 0.1579*** 0.1836*** 0.154*** 0.1564*** 0.1832***Size -0.004*** -0.003*** -0.0021*** -0.004*** -0.003*** -0.002*** -0.004*** -0.003*** -0.0021***Cumul_NEI_TA n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. esimger12 -0.0097*** -0.0079*** -0.0081***mgerregres12 -0.0078*** -0.0038*** -0.0035***esicsumer12 -0.0096*** -0.01*** -0.011*** -0.0108*** -0.0112*** -0.0111***d_ltr212 0.011*** 0.0111*** 0.0059** We regress by OLS the five endogenous variables. We use O.L.S. because we use differences in most of our variables. These resuts are the full resultscorresponding to EQUATIONS 2-1 to 2-3 in our method section. Eight regression results over two tables are presented, instead of 5, because of thesplit of total debt change into its two components (long term public bonds and other financial debts) and because of the reporting of results for marketleverage.

Pooling the regressions of financial policies variables on firm-, industry-specific and business cycle proxies. All regression-coefficients are reported unlike in table 4, which is an extract of this table

Adding ESI Managers only to regressions Adding ESI COnsumers only to regressions Instrumental variable approach

POOLING:

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

212

10. Bibliography  

Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: national characteristics as determinants. Journal of Banking & Finance, 33(10): 1770-1780.

Aggarwal, R., Goodell, J.W., 2010. Financial markets versus institutions in European countries: influence of culture and other national characteristics. International Business Review, 19(5): 502-520.

Antoniou, A., Guney, Y., Paudyal, K., 2006. The determinants of debt maturity structure: evidence from France, Germany and the UK. European Financial Management, 12(2): 161-194.

Antoniou, A., Guney, Y., Paudyal, K.N., 2008. The determinants of capital structure: capital market oriented versus bank oriented institutions. Journal of Financial and Quantitative Analysis, 43: 59-92.

Avramov, D., Chordia, T., 2006. Predicting stock returns. Journal of Financial Economics, 82(387-415)

Baker, M., Wurgler, J., 2002. Market timing and capital structure. Journal of Finance, 57(1): 1-32.

Bancel, F., Mittoo, U.R., 2004. Cross-country determinants of capital structure choice: a survey of European firms. Financial Management, 33(4): 103-132.

Barclay, M.J., Marx, L.M., Smith Jr., C.W., 2003. The joint determination of leverage and maturity. Journal of Corporate Finance, 9(2): 149.

Bates, T.W., Kahle, K.M., Stulz, R., 2009. Why do U.S. firms hold so much more cash than they used to? Journal of Finance, 64(5): 1985-2021.

Benmelech, E., Bergman, N., 2010. Bankruptcy and the Collateral Channel. NBER Working Paper Series w15708.

Berger, P.G., Ofek, E., Yermack, D.L., 1997. Managerial entrenchment and capital structure decisions. Journal of Finance, 52(4): 1411-1438.

Bessler, W., Drobetz, W., Pensa, P., 2008. Do managers adjust the capital structure to market value changes? Evidence from Europe. Journal of Business Economics (Zeitschrift für Betriebswirtschaft), Special Issue: 113-138.

Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovic, V., 2001. Capital structures in developing countries. Journal of Finance, 56(1): 87-130.

Brounen, D., De Jong, A., Koedijk, K., 2006. Capital structure policies in Europe: Survey evidence. Journal of Banking & Finance, 30(5): 1409.

Charreaux, G., 2004. Les théories de la gouvernance: de la gouvernance des entreprises à la gouvernance des systèmes nationaux. Cahier du FARGO, 1040101

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

213

Chen, H., 2010. Credit spreads and Capital Structure. NBER Working Paper No. w16151

Claessens, S., Djankov, S., Lang, L.H.P., 2000. The separation of ownership and control in East Asian Corporations. Journal of Financial Economics, 58(1/2): 81-112.

Claessens, S., Djankov, S., Fan, J.P.H., Lang, L.H.P., 2002. Disentangling the incentive and entrenchment effects of large shareholdings. Journal of Finance, 57(6): 2741-2771.

De Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance, 32(9): 1954-1969.

Deesomsak, R., Paudyal, K., Pescetto, G., 2004. The determinants of capital structure: evidence from the Asia Pacific region. Journal of Multinational Financial Management, 14: 387-405.

Demirguc-Kunt, A., Maksimovic, V., 1999. Institutions, financial markets, and firm debt maturity. Journal of Financial Economics, 54(3): 295-336.

Denis, D.K., McConnell, J.J., 2003. International Corporate Governance. Journal of Financial & Quantitative Analysis, 38(1): 1-36.

Dittmar, A., Mahrt-Smith, J., Servaes, H., 2003. International corporate governance and corporate cash holdings. Journal of Financial and Quantitative Analysis, 38: 111-133.

Faccio, M., Lang, L.H.P., 2002. The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3): 365-395.

Fan, P.H.J.P., Titman, S., Twite, G.J., 2008. An international comparison of capital structure and debt maturity choices. SSRN eLibrary.

Ferreira, E., Serna, M.I.M., Navarro, E., Rubio, G., 2008. Economic sentiment and yield spreads in Europe. European Financial Management, 14(2): 206-221.

Frank, M.Z., Goyal, V.K., 2008. Trade-off and pecking order theories of debt. In Eckbo, E. (Ed.), Handbook of corporate finance: empirical corporate finance (vol. 2). Elsevier: North Holland. Chapter 12.

Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: which factors are reliably important? Financial Management, 38(1): 1-37.

Garvey, G.T., Swan, P.L., 1994. The economics of corporate governance : beyond the Marshallian firm. Journal of Corporate Finance, 1(2): 139-174.

Geyer, A., Kossmeier, S., Pichler, S., 2004. Measuring Systematic Risk in EMU Government Yield Spreads. Review of Finance, 8(2): 171-197.

Giannetti, M., 2003. Do better institutions mitigate agency problems? Evidence from corporate finance choices. Journal of Financial and Quantitative Analysis, 38(1): 185-212.

Gomes, J.F., Schmid, L., 2010. Levered Returns. Journal of Finance, 65(2): 467–494.

Gonzalez, V., Gonzalez, F., 2008. Influence of bank concentration and institutions on capital structure: new international evidence. Journal of Corporate Finance, 14(4): 363-375.

Graham, J., C. Harvey, 2001. The Theory and Practice of Corporate Finance: Evidence from the Field. Journal of Financial Economics, 60: 187-243.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

214

Graham, J.R., 1996. Debt and the marginal tax rate. Journal of Financial Economics, 41(1): 41-73.

Hackbarth, D., Miao, J., Morellec, E., 2006. Capital structure, credit risk, and macroeconomic conditions. Journal of Financial Economics, 82: 519-550.

Hall, G., Hutchinson, P., Michaelas, N., 2004. Determinants of the capital structures of European SMEs. Journal of Business Finance and Accounting, 31: 711-728.

Hambrick, D.C., Mason, P.A., 1984. Upper-echelons: the organization as a reflection of its top managers. Academy of Management Review, 9(2): 193-206.

Harford, J., Mansi, S.A., Maxwell, W.F., 2008. Corporate governance and firm cash holdings in the US. Journal of Financial Economics, 87(3): 535-555.

Hofstede, G., 2003. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

Holderness, C.G., 2010. The myth of diffuse ownership in the United Staes. Review of Financial Studies, 22(4): 1377-1408.

Jensen, M.C., 1989. Eclipse of the public corporation. Harvard Business Review, 67(5): 61-74.

Jensen, M.C., Meckling, W.H., 1976. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, Vol. 3(No. 4): 305 - 360.

Jimenez, G., Salas, V., Saurina, J., 2006. Determinants of collateral. Journal of Financial Economics, 81(2): 255-281.

King, M.R., Santor, E., 2008. Family values: ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, (32): 2423-2432.

Korajczyk, R.A., Levy, A., 2003. Capital structure choice: macroeconomic conditions and financial constraints. Journal of Financial Economics, 68(1): 75-109.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance, 54(2): 471-517.

Lenartowicz, T., Roth, K., 1999. A Framework for culture assessment. Journal of International Business Studies, 30(4): 781-798.

Levy, A., Hennessy, C., 2007. Why does capital structure choice vary with macroeconomic conditions? Journal of Monetary Economics, 54(6): 1545-1564.

Mahajan, A., Tartaroglu, S., 2008. Equity market timing and capital structure: international evidence. Journal of Banking & Finance, 32: 754-766.

Margaritis, D., Psillaki, M., 2009. Capital structure, equity ownership and firm performance. Journal of Banking & Finance, 34: 621–632.

Myers, S.C., 2003. Financing of corporations. In Constantinides, G., Harris, M., and Stulz, R. (Ed.), Handbook of the Economics of Finance: Corporate Finance. Elsevier North Holland

Rajan, R.G., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data. Journal of Finance, 50(5): 1421-1460.

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

215

Reuter, C.H.J., 2009. European Capital Structures: Managerial Insight and Governance Regimes available at SSRN n°1538152.

Ronen, S., Shenkar, O., 1985. Clustering countries on attitudinal dimensions: A review and synthesis. The Academy of Management Review, 10(3): 435.

Ruiz-Mallorqui, M.V., Santana-Martin, D.J., 2010. Dominant institutional owners and firm value. Journal of Banking and Finance, in press

Sarig, O.H., Talmor, E., 1997. In defense of defensive measures. Journal of Corporate Finance, 3(3): 277-297.

Schwartz, S.H., 1999. A Theory of Cultural Values and Some Implications for Work. Applied Psychology: An International Review, 48(1): 23-47.

Shleifer, A., Vishny, R., 1986. Large shareholders and corporate control. Journal of Political Economy, 44: 461-488.

Shleifer, A., Vishny, R.W., 1989. Management entrenchment: the Case of managers specific investments Journal of Financial Economics, 25(1): 123-139.

Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance, 52(2): 737-783.

Wanzenried, G., 2006. Capital structure dynamics in the UK and continental Europe. European Journal of Finance, 12(8): 693–716.

Westbrook, D.A., 2009. Out of crisis: rethinking our financial market. Paradigm. 176 pages.

Zingales, L., 2000. In Search of New Foundations. Journal of Finance, 55(4): 1623.

* * *

Chapter III - Do agency relations mediate the interaction, between firms’ financial policies, and business cycles?

216

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

217

FOURTH CHAPTER:

IS THE RELATION BETWEEN BUSINESS CYCLES

AND LEVERAGE MEDIATED BY STRUCTURAL, INSTITUTIONAL OR CULTURAL FACTORS?

DETAILED TABLE OF CONTENTS

1.   Introduction 219  2.   Literature review and hypotheses 222  

2.1.   Specific dimensions underlying our methodological proceedings 222  2.1.1.   Sampling  on  a  regional  basis   222  2.1.2.   Broader  institutional  antecedents  to  capital  structures   223  2.1.3.   Specific  sources  of  managerial  behavior   224  

2.2.   The broader economic and financial context 225  2.2.1.   Stock  market  conditions   226  2.2.2.   Economic  cycles  and  economic  confidence   228  

2.3.   Culture 230  2.3.1.   Direct  influences   230  2.3.2.   Direct  and  indirect  influences   231  2.3.3.   Embededdness   232  

3.   Sampling, data and method 233  3.1.   Sampling 233  3.2.   Data 233  3.3.   Variables 234  3.4.   Models 237  3.5.   Method-selection 237  3.6.   Controlling for multi-colinearity in context-variables 238  3.7.   Pooling 239  

4.   Results. 240  4.1.   Results’ overview – results restricted to the core-variables 240  4.1.1.   Core  firm-­specific  variables:  limited  contextual  mediation   240  4.1.2.   Dividend  policy  and  leverage:  any  dividend  irrelevance?   241  4.1.3.   Cumulative  net  equity  return:  which  firms  time  the  market?   241  

4.2.   Market timing and re-balancing asset classes 242  4.2.1.   Timing  both  stock  and  debt  markets?   242  4.2.2.   Shareholders’  rights   244  4.2.3.   Creditors’  rights   245  

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

218

4.3.   Pro-, vs. contra-cyclicality in leverage. Sources of managers’ behaviors and resulting capital structures 246  

4.3.1.   Ownership  dispersion:  pro  vs.  contra-­cyclicality   249  4.3.2.   Investors’  rights   249  4.3.3.   The  structural  and  institutional  context   250  4.3.4.   Risk   251  

4.4.   Cultural factors, leverage, market timing and cyclicality in financial policies 252  4.4.1.   A  direct  influence  of  culture  on  leverage?   252  4.4.2.   Culture  and  business  cycle  sensitivity  of  leverage   254  

5.   Conclusion 256  6.   Appendix 1 (chapter IV): data collection. Synthesis from Reuter (2009) 258  7.   Appendix 2 (chapter IV): Pair wise correlations 260  8.   Appendix 3 (chapter IV): Unpooled results: robustness of context-variables’ signs and coefficients 261  9.   Appendix 4 (chapter IV): Results for pooled regressions. 262  

Displaying only signs and significance 263  10.   Appendix 5 (chapter IV): Results pooled by country 264  11.   Bibliography 265 * * *

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

219

Is the relation between business cycles and leverage mediated by structural,

institutional or cultural factors?1

Institutions … are embedded in various types of repositories or “carriers” […there are…] three types of carriers:… cultures, social structures, and routines. Scott (1995: 52) [To] investigate the institutional diversity and the complexity of economies [you need to look] into the nature of the interdependencies of institutions across economic, political, organizational and social domains, as well as that of those institutions linking those domains. Aoki (2001: 3)

1. Introduction  

The investigation of capital structure is one of the most comprehensively studied subjects in

corporate finance. Reviews abound (recently Frank and Goyal, 2008 or Barclay and Smith,

2005). A consensual conclusion is that the capital structure puzzle is not yet resolved (Myers,

2001). Part of the puzzle is brought about by the endogeneity of some variables (Barclay and

Smith, 2005), and another one by the long-lasting effects of market timing (Baker and

Wurgler, 2002), while the stark development of international empirical evidence has brought

forward questioning about potential U.S. specificity in a worldwide perspective (De Jong, et

al., 2008; Frank and Goyal, 2009; King and Santor, 2008; Margaritis and Psillaki, 2009).

In our previous chapter (chapter III) we show that, focusing on European data, the relation

between business cycles and financial policies is mediated by ownership dispersion. In

1 This article results from a complete revision of Reuter (2009), an article presented in spring of 2010, at the following academic conferences: A.I.B., E.F.M.A., C.I.G., the Groupe de Recherche of the C.N.R.S. “Monnaie Banque Finance”, and INFINITI, where it has benefited from commentaries by reviewers and participants. It had been presented in the following research-laboratories as well: CEROS, ESCP and FARGO. The current version is largely differing from the work that was presented, and received commentaries.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

220

European firms with dispersed ownership, a moderate pro-cyclicality of debt-, investment-,

dividend- and cash-policies is evidenced, as in the U.S. (Frank and Goyal, 2009). However in

European firms with concentrated ownership, financial policies are contra-cyclical. More

specifically debt- and dividend policies are contra-cyclical, with no cyclicality in investment

policies. Our specific argument in chapter III is hat this edge is driven by differences in

agency costs, in the benefits of managerial discretion, in incentive alignments, or still in risk

perception and risk-management.

Baker and Wurgler (2002) conclude their study with the note that capital structure can reflect

“the cumulative outcome of past attempts to time the equity market” (Baker and Wurgler:

p. 1). In a related way, our findings in the previous chapter may suggest, that net marginal

cumulative adjustments in capital structures, as triggered by business cycles, could have

lasting effects. However, while market timing has a one-sided cumulative effect on capital

structures, the two underlying and polarized relations between anticipated business cycles,

and capital structures, are more ambiguous, because of the potential effects of ownership

dispersion.

We would like to complement and confront our previous findings. While we confirm the

cumulative pro- and contra-cyclical effects of anticipated business cycles, we show that this

relation can be obtained, not only from ownership dispersion, but also from a very large range

of contextual factors. Economic, institutional, cultural or legal dimensions mediate the

relation between business cycles and leverage simultaneously. This includes, on top of

ownership structures, the quality of the contracting environment, measures of national

cultures, institutional anchoring, and firms’ specific measures, in particular about

transparency and risk.

So, to explain the mediated interaction between business cycles and capital structures, other

complementary mechanisms could be at play, or there could be a more complex causality.

This is consistent with findings by De Jong, et al. (2008), who emphasize that the national

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

221

context can have a double effect on capital structures, directly through country-specific factors

and indirectly through the effect that these country-specific factors have on firm-specific

variables. Further this underlines the inherent complexity in international research. Recently

Stulz (2007) notes that financial scholars should scrutinize more actively what he describes as

a “country puzzle”. In new institutional theorizing, from economics to organization theory the

idea that national systems are structured robustly, with an institutional alignment across

varying levels has been favored for some time now (Aoki, 2001; Greif, 2006; Guillen, 2005;

Williamson, 2000).

In International Business, Leung, Bhagat, Buchan, Erez and Gibson (2005) emphasize the

intertwining between cultural and socio-economic-political [sic] variables, and they stress the

need to untangle these forces. We address these methodological concerns and we illustrate

how contextual dimensions converge to affect capital structures in a specific international

setting.

Additionally, we focus on the relation between managers’ information (on business cycles),

their behaviors and the consequences for firms’ leverage ratios. The informational advantage,

possessed by the manager, is a key dimension in most capital structure theories. How the

manager uses this informational advantage, and with what consequences, varies from one

theory to the next. The financial theory generally rests on motives, taking the relationship

between motives, action and results as unproblematic. One of our chief arguments is that the

context mediates the relationship between manager’s information, his action (financial

policies) and observed results (capital structures). Here we address Earley (2006)’s concern

that “meaning” mediates the relationship between “values”, motives, and “behaviors”, which

implies a need more refined views of causality, and more attention to the context.

In the next section we review the literature and derive our hypotheses. Section 3 is allocated

to data and methods. In section 4, we present our results, and in section 5 we conclude.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

222

2. Literature  review  and  hypotheses  

First we stress a number of specificities in our approach, relating to our sampling procedure,

to the past investigation of institutional antecedents to capital structures, and to the

importance of specifically considering the connection between manager’s motives,

information, decisions and actions.

Second and third, we develop our hypothesizing in relation to economic cycles (section 2.2)

and to culture (section 2.3).

2.1. Specific  dimensions  underlying  our  methodological  proceedings  

2.1.1. Sampling on a regional basis

In this article we study the determinants of leverage, on a large sample of listed firms,

operating across Europe, and we focus on elements of the broader2 institutional context.

While limiting variance in terms of economic, judicial, accounting or institutional

development, this sampling magnifies qualitative differences in institutions as well aw some

aspects of socio-economic systems. Furthermore, it provides for some commonality in the

socio-political, historical or legal context, while it still exhibits some variance in terms of

national cultures (as identified for instance through national culture indices, see below).

Additionally, we note that meaningful regional grouping has been favored for a long time in

International Business (Hofstede, 2003; Ronen and Shenkar, 1985), and we note that this

process circumscribes some of the methodological difficulties posed by broad cross-national

index comparisons (Lenartowicz and Roth, 1999 ; Sivakumar and Nakata, 2001).

2 We use this expression broader institutional context to designate national institutions in their most encompassing definition, including cultural, legal, socio-economic and institutional (in a then narrower definition) dimensions. We acknowledge that this is a simplification, it is aimed at simplifying writing and reading.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

223

2.1.2. Broader institutional antecedents to capital structures

Investigating capital structure with international samples, Rajan and Zingales (1995) note that

much of the empirical variance can be attributed to diverging accounting norms. Chui, et al.

(2002) identify enduring cultural antecedents (mastery and conservatism, related to lower use

of debt). Demirgüç-Kunt and Maksimovic (1999) emphasize the primary importance of

economic development, partly because it drives institutional differentiation, an other key

dimension. Giannetti (2003) suggests that different characteristics, in the national populations

of firms, such as average firm-size, hinder comparisons. Gonzales and Gonzales (2008)

acknowledge the role of macro-financial structures (the relative development of the banking

vs. stock markets). They suggest that structural economic factors, such as the degree of

national banking competition, are important.

Our purpose, in this article, is not to emphasize that such or such dimension, in the broader

institutional context, can be identified as a definite antecedent for capital structures. Rather we

wish to emphasize that there is convergence across a range of contextual factors, structural,

institutional, cultural, which produces differentiation in capital structures, at national levels,

provided our “regional grouping”. To this aim we systematically exploit a specific

methodological approach, the pooling of our sample3. It enables us to emphasize that the

polarity observed in our previous chapter, in the economic cyclicality of capital structures, can

be derived from a range of contextual factors, and not only from the dispersion of ownerhsip.

Past investigation of the relation between business cycles and capital structure is limited: in

traditional financial theorizing, generally based on the “U.S. role model”4, and on dispersed

3 This approach has not been extensively implemented in financial research, potentially due the limited attention, in financial theorizing, for the broader institutional context, as discussed in our additional essay in the annex of the dissertation 4 We owe this expression to De Jong et al. (2006: p. 1956) who note “the conventional theories on capital structure [were] developed using listed firms in the United States as a role model...”, and they call for evidence derived from more diverse sampling, echoing similar remarks in Zingales (2000: p. 1624), Frank and Goyal (2009). Besides, La Porta et al. (1999) show that dispersed ownership is less prevalent worldwide than previously thought, while Holderness (2010) does so for the U.S.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

224

ownership, the anticipation of business cycles is, at best, of limited relevance for the

understanding of capital structures. A primary reason is that debt serves as a disciplining

device (Berger, et al., 1997), or as a signaling device, in a context where agency costs

between managers and shareholders are a primary concern. Another reason can be traced to

the potential importance of stock-market conditions (Baker and Wurgler, 2002; Frank and

Goyal, 2004), which supersede broader economic conditions. This limited, or secondary,

relevance of business cycles for the determination of capital structures does not need to hold

in broader institutional contexts. In fact we show in our previous chapter that contrary

behaviors at the micro-economic level can obscure the analysis of the determinants of

financial policies in aggregate.

2.1.3. Specific sources of managerial behavior

In their survey of capital structure theories, Frank and Goyal (2008: p. 21) note: “no-one has

tried to distinguish among alternative sources of pecking order behaviors [emphasis added]”.

They find this surprising and they emphasize that most capital structure theories are

established on some underlying theoretical conception about managers’ motives and

behaviors, ab abstracto. They note in particular that adverse selection and agency theories are

based on the idea that “the owner manager of the firm knows the true value of the firm’s

assets and growth opportunities. Outside investors can only guess these values” (Ibid, p. 21).

Signaling and market timing theories are based on comparable premises. Similarly, trade-off

theories are based on some given assessment of the opportunity cost of bankruptcy, which is

better known to the manager. Besides, it is different for different stakeholders (such as

creditors and shareholders), and it does not need to be identical for all shareholders; so

Margaritis and Psillaki (2009) note that institutional investors, with block holdings, face

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

225

different sets of constraints in the U.S. vs. Europe, and European institutional investors have,

more often than in the U.S., less liquid or longer-time horizon holdings, implying different

perceptions of opportunity cost of bankruptcy.

Our investigation of context is addressing, in a specific way, the need for emphasis on specific

sources of managers’ behaviors. The broader institutional context is what provides meaning

to the connection between the managerial behavior, and its consequences (with reference to

Earley, 2006, above). Doing so, we try to emphasize varying sources, institutional, cultural,

structural, legal, etc. that foster different types of relations between manager’s specific

information and the consequences of managers’ behaviors. Addressing Leung, et al. (2005)’s

call for a greater attention to overlapping sources of contextual effects, we intend to show that

different aspects of the broader institutional context can mediate simultaneously a given

relationship. Our emphasis is not on the identification of a precise causality, but on attaching

attention to the complexity of these combined contextual influences.

2.2. The  broader  economic  and  financial  context  

In their empirical search for robust determinants of leverage in the U.S., Frank and Goyal

(2009) investigate a number of contextual variables including inflation, T-Bills rates, term-

spreads on government bonds, purchasing managers’ confidence and macro-economic

growth. They note a very high level of pair wise correlation among some of these variables, in

particular among the two former and among the three latter ones. It is noteworthy that these

correlations may be different in our pan-European sample, because of the presence different

governments, different policies, and different sets of reference rates. In fact, we note that in

some cases pair wise correlation across countries, for a given set of variables, dominates,

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

226

while in some others, the pair wise correlation across sets of variables, in a given country

dominates.

In their empirical study, Frank and Goyal (2009) examine the contextual variables, and they

eventually retain expected inflation as part of the six core determinants of leverage (which

additionally include profitability, tangibility, size, Tobin’s Q and the median industry

leverage). They propose theoretically three kinds of mechanisms to explain the robustness of

expected inflation as a core determinant, one related to stock market conditions, one related to

debt market conditions and the third related to macro-economic growth per se. We review

them in turn, deriving our hypotheses, starting with the stock market-related variable, and

then considering interests rates and growth in combination.

2.2.1. Stock market conditions

Frank and Goyal (ibid) suggest a negative relation between stock market conditions and

leverage ratios. They explain that this suggestion can be obtained from a range of theories,

which have consistent predictions. They mention in particular market timing, adverse

selection and the fact that firms do not rebalance their leverage, following temporary stock

price changes (they label this third explanation the relative price of asset classes).

When large block-holders have dominant block-holdings, and household’s direct investment

in shares is limited (as in Europe, Guiso, et al., 2001) smaller investors may monitor and

follow larger ones. This limits the potential for mispricing, and hence this limits the potential

for effective market timing. A similar argument prevails for adverse selection. As a result we

expect the effect of stock market conditions should be mitigated with firms with concentrated

ownership.

H1a (first part). The aggregate relation for (stock) market timing is mediated by ownership

dispersion. The relation holds more strongly, or only, for dispersed ownership.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

227

Furthermore, we suggest that it could be related to the legal context. Weak credit-holder rights

and strong shareholder rights should be conducive to a stronger relation between market-

timing and capital structure. First, with stronger governance (and more accurate valuation on

average), the marginal effect of market timing could be stronger in aggregate (less noise), and

second, the marginal utility of market timing, for ongoing shareholders, managers, and credit

holders, is larger because this is one of the very few possibilities to exploit a positional

advantage. So:

H1a (second part). The aggregate relation for (stock) market timing is mediated by the quality

of investors’ rights. Specifically the relation holds more strongly, or only, for countries with

stronger shareholder rights and for countries with weaker creditor rights.

Overall: H1a. The aggregate relation for (stock) market timing is mediated by ownership

dispersion, and the quality of shareholder- as well as of creditor rights. Specifically the

relation holds more strongly, or only for dispersed ownership, for countries with strong

shareholder rights and for countries with weak creditor rights.

Furthermore, concentrated ownership is often associated with bank-based macro-financial

systems. In these systems, and in particular in continental Europe, strong credit-holders are

often associated as shareholders as well (Shleifer and Vishny, 1997). These credit- and

shareholders may have an inventive to balance the issue of new capital with additional credit,

from which they derive benefits. Similarly, they may take advantage of increases in stock

market prices to suggest and secure new lending. A necessary condition for this behavior is

that credit holders are well protected. Here Welch (2004)’s argument that firms do not re-

balance capital structure following changes in stock prices becomes inadequate, because

lenders may be willing and able to secure additional lending (positive outlook), or refuse to

renew credit lines to reflect the deteriorating conditions. So the argument relating to relative

prices of asset classes is reversed.

H1b. The aggregate relation observed for the importance of stock market timing should be

reversed in countries with strong creditor rights.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

228

2.2.2. Economic cycles and economic confidence

Frank and Goyal (2009) note that leverage ratios are expected to be pro-cyclical for a large

number of reasons, including a relation from business booms to respectively a decrease in

stock prices, a decrease in bankruptcy costs, an increase in collateral values, an increase in

taxable income, an increase in cash.

However they mention: “agency problems are likely to be more severe during downturns as

manager’s wealth is reduced relative to that of shareholders. If debt aligns managers

incentives with those of shareholders, leverage should be counter-cyclical” (Ibid, p. 17).

Our previous chapter investigates the nature of European financial policies in relation to

expected business cycles. Theoretically we rely on the idea that strong ownership

concentration vs. strong dispersion changes the nature of agency relations, reducing agency

costs, reducing the need for disciplining (through debt), changing the nature of monitoring,

decreasing the opportunity costs of entrenchment while increasing the opportunity benefits of

managerial discretion, limiting the need for signaling…

Empirically we show that ownership dispersion in Europe polarizes financial policies: firms

with dispersed ownership exhibit pro-cyclicality of investment, debt, dividend and cash

policies with respects to expected business cycles. Firms with concentrated ownership are

characterized by contra-cyclicality for debt and dividend policies, while investment does not

exhibit any relation to anticipated business cycles.

These results are consistent with Frank and Goyal’s argument on aligning incentives between

managers and shareholders. As a consequence we expect a generalization of our previous

findings from financial policies to observed leverage ratios:

H2a. There is a pro-cyclical [2ai] (contra-cyclical [2aii]) relation between leverage and

economic confidence measures for firms with dispersed (concentrated) ownership.

The theoretical arguments rest on the principle that the pro-cyclicality of debt and investment

are needed as a monitoring device (dispersed ownership companies need to be lean), while it

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

229

is easier to align the interests of some of the shareholders’ with those of managers’ in

concentrated ownership. As a result, we expect that both relations obtained in our previous

chapter should be exacerbated by the quality of shareholder rights, with a stronger pro-

cyclicality in countries with good shareholder protection and a stronger contra-cyclicality in

countries with weak shareholder protection.

H2b(i). The pro and contra-cyclical relations between leverage and economic confidence are

mediated by the quality of shareholders’ rights with a stronger pro-cyclical (contra-cyclical)

relation in countries with stronger (weaker) rights.

H2b(ii). The pro and contra-cyclical relations between leverage and economic confidence is

mediated by the quality of creditors’ rights with a stronger pro-cyclical (contra-cyclical)

relation in countries with weaker (stronger) rights.

When the firm becomes larger, more prominent and more complex, it becomes more

intensively and more accurately scrutinized, so other actors moderate the importance of

manager’s specific information. Furthermore, this general increase in prominence and

complexity should exacerbate the intensity of monitoring, increasing the pro-cyclicality of the

lean companies while moderating contra-cyclical management.

H2c(i). The pro-cyclical relation between leverage and economic confidence is exacerbated

for large firms, for firms operating in multiple businesses, for firms with multiple listings and

for older firms.

H2c(ii). The contra-cyclical relation between leverage and economic confidence is

exacerbated for smaller firms, for firms operating in a single business, listed in a given

country, and younger firms.

When a firm operates in a riskier environment, it becomes less transparent, this should limit

the potential for efficient monitoring, emphasizing the relative importance of incentives’

alignment. As a result:

H2d. The pro-cyclical relation should be weakened in riskier firms while the contra-cyclical

relation should be unchanged.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

230

2.3. Culture  

The relation between culture and the broader institutional context is a complex one in general,

and it is in relation to capital structures.

2.3.1. Direct influences

Chui, et al. (2002) propose a theoretical model explaining a connection between two of

Schwartz’s national culture indices and leverage [see chapter II for details and definitions]:

mastery and conservatism should be related to lower use of debt (Chui, et al., 2002: figure 2,

p. 103). Conservatism is related to firm’s use of debt over three mechanisms, the need to

preserve a public image, the need to foster harmonious working relations, and the need for

security and conformity. Mastery is related to firm’s use of debt over two mechanisms, a

greater on internal locus of control and a greater emphasis on individual success.

However Schwartz’s index is not available for 6 of our 15 countries, so we have to rely on a

proxy. We can rely on Hofstede indices, which are available for our sample, and highly

correlated with Schwartz indices (Schwartz, 1994, Table 7.2 p. 109. Schwartz himself

acknowledges much empirical overlap between the measures he proposed and those of

Hofstede). In particular we note strong and consistent pair wise correlation of the two

theoretically relevant Schwartz indices and three Hofstede indices: conservatism and mastery

exhibit a consistent pair wise correlation with individualism (respectively -56% and -24%),

power distance (respectively 45% and 26%) and masculinity (respectively -2% -to be

discarded- and 56%). Only Uncertainty avoidance exhibits an ambiguous relationship with the

two predictors of leverages (respectively -25% and 25%). As a result we retain the two other

indices as proxies:

H3a. Countries with respectively lower individualism (i), higher power distance (ii) and

higher masculinity (iii) should be characterized by lower leverage.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

231

2.3.2. Direct and indirect influences

There is evidence that culture has both a direct and an indirect influence on varying financial

processes. So De Jong, et al. (2008) emphasize that the national context can have a double

effect on capital structures, directly through country-specific factors and indirectly through the

effect that country-specific factors have on firm-specific variables. A specific and prominent

direct and indirect influence of culture, on financial phenomena is that of law. Stulz and

Williamson (2003), as well as Licht, Goldschmidt and Schwartz (2005), show that national

cultures are a robust, stable and enduring antecedent to financial regulation. We are already

testing these variables in search for direct influences.

Another important direct and indirect influence, which culture has on financial phenomena, is

running through macro-financial structures. Kwok and Tadesse (2006) emphasize that

uncertainty avoidance [chapter II, section 2] is an important determinant of the macro-

financial structures, and countries that have a higher index on uncertainty avoidance tend to

rely more strongly on banking vs. market finance, controlling for financial development.

More recently Aggarwal and Goodell (2009) extend this finding and show that this direct

cultural influence is robust to the inclusion of a very broad range of complementary

economic, financial, institutional and legal variables. Provided the comments made above,

and in the previous chapter, relating to the interrelation between bank-based macro-financial

structures and contra-cyclical policies, we expect that uncertainty avoidance should have an

influence on capital structures. Specifically:

H3b(i). Countries characterized with high uncertainty avoidance should have a stronger

(weaker) contra- (pro-) cyclical relation between business cycles proxies and leverage ratios.

H3b(ii). The reverse should hold for countries with low uncertainty avoidance.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

232

2.3.3. Embededdness

There is evidence of serial correlation of national culture indices with varieties of capitalism

(Schwartz, 2007), with the rule of law, corruption and democratic accountability (Licht,

Goldschmidt and Schwartz, 2005), with economic, demographic and political structures

(Schwartz, 2004) as well as with many other cultural variables (De Jong, 2009)… So, beyond

the general direct and indirect influence that culture can have on a variety of socio-economic,

institutional and political dimensions, it provides a framework that fosters and interacts with

variety of other national dimensions to produce a path-depend system with relative autonomy

and consistency (Aoki, 2001; Hall and Gingerich, 2001; Hofstede, 2003; Schwartz, 2004 [see

in particular sections I.2.3, I.5, II.4.2.1, II.5.3]).

Further, and relating to our topic, it seems that bank-based macro-financing systems and

concentrated ownership exhibits some consistency in their theoretical properties (Carlin and

Mayer, 1999, 2000). So, relying on the conceptual frame of consistent “world-views” (in

particular that opposing “Anglo-Saxon” with other world views: Stulz and Williamson, 2003

see section II.4.2.1) and of the “U.S. Role Model” (citation above in footnote 4), we predict:

H3c. The general pro-cyclicality (contra-cyclicality) of leverage will be observed in countries

that have similar (opposite) positioning on national culture indices, as the U.S., represented

by respectively a high, high, low and low positioning on individualism, masculinity, power

distance and uncertainty avoidance.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

233

3. Sampling,  data  and  method  

3.1. Sampling  

Provided the limited number of studies with the same specific angle (see sections II.1 & III.2),

we intend to provide a database as comprehensive and as complete as possible. One advantage

of this sampling procedure, in line with traditional sampling procedures for qualitative

studies, is to limit variation on a number of dimensions to maximize observable effects on

another. We specifically use the same database as in the previous chapter [chapter III], but we

do not differentiate the data. This enables us to investigate the determinants of leverage in

listed and liquid firms, in a context where the large qualitative institutional variation is

amplified by a limited variation in economic, legal or accounting development. To rephrase,

while this provides us with relatively converging quality standards in judicial, accounting, or

democratic terms, including for instance the existence of a free and lively business press in

most countries, this allows for a large range of rules, in particular legal origins, protection of

creditors, investors, etc. to vary.

3.2. Data  

As emphasized, the database is identical to the one used in the previous chapter, and it is best

described in Reuter (2009). We provide a synthesis in appendix 1.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

234

3.3. Variables  

The table 4 in chapter III provides details on the variables. The table 11 below provides the

statistics on non-differentiated variables used in this chapter and the table 15 in appendix 2

provides the pair wise correlations.

For the most part, our definitions follow those from recent contributions on the study of

capital structures by Antoniou, Guney and Paudyal (2008), Fan, Titman and Twite (2008) or

Frank and Goyal (2009), who provide mostly converging definitions. Whenever appropriate,

we test variables with different definition for robustness.

Variable Obs Mean Std. Dev. Min MaxBook leverage 34 886 22.3 18.1 0 521Market leverage 34 886 26.2 22.5 0 99.3Median book leverage 29 060 0.191 0.046 0.080 0.269Median market leverage 29 060 0.204 0.068 0.041 0.357Profits 38 102 0.107 0.176 -4.35 3.43Tangibility 38 102 0.284 0.212 0 2.560Size 38 102 5.231 2.001 -2.814 12.53Tobin's Q 38 102 1.687 1.526 0.017 41.03Dividend payout 38 099 28.0 31.6 0 100Cumulative net equity c. 38 102 0.092 1.052 -42.74 11.85Risk 38 102 0.491 0.500 0 1managers' confidence 38 102 101.4 8.2 76.4 124.9managers' confidence (I.V.) 34 896 -0.016 6.186 -16.3 16.0Consumers' confidence 37 320 -7.904 12.312 -45.5 25.65Broad stock market index 38 102 684.7 806.8 8.5 5710Inflation 36 810 2.394 1.680 -0.7 17.9Long-term interest rate 36 783 5.667 2.297 3.33 24.01Long-term interest rate (I.V.) 36 718 0.361 2.161 -3.72 20.37Shareholder's rights dummy 38 102 0.579 0.494 0 1Creditors' rights index 38 102 2.196 0.869 1 3Individualism index 38 102 71.8 14.9 27 89Power distance index 38 102 43.6 15.5 11 68Uncertainty avoidance index 38 102 61.5 25.8 23 112Masculinity index 38 102 52.1 20.0 5 79Ownership dispersion dummy 38 102 1.001 0.757 0 2Firms' size dummy 38 102 0.500 0.500 0 1Firms' profitability dummy 38 102 0.530 0.499 0 1Firm's complexity dummy 38 102 0.397 0.489 0 1Firm's listings length dummy 38 102 0.864 0.343 0 1Multiple listing dummy 38 102 0.815 0.388 0 1Individualism dummy 38 102 0.646 0.478 0 1Power distance dummy 38 102 0.400 0.490 0 1Uncertainty avoidance dummy 38 102 0.382 0.486 0 1Masculinity dummy 38 102 0.632 0.482 0 1

Table 11. Variables: statistics (capital structure and broader institutional context)

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

235

Our endogenous variables are standard for the literature and include book- and market

leverages, based on a definition of debt as including short term and long-term financial debts.

Table 4 provides more specificity, while Reuter (2009) provides details about alternatives

considered and used in robustness tests.

Our exogenous variables fall broadly into two main categories. On the one hand, we include

firm-specific variables as identified by the previous literature. As noted upfront, the testing of

varying theoretical frameworks has not yet brought definite evidence, so empirically

grounded approaches have regained favor. Frank and Goyal (2009) establish, in the U.S., over

decades, empirical regularities between firm-specific variables and measures of leverage.

They identify six robust “core factors” (four firm-specific factors), across industries and

across time (profitability, tangibility, size, Tobin’s Q, median industry leverage and expected

inflation)5.

In addition to including the four firm-specific core factors (plus one, the median industry

leverage), we include two other firm-specific variables that provide further details about the

financial policies: the yearly dividend payout ratio and a variable introduced by Baker and

Wurgler (2002: labeled EFWAMB in their original study), the cumulative total net equity

issue6. These variables will turn out useful because the dividend policy is a key dimension of

the contra-cyclicality of financial policies identified in our previous chapter.

On the other hand, we add a large number of contextual variables that fall mainly into five

categories, which we label as the broader economic, structural, legal, institutional or cultural

contexts. Most of these variables are standard in the financial literature. The broader

economic variables include a number of macro-economic variables in relation to business

5 Noteworthy is the fact, however, that Antoniou, Guney and Paudyal (2008: p. 89) identify and list in appendix a large number of studies where the correlation’s signs between leverage and a dozen of firm-specific, or financial, variables (including the 6 core factors) is changing depending on sampling procedures. 6 For technical reason, we do not extend our database beyond either 1990 (starting year) for companies that were already listed at the beginning of our sampling. So our resulting variable is censored as compared to that of Baker and Wurgler (2002)

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

236

economic and financial cycles, including in particular measures of economic sentiment for

managers and for consumers (they are produced and provided by the European Union, see

appendix 2 for details), measure of stock prices, based on broad national indices, measures of

inflation and measures of long-term interest rates mainly based on the 10-year government

bonds for each individual European country included in the sample. Because of the strong pair

wise correlation of some of these variables, we propose two instrumental variables

approaches, one where we regress managers’ economic sentiment on consumers’ economic

sentiment and broad stock market indices, and the other where we regress inflation on long-

term interest rates. In both case we obtain the residuals of the regressions, which are

subsequently implemented in multivariate analyses where several economic variables are

included at once.

Measures of the structural or institutional context include mostly measures based on firm-

specific characteristics. Thses include the measure of ownership dispersion used in our

previous chapter and dummies for the following elements: whether the firm has one national

listing or multiple listings (a dummy), whether the firm has one business line, based on 2-

digits SIC-codes, or more than one (a dummy), whether the company has a total listing-life

below or above the sample’s median (a dummy). In subsequent methodological proceedings,

the sample is pooled based on firm’s profitability and risk, i.e. whether it has average risk and

profitability above or below the sample’s median.

The legal variables include the well-known shareholders’ and creditors’ right indices from

La Porta, Lopez de Silanes, Shleifer and Vishny (1997, 1998); the cultural variables include

the four Hofstede indices. For these six variables, they are included as an exogenous variable

(using the value of the index) in extended regressions, and they are included as a pooling

variable, based on a dummy that classify the firm’s country depending on whether it has an

index value for any of the six variables above or below the country-sample’s average.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

237

3.4. Models  

Following among others Frank and Goyal (2009), we include a lag of one year between our

endogenous and exogenous variables. Our basic specification is:

Leveragei(t+1)=α+β1FSit+β2MILkt+β3EFVjt+β4CVj+β5SVi+β6LDit+D1t+D2k+Dj+ε it (Eq. 1)

Where Leverage stands for either Market or Book leverage, i, j, k and t stand respectively for firms, countries,

industries and years; D1to3 are dummies for these but firms. FS is the set of firm-specific variables, including the

‘core’ variables, dividend payout and net cumulative equity creation. MIL is the median industry leverage. EFV

is the set of five economic and financial variables, proxying for different aspects of business cycles, including

the two instrumental variables. CV is the six country-specific variables relating to the legal and cultural context

(non time-varying). SV are the structural variables (listing-life and business complexity) and LD is a single-

listing dummy. The error is not necessarily presumed normal (see below).

3.5. Method-­selection  

To select the appropriate method we conduct a number of preliminary regressions with for

method selection is:

Leveragei(t+1)=α+β1FSit+β2MILkt+D1t+D2k+Dj +ε it

We start with OLS regressions, with clustered-robust standard errors for countries, industries

or time. Then we implement a range of linear regressions for panel data with fixed effect or

random effects. A Breusch-Pagan test, with a P-value of 0,0007 for a Chi-square distribution

with 2 degrees of freedom, confirms that we should use fixed effects rather than pooled OLS.

Secondly, a Hausman test, with a P-value of 0,0000… for a Chi-square distribution with 2

degrees of freedom, shows that there is a systematic difference between the random effects

and the fixed effects and confirms that we should use panel-data analysis with random effects.

Hence, we subsequently use panel data analysis with random effects controlling further for

year, industry and country effects by the inclusion of dummies in all regressions7.

7 A number of authors mentioned in this article come to similar conclusions. See for instance an earlier version of Antoniou, Guney and Paudyal (2003), Gonzales and Gonzales (2008) or Barclay, Marx and Smith, 2003. We do not report these results here, and they are available from the author on demand. Let us mention that the results of these varying regressions do not

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

238

3.6. Controlling  for  multi-­colinearity  in  context-­variables  

As noted earlier and emphasized by Frank and Goyal (2008) some of the context variables are

highly correlated. For that reason we set up two procedures.

First we use the instrumental variable approach similar to the one in the previous chapter

[chapter III] but on non-differentiated data, and based on two rather than one instrumental

variable approach for the managers’ confidence measures (based on consumers’ confidence

and stock market indices) and for long-term interest rates (based on inflation). This implies

that we use the residuals of monthly regressions of these two variables, managers’ confidence

and interest rates, respectively on consumers’ confidence, stock market indices and on

inflation. These instrumental approaches are implemented following specific analysis as

described in appendix 3 of chapter III.

Second, we implement equation (1) in several steps, including the varying context

dimensions, in addition to the core variables, sequentially and comparatively. More

specifically, we conduct a systematic analysis of the stability in significance, signs and

coefficients, depending on the context variables included into the model. Results are broadly

satisfactory (table 16 in appendix 3).

Furthermore, we proceed to the pooling, described in the next paragraph, in three steps. We

first add to the retained firm- and industry-specific each of the confidence indicators in turn.

We call these the steps (1) and (2), and each regression includes seven plus one variables.

Then we proceed to a regression with a full set of context variables including the two

instrumental variables (step 3). Overall, we note that neither of these proceedings has a

meaningful effect on results. In some cases, the to confidence indicators are affected by the

affect importantly significance (but for the SE-clustered OLS as compared to OLS) or signs. However the coefficients obtained can be significantly different, particularly in the panel data regressions as compared to OLS.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

239

inclusion of other context variables, so we report in a specific column these specific

coefficients that can be compared with those obtained for the two confidence variables

obtained in the full regressions. All these results are provided in the table 17.

3.7. Pooling  

Our hypothesizing primarily suggests a mediating effect of the context on the relation

between leverage and some of its core determinants, including business cycles. As a result we

decide to pool our regressions depending on the dimension of the context considered.

Generally, whatever the variable considered as mediating, we pool the regression into two

subsamples, often based on whether variables are below or above the sample mean. For

ownership dispersion and creditors’ rights we split the sample into three categories because

the relative variance in the sample, and the sample size subsequently obtained, make more

sense.

Our proceedings result into a large range of regressions that need to be compared two by two

(three by three in a very few cases), which is a tedious work. In consequence we present the

full results in a large table in appendix 4 and we present subsets of these results below.

Another table completes the table 17, in appendix 4. Table 18 provides a streamlined

presentation, forgoing specific coefficients, and retaining only signs and significance.

Table 18 is provided for easiness of reading and interpretation.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

240

4. Results.  

4.1. The  effects  of  countries  on  firm-­  and  industry-­specific  variables  are  larger  

than  those  of  the  context  variables.  

As just outlined, we present, in appendix 3 and 4, three tables with un-pooled and pooled

results. This enables a broad overview of our results in comparison to the literature.

Furthermore, we provide in appendix 5 a specific table with results, as restricted to

regressions on core variables, pooled by country. This confirms the insight of De Jong, et al.

(2008) about the variability in coefficients, and even in significance from one country to the

next. As we restrict ourselves only to 15 European countries, the specific interpretation of

these results remains difficult. It is clear that countries mediate more strongly firm- and

industry-specific determinants of leverage than other contextual variables.

For this reason we focus here on exemplifying structural, institutional or cultural

commonalities rather than examining results country by country.

Now, we provide brief commentaries pertaining to the overall results in comparison to the

traditional literature. Then, We review in turn, in sections 4.2 to 4.4, each set of hypotheses

H1 to H3. Tables 12 to 14 with specific results’ extracts from table 17 are then displayed.

4.1.1. Core firm-specific variables: limited contextual mediation

Overall, relating to firm-specific variables, we confirm the robustness of Frank and Goyal

(2008)’s core factors, and our results are consistent with those on the U.S. sample for market

leverage generally, and as a general rule for book-leverage. Relating to the latter, profits fail

to predict leverage in young firms and in firms with negative growth (based on market-

capitalization over the listing period). The Tobin’s Q fails to predict book-leverage, in

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

241

Europe, for firms with negative growth or very high growth (higher than 66% per year on

average), for younger firms, as well as for larger and more profitable firms.

Additionally, the median industry leverage fails to predict book-leverage for younger and

more profitable firms, and it fails to predict market leverage for younger firms and for firms

with negative average growth. Further, median market leverage fails to predict market

leverage for firms in countries with high power distance or high uncertainty avoidance

(countries that are presumed to be bank- rather than market based, see section 2.3.)

4.1.2. Dividend policy and leverage: any dividend irrelevance?

Relating to dividend payouts, they fail to predict both measures of leverage in dispersed

ownership firms, in firms with normal growth (0% to 33%), in more profitable and larger

firms. This result extends further to book-leverage and younger firms or firms with multiple

listings. It seems to provide some sense that the dividend irrelevance proposition (Modigliani

and Miller, 1961) holds but only under certain conditions. These conditions seem to overlap

with those that stock market analysts value for consensual holdings.

4.1.3. Cumulative net equity return: which firms time the market?

The last non-context related variable is the one adapted from Baker and Wurgler, which is

negative for the un-pooled models of book- and market leverage, as they predict, but which is

never consistently negative in the subsequent poolings. Only one pooling, the one based on

size provides consistency: firms of any size seem to engage in market timing, irrespective of

their category. Any other type of pooling (financial, institutional, structural, legal, cultural)

seems to provide contrasted results with firms in the one or the other group. This is now what

we turn to.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

242

4.2. Market  timing  and  re-­balancing  asset  classes  

Our results are consistent with our two hypotheses, H1a and H1b, which cannot be rejected

(table 12). For the un-pooled regressions, we obtain significant and negative relations for both

book- and market leverages, in relation to both the broader stock market indices (BSMI) and

cumulative net equity creation (EFWAMB). In each case the relative effects of BSMI and

EFWAMB is stronger on market leverage as compared to book-leverage. This comparative

result suggests a relative decrease of debt, as a proportion of both assets and market

capitalization, and it suggests limited rebalancing of asset classes and some market timing. As

the effect of BSMI is three times larger for market leverage compared to book leverage8, this

suggests the presence of significant market timing in our sample.

4.2.1. Timing both stock and debt markets?

The pooling of our sample on ownership structures provides for a stark polarization of results.

While EFWAMB is only significant for companies with dispersed ownership, with

coefficients that are much larger than those on the overall sample (-1,4 instead of -0,3 for

market leverage, and -0,8 instead -0,2 for book-leverage), BSMI is significant for market

leverage only with a stronger coefficient than previously found for dispersed ownership (-

0,004 instead of -0,003). This suggests that there is a decrease of debt relative to market

capitalization, while there is no significant difference in comparison to assets. This suggests

some rebalancing of assets classes, possibly in response to (stock-) market timing. Further

exploration should consider whether this class of firms does not pursue some market timing

simultaneously on the stock and debt markets. This would be in line with survey evidence

from financial managers (Graham and Harvey, 2001; Bancel and Mitoo, 2004), as well as in

line with evidence on the differentiation of financial policies depending on ownership

dispersion (Reuter, 2010).

8 for an average Tobin’s Q of 1,7 for our sample, which would imply an average mechanical effect of 58%: the av. M/B=1,7, i.e. a delta of x on D/M should correspond to a delta of (x/1,7) on D/B, that is a delta of 58% of x. This is not reached

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

243

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cumul. net equity cr.

Managers' confi. (IV)

Consumers' confidence

Broad stock market X

Inflation

Long-term inter. rate (IV)

Individualism

Masculinity

Power Distance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. mono(0) line

Mult.(0) vs. sing.(1) listing

Unp

oole

d sa

mpl

e25

878

Mle

v+

-+

+-

-0.0

2***

-0.3

**

-

+-0

.003

***

-+

-+

nsns

+-

nsns

+U

npoo

led

sam

ple

25 8

78B

lev

+-

++

--0

.013

***

-0.2

2 *

ns

ns-0

.001

***

-ns

nsns

nsns

nsns

-ns

+O

wne

rshi

p di

sper

sion

high

ly c

on.

7 18

0M

lev

ns-0

.028

***

n.s.

ns

-0.0

03**

*-

-ns

nsns

nsns

nsns

nsO

wne

rshi

p di

sper

sion

conc

entra

ted

10 8

67M

lev

-0.0

21**

*n.

s.

+-0

.001

**

ns

ns-

nsns

nsns

nsO

wne

rshi

p di

sper

sion

disp

erse

d7

831

Mle

vn.

s.

-1.3

8***

ns+

-0.0

04**

*+

-ns

-+

-ns

nsO

wne

rshi

p di

sper

sion

high

ly c

on.

7 18

0B

lev

-0.0

15**

*n.

s.

nsns

-0.0

01**

*ns

nsns

nsns

nsns

nsns

Ow

ners

hip

disp

ersi

onco

ncen

trate

d10

867

Ble

v-0

.015

***

n.s.

ns

nsn.

s.

ns

ns-

nsns

ns-

nsns

Ow

ners

hip

disp

ersi

ondi

sper

sed

7 83

1B

lev

n.s.

-0

.82*

**ns

nsn.

s.

+

+ns

nsns

nsns

nsns

ns+

Cre

dito

rs' r

ight

slo

w7

429

Mle

v+

-0.0

22**

*-1

.11

**

--

-0.0

05**

*-

--

+ns

nsns

nsC

redi

tors

' rig

hts

inte

rmed

iate

5 14

6M

lev

ns-0

.028

***

n.s.

ns+

0.00

4 *

+

-ns

-ns

nsC

redi

tors

' rig

hts

high

13 3

03M

lev

+-0

.013

***

n.s.

nsns

n.s.

-ns

+ns

ns-

-C

redi

tors

' rig

hts

low

7 42

9B

lev

-0.0

16**

*-1

.44*

**ns

ns-0

.002

***

--

ns+

-+

ns+

nsns

nsC

redi

tors

' rig

hts

inte

rmed

iate

5 14

6B

lev

-0.0

17**

*n.

s.

-

ns0.

009*

**-

ns+

-ns

nsns

nsns

Cre

dito

rs' r

ight

shi

gh13

303

Ble

v-0

.008

*

n.s.

nsns

n.s.

ns-

ns+

nsns

-ns

nsSh

areh

olde

rs' r

ight

slo

w10

471

Mle

v-0

.031

***

n.s.

--0

.003

***

-ns

ns-

nsns

nsns

Shar

ehol

ders

' rig

hts

high

15 4

07M

lev

-0.0

1 **

-0

.4**

*ns

-0.0

04**

*+

+-

ns-

ns-

nsSh

areh

olde

rs' r

ight

slo

w10

471

Ble

v-0

.016

***

0.47

*

nsns

n.s.

ns-

ns+

ns-

nsns

nsns

nsSh

areh

olde

rs' r

ight

shi

gh15

407

Ble

v-0

.01*

**-0

.38*

**ns

ns-0

.001

**

ns

nsns

nsns

ns-

ns-

ns+

+-

++

-

+-

-

+

+-

++

+-

++

Tabl

e 12

. Mai

n re

sults

: mar

ket t

imin

g an

d th

e br

oade

r in

stitu

tiona

l con

text

Boo

kle

vera

geis

repo

rted

ingr

eyan

dm

arke

tlev

erag

eis

repo

rted

inw

hite

.Tab

les

12to

14pr

ovid

eex

tract

sof

Tabl

es17

(coe

ffici

ents

)and

18(s

igns

and

sign

ifica

nce

only

)in

appe

ndix

3.R

egre

ssio

nsar

eco

nduc

ted

com

para

tivel

yon

two

exog

enou

sva

riabl

es:m

arke

tand

book

-leve

rage

s(s

eefo

urth

colu

mn)

;res

ults

fort

hela

ttera

rein

grey

.For

each

exog

enou

sva

riabl

e,th

ere

gres

sion

ispo

oled

base

don

the

varia

ble

prov

ided

inth

efir

stco

lum

n,an

dgi

ven

the

valu

epr

ovid

edin

the

seco

ndco

lum

n.U

npoo

led

base

line

ispr

ovid

edup

fron

t.To

tals

ampl

esi

zeis

prov

ided

fort

heba

se-li

ne:"

unpo

oled

sam

ple"

and

the

pool

ing

alw

ays

add

upto

that

num

ber(

25'8

78fir

m-y

ears

).Fo

rrea

dibi

lity,

only

the

coef

ficie

ntof

the

endo

geno

usva

riabl

eof

inte

rest

is p

rovi

ded,

sign

and

sign

ifica

nce

only

are

pro

vide

d fo

r oth

er v

aria

bles

. All

regr

essi

ons a

re p

anel

-dat

a re

gres

sion

s with

rand

om e

ffets

. Fur

ther

det

ails

pro

vide

d in

Tab

le 1

7

++

+

+

--

+

-+

+-

++

+

Form

emor

y:H

1a.T

heag

greg

ate

rela

tion

for(

stoc

k)m

arke

ttim

ing

ism

edia

ted

byow

ners

hip

disp

ersi

on,a

ndth

equ

ality

ofsh

areh

olde

r-as

wel

las

ofcr

edito

rri

ghts

.Spe

cific

ally

the

rela

tion

hold

sm

ore

stro

ngly

,oro

nly

ford

ispe

rsed

owne

rshi

p,fo

rcou

ntrie

sw

ithst

rong

shar

ehol

derr

ight

san

dfo

rcou

ntrie

sw

ithw

eak

cred

itorr

ight

s.H

1b.T

heag

greg

ate

rela

tion

obse

rved

for t

he im

porta

nce

of st

ock

mar

ket t

imin

g sh

ould

be

reve

rsed

in c

ount

ries

with

str

ong

cred

itor

righ

ts.

--

++

++

+

++

-

+

-

-

-+

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

244

Further, for firms with concentrated ownership, the results are relatively similar to those of

the un-pooled sample, so there is evidence of limited rebalancing of classes of assets, while

EFWAMB is never significantly related with either market or book leverage for firms with

concentrated ownership (more than 2/3 of closely held shares) or intermediate concentration

(between 1/3 and 2/3 of closely held shares). This implies that the limited rebalancing of asset

classes, which we observe, is not complemented by any significant and lasting market timing

policy. Last we observe a market polarization of significance of the dividend payout variables,

which are significant when EFWAMB is not (i.e. for firms with moderate and high

concentration of ownership). This provides further evidence consistent with the findings of

Reuter (2010).

4.2.2. Shareholders’ rights

For European countries with strong shareholders’ rights, results are similar to those of the

unpooled regressions with a discrepancy between BSMI coefficients for book and market

leverages, which is more marked. Results for European countries with lower shareholders’

rights are markedly different. BSMI is not significantly related to book-leverage while it is

negatively with market leverage (consistent with some rebalancing), while there is a positive

correlation for EFWAMB in relation to book, but not market leverage (cumulated net equity

creation is positively related to the debt/asset ratio). This could be consistent with reputational

issues with stronger, more prominent more profitable firms getting stronger debt financing,

and better market valuations. Further evidence is needed to confirm or disconfirm this

suggestion.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

245

4.2.3. Creditors’ rights

Results for European countries with lower creditors’ rights are in line with, and slightly more

marked than, those of the unpooled sample. Results for firms in countries with intermediate or

strong rights, exhibit no significant relation for the EFWAMB variable, while they exhibit an

opposite (positive instead of negative for intermediate cases) or insignificant coefficient (for

firms in countries with strong rights) for BSMI. This would provide support for our

hypotheses H1a (/creditors’ rights) and H1b, suggesting specificity relating to the role and

influence than banks through different channels in European countries with strong creditors’

rights.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

246

4.3. Pro-­,  vs.  contra-­cyclicality  in  leverage.  Sources  of  managers’  behaviors  and  

resulting  capital  structures  

We are testing here four sets of hypothesis pertaining to the influence of the context, on

managers’ behaviors, and capital structures. Results are presented in Table 13 to 4d, relating

specifically to the institutional context (ownership dispersion, section 4.3.1.), the legal context

(section 4.3.2.), the structural and institutional context (varying firm-specific characteristics,

section 4.3.3.), and last, risk (section 4.3.4.).

Broadly, we suggest that each of these dimensions mediate the relationship between the

business cycles and observed capital structure because the broader context provided the

sources for understanding (and interpreting) manager’s behaviors. Generally we suggest that

manager’s behaviors can be polarized into two types, either pro-cyclical, or contra-cyclical.

In aggregate and in Europe, we identify some evidence on both behaviors with a contra-

cyclicality of manager’s confidence and pro-cyclicality of consumers’ confidence for market

leverage only. Other economic variables (inflation and interest rates) are contra-cyclical too

(again for market leverage: higher long-term interest rates, and lower inflation, are associated

with higher market leverage). There is mostly no relation to book-leverage, but for inflation,

which is negative (instead of positive as in Frank and Goyal, 2009)9.

9 This might be due to supply-side or macro-economic effects rather than firm-specific issues. Let us emphasize the relative complexity of the pan-European macro-economic correlation patterns, since 1990, between inflation, growth, exchange rates and interest rates (as compared to the U.S. situation).

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

247

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cumul. net equity cr.

Managers' confidence (IV residual)

Consumers' confidence

Broad stock market X

Inflation

Long-term interest rate (IV residual)

Individualism

Masculinity

Power Distance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. mono(0) line

Mult.(0) vs. sing.(1) listing

Unp

oole

d sa

mpl

e25

878

Mle

v+

-+

+-

--

-0.2

***

0.09

***

--0

.3 *

*

0.5*

**-

+ns

ns+

-ns

ns+

Unp

oole

d sa

mpl

e25

878

Ble

v+

-+

+-

--

n.s.

n.s.

--0

.3 *

*

n.s.

nsns

nsns

nsns

-ns

+O

wne

rshi

p di

sper

sion

high

ly c

on.

7 18

0M

lev

nsns

-0.3

5***

n.s.

-1.5

***

-0.8

**

ns

nsns

nsns

nsns

nsO

wne

rshi

p di

sper

sion

conc

entra

ted

10 8

67M

lev

ns-0

.15*

**0.

05 *

n.

s.

0.

9***

ns-

nsns

nsns

nsO

wne

rshi

p di

sper

sion

disp

erse

d7

831

Mle

vns

-n.

s.

0.

12**

*1.

4***

2.5*

**-

ns-

+-

nsns

Ow

ners

hip

disp

ersi

onhi

ghly

con

.7

180

Ble

vns

n.s.

n.s.

--0

.6**

*-0

.5 *

ns

nsns

nsns

nsns

nsns

Ow

ners

hip

disp

ersi

onco

ncen

trate

d10

867

Ble

vns

n.s.

n.s.

ns-0

.3 *

-0

.5 *

*

nsns

-ns

nsns

-ns

nsO

wne

rshi

p di

sper

sion

disp

erse

d7

831

Ble

vns

-n.

s.

n.

s.

ns

0.7*

**1.

4***

nsns

nsns

nsns

nsns

+

Cre

dito

rs' r

ight

slo

w7

429

Mle

v+

--0

.3**

*-0

.13

**

--2

.6**

*-3

.6**

*-

+ns

nsns

nsC

redi

tors

' rig

hts

inte

rmed

iate

5 14

6M

lev

nsns

n.s.

0.08

**

+

1.3*

**2.

1***

+-

ns-

nsns

Cre

dito

rs' r

ight

shi

gh13

303

Mle

v+

nsn.

s.

n.

s.

ns

2.4*

**3.

1***

-ns

+ns

ns-

-C

redi

tors

' rig

hts

low

7 42

9B

lev

-n.

s.

n.

s.

-

-0.9

***

-0.8

**

ns

+-

+ns

+ns

nsns

Cre

dito

rs' r

ight

sin

term

edia

te5

146

Ble

vns

-0.1

3 **

n.

s.

+

0.5*

**0.

6 **

-

ns+

-ns

nsns

nsns

Cre

dito

rs' r

ight

shi

gh13

303

Ble

vns

n.s.

n.s.

ns1.

1 **

1.

1 **

ns

-ns

+ns

ns-

nsns

Shar

ehol

ders

' rig

hts

low

10 4

71M

lev

ns-0

.35*

**0.

09**

*-0

.6**

*n.

s.

ns

-ns

nsns

nsSh

areh

olde

rs' r

ight

shi

gh15

407

Mle

v-

n.s.

0.1*

**1.

1***

2***

-ns

-ns

-ns

Shar

ehol

ders

' rig

hts

low

10 4

71B

lev

+n.

s.

n.

s.

ns

n.s.

-0.4

**

ns

+ns

-ns

nsns

nsns

Shar

ehol

ders

' rig

hts

high

15 4

07B

lev

-n.

s.

n.

s.

-

n.s.

n.s.

nsns

nsns

-ns

-ns

+Fo

rmem

ory:

H2a

.The

reis

apr

o-cy

clic

al[2

ai](

cont

ra-c

yclic

al[2

aii])

rela

tion

betw

een

leve

rage

and

econ

omic

conf

iden

cem

easu

res

forf

irms

with

disp

erse

d(c

once

ntra

ted)

ow

ners

hip.

H2b

(i).T

hepr

oan

dco

ntra

-cyc

lical

rela

tions

betw

een

leve

rage

and

econ

omic

conf

iden

cear

em

edia

ted

byth

equ

ality

ofsh

areh

olde

rs’r

ight

sw

itha

stro

nger

pro-

cycl

ical

(con

tra-c

yclic

al)r

elat

ion

inco

untri

esw

ithst

rong

er(w

eake

r)rig

hts.

H2b

(ii).

The

pro

and

cont

ra-c

yclic

alre

latio

nsbe

twee

nle

vera

gean

dec

onom

icco

nfid

ence

ism

edia

ted

byth

equ

ality

ofcr

edito

rs’r

ight

sw

itha

stro

nger

pro-

cycl

ical

(con

tra-c

yclic

al)

rela

tion

inco

untri

esw

ithw

eake

r(s

trong

er)

right

s.H

2c(i)

.The

pro-

cycl

ical

rela

tion

betw

een

leve

rage

and

econ

omic

conf

iden

ceis

exac

erba

ted

forl

arge

firm

s,fo

rfirm

sop

erat

ing

inm

ultip

lebu

sine

sses

,for

firm

sw

ithm

ultip

lelis

tings

and

foro

lder

firm

s.H

2c(ii

).Th

eco

ntra

-cyc

lical

rela

tion

betw

een

leve

rage

and

econ

omic

conf

iden

ceis

exac

erba

ted

fors

mal

lerf

irms,

forf

irmso

pera

ting

ina

sing

lebu

sine

ss,l

iste

din

agi

ven

coun

try,a

ndyo

unge

rfir

ms.

H2d

. The

pro

-cyc

lical

rela

tion

shou

ld b

e w

eake

ned

in r

iski

er fi

rms w

hile

the

cont

ra-c

yclic

al re

latio

n sh

ould

be

unch

ange

d.

-+

++

+-

++

--

+-

++

--

+-

++

--

TAB

LE

4b.

Mai

n re

sults

: cyc

lical

ity a

nd le

vera

ge (H

2b)

-+

+-

--

++

+-

++

--

Tabl

e 13

. Mai

n re

sults

: cyc

lical

ity a

nd le

vera

ge

-+

+-

--

++

+

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

248

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cumul. net equity cr.

Managers' confidence (IV residual)

Consumers' confidence

Broad stock market X

Inflation

Long-term interest rate (IV residual)

Individualism

Masculinity

Power Distance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. mono(0) line

Mult.(0) vs. sing.(1) listing

Firm

siz

esm

all

13 1

34M

lev

--

-0.2

6***

n.s.

-0.4

**

n.

s.

ns

nsns

nsns

nsns

nsFi

rm s

ize

larg

e12

744

Mle

vns

--0

.1**

*0.

08**

*n.

s.

0.

9***

-+

ns+

-ns

ns+

Firm

siz

esm

all

13 1

34B

lev

--

-n.

s.

n.

s.

-0

.3 *

n.

s.

ns

nsns

nsns

-ns

nsFi

rm s

ize

larg

e12

744

Ble

vns

ns-

n.s.

0.04

*

n.s.

n.s.

nsns

nsns

ns-

ns+

List

ings

Mul

tiple

4 84

7M

lev

-n.

s.

0.

12 *

*

n.s.

1.6*

**ns

+-

nsns

nsLi

stin

gsSi

ngle

21 0

31M

lev

ns-0

.21*

**0.

06**

*-0

.4**

*n.

s.

ns

+ns

nsns

+Li

stin

gsM

ultip

le4

847

Ble

vns

-n.

s.

n.

s.

ns

n.s.

n.s.

nsns

nsns

nsns

nsns

nsLi

stin

gsSi

ngle

21 0

31B

lev

-ns

n.s.

n.s.

--0

.3**

*-0

.4 *

*

nsns

nsns

nsns

-ns

nsA

geyo

ung

com

pani

es3

235

Mle

vns

+-0

.32*

**n.

s.

n.

s.

n.

s.

ns

nsns

nsns

nsns

nsns

Age

old

com

pani

es22

643

Mle

v+

--0

.18*

**0.

1***

-0.3

**

0.

5***

-+

--

+-

nsns

+A

geyo

ung

com

pani

es3

235

Ble

vns

nsns

ns+

-0.1

8 **

n.

s.

ns

n.s.

n.s.

nsns

nsns

nsns

nsns

nsA

geol

d co

mpa

nies

22 6

43B

lev

+-

--

-n.

s.

n.

s.

-

-0.3

**

-0

.3 *

-

+-

ns+

-+

ns+

Com

plex

itym

ultip

le b

usin

esse

s15

271

Mle

vns

-0.2

2***

0.1*

**-0

.5**

*n.

s.

-

ns+

-ns

nsC

ompl

exity

sing

le b

usin

ess

10 6

07M

lev

--0

.16*

**n.

s.

n.

s.

0.

8 **

ns

-+

-ns

+C

ompl

exity

mul

tiple

bus

ines

ses

15 2

71B

lev

ns-0

.06

**

n.s.

--0

.5**

*-0

.4**

*ns

ns-

nsns

ns-

ns+

Com

plex

itysi

ngle

bus

ines

s10

607

Ble

vns

n.s.

n.s.

nsn.

s.

n.

s.

ns

nsns

nsns

ns-

nsns

Ris

klo

w12

821

MLE

vns

--0

.2**

*0.

14**

*-0

.6**

*n.

s.

ns

-+

-ns

nsR

isk

high

13 0

57M

LEv

-ns

-0.1

6***

n.s.

n.s.

1***

-ns

+ns

nsns

Ris

klo

w12

821

BLE

Vns

--0

.05

**

0.03

**

-

-0.4

***

-0.3

**

-

+ns

nsns

ns-

++

Ris

khi

gh13

057

BLE

V-

nsn.

s.

n.

s.

ns

n.s.

n.s.

nsns

-ns

nsns

nsns

ns+

-+

+-

Boo

kle

vera

geis

repo

rted

ingr

eyan

dm

arke

tlev

erag

eis

repo

rted

inw

hite

.Tab

les

12to

14pr

ovid

eex

tract

sof

Tabl

es17

(coe

ffic

ient

s)an

d18

(sig

nsan

dsi

gnifi

canc

eon

ly)

inap

pend

ix3.

Reg

ress

ions

are

cond

ucte

dco

mpa

rativ

ely

ontw

oex

ogen

ous

varia

bles

:m

arke

tan

dbo

ok-le

vera

ges

(see

four

thco

lum

n);

resu

ltsfo

rth

ela

tter

are

ingr

ey.F

orea

chex

ogen

ous

varia

ble,

the

regr

essi

onis

pool

edba

sed

onth

eva

riabl

epr

ovid

edin

the

first

colu

mn,

and

give

nth

eva

lue

prov

ided

inth

ese

cond

colu

mn.

Unp

oole

dba

selin

eis

prov

ided

upfr

ont.

Tota

lsam

ple

size

ispr

ovid

edfo

rthe

base

-line

:"un

pool

edsa

mpl

e"an

dth

epo

olin

gal

way

sad

dup

toth

atnu

mbe

r(25

'878

firm

-yea

rs).

Forr

eadi

bilit

y,on

lyth

eco

effic

ient

ofth

een

doge

nous

varia

ble

ofin

tere

stis

prov

ided

,sig

nan

dsi

gnifi

canc

eon

lyar

epr

ovid

edfo

rot

her

varia

bles

.All

regr

essi

ons

are

pane

l-dat

are

gres

sion

sw

ithra

ndom

effe

ts.

Furth

er d

etai

ls p

rovi

ded

in T

able

17

TAB

LE

4d.

Mai

n re

sults

: cyc

lical

ity a

nd le

vera

ge (H

2d)

+-

++

--

-+

+

+-

++

--

--

--

++

++

+-

++

-+

+-

---

-+

-

+-

++

-

+-

++

--

+-

++

--

TAB

LE

4c.

Mai

n re

sults

: cyc

lical

ity a

nd le

vera

ge (H

2c)

+-

++

--

-

Tabl

e 13

(con

tinue

d)

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

249

4.3.1. Ownership dispersion: pro vs. contra-cyclicality

We do not delve much on this issue because it largely overlaps with Reuter (2010)’s findings

on financial policies and ownership dispersions. Broadly, we corroborate those findings about

a pro-cyclicality (contra-cyclicality) of financial policies in dispersed (concentrated)

ownership firms. Further the pro-cyclicality of debt, investment management, and market

leverage, seems to be indeed associated with consumers’ confidence, in dispersed ownership

firms. In concentrated ownership firms, there is a marked contra-cyclicality, which is

specifically associated with managers’ confidence. Interestingly results for interest rates and

inflation are polarized too, and the instrumental variable approach suggests a complex

relationship between inflation, long-term interest rates and leverage (both exogenous variables

are either negative or positive simultaneously).

Together with the previous observations on dividend policies, the rebalancing debt and market

capitalization, and market timing, these results suggest that the mechanisms determining

capital structures might generally differ between dispersed and concentrated ownership firms,

consistently with the arguments suggested to derive hypotheses H2a(i) and (ii). In particular

the disciplining motives may be secondary, and overcome, by risk-motives in firms with

concentrated ownership.

4.3.2. Investors’ rights

Our results, consistent with hypotheses H2b(i) and (ii), provide some evidence that the

relation, between business cycles and leverage, is mediated by the quality of investors’ rights.

Relating to shareholders’ rights, there is still no identified correlation between business cycles

and book-leverage, while the positive and significant correlation to consumers’ confidence is

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

250

present for all countries and the negative and significant relation to managers’ confidence is

identified only in countries with lower shareholders’ rights. Other relations with inflation and

interest rates exhibit these opposite results as well. We conclude that there is mixed support

for H2b(i). While the specific positive relation with consumer confidence does not differ

significantly, there is a polarization according to shareholders’ rights for other business cycles

proxies.

Relating to creditors’ rights, results are similar and more pronounced concerning countries

with lower and intermediate levels of creditors’ rights. It is noteworthy that, in countries with

lower creditors’ rights, we identify contra-cyclicality with both managers’ and consumers’

leverage. So we cannot reject H2b(ii).

In countries with strong creditors’ rights no cyclicality at all is identified, which enables us to

reject H2b(ii) for those countries. We suggest that arguments pertaining to credit offering and

the centrality of banks in these countries may complicate the theoretical approach, so that

arguments, similar to those used for deriving hypothesis H1b, may apply to this part of the

sample.

4.3.3. The structural and institutional context

Our range of hypotheses pertain here to the mediating influence that the size, business

complexity, institutional anchoring (listings), and age, can have on firms, in relation to their

respective pro-cyclical lean debt and investment policies (identified in relation to consumers’

confidence: H2c-i) on the one hand, and contra-cyclical financial policies on the other

(identified in relation to managers’ confidence: H2c-ii).

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

251

Broadly evidence is consistent with our hypothesizing and we can not reject either

H2c(i) or (ii), for any of the four mediating influences: book-leverage is mostly not related to

either managers’ or consumer’s confidence, and when it is (5 times out of 16), it is in the

same way than for market-leverage, and much smaller (at to four time smaller). For market

leverage, size, complexity and age mediate the pro-cyclical relationship, which is only

obtained on the upper part of the sample (not significant for the lower end). The pro-cyclical

relationship is twice bigger for firms with multiple listings. The contra-cyclical relationship

identified in connection with market leverage and managers’ confidence is moderated by size

and age (two to three times bigger for the lower end of the sample), and it is mediated by

listings (significant vs. insignificant).

4.3.4. Risk

Last, we note that the pro-cyclical relation between business cycles and leverage is mediated

by firm’s risk: the coefficient for consumers’ confidence, for firms with higher variance in net

income, is insignificant. This suggests that the leverage in these firms is constrained by a

range of factors, and in particular they may be financially constrained. In Diamond (1991)’s

U-shape model of debt, good-quality firms get short term debt because this is cheaper and

they face no risk, medium quality firms get long-term debt because this is cheaper, provided

their risk-profile and bad-quality firms have no choice, so they obtain what they can get.

Similarly, the lack of consumer-driven pro-cyclicality of leverage in volatile firms may be due

to external financing constraints rather than to firms’ policies.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

252

4.4. Cultural  factors,  leverage,  market  timing  and  cyclicality  in  financial  policies  

4.4.1. A direct influence of culture on leverage?

Our first hypothesizing, derived from the seminal contribution by Chui, et al. (2002), is that

we should obtain respectively negative, positive and positive signs for the individualism,

power distance and masculinity variables. Table 14 shows that, in general, and on the

unpooled regressions, the individualism and masculinity indices are mostly negatively and

positively related to market leverage ratios, while the two other indices are mostly not

significant, and there is generally no significance of any of the four indices for book-leverage.

This would provide some evidence partly consistent with H3a (i) and (iii).

However, we note that the coefficient for the individualism index is systematically changed or

affected in the pooling based on masculinity, power distance and uncertainty avoidance.

Similar results are observed for the coefficients for the culture indices in the culture poolings.

The only other significant pooling that exhibits such important moderating or mediating role

on the culture variables is creditors’ rights. As a result, if we cannot reject H3a(i) and (iii) in

aggregate, there is evidence that the relationship is, at best, not linear, and possibly it should

be analyzed with combined culture dimensions. In fact, we use Hofstede’s indices as proxies

for Schwartz indices (based on pair wise correlations), which is not convincing. A valid

underlying cultural model could be that proposed by Chui, et al. (2002) directly based on

Schwartz’s values dimensions.

Third, we emphasize that we work on a restricted sample, with the intent to restrict the

variance in the broader institutional context, based on a regional grouping (see section 2.3 for

details). Even though we have a large number of countries (15 countries), this restricts the

potential effect of culture indices.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

253

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cumul. net equity cr.

Managers' confi. (IV)

Consumers' confidence

Broad stock market X

Inflation

Long-term inter. rate (IV)

Individualism

Masculinity

Power Distance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. mono(0) line

Mult.(0) vs. sing.(1) listing

US positioning on culture indices

Unp

oole

d sa

mpl

e25

878

Mle

v+

-+

+-

--

-+

--

+-

+ns

ns+

-ns

ns+

Unp

oole

d sa

mpl

e25

878

Ble

v+

-+

+-

--

nsns

--

nsns

nsns

nsns

ns-

ns+

Indi

vidu

alis

mco

llect

ivis

tic8

503

Mle

vns

--

--

nsns

+ns

nsns

nsIn

divi

dual

ism

indi

vidu

alis

tic17

375

Mle

v-

ns+

++

+ns

ns-

nsns

ns!

In

divi

dual

ism

colle

ctiv

istic

8 50

3B

lev

+ns

--

--

ns-

-+

nsns

ns-

nsIn

divi

dual

ism

indi

vidu

alis

tic17

375

Ble

v-

nsns

nsns

+ns

nsns

ns-

ns-

ns+

!

Mas

culin

ityfe

min

ine

9 72

2M

lev

-ns

-+

+-

+ns

-ns

nsns

nsM

ascu

linity

mas

culin

e16

156

Mle

vns

-+

-ns

+ns

-ns

--

nsns

Mas

culin

ityfe

min

ine

9 72

2B

lev

-ns

ns-

nsns

ns+

nsns

nsns

ns+

Mas

culin

itym

ascu

line

16 1

56B

lev

ns-

ns-

--

+ns

-ns

-ns

nsns

Pow

er-d

ista

nce

low

15 8

26M

lev

+ns

++

nsns

ns-

-ns

ns!

Po

wer

-dis

tanc

ehi

gh10

052

Mle

vns

-ns

nsns

+-

ns+

ns+

Pow

er-d

ista

nce

low

15 8

26B

lev

nsns

nsns

ns+

ns-

--

nsns

!

Pow

er-d

ista

nce

high

10 0

52B

lev

--

nsns

-ns

+ns

+ns

nsns

Unc

erta

inty

avo

idan

celo

w16

448

Mle

v+

nsns

++

nsns

-+

ns-

-!

U

ncer

tain

ty a

void

ance

high

9 43

0M

lev

ns-

--

ns-

++

-+

++

Unc

erta

inty

avo

idan

celo

w16

448

Ble

vns

nsns

++

+-

-+

ns-

nsns

!

Unc

erta

inty

avo

idan

cehi

gh9

430

Ble

v-

--

--

-ns

+-

+ns

nsns

-

For

mem

ory:

H3b

(i).C

ount

ries

char

acte

rized

with

high

unce

rtain

tyav

oida

nce

shou

ldha

vea

stro

nger

(wea

ker)

cont

ra-

(pro

-)cy

clic

alre

latio

nbe

twee

nbu

sine

sscy

cles

prox

ies

and

leve

rage

ratio

s.H

3b(ii

).Th

ere

vers

esh

ould

hold

forc

ount

ries

with

low

unce

rtain

tyav

oida

nce.

H3c

.The

gene

ralp

ro-c

yclic

ality

(con

tra-c

yclic

ality

)of

leve

rage

will

beob

serv

edin

coun

tries

that

have

sim

ilar

(opp

osite

)po

sitio

ning

onna

tiona

lcul

ture

indi

ces,

asth

eU

.S.,

repr

esen

ted

byre

spec

tivel

ya

high

,hig

h,lo

wan

dlo

wpo

sitio

ning

onin

divi

dual

ism

,mas

culin

ity,p

ower

dist

ance

and

unce

rtain

tyav

oida

nce.

For

H3a

:se

eta

bles

6&

7in

appe

ndix

3.H

3a:c

ount

ries

with

resp

ectiv

ely

low

erin

divi

dual

ism

(i),h

ighe

rpo

wer

dist

ance

(ii)

and

high

erm

ascu

linity

(iii)

shou

ldbe

char

acte

rized

by

low

er le

vera

ge.

+

+-

++

--

-

++

--

+

-+

-

++---

++

-

-+

+-

-

-

+-

+-

+-

++

-

-+

--

+-

++

-

+-

--

Boo

kle

vera

geis

repo

rted

ingr

eyan

dm

arke

tlev

erag

eis

repo

rted

inw

hite

.Tab

les

12to

14pr

ovid

eex

tract

sof

Tabl

es17

(coe

ffic

ient

s)an

d18

(sig

nsan

dsi

gnifi

canc

eon

ly)

inap

pend

ix3.

Reg

ress

ions

are

cond

ucte

dco

mpa

rativ

ely

ontw

oex

ogen

ous

varia

bles

:m

arke

tan

dbo

ok-le

vera

ges

(see

four

thco

lum

n);

resu

ltsfo

rth

ela

tter

are

ingr

ey.F

orea

chex

ogen

ous

varia

ble,

the

regr

essi

onis

pool

edba

sed

onth

eva

riabl

epr

ovid

edin

the

first

colu

mn,

and

give

nth

eva

lue

prov

ided

inth

ese

cond

colu

mn.

Unp

oole

dba

selin

eis

prov

ided

upfr

ont.

Tota

lsam

ple

size

ispr

ovid

edfo

rthe

base

-line

:"un

pool

edsa

mpl

e"an

dth

epo

olin

gal

way

sad

dup

toth

atnu

mbe

r(25

'878

firm

-yea

rs).

Forr

eadi

bilit

y,on

lyth

eco

effic

ient

ofth

een

doge

nous

varia

ble

ofin

tere

stis

prov

ided

,si

gn a

nd si

gnifi

canc

e on

ly a

re p

rovi

ded

for o

ther

var

iabl

es. A

ll re

gres

sion

s are

pan

el-d

ata

regr

essi

ons w

ith ra

ndom

eff

ets.

Furth

er d

etai

ls p

rovi

ded

in T

able

17

Tabl

e 14

. Mai

n re

sults

: cul

ture

, cyc

lical

ity a

nd le

vera

ge (b

ook

leve

rage

in g

rey)

--

+

+-

++

+-

++

-

+

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

254

Fourth and importantly, we note that the significance and polarization of the culture indices is

particularly strong in the pooling based on uncertainty avoidance10. Further we note that

hypotheses H3a(i) to (iii) can be consistently rejected for firms in countries with low

uncertainty avoidances, while none can be rejected in countries with high uncertainty

avoidance. The fourth Hofstede index may be mediating Chui, et al. (2002)’s model of a

cultural influence on capital structure policies.

We emphasize again the importance of uncertainty avoidance as an antecedent for macro-

financial structures (Aggarwal and Goodell, 2009; Kwok and Tadesse, 2006). Most of the

argumentation provided here rests on agency relations, and the relative importance of

disciplining (and entrenchment) vs. incentive alignment (and discretion). As a result we may

be exemplifying some of the underlying mechanisms provided by Shleifer and Vishny (1997),

as they conclude their survey by suggesting that essentially two different governance systems

co-exist worldwide (the U.S. on the one side, Germany and Japan on the other), with agency

relations working essentially differently, but fostering comparable quality levels for

shareholders and creditors.11

4.4.2. Culture and business cycle sensitivity of leverage

So we now turn to business cycle sensitivity of leverage, as polarized by uncertainty

avoidance on the one hand (H3b) and as typified by the US role model (H3c). This provides

some further exploration of the role of culture for the determinant of financial policies and

leverage ratios.

10 The pooling based on individualism and masculinity exhibit a majority of non-significant signs while the pooling based on power-distance exhibit a for a half consistent signs and no opposition for the other half. 11 “concentrated ownership – though large shareholding, takeovers, and bank finance- is a nearly universal method of control that helps investors to get their money back” (ibid: p. 773)… We note, in passing, that the model of the Universal bank, related to strong credit lending, c-o., etc. in Germany, is prevalent across Europe, and in the U.K. too.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

255

Relating to firms in countries with low uncertainty avoidance, we note that there is no contra-

cyclicality of either book- or market leverage, providing evidence that capital structures

follow the traditional evidence established from the U.S. role model. Further, inflation is

consistently positive, as in Frank and Goyal (2009)12. However, firms in countries with high

uncertainty avoidance exhibit a marked contra-cyclicality, based on managers’ confidence,

consumers’ confidence (for book leverage), and inflation. Further we note that the quality of

creditors’ rights is only significant in high-uncertainty avoidance countries.

Additionally, we provide in Table 5 the results of the pooling across the range of culture

variables. In dark grey, we emphasize the culture positioning corresponding to that of the

U.S., on any of the four Hofstede dimensions (high individualism, moderate masculinity, low

power distance and low uncertainty avoidance for the U.S.; only the differing coefficients are

emphasized).

The pooling on individualism suggests no cyclicality for book-leverage, and a limited pro-

cyclicality for market leverage, based on consumers’ confidence and inflation. The pooling on

power-distance suggests no contra-cyclicality, based on book-leverage, and a relation

consistent with our aggregate findings for market leverage, with both pro- and contra-cyclical

effects, and with inflation signed consistently with Frank and Goyal (2009).

As a result we conclude that the poolings, based on the three culture variables that

characterize the U.S. (high individualism, low power distance and uncertainty avoidance),

reveal very limited contra-cyclicality, either based on managers’ confidence, or other macro-

economic indicators, and they reveal pro-cyclicality of consumers’ confidence in two

poolings, with a positive sign across the two poolings. This suggests that the results

traditionally obtained in financial research are best obtained on samples that are culturally

pooled to be similar to the U.S. culture.

12 this was the only core variable not consistently signed as in the broad empirical analysis of capital structures in the U.S.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

256

5. Conclusion  

In traditional financial theorizing, business cycles are of limited, or secondary relevance for

the understanding of capital structure mechanisms. A number of other dimensions are indeed

more important (agency costs, market timing, signaling, etc.), and cycle-proxies are not robust

in empirical studies of leverage (Frank and Goyal, 2009). However in our previous chapter

(chapter III), we emphasize the relevance of two conflicting mechanisms, the one pro-

cyclical, and derived in part from the disciplining role of debt, and the other contra-cyclical,

derived from risk-related considerations and fostered by incentive-alignment.

We suggest that the contra-cyclical effect is underestimated in the literature, because it is of

limited relevance in the paradigmatic “U.S. Role Model” with broad and liquid stock markets,

strong shareholders’ rights, and dispersed ownership. Based on a sample of listed European

firms, we show that both the pro- and contra-cyclical effects are mediated by structural,

cultural and institutional factors. We show that, on contextual dimensions corresponding to

the U.S. context, pro-cyclicality is observed while there is limited or no contra-cyclicality, an

observation that is generally reversed when the contextual dimension does not replicate the

U.S. context. This observation seems to encompass capital structures (cyclicality of leverage,

market timing, dividend policies), and it extends to ownership dispersion, institutional

anchoring, transparency and risk, structural firms’ variables, the quality of the contracting

environment, or still measures of national cultures.

Our contributions relate to a number of elements relating to international capital structures.

First, the relevance and mediating role of the context provides a new angle of approach for the

understanding of capital structures across countries. Indeed, surveys and recent empirical

evidence establish that the determinants of capital structures, across countries, can be derived

from the same set of factors, but with differences in their relative importance and magnitude

(for instance, Bancel and Mitoo, 2004; De Jong, et al., 2008). We obtain similar findings, and

we suggest that accounting for the broader institutional context is important and it may be a

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

257

way forward for international empirical research on capital structure. Effects of this context,

as the one noted here in relation to business cycles, can appear marginal, but provided some

suspicion about a multi-colinearity and endogeneity of firm-specific variables (Barclay and

Smith, 2005), it may lead to strong cumulative effects in time. This reasoning, as applied to

the context, could be compared to the one about a lasting effect of market timing on capital

structures (Baker and Wurgler, 2002).

Second, echoing a large number of relatively more recent studies13, we emphasize the

importance of context. Unlike some of the previous literature, our emphasis is not on the

particular relevance of the one, or the other, aspect of context. Instead, we stress the relevance

of a large number of contextual dimensions simultaneously. It may well be that there is some

convergence across these dimensions: structural, legal, cultural, institutional and economic, to

produce limited diversity (Ragin, 2000). In fact, through this proceeding, we try to tackle

some of the methodological challenges emphasized in International Business about the study

of national cultures (Earley, 2006; Leung, et al., 2005), and we try to underline how structural

and institutional or cultural dimensions intertwine to produce differentiation.

Third and last, we uncover a large number of converging dimensions across the broader

institutional context. They seem to provide some consistence to Shleifer and Vishny (1997)’s

suggestion that there are more than one paradigmatic governance system, with a number of

dimensions to be accounted for, such as the ambiguous interaction between strong creditors’

and shareholders’ rights. Further, our findings echo Carlin and Mayer (1999, 2000)’s

distinction between bank-based and market based countries, because our results are strongly

influenced by uncertainty avoidance, a measure known to be associated with bank-based

macro-financial systems (Aggarwal and Goodell, 2009; Kwok and Tadesse; 2006).

* * *

13 for instance, as cited above: Chui, et al., 2002; De Jong et al., 2006; Demirgüç-Kunt and Maksimovic, 1999; Giannetti, 2003; Gonzales and Gonzales, 2008; King and Santor, 2008; Margaritis and Psillaki, 2009; Rajan and Zingales, 1995).

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

258

Appendix  1  (chapter  IV):  data  collection.  Synthesis  from  Reuter  (2009)  

We collect data from Thomson-One Banker, who provides combined access to Worldscope

and Datastream. We collect data on all Companies that have been or are listed on a stock

exchange on the European soil (64 cities including stock exchanges) since 1990 (an initial

total of 10 633 identified company tickers, dead or alive). Some firms have no financial data

at all over any range of year (34% of total number of collected tickers, many firms in Eastern

Europe). As usually done in these studies, we remove financial companies and

administrations (sic codes 6*** and 9***, 16% of counts). Then to account for partially

missing or discontinued stock-data and to limit sampling biases upfront, we decide to limit

ourselves to liquid companies with a minimal exposure to public ownership. To proxy for

this, we use sales data, the financial item with most availability and we retain companies with

more than Euro: 5million in sales per year and on average over their listing period (11% of

counts removed).

From a total of 4182 companies with exploitable data, we conduct a number of additional

tests bringing the count down from 4182 to 3434 firms in total. These screens, comparable in

size, include successively: removing companies with less than 3 years of data, removing

companies from stock exchanges that have less than 25 stocks listed on average (e.g.

Antwerpen, Bologna, Riga, Prague, etc.), removing companies with missing data on one of

the endogenous variables (i.e. market-cap or asset data; we treat missing debt-data as a 0-

value after successfully comparing shareholders’ equity and total assets). We remove years

with missing data when they are starting or terminating years. Last, we need to account for

discontinued or missing data on some of the item that we need as explanatory variables. We

apply three rules. First we remove companies with more than 25% of items missing (from the

20 items considered from the balance-sheet or income-statement). Second, we try to

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

259

reconstitute remaining missing data through existing balance sheet or income statement

information (for instance missing short term debt is often set to 0 because long-term debt is

equal to total debt). Third we replace a missing value by using the average of proximate years,

if the change between these years is comparable to other-years average.

These 3434 firms provide us with a total of 40 885 firm-years that we check empirically

(check of totals vs. additions of items, observation of pair-wise correlation) and that we check

against previous capital structure studies (observed relation between leverage and exogenous

variables). This check enables us to track down outliers and provides evidence that we need to

winsorize out data as in some previous empirical studies (Antoniou, Guney and Paudyal, 2003

2003; Frank and Goyal, 2009 2009). We try to winsorize the data at 5 and at 3 standard

deviations based on every Balance Sheet and Income Statement item considered. It turns out

that winsorizing the data at 5 fixes issues while it removes only 129 firms-years and 3 firms

entirely (for the most part, it removes the first or last year of quotation for given companies).

For the rest of the paper, we use this database removing further the American-registered

companies (77 Companies, 803 firm-years) and the Swiss-listed firms (131 firms, 1851 firm-

years), for which we do not have sentiment data. For all other companies, but one, the place of

primary listing is the one of registration. Our resulting database contains 3215 firms for 38102

firm-years.

* * *

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

260

6. Appendix  2  (chapter  IV):  Pair  wise  correlations  

(0) B

ook

leve

rage

01

23

45

67

89

1011

1213

1415

1617

1819

2021

2223

2425

2627

2829

3031

32(1

) Mar

ket l

ever

age

76%

(2) M

edia

n bo

ok le

vera

ge14

%16

%Ex

ogen

ous

varia

bles

(3) M

edia

n m

arke

t lev

erag

e10

%19

%81

%(4

) Pro

fits

-5%

-10%

10%

5%(5

) Tan

gibi

lity

23%

21%

22%

17%

11%

(6) S

ize

19%

21%

17%

17%

14%

19%

(7) T

obin

's Q

-11%

-31%

-15%

-24%

6%-1

3%-1

2%(8

) Div

iden

d pa

yout

-3%

-5%

11%

11%

20%

10%

13%

-3%

(9) C

umul

ativ

e ne

t equ

ity c

. 3%

2%4%

3%2%

2%7%

-2%

(10)

Ris

k-7

%-1

1%-1

5%-1

6%-1

2%-1

5%-3

3%13

%-1

8%-4

%Fi

rm s

peci

fic v

aria

bles

(11)

man

ager

s' co

nfid

ence

-4%

-10%

-31%

4%-6

%1%

13%

-1%

2%(1

2) m

anag

ers'

conf

iden

ce (I

.V.)

-1%

-5%

-11%

-26%

4%2%

8%2%

4%70

%(1

3) C

onsu

mer

s' co

nfid

ence

-3%

-7%

-7%

-17%

-3%

1%10

%-3

%-3

%7%

48%

-10%

(14)

Bro

ad st

ock

mar

ket i

ndex

10%

12%

6%-7

%2%

3%-1

1%26

%-1

%-2

6%(1

5) In

flatio

n4%

18%

4%15

%-5

%8%

2%-2

%-3

3%7%

-35%

-8%

(16)

Lon

g-te

rm in

tere

st ra

te-2

%2%

17%

7%13

%-4

%-5

%9%

-3%

-23%

3%-2

6%-2

3%59

%(1

7) L

ong-

term

inte

rest

rate

(I.V

.)-2

%3%

-2%

9%6%

7%-5

%-3

%6%

-1%

-2%

-7%

-1%

-11%

-21%

-2%

86%

Busi

ness

cyc

le, e

cono

mic

and

fina

ncia

l var

iabl

es(1

8) In

divi

dual

ism

inde

x-1

4%-1

7%-9

%-9

%-2

%-5

%-1

%6%

-3%

13%

-2%

6%35

%-5

9%-2

%-3

%-6

%(1

9) P

ower

dis

tanc

e in

dex

5%9%

2%2%

3%-8

%4%

-3%

-2%

4%-1

1%5%

22%

-55%

51%

8%1%

-2%

-37%

(20)

Unc

erta

inty

avo

idan

ce in

dex

11%

17%

8%7%

3%-3

%5%

-7%

2%-1

5%4%

10%

-57%

67%

2%-7

6%80

%(2

1) M

ascu

linity

inde

x-5

%-3

%-3

%5%

-3%

2%2%

5%-5

%33

%-3

9%-1

3%10

%4%

17%

-5%

8%C

ultu

re v

aria

bles

(22)

Ow

ners

hip

disp

ersi

on d

umm

y-3

%-8

%-3

%2%

5%28

%5%

4%3%

2%-3

%5%

20%

-37%

15%

12%

3%41

%-3

0%-4

6%4%

(23)

Firm

s' si

ze d

umm

y3%

-7%

5%7%

14%

10%

69%

10%

11%

5%-2

2%-2

%1%

2%-6

%6%

6%3%

5%-3

%-2

%31

%(2

4) F

irms'

prof

itabi

lity

dum

my

-3%

-18%

15%

10%

37%

19%

16%

8%22

%4%

-17%

-1%

-4%

7%-5

%3%

4%3%

4%-6

%-6

%-6

%10

%22

%(2

5) F

irm's

com

plex

ity d

umm

y-1

0%-1

5%-1

3%-1

6%-5

%-4

%-2

4%9%

-5%

15%

2%5%

8%-1

4%2%

-5%

-7%

20%

-17%

-23%

9%5%

-14%

-3%

(26)

Firm

's lis

tings

leng

th d

umm

y4%

7%11

%13

%9%

12%

16%

-6%

8%6%

-2%

-4%

-5%

7%-1

2%7%

19%

17%

6%-7

%-8

%-7

%15

%13

%11

%-1

1%(2

7) M

ultip

le li

stin

g du

mm

y-4

%-4

%-4

%-3

%-2

4%3%

4%4%

6%3%

1%11

%5%

-2%

13%

7%-9

%-1

5%2%

-15%

-2%

6%-7

%Po

olin

g va

riabl

es(2

8) In

divi

dual

ism

dum

my

10%

15%

6%6%

2%-3

%8%

-6%

-1%

3%-1

8%6%

18%

-51%

56%

11%

3%-2

%-5

4%82

%84

%1%

-35%

-8%

-21%

-7%

7%(2

9) P

ower

dis

tanc

e du

mm

y-7

%-8

%-3

%9%

-8%

2%4%

2%10

%-5

%22

%-2

2%-1

7%12

%7%

3%19

%-3

0%-1

2%87

%10

%-4

%-3

%13

%-8

%-1

2%-2

6%(3

0) U

ncer

tain

ty a

void

ance

dum

my

9%13

%5%

5%4%

-5%

9%-5

%-2

%4%

-16%

6%11

%-4

4%58

%13

%2%

-4%

-45%

88%

80%

-20%

-32%

3%-4

%-2

0%-5

%7%

88%

-39%

(31)

Mas

culin

ity d

umm

y-7

%-1

0%-7

%-6

%-1

%-8

%2%

4%-4

%4%

2%4%

10%

19%

-23%

7%-7

%73

%3%

-47%

-19%

28%

5%12

%5%

29%

-7%

-20%

5%(3

2) S

hare

hold

er's

right

s dum

my

-6%

-10%

-9%

-8%

-1%

-4%

5%2%

5%4%

20%

10%

-28%

11%

3%-5

%40

%10

%-3

4%-1

7%30

%4%

13%

7%25

%-9

%-3

4%-1

3%51

%(3

3) C

redi

tors

' rig

hts i

ndex

-10%

-12%

-4%

-4%

-4%

6%-4

%3%

-2%

14%

-7%

29%

-59%

-5%

-1%

55%

-77%

-69%

42%

30%

1%19

%4%

-11%

-72%

57%

-77%

4%-1

3%O

nly

coef

ficie

nt w

ith si

gnifi

cant

bel

ow th

e 10

% th

resh

old

are

disp

laye

d

Tabl

e 15

. Var

iabl

es: p

air

wis

e co

rrel

atio

ns (c

apita

l str

uctu

re a

nd b

road

er in

stitu

tiona

l con

text

)

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

261

7. Appendix  3  (chapter  IV):  Unpooled  results:  robustness  of  context-­

variables’  signs  and  coefficients  

Onl

yco

reon

e co

ntex

tI.V

.ec

o. c

ycle

cultu

rein

vest

ors'

eco.

cyc

leec

o. c

ycle

all c

onte

xt…

still

plu

sva

riab

les

varia

ble

at a

tim

e1ap

proa

chva

riabl

esva

riabl

esrig

hts

and

cultu

rean

d in

v.' r

ight

sva

riabl

es…

stru

ctur

al v

.M

arke

t Lev

arag

e(3

) Med

ian

mar

ket l

ever

age

35.3

***

incl

. not

repo

rted

37.5

***

37.2

***

35.3

***

35.3

***

37.2

***

37.2

***

37.2

***

37.3

***

(4) P

rofit

s-1

0.04

***

incl

. not

repo

rted

-9.1

7***

-9.1

8***

-10.

04**

*-1

0.04

***

-9.1

8***

-9.1

8***

-9.1

8***

-9.2

2***

(6) S

ize

4.24

***

incl

. not

repo

rted

4.02

***

4.04

***

4.24

***

4.24

***

4.04

***

4.04

***

4.04

***

4.16

***

(5) T

angi

bilit

y17

.3**

*in

cl. n

ot re

port

ed17

.7**

*17

.8**

*17

.3**

*17

.3**

*17

.8**

*17

.8**

*17

.8**

*17

.8**

*(7

) Tob

in's

Q-1

.58*

**in

cl. n

ot re

port

ed-1

.47*

**-1

.46*

**-1

.58*

**-1

.58*

**-1

.46*

**-1

.46*

**-1

.46*

**-1

.45*

**(8

) Div

iden

d pa

yout

-0.0

23**

*in

cl. n

ot re

port

ed-0

.02*

**-0

.02*

**-0

.023

***

-0.0

23**

*-0

.02*

**-0

.02*

**-0

.02*

**-0

.02*

**(9

) Cum

ulat

ive

net e

quity

c.

-0.2

9**

incl

. not

repo

rted

-0.2

8**

-0.2

8**

-0.2

9**

-0.2

9**

-0.2

8**

-0.2

8**

-0.2

8**

-0.3

**(1

1or1

2) m

anag

ers'

conf

.ex

cl.

-0.0

35**

-0.1

98**

*-0

.195

***

excl

.ex

cl.

-0.1

95**

*-0

.195

***

-0.1

95**

*-0

.195

***

(13)

Con

sum

ers'

conf

iden

ceex

cl.

n.s.

0.1*

**0.

09**

*ex

cl.

excl

.0.

09**

*0.

09**

*0.

09**

*0.

09**

*(1

4) B

road

sto

ck m

arke

t Xex

cl.

-0.0

01**

*-0

.003

***

-0.0

03**

*ex

cl.

excl

.-0

.003

***

-0.0

03**

*-0

.003

***

-0.0

03**

*(1

5) In

flatio

nex

cl.

-0.1

91**

*ex

cl.

-0.3

**ex

cl.

excl

.-0

.3**

-0.3

**-0

.3**

-0.3

**(1

7) L

T in

tere

st ra

te (I

.V.)

excl

.-0

.162

**ex

cl.

0.5*

**ex

cl.

excl

.0.

5***

0.5*

**0.

5***

0.5*

**(1

8) In

divi

dual

ism

inde

xex

cl.

n.s.

excl

.ex

cl.

-0.4

***

excl

.-0

.5**

*ex

cl.

-2.6

***

-2.5

***

(21)

Mas

culin

ity in

dex

excl

.n.

s.ex

cl.

excl

.-0

.7**

*ex

cl.

-1**

*ex

cl.

-0.4

*-0

.6**

(19)

Pow

er d

ista

nce

inde

xex

cl.

n.s.

excl

.ex

cl.

0.1*

**ex

cl.

0.2*

**ex

cl.

0.5*

**0.

5***

(20)

Unc

erta

inty

avo

idan

ce X

excl

.n.

s.ex

cl.

excl

.0.

2*ex

cl.

0.4*

**ex

cl.

-0.9

*-0

.7*

(33)

Cre

dito

rs' r

ight

sex

cl.

n.s.

excl

.ex

cl.

excl

.n.

s.ex

cl.

n.s.

24.1

***

22.4

***

(32)

Sha

reho

lder

's rig

hts

excl

.n.

s.ex

cl.

excl

.ex

cl.

n.s.

excl

.n.

s.-4

1.4*

**-3

9.8*

**(2

6) L

istin

gs le

ngth

d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.n.

s.(2

5) C

ompl

exity

d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.n.

s.(2

7) M

ultip

le li

stin

g d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.4.

6***

Nb.

Obs

.27

457

27 4

5725

878

25 8

7827

457

27 4

5725

878

25 8

7825

878

25 8

78

Boo

k le

vera

ge(2

) Med

ian

book

leve

rage

35.6

***

incl

. not

repo

rted

39.4

***

39.2

***

35.6

***

35.6

***

39.2

***

39.2

***

39.2

***

39.6

***

(4) P

rofit

s-6

.79*

**in

cl. n

ot re

port

ed-5

.98*

**-5

.96*

**-6

.79*

**-6

.79*

**-5

.96*

**-5

.96*

**-5

.96*

**-5

.97*

**(6

) Siz

e2.

73**

*in

cl. n

ot re

port

ed2.

48**

*2.

47**

*2.

73**

*2.

73**

*2.

47**

*2.

47**

*2.

47**

*2.

55**

*(5

) Tan

gibi

lity

14.8

***

incl

. not

repo

rted

15.3

***

15.3

***

14.8

***

14.8

***

15.3

***

15.3

***

15.3

***

15.4

***

(7) T

obin

's Q

-0.3

***

incl

. not

repo

rted

-0.2

9***

-0.2

9***

-0.3

***

-0.3

***

-0.2

9***

-0.2

9***

-0.2

9***

-0.2

9***

(8) D

ivid

end

payo

ut-0

.014

***

incl

. not

repo

rted

-0.0

13**

*-0

.013

***

-0.0

14**

*-0

.014

***

-0.0

13**

*-0

.013

***

-0.0

13**

*-0

.013

***

(9) C

umul

ativ

e ne

t equ

ity c

. -0

.33*

**in

cl. n

ot re

port

ed-0

.22

*

-0.2

2 *

-0

.33*

**-0

.33*

**-0

.22

*

-0.2

2 *

-0

.22

*

-0.2

2 *

(1

1or1

2) m

anag

ers'

conf

.ex

cl.

-0.1

45**

*-0

.04

*

-0.0

38 *

ex

cl.

excl

.-0

.038

*

-0.0

38 *

-0

.038

*

n.s.

(13)

Con

sum

ers'

conf

iden

ceex

cl.

n.s.

n.s.

n.s.

excl

.ex

cl.

n.s.

n.s.

n.s.

n.s.

(14)

Bro

ad s

tock

mar

ket X

excl

.-0

.002

***

-0.0

01**

*-0

.001

***

excl

.ex

cl.

-0.0

01**

*-0

.001

***

-0.0

01**

*-0

.001

***

(15)

Infla

tion

excl

.-0

.306

***

excl

.-0

.3 *

*

excl

.ex

cl.

-0.3

**

-0

.3 *

*

-0.3

**

-0

.3 *

*

(17)

LT

inte

rest

rate

(I.V

.)ex

cl.

0.19

3 **

ex

cl.

n.s.

excl

.ex

cl.

n.s.

n.s.

n.s.

n.s.

(18)

Indi

vidu

alis

m in

dex

excl

.n.

s.

ex

cl.

excl

.-0

.2**

*ex

cl.

-0.3

***

excl

.n.

s.

n.

s.

(2

1) M

ascu

linity

inde

xex

cl.

n.s.

excl

.ex

cl.

-0.6

**

ex

cl.

-0.6

**

ex

cl.

-0.4

*

-0.5

**

(1

9) P

ower

dis

tanc

e in

dex

excl

.n.

s.

ex

cl.

excl

.0.

1 **

ex

cl.

0.1

**

excl

.0.

2 *

n.

s.

(2

0) U

ncer

tain

ty a

void

ance

Xex

cl.

n.s.

excl

.ex

cl.

0.2

*

excl

.0.

2 **

ex

cl.

n.s.

n.s.

(33)

Cre

dito

rs' r

ight

sex

cl.

n.s.

excl

.ex

cl.

excl

.n.

s.

ex

cl.

n.s.

n.s.

n.s.

(32)

Sha

reho

lder

's rig

hts

excl

.n.

s.

ex

cl.

excl

.ex

cl.

n.s.

excl

.n.

s.

n.

s.

n.

s.

(2

6) L

istin

gs le

ngth

d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.-1

.9**

*(2

5) C

ompl

exity

d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.n.

s.

(2

7) M

ultip

le li

stin

g d.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.ex

cl.

excl

.1.

5 *

N

b. O

bs.

27 4

5727

457

25 8

7825

878

27 4

5727

457

25 8

7825

878

25 8

7825

878

Add

ing

to th

e co

re v

aria

bles

:Ta

ble

16. M

ain

unpo

oled

res

ults

with

cor

e va

riab

les a

nd v

aryi

ng c

onte

xt v

aria

bles

(cap

ital s

truc

ture

and

bro

ader

inst

itutio

nal c

onte

xt)

Reg

ress

ions

are

cond

ucte

don

mar

ket(

first

)and

book

-leve

rage

s(s

econ

dta

ble)

.Reg

ress

ions

are

base

don

the

unpo

oled

pan-

Euro

pean

sam

ple;

they

are

pane

l-dat

are

gres

sion

sw

ithra

ndom

effe

tsan

ddu

mm

ies.

Sign

ifica

nce

leve

lsar

e**

*1%

,**5

%,*

10%

.Uns

igni

fican

tcoe

ffic

ient

sar

eno

tdis

play

ed("

n.s.

").T

hita

ble

aim

sat

esta

blis

hing

the

robu

stne

ssan

dre

lativ

eef

fect

sof

allc

onte

xtva

riabl

es,

so"e

xcl."

mea

nsth

atth

eva

riabl

ew

asno

tinc

lude

din

the

regr

essi

onco

nduc

ted

inth

eco

lum

n.N

ote

(1):

Reg

ress

ions

are

cond

ucte

dw

ith7

endo

geno

usva

riabl

es,i

.e.t

he6

'cor

eva

riabl

es'a

ndon

lyon

eof

the

5co

ntex

tva

riabl

es.

For

read

ibili

ty,

the

5co

ntex

tva

riabl

esar

edi

spla

yed

inth

isco

lum

n,an

dco

re-v

aria

bles

coef

ficie

nts

are

not

disp

laye

d.Th

eyar

efu

llyco

nsis

tent

with

all

othe

rre

gres

sion

s an

d ex

hibi

t ver

y lim

ited

varia

nce.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

262

8. Appendix  4  (chapter  IV):  Results  for  pooled  regressions.    

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cumul. net equity cr.

Managers' confi. (IV)

Man. conf.1

Consumers' confidence

Cons. conf.1

Broad stock market X

Inflation

Long-term inter. rate (IV)

Individualism

Masculinity

Power Distance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. mono(0) line

Mult.(0) vs. sing.(1) listing

Unp

oole

d sa

mpl

e25

878

Mle

v37

.3**

*-9

.2**

*4.

2***

17.8

***

-1.4

***

-0.0

2***

-0.3

**

-0

.2**

*-0

.15*

**0.

09**

*n.

s.

-0

.003

***

-0.3

**

0.

5***

-1.7

***

0.4*

**n.

s.

n.

s.

12

.2**

*-2

6.7*

**n.

s.

n.

s.

4.

6***

Unp

oole

d sa

mpl

e25

878

Ble

v39

.6**

*-6

***

2.5*

**15

.4**

*-0

.3**

*-0

.013

***

-0.2

2 *

n.

s.

-0

.04

**

n.s.

n.s.

-0.0

01**

*-0

.3 *

*

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

-1.9

***

n.s.

1.5

*

Ow

ners

hip

disp

ersi

onhi

ghly

con

.7

180

Mle

vn.

s.

-1

1.9*

**5.

6***

17.8

***

-1.7

***

-0.0

28**

*n.

s.

-0

.35*

**-0

.32*

**n.

s.

n.

s.

-0

.003

***

-1.5

***

-0.8

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

3.

1 *

O

wne

rshi

p di

sper

sion

conc

entra

ted

10 8

67M

lev

47.6

***

-8.8

***

4.6*

**21

***

-1.4

***

-0.0

21**

*n.

s.

-0

.15*

**-0

.15*

**0.

05 *

n.

s.

-0

.001

**

n.

s.

0.

9***

n.s.

0.3

*

-0.8

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

3.

3 **

O

wne

rshi

p di

sper

sion

disp

erse

d7

831

Mle

v40

.8**

*-8

.6**

*3.

7***

11.4

***

-1.2

***

n.s.

-1.3

8***

n.s.

n.s.

0.12

***

0.09

**

-0

.004

***

1.4*

**2.

5***

-4**

*0.

8***

n.s.

-1.8

*

33.3

***

-75.

3***

n.s.

n.s.

7.4*

**O

wne

rshi

p di

sper

sion

high

ly c

on.

7 18

0B

lev

39.4

***

-6.2

***

2.8*

**16

.6**

*-0

.6**

*-0

.015

***

n.s.

n.s.

-0.0

8***

n.s.

n.s.

-0.0

01**

*-0

.6**

*-0

.5 *

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

O

wne

rshi

p di

sper

sion

conc

entra

ted

10 8

67B

lev

55.1

***

-7**

*2.

5***

18.2

***

-0.2

***

-0.0

15**

*n.

s.

n.

s.

-0

.04

*

n.s.

-0.0

4 *

n.

s.

-0

.3 *

-0

.5 *

*

n.s.

n.s.

-0.6

*

n.s.

n.s.

n.s.

-2.2

**

n.

s.

n.

s.

O

wne

rshi

p di

sper

sion

disp

erse

d7

831

Ble

v19

.3 *

-4

.2**

*2.

7***

10.7

***

-0.2

*

n.s.

-0.8

2***

n.s.

n.s.

n.s.

0.05

*

n.s.

0.7*

**1.

4***

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

5.4*

**Fi

rm si

zesm

all

13 1

34M

lev

50.3

***

-8.7

***

7.7*

**21

.4**

*-1

.3**

*-0

.038

***

-0.5

3***

-0.2

6***

-0.2

***

n.s.

n.s.

-0.0

03**

*-0

.4 *

*

n.s.

n.s.

n.s.

-0.8

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

Fi

rm si

zela

rge

12 7

44M

lev

33.7

***

-13.

8***

5.5*

**13

.1**

*-1

***

n.s.

-1**

*-0

.1**

*-0

.1**

*0.

08**

*n.

s.

-0

.002

***

n.s.

0.9*

**-2

.5**

*0.

5***

-0.6

*

n.s.

22.1

***

-41.

8***

n.s.

n.s.

3.7*

**Fi

rm si

zesm

all

13 1

34B

lev

66.9

***

-5.9

***

3***

17.8

***

-0.4

***

-0.0

25**

*-0

.24

*

n.s.

-0.0

4 *

n.

s.

n.

s.

-0

.001

**

-0

.3 *

n.

s.

n.

s.

n.

s.

-0

.7 *

n.

s.

n.

s.

n.

s.

-1

.5 *

n.

s.

n.

s.

Fi

rm si

zela

rge

12 7

44B

lev

20.3

***

-6.8

***

3.4*

**12

.2**

*n.

s.

n.

s.

-0

.5 *

n.

s.

-0

.04

**

0.04

*

n.s.

-0.0

01 *

n.

s.

n.

s.

n.

s.

n.

s.

-0

.5 *

n.

s.

n.

s.

n.

s.

-2

.9**

*n.

s.

1.

9 *

Pr

ofita

bilit

ylo

w12

266

Mle

v56

.6**

*-4

.9**

*4.

4***

23.3

***

-1.2

***

-0.0

34**

*n.

s.

-0

.22*

**-0

.19*

**0.

09**

*n.

s.

-0

.003

***

n.s.

1.3*

**-2

.2 *

*

0.5

**

-1.1

***

n.s.

19.1

*

-33.

1 *

n.

s.

n.

s.

3.

9***

Prof

itabi

lity

high

13 6

12M

lev

11.4

**

-2

9.9*

**3.

8***

13.8

***

-1.2

***

n.s.

n.s.

-0.1

7***

-0.1

***

0.07

***

n.s.

-0.0

02**

*-0

.8**

*-0

.4 *

*

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

3.8*

**Pr

ofita

bilit

ylo

w12

266

Ble

v63

.9**

*-4

.6**

*1.

9***

17.7

***

-0.4

***

-0.0

19**

*n.

s.

n.

s.

-0

.06

**

n.s.

-0.0

4 *

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

-1

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

Pr

ofita

bilit

yhi

gh13

612

Ble

vn.

s.

-1

2.4*

**3.

6***

13.1

***

n.s.

n.s.

-2.1

2***

-0.0

5 *

n.

s.

0.

03 *

0.

03 *

-0

.001

***

-0.5

***

-0.6

***

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

-2.4

**

n.

s.

2.

7 **

R

isk

low

12 8

21M

LEv

20**

*-7

3.4*

**3.

9***

17.2

***

-2.4

***

n.s.

-1.7

8 **

-0

.2**

*-0

.14*

**0.

14**

*0.

05 *

*

-0.0

03**

*-0

.6**

*n.

s.

-2

.7**

*0.

5***

n.s.

-1.1

*

22.6

***

-46.

2***

n.s.

n.s.

5.1*

**R

isk

high

13 0

57M

LEv

52.9

***

-6.9

***

4.7*

**19

.7**

*-1

.1**

*-0

.043

***

n.s.

-0.1

6***

-0.1

3***

n.s.

n.s.

-0.0

02**

*n.

s.

1*

**-2

**

0.

4 **

-1

***

n.s.

19.7

**

n.

s.

n.

s.

n.

s.

3.

6***

Ris

klo

w12

821

BLE

V27

.6**

*-2

8.1*

**3.

8***

14.2

***

-0.4

***

n.s.

-4.5

4***

-0.0

5 **

-0

.05*

**0.

03 *

*

n.s.

-0.0

01**

*-0

.4**

*-0

.3 *

*

-1.3

*

0.2

*

n.s.

n.s.

n.s.

n.s.

-3**

*1.

4 *

4.

1***

Ris

khi

gh13

057

BLE

V46

.1**

*-5

***

2.4*

**16

.4**

*-0

.2**

*-0

.026

***

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

-0.8

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

Li

stin

gsM

ultip

le4

847

Mle

v34

.2**

*-1

0.1*

**3.

6***

24**

*-1

.3**

*-0

.027

***

-1.4

6***

n.s.

-0.2

4***

0.12

**

n.

s.

-0

.002

*

n.s.

1.6*

**-4

.8**

*0.

9***

-1.3

*

n.s.

44.1

***

-83.

5 **

n.

s.

n.

s.

n.

s.

Li

stin

gsSi

ngle

21 0

31M

lev

37.6

***

-9**

*4.

3***

16.9

***

-1.5

***

-0.0

19**

*n.

s.

-0

.21*

**-0

.15*

**0.

06**

*n.

s.

-0

.002

***

-0.4

***

n.s.

-1.6

**

0.

3 **

-0

.6 *

n.

s.

13

.4 *

n.

s.

n.

s.

n.

s.

13

1.8

**

List

ings

Mul

tiple

4 84

7B

lev

41.5

***

-8.2

***

1.7*

**21

.4**

*-0

.5**

*n.

s.

-1

.02

**

n.s.

-0.0

7 *

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

Li

stin

gsSi

ngle

21 0

31B

lev

38.7

***

-5.4

***

2.7*

**14

.5**

*-0

.2**

*-0

.015

***

n.s.

n.s.

-0.0

3 *

n.

s.

n.

s.

-0

.001

***

-0.3

***

-0.4

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

-2

.1**

*n.

s.

n.

s.

A

geyo

ung

com

pani

es3

235

Mle

vn.

s.

-3

***

3.2*

**20

.8**

*-0

.9**

*-0

.021

**

0.6

4 **

-0

.32*

**-0

.34*

**n.

s.

-0

.12

**

-0.0

04**

*n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

A

geol

d co

mpa

nies

22 6

43M

lev

34.9

***

-11.

3***

4.2*

**17

.3**

*-1

.6**

*-0

.019

***

-0.4

9***

-0.1

8***

-0.1

3***

0.1*

**n.

s.

-0

.002

***

-0.3

**

0.

5***

-3.3

***

0.6*

**-0

.7 *

*

-1.1

**

28

.5**

*-5

5***

n.s.

n.s.

5.6*

**A

geyo

ung

com

pani

es3

235

Ble

vn.

s.

n.

s.

1.

4***

21.8

***

n.s.

n.s.

0.53

*

-0.1

8 **

-0

.16*

**n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

A

geol

d co

mpa

nies

22 6

43B

lev

40.4

***

-7.6

***

2.5*

**14

.6**

*-0

.3**

*-0

.013

***

-0.3

3 **

n.

s.

-0

.03

*

n.s.

n.s.

-0.0

01**

*-0

.3 *

*

-0.3

*

-1.3

**

0.

3 **

-0

.6 *

*

n.s.

10.3

*

-21.

3 *

12

0.9

**

n.s.

2 **

C

ompl

exity

mul

tiple

bus

ines

ses

15 2

71M

lev

31.4

***

-16.

5***

4.2*

**19

.9**

*-1

.9**

*-0

.016

***

n.s.

-0.2

2***

-0.1

9***

0.1*

**n.

s.

-0

.003

***

-0.5

***

n.s.

-2.1

**

0.

4***

-0.9

**

n.

s.

18

.3 *

*

-33.

1 **

n.

s.

n.

s.

5.

6***

Com

plex

itysi

ngle

bus

ines

s10

607

Mle

v37

.8**

*-5

.2**

*4*

**15

.8**

*-1

.1**

*-0

.025

***

-0.3

5 *

-0

.16*

**-0

.09*

**n.

s.

n.

s.

-0

.002

***

n.s.

0.8

**

-3.2

***

0.6*

**n.

s.

-1

.5 *

*

28.5

***

-54.

4***

n.s.

254.

8***

2.6

*

Com

plex

itym

ultip

le b

usin

esse

s15

271

Ble

v28

.8**

*-1

0.7*

**2.

7***

16.6

***

-0.4

***

-0.0

1***

n.s.

-0.0

6 **

-0

.06*

**n.

s.

n.

s.

-0

.001

***

-0.5

***

-0.4

***

n.s.

n.s.

-0.7

**

n.

s.

n.

s.

n.

s.

-1

.9 *

*

n.s.

2.2

**

Com

plex

itysi

ngle

bus

ines

s10

607

Ble

v47

.5**

*-3

.4**

*2.

5***

14.1

***

-0.1

*

-0.0

16**

*n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

-1

.9 *

n.

s.

n.

s.

In

divi

dual

ism

colle

ctiv

istic

8 50

3M

lev

36**

*-1

1.1*

**4.

4***

22.2

***

-1.8

***

-0.0

28**

*n.

s.

-0

.46*

**-0

.42*

**-0

.15*

**n.

s.

-0

.004

***

-2.5

***

-1.8

***

-0.2

***

n.s.

n.s.

0.3*

**n.

s.

n.

s.

n.

s.

n.

s.

3.

5***

Indi

vidu

alis

min

divi

dual

istic

17 3

75M

lev

38.8

***

-8.1

***

3.8*

**14

.9**

*-1

.3**

*-0

.015

***

-0.5

7***

n.s.

n.s.

0.07

***

0.06

***

-0.0

04**

*1.

3***

1.9*

**-0

.3 *

*

0.1

**

n.s.

n.s.

-3.8

*

n.s.

n.s.

n.s.

6.2*

**In

divi

dual

ism

colle

ctiv

istic

8 50

3B

lev

58.1

***

-8.7

***

2***

20.1

***

-0.6

***

-0.0

13**

*0.

85**

*n.

s.

-0

.1**

*-0

.09

**

n.s.

-0.0

01**

*-1

***

-1.1

***

n.s.

-0.2

***

-0.3

***

0.3*

**n.

s.

n.

s.

n.

s.

-1

.6 *

n.

s.

In

divi

dual

ism

indi

vidu

alis

tic17

375

Ble

v34

.1**

*-4

.6**

*2.

6***

12.5

***

-0.2

***

-0.0

12**

*-0

.63*

**n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

0.

5 **

n.

s.

n.

s.

n.

s.

n.

s.

-4

.8 *

*

n.s.

-2.5

***

n.s.

2.9

**

Mas

culin

ityfe

min

ine

9 72

2M

lev

41.4

***

-15.

3***

3.9*

**18

.8**

*-1

.4**

*-0

.017

***

-0.4

**

n.

s.

-0

.17*

**-0

.07

*

-0.1

***

-0.0

06**

*1.

2***

1.7*

**-0

.2**

*0.

3***

n.s.

-0.1

*

n.s.

n.s.

n.s.

n.s.

6.6*

**M

ascu

linity

mas

culin

e16

156

Mle

v35

.1**

*-7

.5**

*4.

2***

17.5

***

-1.4

***

-0.0

21**

*n.

s.

-0

.3**

*-0

.14*

**0.

17**

*n.

s.

-0

.003

***

-1**

*n.

s.

0.

6***

n.s.

-0.3

*

n.s.

-7.2

***

-20.

5***

n.s.

n.s.

3.6*

**M

ascu

linity

fem

inin

e9

722

Ble

v54

.8**

*-7

.1**

*2.

7***

15.2

***

-0.3

***

-0.0

17**

*-0

.42*

**n.

s.

-0

.09*

**n.

s.

-0

.05

**

-0.0

03**

*n.

s.

n.

s.

n.

s.

0.

2 **

n.

s.

n.

s.

n.

s.

n.

s.

-2

.8 *

*

n.s.

3 *

M

ascu

linity

mas

culin

e16

156

Ble

v29

.9**

*-5

.5**

*2.

4***

15.5

***

-0.3

***

-0.0

1***

n.s.

-0.1

1***

n.s.

n.s.

n.s.

-0.0

01 *

*

-0.4

***

-0.3

*

0.5*

**n.

s.

-0

.4**

*n.

s.

-5

.9**

*n.

s.

-1

.6 *

*

n.s.

n.s.

Pow

er-d

ista

nce

low

15 8

26M

lev

51.3

***

-6.6

***

3.6*

**18

.5**

*-1

.3**

*-0

.014

***

n.s.

-0.0

9 **

-0

.14*

**0.

08**

*n.

s.

-0

.003

***

1***

1.6*

**n.

s.

n.

s.

n.

s.

0.

1 **

-4

.6 *

-9

.9**

*n.

s.

n.

s.

4.

9***

Pow

er-d

ista

nce

high

10 0

52M

lev

n.s.

-24.

1***

5.1*

**14

.2**

*-1

.7**

*-0

.024

***

-4.4

***

-0.6

***

-0.1

8***

0.17

***

0.07

**

-0

.003

***

n.s.

n.s.

n.s.

0.3*

**-1

.2**

*0.

7***

n.s.

18.4

***

n.s.

1.8

*

4.5*

**Po

wer

-dis

tanc

elo

w15

826

Ble

v42

.8**

*-4

.9**

*2.

2***

16.6

***

-0.2

***

-0.0

09**

*n.

s.

n.

s.

-0

.04

**

n.s.

n.s.

-0.0

03**

*n.

s.

n.

s.

0.

3 **

n.

s.

-0

.2 *

0.

2***

-5.8

***

-6.8

***

-1.8

**

n.

s.

n.

s.

Po

wer

-dis

tanc

ehi

gh10

052

Ble

v25

.5**

*-1

1.9*

**3.

9***

12.4

***

-0.5

***

-0.0

15**

*-6

.14*

**-0

.25*

**-0

.09*

**n.

s.

n.

s.

-0

.001

***

n.s.

-0.4

**

n.

s.

0.

1 *

-0

.7 *

*

0.4*

**n.

s.

9*

**n.

s.

n.

s.

n.

s.

U

ncer

tain

ty a

void

ance

low

16 4

48M

lev

49.8

***

-6.8

***

3.6*

**17

.9**

*-1

.3**

*-0

.015

***

n.s.

n.s.

-0.1

1***

0.06

**

n.

s.

-0

.002

***

1.3*

**1.

9***

n.s.

n.s.

-1.1

***

0.6*

**n.

s.

5.

8 *

-1

.8 *

*

-1.3

*

4.7*

**U

ncer

tain

ty a

void

ance

high

9 43

0M

lev

n.s.

-21.

8***

5.1*

**15

.5**

*-1

.8**

*-0

.022

***

-3.1

2***

-0.5

7***

-0.3

***

0.06

*

n.s.

-0.0

03**

*-0

.8**

*n.

s.

-1

.3**

*0.

6***

2.1*

**-1

.5**

*35

.3**

*23

.9**

*2.

2 *

2.

1 **

4.

3***

Unc

erta

inty

avo

idan

celo

w16

448

Ble

v42

***

-5.1

***

2.2*

**16

.1**

*-0

.2**

*-0

.01*

**n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

-0

.002

**

0.

7***

0.7

**

0.2

**

-0.0

5 *

-1

.1**

*0.

5***

n.s.

5.8

**

-2.2

***

n.s.

n.s.

Unc

erta

inty

avo

idan

cehi

gh9

430

Ble

v21

.8 *

*

-9.9

***

3.5*

**12

.9**

*-0

.5**

*-0

.013

***

-4**

*-0

.19*

**-0

.14*

**-0

.05

*

-0.0

8***

-0.0

01**

*-0

.5**

*-0

.4 *

-0

.5 *

*

n.s.

0.8

**

-0.5

*

15.5

**

10

.2 *

*

n.s.

n.s.

n.s.

Cre

dito

rs' r

ight

slo

w7

429

Mle

v22

.7**

*-2

1.3*

**4*

**21

.2**

*-1

.5**

*-0

.022

***

-1.1

1 **

-0

.3**

*-0

.29*

**-0

.13

**

n.s.

-0.0

05**

*-2

.6**

*-3

.6**

*-0

.3**

*0.

2 **

-0

.4**

*0.

3***

n.s.

n.s.

n.s.

n.s.

6.4*

**C

redi

tors

' rig

hts

inte

rmed

iate

5 14

6M

lev

n.s.

-11.

1***

4.5*

**12

.6**

*-1

.6**

*-0

.028

***

n.s.

n.s.

n.s.

0.08

**

0.

09 *

*

0.00

4 *

1.

3***

2.1*

**-0

.3**

*0.

2***

0.7

*

-0.5

**

n.

s.

-4

.5 *

n.

s.

n.

s.

2.

8 *

C

redi

tors

' rig

hts

high

13 3

03M

lev

50.4

***

-6.5

***

3.8*

**17

.2**

*-1

.3**

*-0

.013

***

n.s.

n.s.

-0.1

***

n.s.

n.s.

n.s.

2.4*

**3.

1***

-0.2

***

-0.2

***

n.s.

0.3*

**n.

s.

n.

s.

-1

.8 *

-1

.6 *

4.

6***

Cre

dito

rs' r

ight

slo

w7

429

Ble

v35

.1**

*-8

.5**

*2.

1***

17.6

***

-0.5

***

-0.0

16**

*-1

.44*

**n.

s.

-0

.09*

**n.

s.

n.

s.

-0

.002

***

-0.9

***

-0.8

**

n.

s.

0.

2 *

-0

.5**

*0.

3***

n.s.

8.2*

**n.

s.

n.

s.

n.

s.

C

redi

tors

' rig

hts

inte

rmed

iate

5 14

6B

lev

28.9

**

-6

.2**

*3.

6***

12.1

***

-0.4

***

-0.0

17**

*n.

s.

-0

.13

**

-0.0

5 *

n.

s.

n.

s.

0.

009*

**0.

5***

0.6

**

-0.2

***

n.s.

0.6

*

-0.3

*

n.s.

n.s.

n.s.

n.s.

n.s.

Cre

dito

rs' r

ight

shi

gh13

303

Ble

v42

.2**

*-5

.3**

*2.

3***

15.7

***

-0.2

**

-0

.008

*

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

1.1

**

1.1

**

n.s.

-0.1

***

n.s.

0.2*

**n.

s.

n.

s.

-2

**

n.

s.

n.

s.

Sh

areh

olde

rs' r

ight

slo

w10

471

Mle

v34

.7**

*-1

2***

5.1*

**20

.3**

*-2

.1**

*-0

.031

***

n.s.

-0.3

5***

-0.1

9***

0.09

***

n.s.

-0.0

03**

*-0

.6**

*n.

s.

n.

s.

0.

4***

0.3*

**-0

.6**

*n.

s.

n.

s.

n.

s.

n.

s.

4.

1***

Shar

ehol

ders

' rig

hts

high

15 4

07M

lev

37.2

***

-7.8

***

3.7*

**15

.5**

*-1

.2**

*-0

.01

**

-0.4

***

n.s.

n.s.

0.1*

**0.

05 *

*

-0.0

04**

*1.

1***

2***

-0.2

***

0.1

**

0.6

**

n.s.

-5.3

**

n.

s.

-1

.7 *

n.

s.

6.

7***

Shar

ehol

ders

' rig

hts

low

10 4

71B

lev

63.2

***

-9.4

***

2.8*

**18

.1**

*-0

.7**

*-0

.016

***

0.47

*

n.s.

n.s.

n.s.

n.s.

n.s.

n.s.

-0.4

**

n.

s.

0.

2***

n.s.

-0.2

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

Sh

areh

olde

rs' r

ight

shi

gh15

407

Ble

v26

.2**

*-4

.1**

*2.

3***

13.9

***

-0.1

**

-0

.01*

**-0

.38*

**n.

s.

n.

s.

n.

s.

n.

s.

-0

.001

**

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

n.

s.

-5

.6 *

*

n.s.

-2.9

***

n.s.

2.7

*

Tabl

e 17

. Ful

l res

ults

for

the

pool

ed r

egre

ssio

ns. B

ook

leve

rage

is in

gre

y an

d m

arke

t lev

erag

e in

whi

te c

apita

l str

uctu

re a

nd b

road

er in

stitu

tiona

l con

text

)

All

deta

ils a

re p

rovi

ded

at th

e bo

ttom

of t

he n

ext t

able

of t

he a

ppen

dix

3 (ta

ble

7), w

hich

is a

sim

plifi

ed v

ersi

on o

f thi

s tab

le. E

xtra

cts a

re p

rovi

ded

in ta

bles

3 to

5. S

igni

fican

ce le

vels

are

***

1%, *

*5%

, *10

%. U

nsig

nific

ant c

oeffi

cien

ts a

re n

ot d

ispl

ayed

("n.

s.")

AD

DIT

ION

(Not

e1)

.We

incl

ude

the

coef

ficie

ntof

are

gres

sion

base

don

8es

ogen

ous

varia

bles

:the

7co

reva

riabl

es(m

edia

nin

dust

ryle

vera

geto

cum

ulat

ive

nete

quity

)pl

usei

ther

man

ager

s'co

nfid

ence

orco

nsum

ers'

conf

iden

ce.O

nly

the

coef

ficie

nts

onth

eco

nfid

ence

varia

bles

are

disp

laye

dhe

re,a

sa

robu

stne

ss in

dica

tor;

they

are

add

ed in

an

addi

tiona

l col

umn

usin

g gr

ey fo

nt c

olou

r afte

r the

coe

ficie

nt o

btai

ned

for t

he e

quiv

alen

t var

iabl

e in

the

full-

size

regr

esis

ons.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

263

Displaying  only  signs  and  significance  

Pooling variable

Value

Nb. of observations

Exogenous variable

Median industry lev.

Profitability

Size

Tangibility

Tobins' Q

Dividend payout

Cum

ul. net equity cr.

Managers' confi. (IV

)

Consum

ers' confidence

Broad stock m

arket X

Inflation

Long-term inter. rate (IV

)

Individualism

Masculinity

Power D

istance

Uncertainty Avoidance

Shareholders' rights

Creditors' rights

Young (0) vs. old(1) Cies

Multi(0) vs. m

ono(0) line

Mult.(0) vs. sing.(1) listingUnpooled sample 25 878 Mlev + - + + - - - - + - - + - + ns ns + - ns ns +

Unpooled sample 25 878 Blev + - + + - - - ns ns - - ns ns ns ns ns ns ns - ns +Ownership dispersion highly con. 7 180 Mlev ns ns ns - - ns ns ns ns ns ns ns nsOwnership dispersion concentrated 10 867 Mlev ns + ns ns - ns ns ns ns nsOwnership dispersion dispersed 7 831 Mlev ns - ns + + - ns - + - ns nsOwnership dispersion highly con. 7 180 Blev ns ns ns - ns ns ns ns ns ns ns ns nsOwnership dispersion concentrated 10 867 Blev ns ns ns ns ns ns - ns ns ns - ns nsOwnership dispersion dispersed 7 831 Blev ns - ns ns ns + + ns ns ns ns ns ns ns ns +

Firm size small 13 134 Mlev - - ns - ns ns ns ns ns ns ns ns nsFirm size large 12 744 Mlev ns - + ns + - + ns + - ns ns +Firm size small 13 134 Blev - - - ns ns - ns ns ns ns ns ns - ns nsFirm size large 12 744 Blev ns ns - ns + ns ns ns ns ns ns ns - ns +

Profitability low 12 266 Mlev - ns ns + - + - ns + - ns nsProfitability high 13 612 Mlev ns ns - - ns ns ns ns ns ns ns nsProfitability low 12 266 Blev + - + + - - ns ns ns ns ns ns ns ns - ns ns ns ns ns nsProfitability high 13 612 Blev ns - + + ns ns - - + - - - ns ns ns ns ns ns - ns +

Risk low 12 821 MLEv ns - + - ns ns - + - ns nsRisk high 13 057 MLEv - ns ns ns + - ns + ns ns nsRisk low 12 821 BLEV ns - - + - - - - + ns ns ns ns - + +Risk high 13 057 BLEV - ns ns ns ns ns ns ns ns - ns ns ns ns ns ns

Listings Multiple 4 847 Mlev - ns ns + ns + - ns ns nsListings Single 21 031 Mlev ns - - ns ns + ns ns ns +Listings Multiple 4 847 Blev ns - ns ns ns ns ns ns ns ns ns ns ns ns ns nsListings Single 21 031 Blev - ns ns ns - - - ns ns ns ns ns ns - ns ns

Age young companies 3 235 Mlev ns + ns ns ns ns ns ns ns ns ns ns ns nsAge old companies 22 643 Mlev + - + - + - + - - + - ns ns +Age young companies 3 235 Blev ns ns ns ns + - ns ns ns ns ns ns ns ns ns ns ns ns nsAge old companies 22 643 Blev + - - - - ns ns - - - - + - ns + - + ns +

Complexity multiple businesses 15 271 Mlev ns + - ns - ns + - ns nsComplexity single business 10 607 Mlev - ns ns + ns - + - ns +Complexity multiple businesses 15 271 Blev ns - ns - - - ns ns - ns ns ns - ns +Complexity single business 10 607 Blev ns ns ns ns ns ns ns ns ns ns ns ns - ns ns

Individualism collectivistic 8 503 Mlev ns - - - - ns ns + ns ns ns nsIndividualism individualistic 17 375 Mlev - ns + + + + ns ns - ns ns nsIndividualism collectivistic 8 503 Blev + ns - - - - ns - - + ns ns ns - nsIndividualism individualistic 17 375 Blev - ns ns ns ns + ns ns ns ns - ns - ns +Masculinity feminine 9 722 Mlev - ns - + + - + ns - ns ns ns nsMasculinity masculine 16 156 Mlev ns - + - ns + ns - ns - - ns nsMasculinity feminine 9 722 Blev - ns ns ns ns ns + ns ns ns ns ns +Masculinity masculine 16 156 Blev ns - ns - - + ns - ns - ns ns ns

Power-distance low 15 826 Mlev + ns + + ns ns ns - - ns nsPower-distance high 10 052 Mlev ns - ns ns ns + - ns + ns +Power-distance low 15 826 Blev ns ns ns - ns ns + ns - - - ns nsPower-distance high 10 052 Blev - - ns - ns - ns + ns + ns ns ns

Uncertainty avoidance low 16 448 Mlev + ns ns + + ns ns - + ns - -Uncertainty avoidance high 9 430 Mlev ns - - - ns - + + - + + +Uncertainty avoidance low 16 448 Blev ns ns ns + + + - - + ns - ns nsUncertainty avoidance high 9 430 Blev - - - - - - ns + - + ns ns ns

Creditors' rights low 7 429 Mlev + - - - - - - - + ns ns ns nsCreditors' rights intermediate 5 146 Mlev ns ns ns + + + - ns - ns nsCreditors' rights high 13 303 Mlev + ns ns ns ns - ns + ns ns - -Creditors' rights low 7 429 Blev - ns ns - - - ns + - + ns + ns ns nsCreditors' rights intermediate 5 146 Blev ns - ns + - ns + - ns ns ns ns nsCreditors' rights high 13 303 Blev ns ns ns ns ns - ns + ns ns - ns ns

Shareholders' rights low 10 471 Mlev ns - - ns ns - ns ns ns nsShareholders' rights high 15 407 Mlev - ns + + - ns - ns - nsShareholders' rights low 10 471 Blev + ns ns ns ns - ns + ns - ns ns ns ns nsShareholders' rights high 15 407 Blev - ns ns - ns ns ns ns ns ns - ns - ns +

+

-

- +

- + -

Book leverage is reported in grey and market leverage is reported in white. This Table is a simplification of table 6 where the the value of coeficients has beenremoved. When a sign is significant across a pooling, it is displayed only once. No significance ("ns") is always indicated across poolings to make sure which line isconcerned. Precise significance levels are indicated by stars in table 6 and not reproduced here. Regressions are conducted comparatively on two endogenousvariables: market and book-leverages (see fourth column); results for the latter are in grey. For each exogenous variable, the regression is pooled based on a contextvariable, which is provided in the first column, its values being provided in the second column. Unpooled regressions are provided upfront (frist two lines of the table).The total sample size (25'878 firm-years) is provided for the unpooled regressions and the number of variables across poolings always add up to that number. Allregressions are panel-data regressions with random effets and a constant as well as dummies for years, industries and countries (not reproduced). Explained variance isprovided in Table 6 with the coefficients. Endogeous variables are respectively book- and market leverage, that is total short-term and long-term debt divided by theappropriate denominator. All exogenous variables are labelled in a self-sufficient way and scaled by total assets (profitability and tangibility). Specific definitions andsources are provided in table 1.

-

- -

+

+

+

+

+

+

++

-

+

+

+

+

+

+

-

-

+ + +

- +

-

-

++

- + -

-+

+ -

- -

-

-

- + + - -

++-

+

+ - + +

+

+-

-

-

+ +

-

--

+-

--++-

--+

--+

-++-

+ -

-

+

+

+

- + + - -

-

-

+ - + + -

+

-

-+ + + -

- + +

-

--++-

- + + - -

--++-

+

-

+ - + +

+

-

-

-++-

-

+ - + +

+ - + + -

+

+ - + + -

-

+ - + + -

-

+ - + +

++-+

Table 18. Full results for the pooled regressions, displaying only signs and significance for ease of reading

- +

-

-

+ +- + +

-

--

- -

--

+-+

+

-

- + -

-

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

264

9. Appendix  5  (chapter  IV):  Results  pooled  by  country  

RE

SULT

S fo

r M

arke

t le

vera

geM

edia

n in

d.

mar

ket l

ev.

Prof

itabi

lity

Tang

ibili

tyTo

bin'

s Q

Size

Div

iden

d pa

yout

Cum

ul. n

et e

quity

cr

.U

npoo

led

36.6

5***

-9.9

8***

21.1

7***

-1.4

9***

3.78

***

-0.0

2***

-0.3

4**

Aus

tria

nsns

14.9

6*-1

.41*

**6.

8***

nsns

Bel

gium

ns-1

8.96

***

ns-2

.28*

**4.

56**

*ns

nsG

erm

any

34.8

6***

-10.

17**

*36

.77*

**-2

.13*

**3.

73**

*-0

.04*

**ns

Den

mar

k71

.05*

**-3

2.9*

**23

.91*

**-3

.42*

**3.

75**

*ns

nsSp

ain

49.1

8**

-52.

28**

*ns

-2.3

2***

5.7*

**ns

-5.3

5**

Finl

and

72.9

***

-13.

56**

*42

.28*

**-1

.01*

**1.

5***

-0.0

3**

nsFr

ance

36.3

8***

-22.

67**

*30

.07*

**-1

.48*

**3.

89**

*ns

-7.9

***

Gre

at B

ritai

n26

.9**

*-4

.79*

**16

.2**

*-1

.01*

**2.

58**

*ns

-0.7

**G

reac

ens

-42.

63**

*10

.2**

*-0

.67*

6.71

***

nsns

Irel

and

96.1

7***

-36.

77**

*ns

nsns

ns6.

75**

Italy

ns-2

3.84

***

11.1

9**

-3.6

5***

7.33

***

-0.0

4***

-8.2

8***

Net

herla

nds

37.1

4***

-6.4

1*ns

-1.5

4***

3.04

***

nsns

Pola

ndns

-13.

48**

*ns

-3.2

2***

2.98

***

-0.0

7***

nsPo

rtuga

l55

.55*

-49.

46**

*-2

0.2*

*-5

.16*

*6.

52**

*-0

.06*

*ns

Swed

en20

.8**

-5.5

3***

35.5

***

-1.0

2***

2.68

***

nsns

RE

SULT

S fo

r B

ook

leve

rage

Med

ian

ind.

boo

k le

v.Pr

ofita

bilit

yTa

ngib

ility

Tobi

n's

QSi

zeD

ivid

end

payo

utC

umul

. net

equ

ity

cr.

Unp

oole

d43

.39*

**-6

.46*

**18

.12*

**-0

.19*

**2.

26**

*-0

.02*

**-0

.42*

**A

ustri

a87

.82*

**ns

12.1

3*ns

3.13

***

nsns

Bel

gium

-44.

1*-7

.65*

17.7

9***

nsns

ns-3

.36*

**G

erm

any

49.4

6***

-9.9

***

30.4

8***

-0.6

2***

1.46

***

-0.0

2**

nsD

enm

ark

80.1

3***

-23.

31**

*16

.52*

**ns

2.91

***

nsns

Spai

n58

.51*

**-3

3.78

***

8.61

**ns

4.43

***

ns-5

.95*

**Fi

nlan

d87

.56*

**-5

.28*

*36

.38*

**ns

-1.2

3**

-0.0

4***

0.94

**Fr

ance

ns-8

.21*

**20

.33*

**-0

.53*

**2.

6***

ns-1

0.98

***

Gre

at B

ritai

n21

.89*

**-3

.17*

**14

.38*

**ns

1.97

***

ns-0

.65*

*G

reac

e56

.63*

**-2

2.94

***

12.5

4***

ns5.

08**

*ns

nsIr

elan

d10

0.6*

*ns

nsns

2.59

**ns

7.39

***

Italy

ns-1

0.8*

**9.

89**

*ns

3.86

***

-0.0

2***

-4.3

4***

Net

herla

nds

43.0

3***

-6.3

7*19

.34*

**-0

.56*

2.56

***

ns-4

.26*

**Po

land

nsns

8.96

**ns

1.99

***

-0.0

3*ns

Portu

gal

ns-4

5.33

***

nsns

4.67

***

ns-4

.58*

*Sw

eden

34.1

1***

-3.1

3**

26.2

5***

-0.3

7**

2.39

***

nsns

Tabl

e 19

. Res

ults

for

the

regr

essi

ons

on c

ore

vari

able

s, p

oole

d by

cou

ntry

(cap

ital s

truc

ture

and

bro

ader

inst

itutio

nal c

onte

xt)

Reg

ress

ions

are

cond

ucte

dco

mpa

rativ

ely

ontw

oen

doge

nous

varia

bles

:mar

keta

ndbo

ok-le

vera

ges,

that

isto

tals

hort-

term

and

long

-term

debt

divi

ded

byth

eap

prop

riate

deno

min

ator

,and

pool

edby

coun

try.A

llre

gres

sion

sar

epa

nel-d

ata

regr

essi

ons

with

rand

omef

fets

and

aco

nsta

ntas

wel

las

dum

mie

sfo

ryea

rsan

din

dust

ries

(not

repr

oduc

ed).

All

exog

enou

sva

riabl

esar

ela

belle

din

ase

lf-su

ffic

ient

way

and

gene

rally

scal

edby

tota

lass

ets

for

tang

ibili

tyan

dpr

ofita

bilit

y.Sp

ecifi

c de

finiti

ons

and

sour

ces

are

prov

ided

in ta

ble

1.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

265

10. Bibliography  

Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: national characteristics as determinants. Journal of Banking & Finance, 33(10): 1770-1780.

Antoniou, A., Guney, Y., Paudyal, K.N., 2003. The Determinants of Corporate Debt Ownership Structure: Evidence from Market-based and Bank-based Economies. SSRN eLibrary.

Antoniou, A., Guney, Y., Paudyal, K.N., 2003. The determinants of corporate debt ownership structure: evidence from market-based and bank-based economies. Available on SSRN.

Antoniou, A., Guney, Y., Paudyal, K.N., 2008. The determinants of capital structure: capital market oriented versus bank oriented institutions. Journal of Financial and Quantitative Analysis, 43: 59-92.

Aoki, M., 2001. Toward comparative institutional analysis. Cambridge: MIT Press. 467 pages.

Baker, M., Wurgler, J., 2002. Market timing and capital structure. Journal of Finance, 57(1): 1-32.

Bancel, F., Mittoo, U.R., 2004. Cross-country determinants of capital structure choice: a survey of European firms. Financial Management, 33(4): 103-132.

Barclay, M.J., Smith, C.W., 2005. The Capital Structure Puzzle: The Evidence Revisited. Journal of Applied Corporate Finance, 17(1): 8-17.

Barclay, M.J., Marx, L.M., Smith Jr., C.W., 2003. The joint determination of leverage and maturity. Journal of Corporate Finance, 9(2): 149.

Berger, P.G., Ofek, E., Yermack, D.L., 1997. Managerial entrenchment and capital structure decisions. Journal of Finance, 52(4): 1411-1438.

Carlin, W.J., Mayer, C., 2000. How do financial systems affect economic performance? In Vives, X. (Ed.), Corporate governance: theoretical and empirical perspectives. Cambridge: Cambridge University Press. 137-168.

Chui, A., Lloyd, A., Kwok, C., 2002. The determination of capital structure: is national culture a missing piece to the puzzle? Journal of International Business Studies, 33(1): 99-127.

de Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance, 32: 1954-1969.

de Jong, E., 2009. Culture and economics: on values, economics and international business. Routledge. 272 pages.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

266

de Jong, E., Semenov, R., 2006. Cultural determinants of ownership concentration across countries. International Journal of Business Governance and Ethics, 2(1): 145 - 165.

Demirguc-Kunt, A., Maksimovic, V., 1999. Institutions, financial markets, and firm debt maturity. Journal of Financial Economics, 54(3): 295-336.

Diamond, D.W., 1991. Debt maturity structure and liquidity risk. Quarterly Journal of Economics, 106(3): 709-737.

Earley, P.C., 2006. Leading cultural research in the future: a matter of paradigms and taste. Journal of International Business Studies, 37(6): 922-931.

Fan, P.H.J.P., Titman, S., Twite, G.J., 2008. An international comparison of capital structure and debt maturity choices. SSRN eLibrary.

Frank, M.Z., Goyal, V.K., 2004. The effect of market conditions on capital structure adjustement. Finance Research Letters, 1: 47-55.

Frank, M.Z., Goyal, V.K., 2008. Trade-off and pecking order theories of debt. In Eckbo, E. (Ed.), Handbook of corporate finance: empirical corporate finance. Elsevier North Holland

Frank, M.Z., Goyal, V.K., 2009. Capital Structure Decisions: Which Factors Are Reliably Important? Financial Management (Blackwell Publishing Limited), 38(1): 1-37.

Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: which factors are reliably important? Financial Management, 38(1): 1-37.

Giannetti, M., 2003. Do better institutions mitigate agency problems? Evidence from corporate finance choices. Journal of Financial and Quantitative Analysis, 38(1): 185-212.

Gonzalez, V., Gonzalez, F., 2008. Influence of bank concentration and institutions on capital structure: new international evidence. Journal of Corporate Finance, 14(4): 363-375.

Graham, J., C. Harvey, 2001. The Theory and Practice of Corporate Finance: Evidence from the Field. Journal of Financial Economics, 60: 187-243.

Greif, A., 2006. Institutions and the path to the modern economy: lessons from medieval trade. Cambridge University Press. 503 pages.

Guillen, M.F., Suarez, S.L., 2005. The institutional context of multinational activity. In Ghoshal, S. and Westney, D.E. (Ed.), Organization theory and the multinational corporation. New York: Palgrave Macmillan. 360.

Guiso, L., Haliassos, M., Jappelli, T., 2001. Household portfolios. Cambridge: MIT press. 508 pages.

Hofstede, G., 2003. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

Holderness, C.G., 2010. The myth of diffuse ownership in the United Staes. Review of Financial Studies, 22(4): 1377-1408.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

267

King, M.R., Santor, E., 2008. Family values: ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, (32): 2423-2432.

Kwok, C., Tadesse, S., 2006. National culture and financial systems. Journal of International Business Studies, 37(2): 227-247.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance, 54(2): 471-517.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 1997. Legal determinants of external finance. Journal of Finance, 52(3): 1131-50.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political Economy, 106(6): 1113-55.

Lenartowicz, T., Roth, K., 1999. A Framework for culture assessment. Journal of International Business Studies, 30(4): 781-798.

Leung, K., et al., 2005. Culture and international business: recent advances and their implications for future research. Journal of International Business Studies, 36(4): 357-378.

Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2005. Culture, law, and corporate governance. International Review of Law & Economics, 25(2): 229-255.

Margaritis, D., Psillaki, M., 2009. Capital structure, equity ownership and firm performance. Journal of Banking & Finance, 34: 621–632.

Miller, M., Modigliani, F., 1961. Dividend policy, growth, and the valuation of shares. Journal of Business, 34: 411-433.

Myers, S.C., 2001. Capital Structure. Journal of Economic Perspectives, 15(2): 81-102.

Ragin, C.C., 2000. Fuzzy-set social science. Chicago: 370 pages.

Rajan, R.G., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data. Journal of Finance, 50(5): 1421-1460.

Reuter, C.H.J., 2009. European Capital Structures: Managerial Insight and Governance Regimes available at SSRN n°1538152.

Ronen, S., Shenkar, O., 1985. Clustering countries on attitudinal dimensions: A review and synthesis. The Academy of Management Review, 10(3): 435.

Schwartz, S.H., 1994. Beyond individualism/collectivism: new cultural dimensions of values. In Kim, U., et al. (Ed.), Individualism and collectivism: theory, method and application. Newbury Park: Sage

Schwartz, S.H., 2004. Mapping and interpreting cultural differences around the world. In Vinken, H., Soeters, J., and Ester, P. (Ed.), Comparing cultures: dimensions of culture in a comparative perspective. Leiden: Brill Academic Publishers

Schwartz, S.H., 2007. Cultural and individual value correlates of capitalism: a comparative analysis. Psychological Inquiry, 18(1): 52-57.

Chapter IV- Is the relation between business cycles and leverage mediated by structural, institutional or cultural factors?

268

Scott, W.R., 1995. Institutions and Organizations. Thousand Oaks, CA: Sage.

Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance, 52(2): 737-783.

Sivakumar, K., Nakata, C., 2001. The stampede toward Hofstede's framework: avoiding the sample design pit in cross-cultural research. Journal of International Business Studies, 32(3): 555.

Stulz, R.M., 2007. The Limits of Financial Globalization. Journal of Applied Corporate Finance, 19(1): 8-15.

Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): 313-349.

Welch, I., 2004. Capital structure and stock returns. Journal of Political Economy, 112(106-131)

Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature, 38(3): 595-613.

Zingales, L., 2000. In Search of New Foundations. Journal of Finance, 55(4): 1623-1653.

* * *

Annex: Embedding culture and finance: divorce or a fresh start?

269

Culture, finance, and national institutions

 Three  essays  in  corporate  finance  

   

by  Charles-­‐Henri  J.  REUTER    

Annex

Annex: Embedding culture and finance: divorce or a fresh start?

270

Annex: Embedding culture and finance: divorce or a fresh start?

271

ANNEX

EMBEDDING CULTURE AND FINANCE: DIVORCE

OR A FRESH START?

DETAILED TABLE OF CONTENTS 1.   Introduction 273  2.   Culture, human sciences and contextual analysis 274  

2.1.   Institutional embededdness across human sciences… 275  2.2.   The law & finance approach, illustrating embededdness. 277  

3.   Culture research in Finance and what it teaches us 279  3.1.   The rising interest for “National Culture Indices” 279  3.2.   Novel directions and their interconnectedness. 281  

4.   Culture research in Finance, a conceptual framework 283  4.1.   Defining culture? 283  4.1.1.   Definitional  tactics   283  4.1.2.   Defining  culture  in  economics.   284  

4.2.   Culture: from defining to conceptualizing it 287  4.3.   A typology of research frames 288  4.4.   A practical illustration 289  4.5.   Range of topics and contextualization 290  

5.   Finance, financial cultures and rationality 293  5.1.   Culture, rationality & Behavioral Finance 293  5.1.1.   Rationality,  lack  of  rationality  and  sources  for  logic  in  action   294  5.1.2.   Individual  consistency  or  inconsistency  and  behavior   295  5.1.3.   Mental  schemes  in  individual  vs.  collective  behaviors   295  

5.2.   Finance as a Culture and the performativity of markets 296  6.   Conclusion 298  7.   Bibliography 300  

* * *

Annex: Embedding culture and finance: divorce or a fresh start?

272

Annex: Embedding culture and finance: divorce or a fresh start?

273

Embedding Culture and Finance: divorce or a fresh start?

1. Introduction  Exploring further the intimate relations between culture and finance is an important task at

hand for scholars in management, in economics and at the intersection, in finance. Much can

be learned and cross-fertilization will occur as a result. Goals include epistemological

progress, better understanding of the meaning of national embededdness or exploring further

the meaning of culture, a central concept in International Business. Finance, as a profession,

as a technique, as a lobby, occupies a special place in the “modern” world. It generates norms

and constraints for business or politics. Some concepts such as shareholder value or corporate

governance, originally derived from financial research, are now central in policy making and

in public discourse. Yet can we consider ‘Culture and Finance’ as a potential research-field?

Can we characterize one or more accepted theoretical background(s) to derive testable

hypothesis for academic research? In the “Survey” (chapter II), it is answered negatively to

both questions. Why?

Further: following the recent financial crisis, critiques coming from a variety of backgrounds

(professionals, press, politics…), put blame on individual and collective irrationality. Why do

we need to differentiate between individual and collective irrationality in Finance? Can

rational individuals be collectively irrational? Or conversely? Behavioral Finance, among

others, has scrutinized the relevance of individuals in financial research. Who has done so

with collectives, i.e. groups of individuals bound together by some specific culture?

Annex: Embedding culture and finance: divorce or a fresh start?

274

Here we proceed in four steps. First, we connect the epistemology of culture research to the

analysis of context. Doing so we emphasize how contextual analysis has developed across

scientific disciplines, but in finance. Why has Finance failed to look at context? Second we

scrutinize “culture-ventures” within financial research and we untangle commonalities (in

particular articulated around the “country puzzle”) and sources of contradictions. Third, we

introduce a typology to distinguish between kinds of culture research within finance. Doing

so we exemplify the view of culture as “layers” and “toolkit” that is becoming prominent

across disciplines and in International Business Research. Fourth, we reflect on “financial

cultures” and the central role ensued to rationality. Our abstract provides a conclusion.

2. Culture,  human  sciences  and  contextual  analysis  Recent contributions in the JIBS exemplify the past benefits of the country approach to

culture (Hofstede, 1980, 2003). Yet they underline its limits, methodological, conceptual or

practical (Au, 1999; Earley, 2006; Lenartowicz and Roth, 1999; Leung, et al., 2005; Smith,

2006 or Tung, 2008), emphasizing how complementary angles, based on the idea that

individuals relate to varying culture layers, could prove useful...

Understanding culture has been the primary purpose of a number of researchers across human

sciences, particularly in the “social sciences” (anthropology, sociology…). Denys Cuche

(2004: p. 1), in his synthesis, starts by noting that the concept is central for thinking in the

social sciences because it enables to fathom diversity in the human race... So, back in the

50’s, Kroeber and Kluckhohn (1952) listed more than two hundred definitions. Similarly, still

half a century earlier the understanding of culture and its relation to the economic and

political structured the controversy surrounding the debate between Karl Marx and Max

Weber. We provide in appendix a brief perspective on variations in culture definitions across

Annex: Embedding culture and finance: divorce or a fresh start?

275

time and civilizations, and we suggest in this appendix, that defining culture provide more

understanding about those that define it, than on what culture is per se. We come back to this

issue in section 3 of this essay.

Smelser and Swedberg (1994: p. 4) in introduction to their “Handbook of Economic

Sociology” note that while sociology is interested in “describing and explaining, rarely

predicting”, economics is interested in “predicting and explaining, rarely describing”. This

difference in polarized in the opposition “Culture and Finance”. In short “culture-research”

has epistemological roots that are opposite to that of financial research, and this opposition is

fertile. Culture provides a specific angle of approach, through its attention to the mechanisms

that bind groups of people together, through its attention to institutions and to embededdness1.

Let us provide in the next two sections a brief outlook on the rise of attention to contextual

analysis (culture, institutions, embededdness) across human sciences, but in finance.

2.1. Institutional  embededdness  across  human  sciences…    

In economics, a number of heterodox approaches have laid ground for culture research

(Throsby, 2001: chapter 1), particularly in the field of development economics (Throsby,

2001: chapter 4, Bardhan, 2005)2. At the same time a large number of economists have started

to borrow from related disciplines, emphasizing the importance of ‘systems of shared beliefs’

(i.e. culture, Aoki, 2001; Greif, 2006) while an institutional stream of research progressively

gained a larger and more prominent audience within economics (North, 1990). Many

researchers evolving in these circles have favored changes or additions in mainstream

economic methods and it is no surprise that economic historians, or game-theory economists

are strong contributors. Further, a reflection on the epistemological rooting and validity of 1 In sociology, Jepperson and Swindler (1994) stress the intimate interconnection between culture and institutions, and so do North (1990) or Williamson (2000) in economics. 2 Not coincidentally this interest corresponds to the collapse of the communist experience as the Marxist theory stipulates that culture is caused by economic interactions (Guiso, Sapienza, and Zingales 2007a: p. 6).

Annex: Embedding culture and finance: divorce or a fresh start?

276

contemporaneous neo-classical methodologies surfaced, casting light on the relation between

methodological choices and mainstream economic theorizing (McCloskey, 1983, 2002).

An interest for institutions has grown very strong in other human sciences, including

sociology, organization theory, management, or political sciences (Immergut, 1998). On the

sociological and organization theory side a lasting influence has been exercised on

management research. Renewed interest for institutional approaches emphasize the

importance of societal context, networks and institutions while they underline the need for

new, combined methodologies or converging bodies of evidence (DiMaggio and Powell,

1983, Granovetter, 1985; Scott, 1995; Whitley 2000).

In fact a number of autonomous streams of research have emerged across human sciences

fostering combined disciplinary approaches, including the North-American New Economic

Sociology (Smelser and Swedberg, 1994), the Comparative Study of Business Systems

(Hamilton and Biggart, 1988; Whitley 1992), or Socio-economics (Etzioni, 1990).

Comparable research trajectories can be regionally observed, as within French-speaking

regions (Lévesque, et al. , 2001).

Of course, I should add the specific field of cross-cultural psychology that arose as a result of

psychologists taking interest in the seminal works of Hofstede (1980). Triandis (2004)

underlines that a turning point occurred in the mid 90’s: when some mainstream psychologists

“became converted” and realized that individuals had consistent and important emotional and

cognitive differences across cultures (Markus and Kitayama, 1991). This line of research has

sustained consistent research in psychology, with dedicated journals, working initially on self-

perceptions or relation with others (extending in particular the investigation on

individualism/collectivism across nations).

Further, this line of research had lasting implications in management research and Hofstede’s

seminal contribution was augmented (Hofstede, 2003, 1988, 1991). It was subsequently

refined by a number of researchers (Franke, et al. 1991; House, et al. 1999; Inglehart, 1997;

Annex: Embedding culture and finance: divorce or a fresh start?

277

Schwartz, 1994). These progresses are sometimes ignored outside the field of International

Management Research, particularly in Finance, and newer and stronger indices based on the

survey of individual values nation-wide across countries are available. A recent gathering of

main contributors occurred in Holland in 2003 that included among others Hofstede, Triandis,

Schwartz or Inglehart. A book compiling contributions was proposed, it suggests to label

these research efforts as “dimensionalist” (Vinken, et al., 2004).

So, much has recently occurred in relation to embededdness across human sciences,

suggesting that context matters and that causal relationships are most likely to be moderated

by context. In fact recent advances in medicine now often emphasize a “context and

triggering” view of causality. Finance still lacks heterodox approaches of this kind.

2.2. The  law  and  finance  approach,  illustrating  embededdness.    

A specific case at hand is that of the law and finance research program. The law and finance

approach is derived from an earlier intellectual movement, “law and economics”, popular and

strong in the 60’s. It relies primarily on cross-sectional empirics based on legal origin and

scores measuring different types of rights (Djankov, et al., 2007; Djankov, et al., 2008;

La Porta, et al., 1997, 1998). This research program has derived from these cross-sectional

analyses strong empirical associations between the Anglo-American system of law, the

efficiency of financial markets and varying financial, economic or political dimensions,

including in particular the protection of corporate shareholders (La Porta, et al., 1998, 2000),

the valuation of firms (La Porta, et al., 1999), the efficiency of financial institutions (Botero,

et al., 2003), the regulation of economic activity (corporate governance, La Porta, et al., 2000;

the regulation of labor, Botero, et al., 2004; the regulation of entry, Djankov, et al., 2002),

growth and macro-economic performance (La Porta, et al., 1997, Gleaser, et al., 2004). The

program has then derived a number of normative conclusions pointing to the superiority of

Annex: Embedding culture and finance: divorce or a fresh start?

278

the Anglo-American legal system, suggesting that common-law systems are intrinsically

superior in delivering both economic and political freedom (La Porta, et al., 2004) and

recommending systematic reform of corporate governance rules and law.

The success of this research program in finance is significant in many ways. First, its limited

attention to contributions coming from adjacent disciplines (law, political sciences or

economics in particular) exemplifies the weaknesses of heterodox financial approaches. So,

the law and finance approach has prospered while many contributors in political sciences or

in law were attracting attention to the caveats of this approach. This included methodological

issues (Spamann, 2006) and conceptual issues (Romano, 1993; Merryman and Pérez-

Perdomo, 2007). Further Coffee (2001) and Roe (2006) were attracting attention to the

embededdness of legal and socio-political variables. Similarly, the reliability and significance

of cross-sectional comparisons, the chief methodological focus of law and finance analysis,

has been questioned as early as 2000 in development economics (Atkinson and Brandolini,

2001; Rodrik, 2005 for an outline).

Second, a key argument in the idea that “English law” works better, is the idea that colonial

inheritance provided former British colonies with stronger legal frameworks. Yet, in most of

the studies investigating the functioning and development of capital market, a 49-countries

sample is retained. In that sample, there is no former French colony. Further French-law

countries voluntarily embraced French law, more than 150 years ago, often over a Spanish

connection, whereas the rule was passed over less than 50 years ago in former British

colonies. So the methodological sophistication and the strength of empirical regularities that

guided normative recommendations did economize on conceptual analysis or historical

evidence.

Annex: Embedding culture and finance: divorce or a fresh start?

279

The law and Finance program is exemplary of the methodological and epistemological

strength and limits of financial research with regards to contextual analysis. Institutional

embededdness may condition the interpretation of the empirical regularities. One difficulty,

though, is to distinguish between various dimensions in embededdness such as culture,

institutions, and networks. So, for instance, in economics, North (1990), defines Culture as

the “antecedents” of the “informal constraints on the economy”3. Williamson (2000) defines

four level of “social analyzing” and emphasizes the strong constraining effect of the first level

on the three other and lower levels (“economic analyzing”). He stresses the importance of

alignment in frameworks, from upper to lower levels, and refers for the upper-level to

DiMaggio (culture), Granovetter (networks) and Smelser and Swedberg (embededdness).

In finance, two specific streams of analysis have come to fertilize neo-classical foundations:

game theory (Brennan, 1995) and psychology (Kahneman and Tversky, 1979). Neither

attends to context or embededdness.

3. Culture  research  in  Finance  and  what  it  teaches  us  The “Survey” (chapter II) underlines how the common thread to culture-research within

finance rests on country comparisons. He stresses the dichotomy between ‘dimensionalist’

efforts (see end of section 1.1) and other efforts.

3.1. The  rising  interest  for  “National  Culture  Indices”    

Dimensionalist research has a long track record in specific disciplines, particularly in

International Business research (Baskerville, 2003). Yet recent reviews and perspectives in

the Journal of International Business Studies stress the importance of intra-national cultural

variation (Au, 1999; Tung, 2008). Other research stresses the conflicting influence of varying

3 (a “most important” and “pervasive part”

Annex: Embedding culture and finance: divorce or a fresh start?

280

culture layers. (Earley, 2006; Leung, et al., 2005). Further, Earley underlines other promising

options than the focus on expressed values used by the national culture approach, including in

particular a potential focus on ‘meaning’ (which mediates the relation between values and

behavior) or a practical focus on behaviors.

Specifically, in Finance, Breuer and Quinten (2009) provide an overview of the financial

literature that has implemented dimensionalist frameworks. This research, through

comparatively recent, covers now a wide range of topics, including: accounting standards,

capital structure, cash holdings, financial architecture, inflation, foreign direct investments,

mergers, investors’ protection, life insurance holdings, stock market analysis, stock ownership

patterns, stock market development. Here the majority of papers are working papers and none

is published in a peer-reviewed financial journal. So, the lack of porosity of mainstream

financial research for dimensionalism is itself a finding, and it is probably one of the reasons

why Breuer and Quinten suggests that “cultural Finance” should be established independently

as a field, just like Behavioral Finance. Again the core-issue is about epistemology.

Let us remind, in concluding, that Hofstede (1980) excluded4 financial and accounting

departments from his original factorial analysis, from which he derived his four seminal

indices: “uncertainty avoidance”, “power-distance”, “individualism” and “masculinity”. It is

intriguing that so few scholars would point that out. So while many discussed the overlap

between national and organizational cultures (Hofstede, et al., 1993, 1990), very few focused

on professional cultures. This may just be one of the important issues at stake here, in Culture

and Finance, and a key cultural layer.

4 due to the lack of variance in individual answers across countries compared to those within the servicing or marketing departments

Annex: Embedding culture and finance: divorce or a fresh start?

281

3.2. Novel  directions  and  their  interconnectedness.    

Let us outline some other paths that culture research has taken in finance5. A first research

direction of research investigates the relation between national cultures, religions and

institutions (Stulz and Williamson, 2003). A second one distinguishes between various kinds

of trust6 and their importance for differences in financial phenomena such as differential

stock-market participation (across countries and socio-economic classes, Guiso, et al., 2007c),

or as financial exchange between countries (Guiso, et al., 2007b). Third, Statman (2007,

2008) focuses on varying perceptions of time, risk or uncertainty in relation to other culture

‘layers’ and in relation to stock-investing. His research has precedence and the relation

between risk-perception and risky behavior is robust and has been a primary object of

research Behavioral Finance

All of these approaches intend to unpack specific culture influences through specific cultural

lenses. They need not be opposed and in fact, they present large overlaps, often in relation to

national and institutional influences. Stulz and Williamson (2003) provide a discussion on

how Protestantism relates to decentralization and individualism, with a suggested relation to

societal trust. Further they show how the religious spirit and the related norms interact with

national institutions to foster the development of cross-national differentiation. In relation to

the opposite connection, both Guiso, et al. (2007c) and Sapienza, et al. (2007) remind us on

how religious people have differential trust-level (in a given country, for instance the U.S.).

Further, Statman (2007) relates social capital (societal trust), individualism, culture and

religion, contrasting them with various socio-economic categories (such as revenue,

education, age, gender). In addition, the “trust-literature” in economics identifies in aggregate,

and in particular at the level of regions or nations, lasting differences in the way individuals

5 See chapter II 6 Such as interpersonal trust, societal trust, rational trust, generalized trust…

Annex: Embedding culture and finance: divorce or a fresh start?

282

relate to their community. In this sense they identify lasting differences in the way societal

trust structures socio-economic and political interactions (with references to Banfield, 1958;

Putnam, 1993). All these research-paths have very different roots, but they exhibit

convergence, particularly when countries become the focus of attention.

Additionally, some authors in finance use Culture in a way directly related to the idea of

culture layers (as in Leung, et al., 2005). Guiso, et al. (2007c) show that national and socio-

economic sources for trust combine or conflict to explain individual stock-holding patterns.

Guiso, et al. (2001) refer to a rising “culture of stockholding” across continental Europe,

spreading from the top socio-economic ladders to intermediate ones in the 90’s. Similarly

Grinblatt and Keloharju (2007) shows how personal emotional bias mingle with certain

culture layers (such as generations, genders, education) to influence the average trading

volume by trader’s portfolio.

Last and importantly, we need to draw attention on research outside the financial discipline

but on financial issues. In particularly McKenzie and Millo (2003) presents finance as a

“performative culture”, and extend a longer tradition of research in American economic

sociology. We come back to this literature later, and in particular section 4.2

So, what does the recent interest for culture in finance teaches us? It exemplifies recent

development in International Business, emphasizing the practical importance of nations

(institutional embededdness), while questioning the practical relevance of the monolithic idea

of national culture. So it suggests that the “layer” approach could be favored with success,

while it leaves open room for the consideration of the professional layer, i.e. a “financial

culture”, whether “performative” or not. Overall, we suggest that culture research in finance

may not develop further without some conceptualizing.

Annex: Embedding culture and finance: divorce or a fresh start?

283

4. Culture  research  in  Finance,  a  conceptual  framework  4.1. Defining  culture?  7  

Schematically four main categories of meaning can be retained for culture, only one of them

being of interest here. The category of interest is the sociological meaning of “culture” that

broadly refers to values, norms or beliefs that are shared by some human group, reflecting its

specificity (illustrations in appendix 1). The three “residual” categories are culture as in

“agriculture” or “cell culture”, “culture as general knowledge” or “education” and “culture

as art’.

4.1.1. Definitional tactics

More specific and accurate definitions abound, and it is customary to point to the seminal

book by Kroeber and Kluckhohn (1952) that lists 251 definitions and statements about

culture. It is less customary to remind that their survey is preceded by “a general history of

the word culture”. And so the contemporaneous book by Williams (1958) focuses on tracing

the history of the world culture in England over two centuries, showing how the concept

emerged in common language, gained ground, changed, and reflected the related emergence

of industrial society, democracy (and the “masses”) as well as art, we understand it today. In

turn, all this paralleled a profound transformation in social life, providing ground for various

conceptions of the cement that unite men and women together (i.e. varying cultures).

Meanings of culture in its societal sense are not only subject to time, they are subject to

geography as well. So Elias (1982 [1939]) reminds us how enduring conceptions of culture

and civilization were central in the history of France and Germany, and their oppositions,

while the Chicago school in the U.S. focused, from the beginning of the 20th century onward,

on cultural differences, and on the process of interethnic mingling known as the “melting pot”

7 This section rephrases elements of the second chapter and develops the argumentation

Annex: Embedding culture and finance: divorce or a fresh start?

284

(Cuche, 2004: chapter 3), i.e. it focused on cultural change and cultural recombination, which

comes as a contrast to the previous idea of culture as primary enduring.

The point of these commentaries is provide a sense of how the definitions of culture may be

central in the representation of the political, intellectual and scientific environment in which

they are proposed. In fact culture-definitions may be a reflection of the objectives of the

actors and institutions that propose and use them, rather than a description of given, stable,

analytic socio-economic reality. This is notwithstanding profound differences across Western

languages in the semantics of a word, with common Latin roots, in English (culture), German

(Kultur), French (culture), etc. Hence we wish to suggest the idea of “definitional tactics”.

4.1.2. Defining culture in economics.

Let us focus on economics because it has been closely associated with finance (Brennan, AFA

presidential address, 1995). In economics, the interest for culture is ancient, and three recent

contributions about “culture and economics”, including two books, attest of its timeliness

(Throsby, 2001; Guiso, et al., 2007; De Jong, 2009∆∆∆). Culture- conceptualizations retained

by these scholars reflect the difficulty in agreeing on a definition. To some extend, they

reflect epistemological postures as well. So, on the one hand, Guiso, et al. focus on the idea of

temporal stability to identify cultural elements (see their definition in appendix 1). Doing so,

they echo a conception expressed earlier by Williamson (2000), who defines culture as one of

the elements of the more general “institutions of embededdness”, elements that result in

lasting constraints on the economy (they operate “on the order of centuries or millennia”: p.

596).

Williamson (2000)’s general assessment on the relevance of these “institutions of

embededdness“ for economic analysis is mixed. He reminds us that economics is the science

Annex: Embedding culture and finance: divorce or a fresh start?

285

of choices and scarcity, and hence that these influences are enduring and should be treated as

external and non-relevant for economic analyzing. Nevertheless, he concludes that they are

“an important but underdeveloped part of the story [that is the research program ahead for

New Institutional Economics]” (2000: p. 610); so he emphasizes the need to align economic

analysis with “upper-level societal constraints” and he underlines the promises attached to

this research-direction. This provides us with preliminary material to reflect on whether

financial research should tackle culture-issues at all: while it may not be within the prima

facie goals for financial research, it may be currently becoming a necessity.

Relating to another definition of “culture in economics”, Throsby characterizes culture as a

process, delineated by the specificity of some human group vis-à-vis other groups (appendix

1). So, he retains a sociological rather than economic conception for culture, where he allows

for fluidity and change (which was previously excluded by definition). Process-wise, instead

of focusing on durability to identify cultural elements, he focuses on cross-sectional

homogeneity. Behavior is important here, as well as meaning. Both can be observed, and their

relative homogeneity (within a group) and heterogeneity (across groups) lead to the

identification of a culture. Now, concerning a third definition in economics, De Jong

(2009)∆∆∆ retains the motivational aspect of cultures, through the concept of values (it is the

basis for the dimensionalist approach, see specific details in section 2.1), following the path

chosen in cross-cultural psychology8. In this sense, cultures are enduring like with the first

definition; but another important distinction emerges...

Defining culture is not only difficult, it engaging too. DiMaggio (1994, 1997) reminds us how

the cognitive aspects of culture provide ground for “enabling” views on culture: culture not

only constrains, but structures and fosters behaviors, communication, decision-making, etc.

8 Another classic distinction contrasts conscious (symbols, discourses) vs. unconscious aspects of culture. It is an important distinction in the social sciences literature, as the popular and pervasive “iceberg conception” of culture attests.

Annex: Embedding culture and finance: divorce or a fresh start?

286

DiMaggio emphasizes how this distinction is central in the epistemological structuring of

scientific disciplines and in the choice of methodologies9. In fact he concludes that economics

itself can be seen as a “cultural system” (1994: p. 46 and 47), a point that has been recently

transposed to finance with force (MacKenzie and Millo, 2003, a seminal paper for what is

becoming known as the social studies of finance, below S.S.F.). Let us quote Larry Summers

(1985: p. 1)° in comparison: “the fields of economics and finance are allied. Indeed finance is

sometimes defined as a subfield of economics [… ] yet we are increasingly witnessing the

development of two cultures [emphasis added]”. Larry Summers’ statement may seem

anecdotic, because it is not clear whether it assimilates finance to something that could be

described as a “cultural system”. Yet it provides ground for more than just familiarity- or

knowledge of the financial field and theory. The distinction with “economic culture” suggests

some form of relative coherence that distinguishes the two sub-fields in theory and in

practice. This coherence excludes anecdotic views. In fact, this statement may seem anecdotic

precisely because of epistemological postures. The idea of a coherent whole, guiding and

structuring behaviors, is an “enabling” view of culture, so that the idea of a financial culture

may seem straightforward to a sociologist, while it is fuzzy and apparently anecdotic for a

“constraining” view of culture, as in economics. 10

9 DiMaggio calls attention to the dichotomy between enabling and constraining views of culture. He notes a “rough analytic distinction between forms of culture that are characteristically constitutive (categories, scripts, conceptions of agency, notions of techniques) and forms that are predominantly regulatory (norms, values, routines)”. Then he underlines that most economists are in the second category (emphasizing regulatory aspects), treating “economic behavior as analytically distinct from culture, and stress[ing] the ways in which norms and conventions constrain the individual’s […] pursuit of self-interest”, while most anthropologists are in the first category (enabling views). This distinction is emphasized in concluding his introductory chapter on defining culture (De Jong, 2009∆∆∆: introduction) We will illustrate this distinction enabling / constraining with an ad-hoc example with “cultural affinity hypothesis” in section 3.3 10 In their account of the reticence of the economic field, Guiso, et al. emphasize “in the last fifty years, most economists have been reluctant to rely on culture as a possible explanatory variable” (introduction). They attend to this lack of past interest, putting some of the blame on the “inevitable grittiness” of cultural variables, hardly reconcilable with “the parsimony of deductive models”, their “formal elegance”, the priority for “logical consistency over empirical relevance” and the methodological needs for “designing testable hypothesis” characterizing the economic field. Similar commentaries can be found in De Jong∆∆∆ (chapter 1&2). Together with DiMaggio’s Culture and the Economy, they set the epistemological context to whether or not culture variables should be considered or not in finance.

Annex: Embedding culture and finance: divorce or a fresh start?

287

So, provided the complexity and stakes involved in a definition of “culture”, the “Survey”

(chapter II) favors a grounded approach, exploring the academic literature in finance and

looking for definitions emerging from a systematic screening process. Further he provides in

conclusion a sample of definitions as obtained from the varying authors he cites, giving a

sense of the breadth and divergence of the underlying conceptual orientations.

4.2. Culture:  from  defining  to  conceptualizing  it  

In economics Guiso, et al. (2007a) propose to define culture as “those customary beliefs,

values, and social constraints that ethnic, religious and social groups transmit fairly

unchanged from generation to generation”. Throsby (2001) proposes to define it as “a set of

values, beliefs, traditions, customs, etc. which serve to identify and bind a group together”11.

The contrast between the two definitions is deeper than expected from a quick reading,

because Throsby’s idea of “binding together” is essential. To allow for rich

conceptualizations of culture, akin to those favored by newer cognitive advances (i.e. akin to

culture layers, see DiMaggio, 1997), on must allow for overlap, loose organization, change,

and contingency, elements that Guiso, et al. exclude by definition (“unchanging …beliefs,

values and social constraints”12). Similarly Williamson (2000) use varying time-horizons to

characterize his 4 levels of social analyzing (see above), and characterizes the upper level as

“unchanging … in terms of millennia”

So Throsby replaces this condition of continuity for any given norm, value… by a focus on

the group that is bound together by some norm, value… (i.e. a culture). Then the primary

issue is not to identify ex-ante some set of values, beliefs or customs, but to show ex-post that

11 Chapter II , appendix 1 12 It I in fact surprising that Guiso, et al. should include this condition of continuity here while they use the idea of a rising “equity culture” as mentioned earlier.

Annex: Embedding culture and finance: divorce or a fresh start?

288

they have practical relevance as the culture within a specific and homogeneous group of

people to some purpose over a defined period of time.

As a result, trying to shy away from controversies about what culture is (or should be) we

suggest that finance researchers should implement backward definitions of culture. They

should define culture from the financial phenomena being studied, and they should relate it to

collective behaviors grounded in some specific culture, relevant to that financial phenomenon.

In particular, this implies that culture research in finance can only develop while borrowing

from adjacent disciplines (specific culture definitions, with appropriate depth), while

developing conceptualization and approaches based on “converging bodies” of evidence.

While this does not preclude neo-classical methodologies (and quantitative proceedings), it

requires their extension13.

4.3. A  typology  of  research  frames  

Paul DiMaggio (1994)14 suggests that “to establish a cultural effect, one must meet two

conditions. First, one must demonstrate that individuals or collective actors with some

specific kind of culture behave differently than others without it […] Second, one must

demonstrate that such differences do more than mediate structural or material influences

[…]”. This provides us with two axes for our typology, a focus on actors (provided a

condition of homogeneity on the trait being considered) and on a mechanism (relating the trait

to the financial outcome, hence our suggestion of a ‘backward definition’ process).

13 In economics, see Throsby (2001, p. 29-30) on methodological requirements for studying culture. 14 In economic, he is mentioned as an authoritarian source on culture by Williamson (2000) or Guiso, et al. (2007a)

Annex: Embedding culture and finance: divorce or a fresh start?

289

Regarding the definition of traits, we suggest to focus on the different types of actors involved

in financial phenomena, with a primary focus on the most common dichotomy used in

Behavioral Finance: corporate- vs. market agents (Shefrin, 1999, Baker, et al., 2005).

From these considerations, we propose a two-axis matrix for studying culture in Finance,

where the first axis pertains to the type of actors being studied, from most specific groups of

individuals (and broadly more homogeneous), to collectives and to abstract notion of agency

(as in the study of the antecedents of financial laws for instance). The second axis separates

out the analysis of situations where a cultural (or cognitive or behavioral) effect is presumed

to be direct from situations where it is embedded in institutions, in networks, in polities or

private organizations. The purpose of this dichotomy is to attract attention to different

methodological needs, to the potential importance of different layers of culture, to their inter-

relation, to different type of content to be provided to ‘culture’, and still to different time-

horizons.

4.4. A  practical  illustration  

Financial studies involving homogeneous, educated and informed groups of actors, such as

traders or most stock-market participants are unlikely to be able to rely on dimensionalist

frameworks. Indeed, dimensionalists’ frameworks are derived from national populations and

traders are not representative within these populations. Second the sub-population of traders

may not have the same relationship to overall populations across nations. Simplifying for the

sake of illustration, the political spectrum of American traders might be loosely split between

democrats and republicans (see where the financing comes from) while that of French traders

may be more conservative (in national context).

Annex: Embedding culture and finance: divorce or a fresh start?

290

Third the activation of unconscious cultural frameworks (and in particular national culture

reference-frames) might have less or a differential pattern of activation among traders. So

almost all traders are English-speakers (often bilingual), they read the Anglo-American

business press; their information and interpretative frames differ from those of their co-

nationals. They often live or have lived abroad. If views of culture as more or less conscious

are retained15, then the fact that they are bilingual means that they must have experience some

degree of acculturation in national culture term.

So in this example we suggest that national cultures are less promising than other layers of

culture or other behavioral biases and that a direct methodological approach for instance

through surveys is promising because the presumption is that of a relative homogeneity of the

sub-population being considered.

4.5. Range  of  topics  and  contextualization  

We propose to apply this two-axis matrix to a range of financial topics that can be considered

as promising. This range of topics can be collected ad-hoc from a number of sources

including in particular Stulz (2007) who identifies nine areas of financial research that are

affected by what he labels a “country puzzle”16.

Other topics can be identified in the Behavioral Finance literature, in connection with what

Breuer and Quinten (2008) term “Cultural Finance”, or still in the sociology of markets or

finance (including for instance Zuckerman, 2004; McKenzie and Millo, 2003; Podolny, 1993;

Mizruchi and Stearns, 1994).

15 culture in opposition to symbols, ideologies, social prescriptions, see Jepperson and Swindler, 1994 16 “we would expect countries, per se, to matter little in finance [and that in fact] a firm’s country of incorporation is a more important determinant of its financial policies than its industry”. He concludes that these “national puzzles […] have become paradoxes that are explored in many papers”. Nine research-areas are mentioned: portfolio choice, savings and investments, consumption, stock returns, the size of stock markets, corporate ownership, firm size, capital structure and governance.

Annex: Embedding culture and finance: divorce or a fresh start?

291

TABLE 1: Financial research-topics grouped according to aggregation mechanism

A key issue in dealing with this list of topics relate to the mechanism relating culture to

collective behaviors to finance: either through a direct behavioral rule, aggregated over

individuals, or offer and demand aggregation (atomistic markets), through more complex

market or contracting mechanisms, with or without path-dependency, institutional co-

determination, institutional embededdness…. As an illustration, we provide a preliminary

table built in this way, which purpose is to identify the range of topics grouped according to

the primary type of underlying aggregation mechanisms.

The next conceptual step is to systematize the type of analysis (volunteered in illustration

above in the case of traders), to all groups of financial agents depending on mechanisms.

Annex: Embedding culture and finance: divorce or a fresh start?

292

Then it is possible to provide a systematic assessment of what sort of culture layer is the more

promising, and what type of methods are theoretically needed (quantitative, qualitative,

theoretical, conceptual), whether existing definitions of culture can be used, from which

discipline and whether the presumption is that bodies of evidence (rather than research

papers) will be required to explore a topic. So for instance Roe (2006)’s account on corporate

governance practices across countries include thick description, ethnographic work, surveys,

secondary data analysis and quantitative proceedings. Similar combination of methods are

suggested in multiple chapters of Throsby (2001)’s, Aoki (2001)’s or Greif (2006)’s books in

relation to culture and “systems of shared beliefs”. The study of management practices and

economic relations in the Comparative Study of Business Systems follows the same

methodological path (Guillén, 1994; Djelic, 2001; Redding, 2005; Whitley, 1992, 2000…).

TABLE 2: Conceptualizing culture, and culture’s consequence: a typology of methods and

types of culture depending on financial subjects

Annex: Embedding culture and finance: divorce or a fresh start?

293

Our conceptual proposition follows a tradition in the Management literature (for instance

Payne, et al., 2008 relating to social capital). Our purpose is not to be comprehensive but to

provide a conceptual grid to facilitate exchanges among scholars and to guide understanding.

So relating to table 2, dimensionalist researchers are pursuing issues deeply embedded in

institutional contexts, they do not consider the behavior of actors per se and they situated in

the lower-right part of the matrix. Correspondingly, they tackle macro-financial issues. On the

other hand, the realm of behavioral finance, focusing on direct cognitive effects relates to

specific groups of actors and is situated in the upper-left quadrant.

5. Finance,  financial  cultures  and  rationality  So, we outline the intrinsic complexity of culture studies and the specificity that Finance, i.e.

the financial discipline, the financial techniques and financial actors bring to this research.

Our typology should provide a tool to manage this complexity and untangle influences. Now

we would like to delve into two specific subjects brought forward by this typology before

concluding; they both pertain to the inter-relation between finance, culture and rationality.

5.1. Culture,  rationality  and  Behavioral  Finance    

While reviews on Behavioral Finance by Baker, et al. (2005) or by Shefrin (1999) do not

consider culture, nor mention any culture-related words, Shiller (1997) or Greenfich (2005)

link culture and behavioral biases. Additionally, Charreaux (2005, 2002) suggests an overlap

between the two lines of research.

Annex: Embedding culture and finance: divorce or a fresh start?

294

5.1.1. Rationality, lack of rationality and sources for logic in action

As derived from the reviews just mentioned, Behavioral finance is mainly concerned with

explaining irrational behaviors in financial decision-making. Shefrin mentions that a key

element in definition relates to “the errors made” by managers, errors derived from “cognitive

imperfections or emotional influences”. Baker, et al. note “this research replaces the

traditional rationality assumptions with potentially more realistic behavioral assumptions”.

Both definitions emphasize lack of rationality as a departure-point for Behavioral Finance

studies.

In contrast culture-research does not imply irrationality. Seminal works exemplify strong

logics in action. Durkheim (1912) insists on systems of integrated beliefs, while Weber

(1914) underline the co-existence of varying forms of rationality, deductive, inductive,

teleological- and value-rationality. In his review, Shiller defines the purpose of Behavioral

Finance as “understanding anomalies in financial markets”, shifting focus away from errors

in individual behavior to anomalies in aggregate patterns (offer and demand, collective

behaviors…). Shiller notes further that “one difficulty [with cultural biases is] that of

disentangling the ‘rational’ reasons for the imitation of others from the purely psychological

ones”.

So while Behavioral Finance is often concerned with a lack of rationality, culture-studies

emphasize different and potentially conflicting sources for rationality, for logic in individual

behavior. In particular, value-rationality as derived from Weber is central to many researchers

in culture. Indeed, the primacy of values is utilized to derive indices of national culture in the

dimensionalist literature (Vinken, et al., 2004; Breuer and Quinten, 2009) or in political

sciences (Inglehart, 1997). Emotion-research and research on motivation in organizational

behavior, sociology and social psychology have built on similar grounds (McClelland 1953).

Annex: Embedding culture and finance: divorce or a fresh start?

295

5.1.2. Individual consistency or inconsistency and behavior

Additionally, “cognitive imperfections” are generally considered relatively stable across time

and personalities, implying that under identical clinical conditions in experiences, the same

imperfections will be reliably observed. There is consistency in the behavioral biases across

persons and time, and this individual consistency lead to inconsistency in social outcomes, or

errors observed on markets.

In contrast, culture characterizes reliability of individual behaviors within a group, even

though inconsistencies would be individually expected. A key difference is the nature of the

scientific methods retained, clinical and experimental designs being common in psychology,

while most of culture studies focus on experiences in vivo, or observations (anthropological

methods, grounded theory). Moreover, sources of cultural homogeneity within existing

groups include age (generational gap), religion, gender, social status, ethnicity, etc. These

sources of intra-group homogeneity enable to think the diversity across groups, implying

inconsistency in cognitive schemata across persons, time, in contrast to the psychological

baseline.

5.1.3. Mental schemes in individual vs. collective behaviors

Another key difference between behavioral finance and culture-study is the focus of interest.

Behavioral finance centers on individual’s psyche while culture-research focuses on groups;

DiMaggio (1997) emphasize that culture is “supra-individual”. In the dimensionalist

literature, an explicit distinction is made between individual and cultural levels of analysis

(Hofstede, 2003; Schwartz, 1994). In fact Schwartz (1999) provides different sets of values

for the individual and collective levels.

Annex: Embedding culture and finance: divorce or a fresh start?

296

While the relation from cognition to behavior is straightforward in psychology, it is more

complex for culture studies. So DiMaggio (1997) emphasize a distinction between common

and collective beliefs: culture is common when it is “an aggregate of individual beliefs or

preferences” and collective when it is a “shared representations of individual beliefs” (p.

272). Similarly, the Globe project departs from a distinction between modal values vs. modal

beliefs: the world can be viewed “as it is” or as “how it should be” (House, et al. 1999 p. 14).

Still, Shiller (1997) mentions culture in two ways, both as an element of national

differentiation and as a “global culture”, that is an element of world-wide homogenization.

Doing so, he stresses the importance of culture change, the fluidity of culture-dynamics and

the differentiation between the observation of behaviors and their cultural determinants.

5.2. Finance  as  a  Culture  and  the  performativity  of  markets  

Finance, unlike for instance Economics, is a specific object for culture research: it is, first, the

abstraction and complement of concrete exchanges (i.e. “Trade domain”17), through its

liquidity function. Second, through its value function, finance (money) structures the social

domain (Zelizer, 1994), orients value creation and redistribution (inflation). Third, it is a

profession and a concrete socio-economic agent (a network and an organizational field),

which members have achieved central positions (Fligstein, 1993). As a network, or a lobby it

has lately achieved a position of power, comparable, at least, to that of the most powerful

states. Hence and last, it has become a source of legitimacy for the Organization Domain,

generating norms applicable to a wide range of Organizations. As such, the meaning

“financial culture” is more concrete and broader than that of “economic culture” (as defined

for instance in DiMaggio, 1994).

17 Aoki, 2001 2001 for the terminology. Aoki (p. 23) distinguishes various domains: Commons, Trade, Organization, Polity, Social Exchange and then the “Organizational Field”,

Annex: Embedding culture and finance: divorce or a fresh start?

297

Recent progresses in the social study of finance underline the fact that Financial Culture is

performative (MacKenzie and Millo, 2003) and suggestions arise that Finance should be re-

embedded18. Yet, this performativity itself is a sign that Finance, its technique, its culture, its

professionals, its networks may have become central in the economic and political spheres

and that in converse, the social may be dis-embedding itself (as in the past, Polanyi, 1944,

1957?). In particular the sheer power of the financial lobby, its political connectedness, its

economic embededdness, may isolate social preoccupations from the other preoccupations;

the sheer efficiency of financial markets (but for their spectacular and ever more frequent

accidents), must have some role in their “performativity”. The epistemology if financial

research, including

Recently Mayer (2008) reminds us how market operators or investors were part of close-knit

local communities before the world war. This influenced the way trust was established and

maintained, and this influenced the time-horizon of actors. In cultural terms, the “professional

layer” of financial culture was rooted into local communities, diverse personal histories and

backgrounds and into a strong interdependence between the financial, the industrial and the

economic (Mayer emphasizes how market operators were local business leaders as well).

Today the strong integration of financial actors, the professionalization of financial services,

the leveling up and specialization of financial curricula has changed the way the “financial

culture” is created, shaped, transmitted. If financial market prices are considered as the result

of the clearing of offer and demand by market-actors, the “cultural” profile of these market-

actors must have an influence. Does the greater specialization of the financial community

have an influence on the price-discovery mechanism?... Modeling the price mechanism

enables to capture the quality and accuracy of information around some norm. Then an issue

is whether this norm is reflecting that of the financial community’s consensus versus that of

18 see Social Study Finance of Association’s call for contributions: http://ssfa.free.fr/hoprubrique.php?id_rub=39

Annex: Embedding culture and finance: divorce or a fresh start?

298

the wider society (then which one? which nation? which social classes?), or that of the

underlying “true economic values”. Greater specialization and integration of professional

communities has an effect on society’s functioning (Durkheim, 1897/1967on the interplay

between mechanic and organic solidarity)

6. Conclusion  We provide a perspective on the contradiction provided by a rising culture-interest, and, the

lack of a successful culture-paradigm in financial research. We emphasize how “culture-

research” and financial research rest on strikingly different epistemological roots. Culture is

concerned with conditionality, conceptuality, ends (values), qualitative evidence, limited

diversity, fades, and it is guided by understanding. Culture largely evades precise definition

and measurement. Finance is established on deductive and instrumental rationality, on

modeling, on measurement, on a-temporality, on an appeal for marginal analysis and it is

oriented by quantitative prediction.

To overcome this tension we propose a conceptual tool to think about culture in financial

research. Chiefly we distinguish between the types of actors being involved in financial

phenomena and their expected behavioral and cognitive homogeneity. Then we recommend

defining culture ‘backward’, from primo the mechanism relating it to the financial outcome,

and segundo the homogeneity within the group regarding that mechanism.

This conceptual framework can be of use to culture-scholars across disciplines. In

management, it exemplifies how the view of culture as “toolkits”, “layers” and “repertoires”

can be intellectually powerful and empirically fruitful within a specific field of inquiry

(finance). In finance, we contribute to the understanding of the tensions around “culture-

research”, and, outlining the epistemological roots to that tension, we exemplify one path that

Annex: Embedding culture and finance: divorce or a fresh start?

299

a re-foundation (Zingales, 2000) could engage into. Overall, we provide a grid to understand,

and partly reconcile, the complementary foundations of culture research about financial issues

within finance, or in sociology, economics, psychology…

Eventually we reflect on how the culture-approach, unlike Behavioral Finance, can be

compatible with a view of Finance as primarily, conditionally and deductively rational. In a

way reminiscent of Simon (1945/1976) suggesting that complexity in Organizations brings

limits to rationality, we suggest, here, that culture brings conditionality and temporality to the

scientific analysis and quantitative modeling of financial phenomena (culture understood here

as the grip exerted by collectives on individual cognitions, behaviors and the rules structuring

their interactions).

* * *

Annex: Embedding culture and finance: divorce or a fresh start?

300

7. Bibliography  

Aoki, M., 2001. Toward comparative institutional analysis. Cambridge: MIT Press. 467 pages.

Atkinson, A.B., Brandolini, A., 2001. Promise and Pitfalls in the Use of "Secondary" Data-Sets: Income Inequality in OECD Countries as a Case Study. Journal of Economic Literature, 39(3): p. 771-779.

Au, K.Y., 1999. Intra-cultural variation: evidence and implications for international business. Journal of International Business Studies, 30(4): p. 799-812.

Baker, M., Ruback, R.S., Wurgler, J., 2005. Behavioral Corporate Finance: a survey. In Eckbo, E. (Eds.), Handbook in corporate finance: Empirical Corporate Finance. Harvard Business School Press

Banfield, E., 1958. The Moral Basis of a Backward Society. New York: Free Press.pages.

Bardhan, P.K., 2005. Institutions matter, but which ones? Economics of Transition, 13(3): p. 499-532.

Baskerville, R.F., 2003. Hofstede never studied culture. Accounting, Organizations and Society, 28(1): p. 1-14.

Botero, J.C., La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Volokh, A., 2003. Judicial Reform. World Bank Research Observer, 18(1): p. 61.

Botero, J.C., Djankov, S., La Porta, R., Lopez-De-Silanes, F., Shleifer, A., 2004. The Regulation of Labor. Quarterly Journal of Economics, 119(4): p. 1339-1382.

Brennan, M.J., 1995. Corporate Finance over the past 25 years. Financial Management, 24(2): p. 9-22.

Breuer, W., Salzmann, A., Cultural Dimensions of Corporate Governance Systems, in Working Paper (available from SSRN). 2008.

Breuer, W., Quinten, B., Cultural Finance, in Working Paper (available from SSRN). 2009.

Charreaux, G., 2002. Variation sur le thème: 'à la recherche de nouvelles fondations pour la finance et la gouvernance d'entreprise'. Finance Contrôle Stratégie, 5(3): p. 5-68.

Charreaux, G., 2005. Pour une gouvernance d'entreprise 'comportementale'. Une réflexion exploratoire. Revue Française de Gestion, 4(157): p. 215-238.

Coffee, J.C., Do Norms Matter? A Cross-Country Examination of the Private Benefits of Control, in Columbia Law and Economics Working Paper. 2001.

Cuche, D., 2004. La Notion de culture dans les sciences sociales. Paris: La Découverte - Repères. 124 pages.

De Jong, E., 2009. Culture and economics: on values, economics and international business. Routledge.pages.

DiMaggio, P., 1997. Culture and Cognition. Annual Review of Sociology, 23: p. 263-287.

Annex: Embedding culture and finance: divorce or a fresh start?

301

DiMaggio, P.J., 1994. Culture and Economy. In Smelser, N.J. and Swedberg, R. (Eds.), The Handbook of Economic Sociology. Princeton: Princeton University Press

DiMaggio, P.J., Powell, W.W., 1983. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2): p. 147-160.

Djankov, S., Mcliesh, C., Shleifer, A., 2007. Private Credit in 129 Countries. Journal of Financial Economics, 84(2): p. 299-329.

Djankov, S., Porta, R.L., Lopez-de-Silanes, F., Shleifer, A., 2002. The regulation of entry. The Quarterly Journal of Economics:

Djankov, S., La Porta, R., Lopez De Silanes, F., Shleifer, A., 2008. The Law and Economics of Self-Dealing. Journal of Financial Economics, 88(3): p. 430-465.

Djelic, M.-L., 2001. Exporting the American Model: The Postwar Transformation of European Business. Oxford: Oxford University Press. 328 pages.

Durkheim, E., 1912. The elementary form of religious life. Reprinted by Oxford University Press.pages.

Durkheim, É., 1897/1967. De la division du travail social. Paris: Les Presses universitaires de France. 416 pages.

Earley, P.C., 2006. Leading cultural research in the future: a matter of paradigms and taste. Journal of International Business Studies, 37(6): p. 922-931.

Elias, N., 1982 [1939]. The civilizing process (vol. II state formation and civilization). Oxford:pages.

Etzioni, A., 1990. Moral Dimension : Toward a New Economics. Free Press. 314 pages.

Fligstein, N., 1993. The Transformation of Corporate Control. Harvard University Press. 408 pages.

Franke, R.H., Hofstede, G., Bond, M.H., 1991. Cultural Roots of Economic-Performance - a Research Note. Strategic Management Journal, 12: p. 165-173.

Glaeser, E.L., Porta, R.L., Lopez-de-Silanes, F., Shleifer, A., 2004. Do Institutions Cause Growth? Journal of Economic Growth:

Granovetter, M., 1985. Economic action and social structure: the problem of embeddedness. American Journal of Sociology, 91(3): p. 481-510.

Greenfich, P., Behavioral Finance Definitions : Main Concepts, in http ://perso.wanadoo.fr/greenwich/bfdef.htm. 2005.

Greif, A., 2006. Institutions and the path to the modern economy: lessons from medieval trade. Cambridge University Press. 503 pages.

Grinblatt, M., Keloharju, M., Sensation seeking, overconfidence, and trading Activity, in Journal of Finance (forthcoming). 2007.

Guillen, M.F., 1994. Models of management. Work, authority and organization in a comparative perspective. Chicago: University of Chicago Press. 424 pages.

Annex: Embedding culture and finance: divorce or a fresh start?

302

Guiso, L., Haliassos, M., Jappelli, T., 2001. Household Portfolios. Cambridge: MIT press. 508 pages.

Guiso, L., Sapienza, P., Zingales, L., 2007. Does culture affect economic outcomes? Journal of Economic Perspectives, 20(2): p. 23-48.

Guiso, L., Sapienza, P., Zingales, L., 2007a. Does culture affect economic outcomes? Journal of Economic Perspectives, 20(2): p. 23-48.

Guiso, L., Sapienza, P., Zingales, L., Culture biases in economic exchange, in NBER Working paper series. 2007b.

Guiso, L., Sapienza, P., Zingales, L., Trusting the Stock Market, in ECGI - Finance Working Paper. 2007c.

Hamilton, G.G., Biggart, N.W., 1988. Market, Culture, and Authority: A Comparative Analysis of Management and Organization in the Far East. The American Journal of Sociology, 94(Supplement): p. S52-S94.

Hofstede, G., 1980. Culture's consequence. New-York: Sage publications.pages.

Hofstede, G., 2003. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

Hofstede, G., Bond, M.H., 1988. The Confucius Connection - from Cultural Roots to Economic-Growth. Organizational Dynamics, 16(4): p. 5-21.

Hofstede, G., Hofstede, G.J., 1991. Cultures and organizations: software for the mind. New-York: McGraw-Hill. 300 pages.

Hofstede, G., Bond, M.H., Luk, C.-l., 1993. Individual perceptions of organizational cultures: A methodological treatise on levels of analysis. Organization Studies, 14(4): p. 483.

Hofstede, G., Neuijen, B., Ohayv, D.D., Sanders, G., 1990. Measuring Organizational Cultures: A Qualitative and Quantitative Study Across Twenty Cases. Administrative Science Quarterly, 35(2): p. 286.

House, R.J., Hanges, P.J., Ruiz-Quintanilla, S.A., Dorfman, P.W., Javidan, M., Dickson, M., Gupta, V., 1999. Cultural influences on leadership and organizations: project Globe. In Mobley, W.H. (Eds.), Advances in global leadership. Stamford, CT: JAI Press. 171-233.

Immergut, E.M., 1998. The Theoretical Core of the New Institutionalism. Politics & Society, 26(No. 1): p. 5 - 34.

Inglehart, R., 1997. Modernization and postmodernization: cultural economic and political change. Princeton: Princeton University Press. 440 pages.

Jepperson, R.L., Swindler, A., 1994. What Properties of Culture should we Measure? Poetics, 22: p. 359-371.

Kahneman, D., Tversky, A., 1979. Prospect theory: An analysis of decision under risk. Econometrica, 47: p. 263-291.

Kroeber, A.L., Kluckhohn, C., 1952. Culture: a critical review of concepts and definitions. Harvard: Peabody Museum of American Archaeology and Ethnology.pages.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 1997. Legal determinants of external finance. Journal of Finance, 52(3): p. 1131-50.

Annex: Embedding culture and finance: divorce or a fresh start?

303

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political Economy, 106(6): p. 1113-55.

La Porta, R., Lopez-de-Silanes, F., Schleifer, A., Vishny, R., Investor Protection and Corporate Valuation., in Working Papers -- Yale School of Management's Economics Research Network. 1999. p. 1.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000. Investor protection and corporate governance. Journal of Financial Economics, 58(1/2): p. 3-27.

La Porta, R., Lopez-de-Silanes, F., Pop-Eleches, C., Schleifer, A., 2004. Judicial check and balances. Journal of Political Economy, 112(2): p. 445-470.

Lenartowicz, T., Roth, K., 1999. A Framework for Culture Assessment. Journal of International Business Studies, 30(4): p. 781-798.

Leung, K., Bhagat, R.S., Buchan, N.R., Erez, M., Gibson, C.B., 2005. Culture and International Business: Recent Advances and Their Implications for Future Research. Journal of International Business Studies, 36(4): p. 357-378.

Lévesque, B., Bourque, G.L., Forgues, E., 2001. La nouvelle sociologie économique. Desclée de Brouwer. 268 pages.

MacKenzie, D., Millo, Y., 2003. Constructing a market, performing theory: the historical sociology of a financial derivatives exchange. American Journal of Sociology, 109(1): p. 107-145.

Markus, H.R., Kitayama, S., 1991. Culture and Self: Implications for Cognition, Emotion and Motivation. Psychological Review, 98: p. 225-253.

Mayer, C., 2008. Trust in Financial Markets. European Financial Management, 14(4): p. 617-632.

McClelland, D.C., Atkinson, J.W., Clark, R.A., Lowell, E.L., 1953. The achievement motive. New-York: John Wiley & Sons (second. ed. Dec. 1976). 410 pages.

McCloskey, D.D., 1983. The rhetoric of economics. Journal of Economic Literature, 21: p. 481-517.

McCloskey, D.D., 2002. The Secret Sins of Economics. Chicago: Prickly Paradigm Press.pages.

Merryman, J.H., Pérez-Perdomo, R., 2007. The civil law tradition: an introduction to the legal systems of Europe and Latin America. Stanford: Stanford University Press. 173 pages.

Mizruchi, M.S., Stearns, L.B., 1994. Money, Banking and Financial Markets. In Smelser, N.J. and Swedberg, R. (Eds.), The Handbook of Economic Sociology. Princeton University Press

North, D., 1990. Institutions, institutional change and economic performance. Cambridge: Cambridge University Press. 159 pages.

Payne, G.T., Moore, C.B., Griffis, S., Autry, C., Social Capital Research in Management: A Review of Conceptualizations, Measurements, and Implications, in Invited for revision for special annual review issue of Journal of Management. (Invitation Received: 3- 13-09). 2008.

Annex: Embedding culture and finance: divorce or a fresh start?

304

Podolny, J.M., 1993. A Status-Based Model of Market Competition. American Journal of Sociology, 98(4): p. 829.

Polanyi, K., 1944, 1957. The Great Transformation: The Political and Economic Origins of Our Time. Boston:pages.

Putnam, R., 1993. Making Democracy Work. Princeton University Press, Princeton, NJ.pages.

Redding, G., 2005. The thick description and comparison of societal systems of capitalism. Journal of International Business Studies, 36(2): p. 123-155.

Rodrik, D., Why We Learn Nothing From Regressing Economic Growth on Policies, in Harvard University Working Paper. 2005.

Roe, M.J., 2006. Political Determinants of Corporate Governance: Political Context, Corporate Impact. New-York: Oxford University Press. 256 pages.

Romano, R., 1993. A Cautionary Note on Drawing Lessons from Comparative Corporate Law. Yale Law Journal, 102(8): p. 2021-2037.

Sapienza, P., Toldra, A., Zingales, L., Understanding Trust, in NBER Working Paper. 2007.

Schwartz, S.H., 1994. Beyond individualism/collectivism: new cultural dimensions of values. In Kim, U., et al. (Eds.), Individualism and Collectivism: Theory, Method and Application. Newbury Park: Sage

Schwartz, S.H., 1999. A Theory of Cultural Values and Some Implications for Work. Applied Psychology: An International Review, 48(1): p. 23-47.

Scott, W.R., 1995. Institutions and Organizations. Thousand Oaks, CA: Sage.pages.

Shefrin, H., 1999. Behavioral Corporate Finance. In Shefrin, H. (Eds.), Beyond Greed and Fear: Understanding Behavioral Finance and the Psychology of Investing. Harvard Business School Press

Shiller, R.J., Human behavior and the efficiency of the financial system, in Handbook of macroeconomics, Taylor, J.B. and Woodford, M., Editors. 1997.

Simon, H.A., 1945/1976. Administrative Behavior: a study of decision-making processes in administrative organization. New York: The Free Press. 1364 pages.

Smelser, N.J., Swedberg, R., 1994. The Handbook of Economic Sociology. Princeton University Press.pages.

Smelser, N.J., Swedberg, R., 1994. The sociological perspective on the economy. In Smelser, N.J. and Swedberg, R. (Eds.), The Handbook of Economic Sociology.

Smith, P.B., 2006. When elephants fight, the grass gets trampled: the GLOBE and Hofstede projects. Journal of International Business Studies, 37(6): p. 915-921.

Spamann, H., 2006. On the Insignificance and/or Endogeneity of La Porta et al.'s 'Anti-Director Rights Index' under Consistent Coding. SSRN eLibrary.

Statman, M., 2007. local ethics in a global world. Financial Analysts Journal, 63(3): p. 32-41.

Statman, M., 2008. Countries and Culture in Behavioral Finance. CFA Institute Conference Proceedings Quarterly, 25(3): p. 38-41.

Annex: Embedding culture and finance: divorce or a fresh start?

305

Stulz, R.M., 2007. The Limits of Financial Globalization. Journal of Applied Corporate Finance, 19(1): p. 8-15.

Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): p. 313-349.

Summers, L.H., 1985. On economics and finance. Journal of Finance, 40(3): p. 633-635.

Throsby, C.D., 2001. Economics and culture. Cambridge University Press. 208 pages.

Triandis, H.C., 2004. The many dimensions of culture. Academy of Management Executive, 18(1): p. 88-93.

Tung, R.L., 2008. The cross-cultural research imperative: the need to balance cross-national and intra-national diversity. Journal of International Business Studies, 39(1): p. 41-46.

Vinken, H., Soeters, J., Ester, P., 2004. Comparing Cultures: Dimensions of Culture in a Comparative Perspective. Leiden: Brill Academic Publishers.pages.

Weber, M., 1914. Economy and Society: An Outline of Interpretive Sociology. pages.

Whitley, R., 1992. European Business Systems: Firms and Markets in Their National Contexts. Sage. 302 pages.

Whitley, R., 2000. Divergent Capitalisms: The Social Structuring and Change of Business Systems. Oxford University Press.pages.

Williams, R., 1958. Culture and society. London:pages.

Williamson, O.E., 2000. The New Institutional Economics: Taking Stock, Looking Ahead. Journal of Economic Literature, 38(3): p. 595-613.

Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature, 38(3): p. 595-613.

Zelizer, V.A., 1994. The Social Meaning of Money. 304 pages.

Zingales, L., 2000. In Search of New Foundations. Journal of Finance, 55(4): p. 1623-1653.

Zuckerman, E.W., 2004. Structural Incoherence and Stock Market Activity. American Sociological Review, 69: p. 405-432.

* * *

Annex: Embedding culture and finance: divorce or a fresh start?

306

307

Culture, finance, and national institutions

 Three  essays  in  corporate  finance  

   

by  Charles-­‐Henri  J.  REUTER    

Bibliography, tables and official abstracts

Bibliography   309  

Tables  of  Figures   325  

Index   327  

Table  of  contents   331  

Official  abstracts   336  

 

308

309

Bibliography

Abbott, A., 1988. The system of professions: an essay on the division of expert labor.

Chicago: The University of Chicago Press.

Adler, N.J., Graham, J.L., 1989. Cross-cultural interaction: the international comparison fallacy? Journal of International Business Studies, 20(3): 515-537.

Aggarwal, R., Goodell, J.W., 2008. Equity premia in emerging markets: national characteristics as determinants. Journal of Multinational Financial Management, 18(4): 389-404.

Aggarwal, R., Goodell, J.W., 2009. Markets and institutions in financial intermediation: national characteristics as determinants. Journal of Banking & Finance, 33(10): 1770-1780.

Aggarwal, R., Goodell, J.W., 2010. Financial markets versus institutions in European countries: influence of culture and other national characteristics. International Business Review, 19(5): 502-520.

Aggarwal, R., Kearney, C., Lucey, B., 2009. Is gravity a cultural artifact? The role of cultural distance in foreign portfolio investment. Infiniti conference proceedings.

Aglietta, M., 2004. Dérives du capitalisme financier. Paris: Albin Michel. 394 pages.

Alessandrini, P., Calcagnini, G., Zazzaro, A., 2008. Asset restructuring strategies in bank acquisitions: does distance between dealing partners matter? Journal of Banking & Finance, 32(5): 699-713.

Amable, B., 2005. Les cinq capitalismes. Diversité des systèmes économiques et sociaux dans la mondialisation. Paris: Seuil. 374 pages.

Anderson, C.W., Fedenia, M., Hirschey, M., Rantala, H., 2007. There's no place like home: cultural influences on international diversification by institutional investors. Working paper (available from SSRN)

Antoniou, A., Guney, Y., Paudyal, K.N., 2003. The determinants of corporate debt ownership structure: evidence from market-based and bank-based economies. Available on SSRN.

Antoniou, A., Guney, Y., Paudyal, K., 2006. The determinants of debt maturity structure: evidence from France, Germany and the UK. European Financial Management, 12(2): 161-194.

310

Antoniou, A., Guney, Y., Paudyal, K.N., 2008. The determinants of capital structure: capital market oriented versus bank oriented institutions. Journal of Financial and Quantitative Analysis, 43: 59-92.

Aoki, M., 2001. Toward comparative institutional analysis. Cambridge: MIT Press. 467 pages.

Arrow, K., 1972. Gifts and exchanges. Philosophy and Public Affairs, 1(343-362)

Atkinson, A.B., Brandolini, A., 2001. Promise and Pitfalls in the Use of "Secondary" Data-Sets: Income Inequality in OECD Countries as a Case Study. Journal of Economic Literature, 39(3): 771-779.

Au, K.Y., 1999. Intra-cultural variation: evidence and implications for international business. Journal of International Business Studies, 30(4): 799-812.

Avramov, D., Chordia, T., 2006. Predicting stock returns. Journal of Financial Economics, 82(387-415)

Baker, M., Wurgler, J., 2002. Market timing and capital structure. Journal of Finance, 57(1): 1-32.

Baker, M., Wurgler, J., 2004. A catering theory of dividends. Journal of Finance, 59(3): 1125-1165.

Baker, M., Ruback, R.S., Wurgler, J., 2005. Behavioral Corporate Finance: a survey. In Eckbo, E. (Ed.), Handbook in corporate finance: Empirical Corporate Finance. Harvard Business School Press

Bancel, F., 2007. La cotation des titres Europeens aux Etats-Unis: une approche critique (The listing of European companies in the US: a critical approach). Revue d'Economie Financiere, 89: 143-162.

Bancel, F., Mittoo, U.R., 2004. Cross-country determinants of capital structure choice: a survey of European firms. Financial Management, 33(4): 103-132.

Banfield, E., 1958. The Moral Basis of a Backward Society. New York: Free Press.

Barclay, M.J., Smith, C.W., 2005. The Capital Structure Puzzle: The Evidence Revisited. Journal of Applied Corporate Finance, 17(1): 8-17.

Barclay, M.J., Marx, L.M., Smith Jr., C.W., 2003. The joint determination of leverage and maturity. Journal of Corporate Finance, 9(2): 149.

Bardhan, P.K., 2005. Institutions matter, but which ones? Economics of Transition, 13(3): 499-532.

Baskerville, R.F., 2003. Hofstede never studied culture. Accounting, Organizations and Society, 28(1): 1-14.

Bates, T.W., Kahle, K.M., Stulz, R., 2009. Why do U.S. firms hold so much more cash than they used to? Journal of Finance, 64(5): 1985-2021.

Beckerman, W., 1956. Distance and the pattern of intra-European trade. Review of Economics and Statistics, 38(1): 31-40.

Benmelech, E., Bergman, N., 2010. Bankruptcy and the Collateral Channel. NBER Working Paper Series w15708.

311

Berger, P.G., Ofek, E., Yermack, D.L., 1997. Managerial entrenchment and capital structure decisions. Journal of Finance, 52(4): 1411-1438.

Bessler, W., Drobetz, W., Pensa, P., 2008. Do managers adjust the capital structure to market value changes? Evidence from Europe. Journal of Business Economics (Zeitschrift für Betriebswirtschaft), Special Issue: 113-138.

Beugelsdijk, S., Frijns, B., 2009. A cultural explanation of the foreign bias in international asset allocation. Australasian Finance and Banking conference (available from SSRN)

Black, B.S., Gilson, R.J., 1998. Venture capital and the structure of capital markets: Banks versus stock markets. Journal of Financial Economics, 47(3): 243.

Booth, L., Aivazian, V., Demirguc-Kunt, A., Maksimovic, V., 2001. Capital structures in developing countries. Journal of Finance, 56(1): 87-130.

Botero, J.C., et al., 2003. Judicial Reform. World Bank Research Observer, 18(1): 61.

Botero, J.C., et al., 2004. The Regulation of Labor. Quarterly Journal of Economics, 119(4): 1339-1382.

Brennan, M.J., 1995. Corporate finance over the past 25 years. Financial Management, 24(2): 9-22.

Breuer, W., Salzmann, A., 2008. Cultural Dimensions of Corporate Governance Systems. Working Paper (available from SSRN)

Breuer, W., Quinten, B., 2009. Cultural Finance. Working Paper (available from SSRN)

Brounen, D., De Jong, A., Koedijk, K., 2006. Capital structure policies in Europe: Survey evidence. Journal of Banking & Finance, 30(5): 1409.

Burnett, J.J., Palmer, B.A., 1984. Examining life insurance ownership through demographic and psychographic characteristics. Journal of Risk & Insurance, 51(3): 453-467.

Calomiris, C.W., Kahn, C.M., Longhofer, S.D., 1994. Housing-finance intervention and private incentives: helping minorities and the poor. Journal of Money, Credit & Banking, 26(3): 634-674.

Carlin, W.J., Mayer, C., 2000. How do financial systems affect economic performance? In Vives, X. (Ed.), Corporate governance: theoretical and empirical perspectives. Cambridge: Cambridge University Press. 137-168.

Chang, K., Noorbakhsh, A., 2007. The Effects of National Culture on Corporate Cash Holdings beyond Corporate Governance and Financial Development Factors. Working Paper (available from SSRN)

Charreaux, G., 1996. Pour une véritable théorie de la latitude managériale et du gouvernement des entreprises. Revue Française de Gestion, 111(special issue): 50-64.

Charreaux, G., 1998. Le role de la confiance dans le systeme de gouvernance des entreprises. (Part Played by Trust in the Corporate Governance. With English summary.). Economies et Societes, 32(8-9): 47-65.

Charreaux, G., 2002. Variation sur le thème: 'à la recherche de nouvelles fondations pour la finance et la gouvernance d'entreprise'. Finance Contrôle Stratégie, 5(3): 5-68.

312

Charreaux, G., 2004. Les théories de la gouvernance: de la gouvernance des entreprises à la gouvernance des systèmes nationaux. Cahier du FARGO n° 1040101.

Charreaux, G., 2005. Pour une gouvernance d'entreprise 'comportementale'. Une réflexion exploratoire. Revue Française de Gestion, 4(157): 215-238.

Charreaux, G., Schatt, A., 2006. La recherche française en gouvernance d'entreprise: un panorama. Banque et marchés, 82: 7-23.

Charreaux, G., Schatt, A., 2006. La recherche française en finance: une perspective à travers les travaux des enseignants-chercheurs en gestion sur la période 1994-2003. Revue Finance Contrôle Stratégie, 9(1): 239-271.

Chen, H., 2010. Credit spreads and Capital Structure. NBER Working Paper No. w16151

Chenhall, R.H., 2003. Management control systems design within its organizational context: findings from contingency-based research and directions for the future. Accounting, Organizations & Society, 28(2): 127-168.

Chui, A., Kwok, C., 2008. National culture and life insurance consumption. Journal of International Business Studies, 39(1): 88-101.

Chui, A., Lloyd, A., Kwok, C., 2002. The determination of capital structure: is national culture a missing piece to the puzzle? Journal of International Business Studies, 33(1): 99-127.

Chui, A.C.W., Titman, S., Wei, K.C.J., 2010. Individualism and momentum around the world. Journal of Finance, 54(1): 2143-2184.

Claessens, S., Djankov, S., Lang, L.H.P., 2000. The separation of ownership and control in East Asian Corporations. Journal of Financial Economics, 58(1/2): 81-112.

Claessens, S., Djankov, S., Fan, J.P.H., Lang, L.H.P., 2002. Disentangling the incentive and entrenchment effects of large shareholdings. Journal of Finance, 57(6): 2741-2771.

Clement, M.B., Rees, L., Swanson, E.P., 2003. The influence of culture and corporate governance on the characteristics that distinguish superior analysts. Journal of Accounting, Auditing & Finance, 18(4): 593-618.

Coffee, J.C., 2001. Do Norms Matter? A Cross-Country Examination of the Private Benefits of Control. Columbia Law and Economics Working Paper

Conn, R.L., Cosh, A., Guest, P.M., Hughes, A., 2005. The impact on U.K. acquirers of domestic, cross-border, public and private acquisitions. Journal of Business Finance & Accounting, 32(5/6): 815-870.

Cuche, D., 2004. La notion de culture dans les sciences sociales. Paris: La Découverte - Repères. 124 pages.

d'Iribarne, P., 1997. The usefulness of an ethnographic approach to the international comparisons of organizations. International Studies of Management and Organization, 26(4): 30-47.

De Jong, A., Kabir, R., Nguyen, T.T., 2008. Capital structure around the world: the roles of firm- and country-specific determinants. Journal of Banking & Finance, 32(9): 1954-1969.

313

De Jong, E., 2002. Why are price stability and statutory independence of central banks negatively correlated? The role of culture. European Journal of Political Economy, 18(4): 675.

De Jong, E., 2009. Culture and economics: on values, economics and international business. Routledge. 272 pages.

De Jong, E., Semenov, R., 2002. Cross-country differences in stock market development: a cultural view. Working paper (available from SSRN)

De Jong, E., Semenov, R., 2006. Cultural determinants of ownership concentration across countries. International Journal of Business Governance and Ethics, 2(1): 145-165.

Deesomsak, R., Paudyal, K., Pescetto, G., 2004. The determinants of capital structure: evidence from the Asia Pacific region. Journal of Multinational Financial Management, 14: 387-405.

Demirguc-Kunt, A., Maksimovic, V., 1999. Institutions, financial markets, and firm debt maturity. Journal of Financial Economics, 54(3): 295-336.

Denis, D.K., McConnell, J.J., 2003. International Corporate Governance. Journal of Financial & Quantitative Analysis, 38(1): 1-36.

Desender, K.A., Castro, C.E., Escamilla de Leon, S.A., 2011 (forthcoming). Earnings Management and Cultural Values. American Journal of Economics and Sociology,

Diamond, D.W., 1991. Debt maturity structure and liquidity risk. Quarterly Journal of Economics, 106(3): 709-737.

DiMaggio, P.J., 1994. Culture and economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology. Princeton: Princeton University Press. 835 pages: p. 27-57.

DiMaggio, P.J., 1997. Culture and cognition. Annual Review of Sociology, 23: 263-287.

DiMaggio, P.J., Powell, W.W., 1983. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2): 147-160.

Dittmar, A., Mahrt-Smith, J., Servaes, H., 2003. International corporate governance and corporate cash holdings. Journal of Financial and Quantitative Analysis, 38: 111-133.

Djankov, S., Mcliesh, C., Shleifer, A., 2007. Private Credit in 129 Countries. Journal of Financial Economics, 84(2): 299-329.

Djankov, S., Porta, R.L., Lopez-de-Silanes, F., Shleifer, A., 2002. The regulation of entry. The Quarterly Journal of Economics,

Djankov, S., La Porta, R., Lopez De Silanes, F., Shleifer, A., 2008. The Law and Economics of Self-Dealing. Journal of Financial Economics, 88(3): 430-465.

Djelic, M.-L., 2001. Exporting the American Model: The Postwar Transformation of European Business. Oxford: Oxford University Press. 328 pages.

Durkheim, E., 1897/1967. De la division du travail social. Paris: Les Presses universitaires de France. 416 pages.

314

Durkheim, E., 1912. The elementary form of religious life. Reprinted by Oxford University Press.

Dyck, A., 2001. Privatization and Corporate Governance: Principles, Evidence, and Future Challenges. World Bank Research Observer, 16(1): 59-84.

Earley, P.C., 2006. Leading cultural research in the future: a matter of paradigms and taste. Journal of International Business Studies, 37(6): 922-931.

Earley, P.C., Singh, H., 1995. International and intercultural management research: what's next. The Academy of Management Journal, Vol. 38(2): 327 - 340.

Elias, N., 1982 [1939]. The civilizing process (vol. II state formation and civilization). Oxford:

Etzioni, A., 1990. Moral Dimension : Toward a New Economics. Free Press. 314 pages.

Faccio, M., Lang, L.H.P., 2002. The ultimate ownership of Western European corporations. Journal of Financial Economics, 65(3): 365-395.

Fan, P.H.J.P., Titman, S., Twite, G.J., 2008. An international comparison of capital structure and debt maturity choices. SSRN eLibrary.

Ferreira, E., Serna, M.I.M., Navarro, E., Rubio, G., 2008. Economic sentiment and yield spreads in Europe. European Financial Management, 14(2): 206-221.

Fidrmuc, J.P., Jacob, M., 2008. A cultural explanation of agency model of dividends. Working Paper (available from SSRN)

Fligstein, N., 1993. The Transformation of Corporate Control. Harvard University Press. 408 pages.

Fligstein, N., 2001. The architecture of markets: an economic sociology of twenty-first century capitalist societies. Princeton: Princeton University Press. 288 pages.

Frank, M.Z., Goyal, V.K., 2004. The effect of market conditions on capital structure adjustement. Finance Research Letters, 1: 47-55.

Frank, M.Z., Goyal, V.K., 2008. Trade-off and pecking order theories of debt. In Eckbo, E. (Ed.), Handbook of corporate finance: empirical corporate finance (vol. 2). Elsevier: North Holland. Chapter 12.

Frank, M.Z., Goyal, V.K., 2009. Capital structure decisions: which factors are reliably important? Financial Management, 38(1): 1-37.

Franke, R.H., Hofstede, G., Bond, M.H., 1991. Cultural Roots of Economic-Performance - a Research Note. Strategic Management Journal, 12: 165-173.

Gande, A., Parsley, D.C., 2005. News spillovers in the sovereign debt market. Journal of Financial Economics, 75(3): 691-734.

Garvey, G.T., Swan, P.L., 1994. The economics of corporate governance : beyond the Marshallian firm. Journal of Corporate Finance, 1(2): 139-174.

Geyer, A., Kossmeier, S., Pichler, S., 2004. Measuring Systematic Risk in EMU Government Yield Spreads. Review of Finance, 8(2): 171-197.

315

Giannetti, M., 2003. Do better institutions mitigate agency problems? Evidence from corporate finance choices. Journal of Financial and Quantitative Analysis, 38(1): 185-212.

Glaeser, E.L., Porta, R.L., Lopez-de-Silanes, F., Shleifer, A., 2004. Do Institutions Cause Growth? Journal of Economic Growth,

Gleason, K.C., McNulty, J.E., Pennathur, A.K., 2005. Returns to acquirers of privatizing financial services firms: an international examination. Journal of Banking & Finance, 29(8): 2043-2065.

Gleason, K.C., Mathur, L.K., Mathur, I., 2000. The interrelationship between culture, capital structure, and performance: evidence from European retailers. Journal of Business Research, 50(2): 185-191.

Gomes, J.F., Schmid, L., 2010. Levered Returns. Journal of Finance, 65(2): 467–494.

Gonzalez, V., Gonzalez, F., 2008. Influence of bank concentration and institutions on capital structure: new international evidence. Journal of Corporate Finance, 14(4): 363-375.

Graham, J., Harvey, C., 2001. The Theory and Practice of Corporate Finance: Evidence from the Field. Journal of Financial Economics, 60: 187-243.

Graham, J.R., 1996. Debt and the marginal tax rate. Journal of Financial Economics, 41(1): 41-73.

Granovetter, M., 1985. Economic action and social structure: the problem of embeddedness. American Journal of Sociology, 91(3): 481-510.

Greenfich, P., 2005. Behavioral Finance Definitions : Main Concepts. author's website available at http:// perso.wanadoo.fr/ greenwich/ bfdef.htm

Greif, A., 2006. Institutions and the path to the modern economy: lessons from medieval trade. Cambridge University Press. 503 pages.

Grinblatt, M., Keloharju, M., 2001. How distance, language, and culture influence stockholdings and trades. Journal of Finance, 56(3): 1053.

Grinblatt, M., Keloharju, M., 2007. Sensation seeking, overconfidence, and trading Activity. Journal of Finance (forthcoming)

Grinold, R., Rudd, A., Stefek, D., 1989. Global factors: fact or fiction? Journal of Portfolio Management, 16(1): 79-88.

Guillen, M.F., 1994. Models of management. Work, authority and organization in a comparative perspective. Chicago: University of Chicago Press. 424 pages.

Guillen, M.F., 2001. Is globalization civilizing, destructive or feeble? A critique of five key debates in the social science literature. Annual Review of Sociology, Vol. 27: 235 - 260.

Guillen, M.F., Suarez, S.L., 2005. The institutional context of multinational activity. In Ghoshal, S. and Westney, D.E. (Ed.), Organization theory and the multinational corporation. New York: Palgrave Macmillan. 360.

Guiso, L., Haliassos, M., Jappelli, T., 2001. Household portfolios. Cambridge: MIT press. 508 pages.

316

Guiso, L., Sapienza, P., Zingales, L., 2007. Trusting the Stock Market. ECGI - Finance Working Paper

Guiso, L., Sapienza, P., Zingales, L., 2007. Does culture affect economic outcomes? Journal of Economic Perspectives, 20(2): 23-48.

Guiso, L., Sapienza, P., Zingales, L., 2009. Culture biases in economic exchange? Quarterly Journal of Economics, 124(3): 1095-1131.

Hackbarth, D., Miao, J., Morellec, E., 2006. Capital structure, credit risk, and macroeconomic conditions. Journal of Financial Economics, 82: 519-550.

Hall, D.P.A., Soskice, D.W., 2001. Varieties of capitalism: the institutional foundations of comparative advantage. Oxford: Oxford University Press. 560 pages.

Hall, G., Hutchinson, P., Michaelas, N., 2004. Determinants of the capital structures of European SMEs. Journal of Business Finance and Accounting, 31: 711-728.

Hambrick, D.C., Mason, P.A., 1984. Upper-echelons: the organization as a reflection of its top managers. Academy of Management Review, 9(2): 193-206.

Hamilton, G.G., Biggart, N.W., 1988. Market, Culture, and Authority: A Comparative Analysis of Management and Organization in the Far East. The American Journal of Sociology, 94(Supplement): S52-S94.

Harford, J., Mansi, S.A., Maxwell, W.F., 2008. Corporate governance and firm cash holdings in the US. Journal of Financial Economics, 87(3): 535-555.

Harrison, G., McKinnon, J., 1999. Cross-cultural research in management control systems design: a review of the current state. Accounting, Organizations & Society, 24(5): 483-506.

Harzing, A.W.K., 2004. The role of culture in entry mode studies: from negligence to myopia. Advances in International Management, 15: 75-127.

Hilary, G., Hui, K.W., 2009. Does religion matter in corporate decision making in America? Journal of Financial Economics, 93(3): 455-473.

Hofstede, G., 1980. Culture's consequence. New-York: Sage publications.

Hofstede, G., 1994. Preface. In Kim, U., et al. (Ed.), Individualism and collectivism. Theory, methods and applications. New-York: Sage publications. 348 pages: p. ix-xiii.

Hofstede, G., 2001. Culture's consequences: comparing values, behaviors, institutions and organizations across nations. New-York: Sage publications. 616 pages.

Hofstede, G., Bond, M.H., 1988. The Confucius Connection - from Cultural Roots to Economic-Growth. Organizational Dynamics, 16(4): 5-21.

Hofstede, G., Hofstede, G.J., 1991. Cultures and organizations: software for the mind. New-York: McGraw-Hill. 300 pages.

Hofstede, G., Bond, M.H., Luk, C.-l., 1993. Individual perceptions of organizational cultures: A methodological treatise on levels of analysis. Organization Studies, 14(4): 483.

Hofstede, G., Neuijen, B., Ohayv, D.D., Sanders, G., 1990. Measuring Organizational Cultures: A Qualitative and Quantitative Study Across Twenty Cases. Administrative Science Quarterly, 35(2): 286.

317

Holderness, C.G., 2010. The myth of diffuse ownership in the United States. Review of Financial Studies, 22(4): 1377-1408.

Hollingsworth, J.R., Boyer, R., 1997. Contemporary capitalism: the embedness of institutions. Cambridge: Cambridge University Press.

House, R.J., et al., 1999. Cultural influences on leadership and organizations: project Globe. In Mobley, W.H. (Ed.), Advances in global leadership. Stamford, CT: JAI Press. 171-233.

Huang, R., 2009. Tolerance for uncertainty and the growth of informationally opaque industries. Journal of Development Economics, 87(2): 333-353.

Hunter, W.C., Walke, M.B., 1996. The cultural affinity hypothesis and mortgage lending decisions. The Journal of Real Estate Finance and Economics, 13(1): 57-70.

Immergut, E.M., 1998. The theoretical core of the new institutionalism. Politics & Society, 26(1): 5 - 34.

Inglehart, R., 1997. Modernization and postmodernization: cultural economic and political change. Princeton: Princeton University Press. 440 pages.

Jensen, M.C., 1989. Eclipse of the public corporation. Harvard Business Review, 67(5): 61-74.

Jensen, M.C., Meckling, W.H., 1976. Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Journal of Financial Economics, Vol. 3(No. 4): 305 - 360.

Jepperson, R.L., Swidler, A., 1994. What properties of culture should we measure? Poetics, 22(4): 359-371.

Jimenez, G., Salas, V., Saurina, J., 2006. Determinants of collateral. Journal of Financial Economics, 81(2): 255-281.

Kagitcibasi, C., 1997. Individualism and collectivism. In Berry, J., Segall, S., and Kagitçibasi, C. (Ed.), Handbook of cross-cultural psychology. Needham Heights: Allyn and Bacon. 1-49.

Kahneman, D., Tversky, A., 1979. Prospect theory: An analysis of decision under risk. Econometrica, 47: 263-291.

Kimbro, M.B., 2002. A cross-country empirical investigation of corruption and its relationship to economic, cultural, and monitoring institutions: an examination of the role of accounting and financial statements quality. Journal of Accounting, Auditing & Finance, 17(4): 325-349.

King, M.R., Santor, E., 2008. Family values: ownership structure, performance and capital structure of Canadian firms. Journal of Banking & Finance, (32): 2423-2432.

Kirkman, B.L., Lowe, K.B., Gibson, C.B., 2006. A quarter century of Culture's Consequences: a review of empirical research incorporating Hofstede's cultural values framework. Journal of International Business Studies, 37(3): 285-320.

Klep, P.M., 2001. 'Reculturalisation' in economic and social history? Deel, 64: 6-27.

318

Kogut, B., Singh, H., 1988. The effect of national culture on the choice of entry mode. Journal of International Business Studies, 19(3): 411-432.

Korajczyk, R.A., Levy, A., 2003. Capital structure choice: macroeconomic conditions and financial constraints. Journal of Financial Economics, 68(1): 75-109.

Kroeber, A.L., Kluckhohn, C., 1952. Culture: A Critical Review of Concepts and Definitions. Cambridge, MA: Peabody Museum.

Kwok, C., Tadesse, S., 2006. National culture and financial systems. Journal of International Business Studies, 37(2): 227-247.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., 1999. Corporate ownership around the world. Journal of Finance, 54(2): 471-517.

La Porta, R., Lopez-De-Silanes, F., Shleifer, A., Vishny, R.W., 1997. Legal determinants of external finance. Journal of Finance, 52(3): 1131-50.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R.W., 1998. Law and finance. Journal of Political Economy, 106(6): 1113-55.

La Porta, R., Lopez-de-Silanes, F., Shleifer, A., Vishny, R., 2000. Investor protection and corporate governance. Journal of Financial Economics, 58(1/2): 3-27.

La Porta, R., Lopez-de-Silanes, F., Pop-Eleches, C., Schleifer, A., 2004. Judicial check and balances. Journal of Political Economy, 112(2): 445-470.

Lenartowicz, T., Roth, K., 1999. A Framework for culture assessment. Journal of International Business Studies, 30(4): 781-798.

Leung, K., et al., 2005. Culture and international business: recent advances and their implications for future research. Journal of International Business Studies, 36(4): 357-378.

Lévesque, B., Bourque, G.L., Forgues, E., 2001. La nouvelle sociologie économique. Desclée de Brouwer. 268 pages.

Levy, A., Hennessy, C., 2007. Why does capital structure choice vary with macroeconomic conditions? Journal of Monetary Economics, 54(6): 1545-1564.

Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2005. Culture, law, and corporate governance. International Review of Law & Economics, 25(2): 229-255.

Licht, A.N., Goldschmidt, C., Schwartz, S.H., 2007. Culture rules: the foundations of the rule of law and other norms of governance. Journal of Comparative Economics, 35(4): 659-688.

MacKenzie, D., Millo, Y., 2003. Constructing a Market, Performing Theory: The Historical Sociology of a Financial Derivatives Exchange. American Journal of Sociology, 109(1): 107-145.

Mahajan, A., Tartaroglu, S., 2008. Equity market timing and capital structure: international evidence. Journal of Banking & Finance, 32: 754-766.

Margaritis, D., Psillaki, M., 2009. Capital structure, equity ownership and firm performance. Journal of Banking & Finance, 34: 621–632.

319

Markus, H.R., Kitayama, S., 1991. Culture and Self: Implications for Cognition, Emotion and Motivation. Psychological Review, 98: 225-253.

Mayer, C., 1990. Financial systems, corporate finance, and economic development. Hubbard, R. (Ed.), Asymmetric Information, Corporate Finance and Investment. Chicago: Chicago University Press

Mayer, C., 2008. Trust in Financial Markets. European Financial Management, 14(4): 617-632.

McClelland, D.C., Atkinson, J.W., Clark, R.A., Lowell, E.L., 1953. The achievement motive. New-York: John Wiley & Sons (second. ed. Dec. 1976). 410 pages.

McCloskey, D.D., 1983. The rhetoric of economics. Journal of Economic Literature, 21: 481-517.

McCloskey, D.D., 2002. The Secret Sins of Economics. Chicago: Prickly Paradigm Press.

McSweeney, B., 2002. Hofstede's model of national cultural differences and their conserquences: a triumph of faith - a failure of analysis. Human Relations, 55(1): 89-118.

Merryman, J.H., Pérez-Perdomo, R., 2007. The civil law tradition: an introduction to the legal systems of Europe and Latin America. Stanford: Stanford University Press. 173 pages.

Mian, A., 2006. Distance constraints: the limits of foreign lending in poor economies. Journal of Finance, 61(3): 1465-1505.

Miller, M., Modigliani, F., 1961. Dividend policy, growth, and the valuation of shares. Journal of Business, 34: 411-433.

Mizruchi, M.S., Stearns, L.B., 1994. Money, Banking and Financial Markets. In Smelser, N.J. and Swedberg, R. (Ed.), The Handbook of Economic Sociology. Princeton University Press

Morosini, P., Shan, S., Singh, H., 1998. National cultural distance and cross-border acquisition performance. Journal of International Business Studies, 29(1): 137-158.

Myers, S.C., 2001. Capital Structure. Journal of Economic Perspectives, 15(2): 81-102.

Myers, S.C., 2003. Financing of corporations. In Constantinides, G., Harris, M., and Stulz, R. (Ed.), Handbook of the Economics of Finance: Corporate Finance. Elsevier North Holland

North, D., 1990. Institutions, institutional change and economic performance. Cambridge: Cambridge University Press. 159 pages.

O'Barr, W.M., Conley, J.M., 1992. Managing relationships: the culture of institutional investing. Financial Analysts Journal, 48(5): 21-27.

Outreville, J.F., 2008. Foreign affiliates of the largest insurance groups: location-specific advantages. Journal of Risk and Insurance, 75(2): 463-491.

Oyserman, D., 2002. Values, psychological perspectives. In Smelser and Baltes (Ed.), International Encyclopedia of the social and behavioral sciencese. 16150-16153.

Payne, G.T., Moore, C.B., Griffis, S., Autry, C., 2008. Social Capital Research in Management: A Review of Conceptualizations, Measurements, and Implications. Invited

320

for revision for special annual review issue of Journal of Management. (Invitation Received: 3- 13-09).

Pirouz, D.M., 2004. National culture and global stock market volatility. Working paper (available from SSRN)

Podolny, J.M., 1993. A Status-Based Model of Market Competition. American Journal of Sociology, 98(4): 829.

Polanyi, K., 1944, 1957. The Great Transformation: The Political and Economic Origins of Our Time. Boston:

Portes, R., Rey, H., Oh, Y., 2001. Information and capital flows: the determinants of transactions in financial assets. European Economic Review, 45(4-6): 783-796.

Putnam, R., 1993. Making Democracy Work. Princeton University Press, Princeton, NJ.

Ragin, C.C., 2000. Fuzzy-set social science. Chicago: Chicago University. 370 pages.

Rajan, R.G., Zingales, L., 1995. What do we know about capital structure? Some evidence from international data. Journal of Finance, 50(5): 1421-1460.

Rajan, R.G., Zingales, L., 2003. The great reversals: the politics of financial development in the twentieth century. Journal of Financial Economics, 69(1): 5.

Ramirez, A., Tadesse, S., 2007. Corporate cash holdings, national culture and multinationality. William Davidson Institute Working Papers

Redding, G., 2005. The thick description and comparison of societal systems of capitalism. Journal of International Business Studies, 36(2): 123-155.

Reuter, C.H.J., 2009. European Capital Structures: Managerial Insight and Governance Regimes available at SSRN n°1538152.

Rodrik, D., 2005. Why We Learn Nothing From Regressing Economic Growth on Policies. Harvard University Working Paper

Roe, M.J., 2006. Political Determinants of Corporate Governance: Political Context, Corporate Impact. New-York: Oxford University Press. 256 pages.

Romano, R., 1993. A Cautionary Note on Drawing Lessons from Comparative Corporate Law. Yale Law Journal, 102(8): 2021-2037.

Ronen, S., Shenkar, O., 1985. Clustering countries on attitudinal dimensions: A review and synthesis. The Academy of Management Review, 10(3): 435.

Rosati, S., Secola, S., 2006. Explaining cross-border large-value payment flows: evidence from 'TARGET' and 'EURO1' data. Journal of Banking & Finance, 30(6): 1753-1782.

Ruiz-Mallorqui, M.V., Santana-Martin, D.J., 2011. Dominant institutional owners and firm value. Journal of Banking and Finance, 35(1): 118-129.

Sapienza, P., Toldra, A., Zingales, L., 2007. Understanding Trust. NBER Working Paper

Sarig, O.H., Talmor, E., 1997. In defense of defensive measures. Journal of Corporate Finance, 3(3): 277-297.

Sarkissian, S., Schill, M.J., 2004. The overseas listing decision: new evidence of proximity preference. Review of Financial Studies, 17(3): 769-809.

321

Schmeling, M., 2009. Investor sentiment and stock returns: Some international evidence. Journal of Empirical Finance, 16(3): 394-408.

Schwartz, S.H., 1994. Beyond individualism/collectivism: new cultural dimensions of values. In Kim, U., et al. (Ed.), Individualism and collectivism: theory, method and application. Newbury Park: Sage

Schwartz, S.H., 1999. A Theory of Cultural Values and Some Implications for Work. Applied Psychology: An International Review, 48(1): 23-47.

Schwartz, S.H., 2004. Mapping and interpreting cultural differences around the world. In Vinken, H., Soeters, J., and Ester, P. (Ed.), Comparing cultures: dimensions of culture in a comparative perspective. Leiden: Brill Academic Publishers

Schwartz, S.H., 2007. Cultural and individual value correlates of capitalism: a comparative analysis. Psychological Inquiry, 18(1): 52-57.

Schwartz, S.H., Ros, M., 1995. Values in the West: a theoretical and empirical challenge to the individualism-collectivism dimension. World Psychology, 1(2): 91-122.

Scott, W.R., 1995. Institutions and Organizations. Thousand Oaks, CA: Sage.

Shao, L., Kwok, C.C.Y., Guedhami, O., 2008. Is National Culture a Missing Piece of the Dividend Puzzle? Working Paper (available from SSRN)

Shefrin, H., 1999. Behavioral Corporate Finance. In Shefrin, H. (Ed.), Beyond Greed and Fear: Understanding Behavioral Finance and the Psychology of Investing. Harvard Business School Press

Shenkar, O., 2001. Cultural distance revisited: towards a more rigorous conceptualization and measurement of cultural differences. Journal of International Business Studies, 32(3): 519-535.

Shiller, R.J., 1999. Human behavior and the efficiency of the financial system. In Taylor, J.B. and Woodford, M. (Ed.), Handbook of macroeconomics. Amsterdam: Elsevier. 1305-40.

Shleifer, A., Vishny, R., 1986. Large shareholders and corporate control. Journal of Political Economy, 44: 461-488.

Shleifer, A., Vishny, R.W., 1989. Management entrenchment: the Case of managers specific investments Journal of Financial Economics, 25(1): 123-139.

Shleifer, A., Vishny, R.W., 1997. A survey of corporate governance. Journal of Finance, 52(2): 737-783.

Siegel, J.I., Licht, A.N., Schwartz, S.H., 2006. Egalitarianism and international investment. Working paper (available from SSRN)

Siegel, J.I., Licht, A.N., Schwartz, S.H., 2008. Egalitarianism, cultural distance, and foreign direct investment: a new approach. Working Paper (available from SSRN)

Simon, H.A., 1945/1976. Administrative Behavior: a study of decision-making processes in administrative organization. New York: The Free Press. 1364 pages.

Sivakumar, K., Nakata, C., 2001. The stampede toward Hofstede's framework: avoiding the sample design pit in cross-cultural research. Journal of International Business Studies, 32(3): 555.

322

Smelser, N.J., Swedberg, R., 1994. The Handbook of Economic Sociology. Princeton University Press.

Smelser, N.J., Swedberg, R., 1994. The sociological perspective on the economy. In Smelser, N.J. and Swedberg, R. (Ed.), The handbook of economic sociology.

Smith, P.B., 2006. When elephants fight, the grass gets trampled: the GLOBE and Hofstede projects. Journal of International Business Studies, 37(6): 915-921.

Smith, P.B., Schwartz, S.H., 1997. Values. In J.W. Berry, Segall, M.H., and Kagitcibasi, C. (Ed.), Handbook of cross-cultural psychology. Needham Heights, MA: Allyn and Bacon. 79-116.

Smith, P.B., Dugan, S., Trompenaars, F., 1996. National culture and the values of organizational employees. Journal of Cross-Cultural Psychology, 27(2): 231-264.

Smith, P.B., Peterson, M.F., Schwartz, S.H., 2002. Cultural values, sources of guidance, and their relevance to managerial behavior. Journal of Cross-Cultural Psychology, 33(2): 188-208.

Spamann, H., 2006. On the Insignificance and/or Endogeneity of La Porta et al.'s 'Anti-Director Rights Index' under Consistent Coding. SSRN eLibrary.

Statman, M., 2007. local ethics in a global world. Financial Analysts Journal, 63(3): 32-41.

Statman, M., 2008. Countries and culture in behavioral finance. CFA Institute, 25(3): 38-44.

Stonehill, A., et al., 1975. Financial goals and debt ratio determinants: a survey of practice in five counries. Financial Management, 4(3): 27-41.

Streeck, W., 1992. Social institutions and economic performance : studies on industrial relations in advanced european capitalist countries. London: Sage.

Stulz, R.M., 2005. The limits of financial globalization. Journal of Finance, 60(4): 1595-1638.

Stulz, R.M., 2007. The limits of financial globalization. Journal of Applied Corporate Finance, 19(1): 8-15.

Stulz, R.M., Williamson, R., 2003. Culture, openness, and finance. Journal of Financial Economics, 70(3): 313-349.

Summers, L.H., 1985. On economics and finance. Journal of Finance, 40(3): 633-635.

Throsby, C.D., 2001. Economics and culture. Cambridge University Press. 208 pages.

Todd, E., 1990. L'invention de l'Europe. Paris: Seuil. 537 pages.

Triandis, H.C., 2004. The many dimensions of culture. Academy of Management Executive, 18(1): 88-93.

Tung, R.L., 2008. The cross-cultural research imperative: the need to balance cross-national and intra-national diversity. Journal of International Business Studies, 39(1): 41-46.

Vinken, H., Soeters, J., Ester, P., 2004. Comparing cultures: dimensions of culture in a comparative perspective. Leiden: Brill Academic Publishers.

Wanzenried, G., 2006. Capital structure dynamics in the UK and continental Europe. European Journal of Finance, 12(8): 693–716.

323

Weber, M., 1914. Economy and society: An Outline of Interpretive Sociology.

Weber, Y., Shenkar, O., Raveh, A., 1996. National and corporate cultural fit in mergers/acquisitions : an exploratory study. Management Science, 42(8): 1215-1227.

Welch, I., 2004. Capital structure and stock returns. Journal of Political Economy, 112(106-131)

Westbrook, D.A., 2009. Out of crisis: rethinking our financial market. Paradigm. 176 pages.

Whitley, R., 1992. European Business Systems: Firms and Markets in Their National Contexts. Sage. 302 pages.

Whitley, R., 2000. Divergent Capitalisms: The Social Structuring and Change of Business Systems. Oxford University Press.

Williams, R., 1958. Culture and society. London:

Williamson, O.E., 2000. The new institutional economics: taking stock, looking ahead. Journal of Economic Literature, 38(3): 595-613.

Wiltbank, R., Dew, N., Read, S., Sarasvathy, S.D., 2006. What to do next the case for non predictive strategy. Strategic Management Journal, 27(10): 981-998.

Zelizer, V.A., 1994. The social meaning of money. 304 pages.

Zingales, L., 2000. In search of new foundations. Journal of Finance, 55(4): 1623.

Zuckerman, E.W., 2004. Structural incoherence and stock market activity. American Sociological Review, 69: 405-432.

* * *

324

325

Figures and tables

Table 1. List of journals screened and results obtained 89 Table 2. General Database from the main screenings 91

Figure 1. Trend in the identified articles 93 Figure 2. Theoretical frameworks and culture definitions in the identified articles 95

Table 3. Dimensionalist database from main screenings and double review 100 Table 4 to 10 (but for 7 & 9). Results for the study of financial policies in Europe

Table 4. Sources, definitions and variables 181 Table 5. Statistics 184

Table 6. Main results - streamlined 189 Table 7. Ownership dispersion across Europe and sample-size (for chapters III & IV) 205

Table 8. Pair wise correlations 206 Table 9. Presenting and analyzing E.S.I. measures (for chapters III & IV) 209

Table 10. Main results - all coefficients reported 210 Table 11 to 19. Results for the study of capital structure and broader institutional context

Table 11. Statistics 234 Table 12. Results – streamlined: market timing and the broader institutional context 243

Table 13. Results – streamlined: cyclicality and leverage 247 Table 14. Results – streamlined: culture, cyclicality and leverage 253

Table 15. Pair wise correlations 260 Table 16. Results – full results on the baseline, un-pooled models, plus selected

context variables 261 Table 17. Results – full results for the pooled models - all coefficients reported 262

Table 18. Results – full results for the pooled models – signs and significance only 263 Table 19. Results – baseline models, pooled by country 264

* * *

326

 

327

Index for authors’ names

The abbreviation et al. has been used for any article co-authored by three researchers or more. These references are indexed separately, under the name of the first author, et al. A cross-reference to the first author is provided for all other co-authors. For ambiguous articles, the year is provided in parenthesis, and classified separately.

A Adler : 129 Aggarwal : 98,   100,   101,   104,   131,   167,   200,  

231,  254,  257 Aggarwal, et al. : 98,  100,  101,  102 Aivazian : cf.  Booth, et al. Alessandrini, et al. : 115 Amable : 71 Anderson, et al. : 103 Antoniou, et al. : 166,   183,   186,   200,   234,  

235,  237,  259 Aoki : 66,   71,   143,   219,   221,   232,   275,   292,  

296 Arrow : 117,  118 Atkinson : 278 Au : 130,  274,  280 Autry : cf.  Payne, et al. Avramov : 161

B Baker : 72,  185,  187,  190,  219,  220,  224,  235,  

241,  257,  289,  293,  294 Baker, et al. : 289,  293,  294 Bancel : 60,  66,  162,  193,  242,  256 Banfield : 117,  282 Barclay : 186,  219,  237,  257 Barclay, et al. : 187,  238 Bardhan : 275 Baskerville : 63,  99,  129,  131,  279 Bates, et al. : 197 Beckerman : 112 Benmelech : 161

Berger, et al. : 163,  168,  224 Bergman : 161 Bessler, et al. : 166 Beugelsdijk : 101 Biggart : 71,  276 Bond : cf.  Franke, et al. Booth, et al. : 166 Botero, et al. : 120,  277 Bourque : cf.  Lévesque, et al. Boyer : 71 Brandolini : 278 Brennan : 85,  117,  279,  284 Breuer : 80,  99,  102,  105,  132,  144,  280,  290,  

294 Brounen, et al. : 162,  193 Burnett : 123

C Calcagnini : cf.  Alessandrini, et al. Calomiris, et al. : 124 Carlin : 71,  232,  257 Castro : cf.  Desender, et al. Charreaux : 61,  71,  90,  161,  171,  293 Chen : 161 Chenhall : 129 Chordia : 161 Chui : 105,  106,  107,  132,  134,  149,  223,  230,  

252,  254,  257 Chui, et al. : 106,   107,   132,   134,   223,   230,  

252,  254,  257 Chui, et al. (2002) : 149,  223,  230,  252,  254,  

257 Chui, et al. (2010) : 106,  107,  132,  134

328

Claessens, et al. : 170,  175,  176 Claessens, et al. (2000) : 170,  175,  176 Claessens, et al. (2002) : 170 Clark : cf.  McClelland, et al. Clement, et al. : 108 Coffee : 278 Conley : 125,  126,  144 Conn, et al. : 101,  131 Cosh : cf.  Conn, et al. Cuche : 61,  79,  86,  127,  136,  147,  274,  284

D De Jong A. : cf.  Brounen, et al. De Jong A., et al. : 165,  195,  200,  201,  203,  

219,  220,  231,  240,  256 De Jong E. : 59,   80,   86,   87,   88,   98,   99,   105,  

132,  133,  135,  143,  148,  165,  195,  200,  201,  203,  219,  220,  223,  231,  232,  240,  256,  257,  284,  285,  286

Deesomsak, et al. : 166 Demirguc-Kunt : 166;  cf.  Booth, et al. Denis : 167 Desender, et al. : 144 Dew : cf.  Wiltbank, et al. Diamond : 251 DiMaggio : 63,  70,  86,  87,  88,  123,  137,  144,  

147,  276,  279,  285,  286,  287,  288,  295,  296

Dittmar, et al. : 198 Djankov, et al. : 120,  277 Djankov, et al. (2002, 2008) : 120,  177,  277 Djankov, et al. (2007) : 177 Djelic : 292 Drobetz : cf.  Bessler, et al. Durkheim : 137,  145,  294,  298

E Earley : 99,  130,  131,  146,  221,  225,  257,  274,  

280 Elias : 283 Escamilla de Leon : cf.  Desender, et al. Ester : cf.  Vinken, et al. Etzioni : 276

F Faccio : 163,  175,  176 Fan : cf.  Claessens, et al.  (2002):  170 Fan, et al. : 166,  183,  200 Fedenia : cf.  Anderson, et al. Ferreira, et al. : 177,  208 Fligstein : 66,  71,  138,  296

Forgues : cf.  Lévesque, et al. Frank : 65,   66,   68,   164,   165,   183,   185,   186,  

187,  203,  219,  220,  223,  224,  225,  226,  228,  234,  235,  237,  238,  240,  246,  255,  256,  259

Franke, et al. : 277 Frijns : 101

G Gande : 112,  113 Garvey : 171 Geyer, et al. : 177,  208 Giannetti : 166,  223,  257 Gibson : cf.  Kirkman, et al. Glaeser, et al. : 120,  278 Gleason, et al. : 113,  116 Gleason, et al. (2000) : Table Gleason, et al. (2005) : 113,  116 Goldschmidt : cf.  Licht, et al. Gomes : 161 Goodell : 100,   104,   131,   167,   200,   231,   254,  

257 Goyal : 65,   66,   68,   164,   165,   183,   185,   186,  

187,  203,  219,  220,  223,  224,  225,  226,  228,  234,  235,  237,  238,  240,  246,  255,  256,  259

Graham : 129,  161,  162,  193,  242 Granovetter : 86,  144,  276,  279 Greenfich : 293 Greif : 71,  80,  136,  143,  221,  275,  292 Griffis : cf.  Payne, et al. Grinblatt : 80,  120,  122,  140,  282 Grinold, et al. : 125,  126 Guedhami : cf.  Shao, et al. Guest : cf.  Conn, et al. Guillen : 133,  136,  221 Guiso, et al. : 86,  87,  117,  118,  119,  127,  134,  

135,  226,  281,  282,  284,  286,  287,  288 Guiso, et al. (2001) : 127,  226,  282 Guiso, et al. (2007, 2009) : 86,  87,  117,  118,  

119   134,   135,   281,   282,   284,   286,   287,  288

Guney : cf.  Antoniou, et al.

H Hackbarth, et al. : 161 Haliassos : cf.  Guiso, et al.:  127,  226,  282 Hall : 71,  136,  166,  232 Hall, et al. : 166 Hambrick : 171 Hamilton : 71,  276

329

Harford, et al. : 198 Harrison : 129 Harvey : 162,  193,  242 Harzing : 90,  129 Hennessy : 164 Hilary : 120,  123,  126 Hirschey : cf.  Anderson, et al. Hofstede : 59,   80,   81,   84,   92,   96,   97,   98,   99,  

101,  103,  104,  105,  106,  109,  110,  121,  128,  129,  130,  131,  135,  142,  146,  148,  166,  222,  230,  232,  236,  252,  254,  255,  274,  276,  277,  280,  295;  cf.  Franke, et al.

Hofstede, et al. : 280 Holderness : 59,  162,  167,  175,  176,  200,  223 Hollingsworth : 71 House, et al. : 97,  277,  296 Huang : 104 Hughes : cf.  Conn, et al. Hui : 120,  123,  126 Hunter : 124 Hutchinson : cf.  Hall, et al.

I Immergut : 71,  276 Inglehart : 96,  97,  98,  136,  146,  147,  277,  294

J Jappelli : cf.  Guiso, et al.:  127,  226,  282 Jensen : 172,  195,  202 Jepperson : 64,  147,  275,  290 Jimenez, et al. : 161

K Kabir : cf.  De  Jong, et al. Kagitcibasi : 128 Kahle : cf.  Bates, et al. Kahn : cf.  Calomiris, et al. Kahneman : 279 Kearney : cf.  Aggarwal, et al. Keloharju : 80,  120,  122,  140,  282 Kimbro : 108 King : 66,   162,   164,   165,   170,   177,   200,   219,  

257 Kirkman, et al. : 99 Kitayama : 145,  276 Klep : 63 Kluckhohn : 85,  99,  146,  274,  283 Koedijk : cf.  Brounen, et al. Kogut : 98 Korajczyk : 161

Kossmeier : cf.  Geyer, et al. Kroeber : 85,  274,  283 Kwok : 104,  105,  149,  231,  254,  257;  cf.  Shao,

et al.;   cf.   Chui, et al.   (2002):   149,   223,  230,  252,  254,  257

L La Porta : cf.  Glaeser, et al. La Porta : cf.  Djankov, et al.  (2002,  2008):  120,  

177,  277 La Porta, et al. : 80,  120,  167,  169,  172,  175,  

203,  277 La Porta, et al. (1997, 1998, 2000) : 86,  87,  

117,   118,   119   134,   135,   281,   282,   284,  286,  287,  288

La Porta, et al. (1999) : 86,  87,  117,  118,  119  134,  135,  281,  282,  284,  286,  287,  288

La Porta, et al. (2004) : 86,  87,  117,  118,  119  134,  135,  281,  282,  284,  286,  287,  288

Lang : 163,  175,  176 Lenartowicz : 130,  166,  222,  274 Leung, et al. : 130,   131,   136,   225,   257,   274,  

280,  282 Lévesque, et al. : 276 Levy : 161,  164 Licht : cf.  Siegel, et al. Licht, et al. : 109,  120,  121,  133,  135 Lloyd : cf.   Chui, et al.   (2002):   149,   223,   230,  

252,  254,  257 Longhofer : cf.  Calomiris, et al. Lopez De Silanes : cf.   Djankov, et al.   (2002,  

2008):  120,  177,  277 Lopez-De-Silanes : cf.   Glaeser, et al.;   cf.   La  

Porta, et al.   (1999):  120,  167,  169,  172,  175,  203,  277;  cf.  La  Porta, et al.  (2004):  120,  278;  cf.  La  Porta, et al.  (1997,  1998,  2000):  80,  120,  277

Lowe : cf.  Kirkman, et al. Lowell : cf.  McClelland, et al. Lucey : cf.  Aggarwal, et al.

M MacKenzie : 87,  137,  138,  286,  296 Mahajan : 165,  166 Mahrt-Smith : cf.  Dittmar, et al. Maksimovic : 166,  223,  257;  cf.  Booth, et al. Mansi : cf.  Harford, et al. Margaritis : 162,  200,  219,  224,  257 Markus : 145,  276 Marx : cf.  Barclay, et al. Mason : 171 Maxwell : cf.  Harford, et al.

330

Mayer : 71,  232,  257,  297 McClelland : 294 McCloskey : 276 McConnell : 167 McKinnon : 129 Mcliesh : cf.  Djankov, et al.  (2007):  177 McSweeney : 129 Meckling : 172 Merryman : 278 Mian : 112 Miao : cf.  Hackbarth, et al. Michaelas : cf.  Hall, et al. Miller : 241 Millo : 87,  137,  138,  282,  286,  290,  296 Mizruchi : 290 Modigliani : 241 Moore : cf.  Payne, et al. Morellec : cf.  Hackbarth, et al. Morosini, et al. : 102 Myers : 161,  165,  172,  219

N Nakata : 130,  222 Navarro : cf.  Ferreira, et al. Neuijen : cf.  Hofstede, et al. Nguyen : cf.  De  Jong, et al. North : 59,  80,  133,  144,  145,  275,  279

O Ofek : cf.  Berger, et al. Oh : cf.  Portes, et al. Ohayv : cf.  Hofstede, et al. Outreville : 101,  131 Oyserman : 99

P Palmer : 123 Parsley : 112,  113 Paudyal : cf.  Deesomsak, et al.  &  cf.  Antoniou,

et al. Payne, et al. : 293 Pensa : cf.  Bessler, et al. Pérez-Perdomo : 278 Pescetto : cf.  Deesomsak, et al. Peterson : cf.  Smith, et al. Pichler : cf.  Geyer, et al. Pirouz : 108 Podolny : 290 Polanyi : 297

Pop-Eleches : cf.  La  Porta, et al.   (2004):  120,  278

Portes, et al. : 111,  114 Powell : 70,  276 Psillaki : 162,  200,  219,  224,  257 Putnam : 117,  282

Q Quinten : 80,  99,  102,  132,  144,  280,  290,  294

R Ragin : 64,  69,  257 Rajan : 60,  166,  223,  257 Ramirez : 80,  99 Rantala : cf.  Anderson, et al. Raveh : cf.  Weber, et al. Read : cf.  Wiltbank, et al. Redding : 292 Rees : cf.  Clement, et al. Rey : cf.  Portes, et al. Rodrik : 278 Roe : 278,  292 Romano : 278 Ronen : 166,  222 Ros : 128 Rosati : 112,  113 Roth : 130,  166,  222,  274 Ruback : cf.  Baker, et al. Rubio : cf.  Ferreira, et al. Rudd : cf.  Grinold, et al. Ruiz-Mallorqui : 72,  163,  172

S Salas : cf.  Jimenez, et al. Salzmann : 105 Sanders : cf.  Hofstede, et al. Santana-Martin : 72,  163,  172 Santor : 66,  162,  164,  165,  170,  177,  200,  219,  

257 Sapienza : cf.   Guiso, et al.;   cf.   Guiso, et al.  

(2007,  2009):  86,  87,  117,  118,  119  134,  135,  281,  282,  284,  286,  287,  288

Sapienza, et al. : 281 Sarasvathy : cf.  Wiltbank, et al. Sarig : 170 Sarkissian : 112,  113,  114 Saurina : cf.  Jimenez, et al. Schatt : 61,  90 Schill : 112,  113,  114 Schleifer : cf.  La  Porta, et al.  (2004):  120,  278

331

Schmeling : 106,  107,  108,  132,  134 Schmid : 161 Schwartz : 96,  97,  98,  101,  103,  105,  109,  110,  

128,  135,  145,  146,  149,  166,  230,  231,  232,   252,   277,   295;   cf.   Licht, et al.,   cf.  Siegel, et al.  &  cf.  Smith, et al.

Scott : 71,  147,  219,  276 Secola : 112,  113 Semenov : 80,  105 Serna : cf.  Ferreira, et al. Servaes : cf.  Dittmar, et al. Shan : cf.  Morosini, et al. Shao, et al. : 0 Shefrin : 289,  293,  294 Shenkar : 129,  166,  222;  cf.  Weber, et al. Shiller : 80,  136,  293,  294,  296 Shleifer : 72,   162,   168,   203,   227,   236,   254,  

257;  cf.  Glaeser, et al.;  cf.  Djankov, et al.  (2002,   2008):   120,   177,   277;   cf.   La  Porta, et al.   (1999):  120,  167,  169,  172,  175,  203,  277;  cf.  La  Porta, et al.   (1997,  1998,  2000):  80,  120,  277

Siegel, et al. : 98,  101,  102,  118,  119 Simon : 299 Singh : 98,  99;  cf.  Morosini, et al. Sivakumar : 130,  222 Smelser : 86,  144,  275,  276,  279 Smith : 97,   98,   128,   130,   186,   219,   237,   257,  

274;  cf.  Barclay, et al. Smith, et al. : 97,  128 Soeters : cf.  Vinken, et al. Soskice : 71,  136 Spamann : 278 Statman : 80,  134,  135,  281 Stearns : 290 Stefek : cf.  Grinold, et al. Stonehill, et al. : 72,  125,  126 Streeck : 71 Stulz : 59,   61,   62,   69,   80,   120,   121,   125,   133,  

134,  135,  136,  139,  140,  145,  221,  231,  232,  281,  290;  cf.  Bates, et al.

Suarez : 133 Summers : 87,  127,  286 Swan : 171 Swanson : cf.  Clement, et al. Swedberg : 86,  144,  275,  276,  279

T Tadesse : 80,  99,  104,  231,  254,  257 Talmor : 170

Tartaroglu : 165,  166 Throsby : 86,   145,   275,   284,   285,   287,   288,  

292 Titman : cf.   Fan, et al.;   cf.   Chui, et al.   (2010):  

106,  107,  132,  134 Todd : 119 Toldra : cf.  Sapienza, et al. Triandis : 96,  98,  145,  276,  277 Tung : 130,  274,  280 Tversky : 279 Twite : cf.  Fan, et al.

V Vinken, et al. : 96,  99,  128,  277,  294 Vishny : 72,  162,  168,  203,  227,  236,  254,  257;  

cf.   La   Porta, et al.   (1997,   1998,   2000):  80,  120,  277

W Walke : 124 Wanzenried : 166 Weber : 80,  102,  145,  274,  294 Weber, et al. : 102 Wei : cf.  Chui, et al.  (2010):  106,  107,  132,  134 Welch : 227 Westbrook : 138,  168 Whitley : 71,  276,  292 Williams : 144,  283 Williamson, O. : 86,  106,  119,  125,  133,  144,  

221,  275,  279,  284,  287,  288 Williamson, R. : 60,   80,   120,   121,   135,   140,  

144,  145,  231,  232,  281 Wiltbank, et al. : 90 Wurgler : 72,   185,   187,   190,   219,   220,   224,  

235,  241,  257;  cf.  Baker, et al.

Y Yermack : cf.  Berger, et al.

Z Zazzaro : cf.  Alessandrini, et al. Zelizer : 296 Zingales : 60,  61,  117,  119,  128,  144,  165,  166,  195,  202,  223,  257,  275,  298;  cf.  Guiso, et al.  &  cf.  Sapienza, et al.;  cf.  Guiso, et al.  (2007,  2009):  86,  87,  117,  118,  119  134,  135,  281,  282,  284,  286,  287,  288

332

333

Table of contents for the English document

Title,  epigraph  and  exergue   51  

Abstract   53  

First  chapter:  Introduction   57  1. Genesis 59 2. The Survey and implications 62

2.1. The Survey 62 2.2. Dimensionalism and further dissertation-essays 63 2.3. National cultures, national institutions and national structures 63

3. Two empirical essays on capital structure 65 3.1. European listed firms and pooling: methodological stakes 65 3.2. Do agency relations mediate the interaction between firms’ financial

policies and business cycles? 67 3.3. Is the relation between business cycles, and leverage, mediated by

structural, institutional or cultural factors? 68 4. A conceptual addition 70 5. Conclusion: positioning, tentative interpretation and directions 71

5.1. Positioning 71 5.2. Tentative interpretation 71 5.3. Conclusion and directions 73

6. Bibliography 74

Second  chapter:  «  A  survey  of  culture  and  finance  »   77   Abstract 79 1. Introduction 80

1.1. Research question 80 1.2. Organization of the Survey 83

2. Background, stakes and methodology 85 2.1. Defining culture? 85 2.2. Screening and identification process 88 2.3. Classification, trends and first results 93

334

3. Dimensionalist approaches and dimensionalist successes 96 3.1. Some background on the dimensionalist approach 96 3.2. Empirical successes for dimensionalism in finance 101

4. Further culture-research in finance 111 4.1. Gravity- and other models of trade, as applied to financial flows 111 4.2. Three specific and successful path for culture-research in finance 120 4.3. Anecdotic and less anecdotic references to culture, uncovered in the

screenings 124 5. Emerging culture-views in finance: synthesis 128

5.1. Dimensionalism in management research, critiques and trajectories 128 5.2. Dimensionalism as an emerging framework for financial research? 131 5.3. National embededdness in finance 132 5.4. Culture as layers and considering financial cultures 136

6. Conclusion: shouldn’t financial research explore culture further? 139 7. Appendix 1 (chapter II). Intuitive breadth of culture definitions 143 8. Appendix 2 (chapter II). Definitions for national culture indices 148 9. Appendix 3 (chapter II). Institutional control variables in the dimensionalist

literature presented in our database 150 10. Bibliography 152

Third  chapter:  do  agency  relations  mediate  the  interaction  between  firms’  financial  policies  and  business  cycles?,   159  

1. Introduction 161 2. Financial policies, ownership and business cycles in institutional

perspective 165 2.1. Sampling European listed firms 165 2.2. Lean financial policies (dispersed ownership firms) 167 2.3. Financial smoothing (firms with concentrated ownership or under

control) 170 3. Sampling, data and methods 175

3.1. Pooling depending on the concentration of ownership 175 3.2. Firm-, industry- and country-specific data 176 3.3. Variables 183 3.4. Methodological proceedings 186 3.5. Empirical models 187

4. Results 188 4.1. Short notes about overall results 190 4.2. Interactions between financial policies and business cycles (1): debt 191 4.3. Interactions … (2): dividends, cash holdings and investments 194 4.4. Interactions …(3): varying proxies for anticipated business cycles 196

5. Conclusion. 200 6. Appendix 1 (chapter III): average ownership concentration in Europe 1990-

2007. 204 7. Appendix 2 (chapter III): pair wise correlations 206 8. Appendix 3 (chapter III): the E.S.I. measures 207 9. Appendix 4 (chapter III): main results with all coefficients reported 210 10. Bibliography 212

335

Fourth  chapter:  Is  the  relation  between  business  cycles  and  leverage  mediated  by  structural,  institutional  or  cultural  factors?,   217  

1. Introduction 219 2. Literature review and hypotheses 222

2.1. Specific dimensions underlying our methodological proceedings 222 2.2. The broader economic and financial context 225 2.3. Culture 230

3. Sampling, data and method 233 3.1. Sampling 233 3.2. Data 233 3.3. Variables 234 3.4. Models 237 3.5. Method-selection 237 3.6. Controlling for multi-colinearity in context-variables 238 3.7. Pooling 239

4. Results. 240 4.1. The effects of countries on firm- and industry-specific variables are

larger than those of the context variables. 240 4.2. Market timing and re-balancing asset classes 242 4.3. Pro-, vs. contra-cyclicality in leverage. Sources of managers’ behaviors

and resulting capital structures 246 4.4. Cultural factors, leverage, market timing and cyclicality in financial

policies 252 5. Conclusion 256 6. Appendix 1 (chapter IV): data collection. Synthesis from Reuter (2009) 258 7. Appendix 2 (chapter IV): Pair wise correlations 260 8. Appendix 3 (chapter IV): Unpooled results: robustness of context-variables 261 9. Appendix 4 (chapter IV): Results for pooled regressions. 262 10. Appendix 5 (chapter IV): Results pooled by country 264 11. Bibliography 265

Annex:  Embedding  Culture  and  Finance:  divorce  or  a  fresh  start?   269  1. Introduction 273 2. Culture, human sciences and contextual analysis 274

2.1. Institutional embededdness across human sciences… 275 2.2. The law & finance approach, illustrating embededdness. 277

3. Culture research in Finance and what it teaches us 279 3.1. The rising interest for “National Culture Indices” 279 3.2. Novel directions and their interconnectedness. 281

4. Culture research in Finance, a conceptual framework 283 4.1. Defining culture? 283 4.2. Culture: from defining to conceptualizing it 287 4.3. A typology of research frames 288 4.4. A practical illustration 289 4.5. Range of topics and contextualization 290

5. Finance, financial cultures and rationality 293 5.1. Culture, rationality & Behavioral Finance 293 5.2. Finance as a Culture and the performativity of markets 296

6. Conclusion 298 7. Bibliography 300

336

Bibliography,  tables  and  additions   307  Bibliography 309 Figures and tables 325 Index 327 Table of contents 333 Official short summaries 337 * * *

337

     

Official abstracts

Résumés protocolaires      

338

     English

This thesis is composed of three essays. The first is composed of a screening process that has been performed on peer-reviewed journals in finance, to investigate the recent rise in interest for cultural approaches. The aim has been to let definitions emerge in order to build a field-based analysis about culture in finance. The results include the following: the use of culture is mainly connected to the country puzzle; it concerns a very large range of topics; there is a marked polarization in approaches; and, finally, no consensual framework emerges from the screening process. As a result, the focus has been on a subject where the country puzzle is looming large: i.e. capital structures. A sample has been taken, consistent with the first essay findings and recommendations. The variability of the sample is limited in terms of: economic development, and judicial, political, institutional quality, and has focused upon European listed firms over a 20-year period.

In the second essay, the interactions between firms’ financial policies and expected business cycles have been investigated. Two conflicting mechanisms have been evidenced, finding: firms with dispersed ownership lead pro-cyclical policies, while firms with concentrated ownership lead contra-cyclical policies. The theoretical considerations unfold from the idea that ownership dispersion implies a different mix in agency relations, and entails specificities in agency costs and the benefits of managerial discretion, while it fosters differing needs for such things as, disciplining through debt and signaling.

The third essay shows that, both, the pro- and contra-cyclical mechanisms are mediated by structural, cultural and institutional factors. Specifically, it is demonstrated that the pro-cyclical effects are observed in contexts similar akin the “U.S Role Model”; furthermore, the effects encompass varying dimensions of capital structures, such as, cyclicality of leverage, market timing or still dividend policies, and they extend to varying contextual dimensions, including ownership dispersion, institutional anchoring, transparency, risk, structural variables for firms, the contracting environment, and measures of national cultures.      

Français Notre thèse compte trois articles. Premièrement, nous analysons de manière systématique les articles publiés dans les journaux à comité de lecture en finance qui font référence à la culture. Il s’agit de faire émerger les définitions en usage, à partir d’une analyse du terrain académique. Nos résultats sont les suivants : l’usage est généralement lié à l’énigme pays, il concerne un champs d’investigation large, il est marqué par une forte polarisation disciplinaire, et aucun cadre théorique accepté n’est identifiable. En conséquence nous nous attachons à l’étude d’un sujet où l’énigme pays prévaut : l’analyse des structures de capitaux. Notre échantillonnage - les sociétés européennes cotées sur 20 ans - procède directement de nos conclusions : nous nous efforçons de circonscrire les effets du développement économique, et de la variété qualitative dans les institutions légales, politiques, socio-économiques.

Deuxièmement, nous analysons les interactions entre la politique financière des entreprises et les cycles d’affaire anticipés. Nous mettons en évidence deux mécanismes opposés, l’un pro-cyclique et l’autre contra-cyclique, caractérisant respectivement les sociétés à actionnariat dispersé et concentré. Notre argumentation se fonde sur les corollaires de la dispersion de l’actionnariat : elle entraine un équilibre d’agence particulier, une variation dans les coûts d’opportunité de l’agence et de la latitude managériale, un rôle disciplinaire spécifique pour l’endettement, un rapport de signaux altéré.

Troisièmement, nous démontrons que ce mécanisme principalement financier, - de polarisation de la politique financière en fonction des cycles d’affaires anticipés - est associé à un contexte institutionnel et à des comportements financiers différents. En particulier nous démontrons que la même polarisation empirique, observée pour la cyclicité du levier d’endettement des entreprises, est observée en relation aux politiques de dividende, au « market timing », concerne l’ancrage institutionnelle des sociétés, leur profil de transparence, certaines de leurs caractéristiques structurelles, l’environnement contractuel, ainsi que des variables culturelles.