Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
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.