The Role of Banks in Monitoring

THE ROLE OF BANKS IN
MONITORING FIRMS
Does the Anglo-American approach to the relationship between banks and

firms have significant weaknesses compared with the German and Japanese
approach?
This book addresses issues in the current literature on corporate finance
using historical evidence. In particular it looks at the role of universal banks
in relaxing the credit constraints of firms, supervising managers and
stabilising share prices. The key issue is whether Anglo-American asset-
based financing is more effective than the main-bank approach used in
Germany and Japan. Earlier studies have found that firms with a close
relationship with a major bank have high market value compared to book
value, although it is difficult to determine whether this is cause or effect.
The case of the Crédit Mobilier—the first universal bank—is interesting
because the bank failed. Had it been the case that links with the bank
brought about high and stable share prices or relaxed credit constraints, the
bank’s bankruptcy should have precipitated the loss of these benefits. In
fact, the bankruptcy had almost no effect on the share prices or the
investment behaviour of the relevant firms, casting doubts on the benefits of
powerful banks.
Elisabeth Paulet is Research Assistant at the European University Institute

of Florence. Her main interest is in banking systems from the nineteenth
century to the present day.
ROUTLEDGE EXPLORATIONS IN
ECONOMIC HISTORY
1. ECONOMIC IDEAS AND GOVERNMENT POLICY

Contributions to contemporary economic history
Sir Alec Cairncross
2. THE ORGANIZATION OF LABOUR MARKETS
Modernity, culture and governance in Germany, Sweden, Britian and Japan
Bro Stråth
3. CURRENCY CONVERTIBILITY IN THE TWENTIETH CENTURY
The gold standard and beyond
Edited by Jorge Braga de Macedo, Barry Eichengreen and Jaime Reis
4. BRITAIN’S PLACE IN THE WORLD
A historical enquiry into import controls 1945–1960
Alan S.Milward and George Brennan
5. FRANCE AND THE INTERNATIONAL ECONOMY
From Vichy to the Treaty of Rome
Frances M.B.Lynch
6. MONETARY STANDARDS AND EXCHANGE RATES
Edited by M.C.Marcuzzo, L.Officer and A.Rosselli
7. PRODUCTION EFFICIENCY IN DOMESDAY ENGLAND, 1086
John McDonald
8. FREE TRADE AND ITS RECEPTION 1815–1960
Freedom and trade: Volume I
Edited by Andrew Marrison
9. CONCEIVING COMPANIES
Joint-stock politics in Victorian England
Timothy L.Alborn
10. THE BRITISH INDUSTRIAL DECLINE RECONSIDERED
Edited by Jean-Pierre Dormois and Michael Dintenfass
11. THE CONSERVATIVES AND INDUSTRIAL EFFICIENCY, 1951–1964
Thirteen wasted years?
Nick Tiratsoo and Jim Tomlinson
12. PACIFIC CENTURIES
Pacific and Pacific Rim economic history since the 16th century
Edited by Dennis O.Flynn, Lionel Frost and A.J.H.Latham
13. THE POSTMODERN CHINESE ECONOMY
Structural equilibrium and capitalist sterility
Gang Deng
14. THE ROLE OF BANKS IN MONITORING FIRMS
The case of the Crédit Mobilier
Elisabeth Paulet
THE ROLE OF BANKS
IN MONITORING
FIRMS
Elisabeth Paulet
London and New York

First published 1999
by Routledge
11 New Fetter Lane, London EC4P 4EE
This edition published in the Taylor & Francis e-Library, 2002.
Simultaneously published in the USA and Canada
by Routledge
29 West 35th Street, New York, NY 10001
© 1999 Elisabeth Paulet
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
Paulet, Elisabeth, 1962–
The role of banks in monitoring firms: the case of the Crédit
Mobilier/Elsabeth Paulet.
p. cm.
Includes bibliographical references.
1. Corporations—Finance—Case studies. 2. Business enterprises—
Finance—Case studies. 3. Banks and banking—Case studies.
4. Crédit Mobilier (France) —History. 5. Credit—Case studies.
I. Title.
HG4026.P329 1999
658.15´224–dc21 98–30729
CIP
ISBN 0-415-19539-X (Print Edition)
ISBN 0-203-21180-4 Master e-book ISBN
ISBN 0-203-21192-8 (Glassbook Format)
It is necessarily part of the business of a banker to
profess a conventional respectability which is more than
human. Lifelong practices of this kind make them
the most romantic and the least realistic of men.
John Maynard Keynes

CONTENTS
List of figures ix
List of tables x
Preface xiii
Acknowledgements xiv
INTRODUCTION 1
1 AGENCY THEORY AND MONITORING

A theoretical and empirical interpretation 7
1.1 Introduction 7
1.2 Presentation of the financial theories 8
1.3 The supervisory role of the bank in agency literature 15
1.4 General conclusion 24
2 THE CRÉDIT MOBILIER AND THE FRENCH STOCK

EXCHANGE 1853–1914
An empirical perspective 26
2.1 Introduction 26
2.2 Financial capitalism and data on the Crédit Mobilier in
France, 1853–1914 29
2.3 Data available over the period of the two Crédit Mobiliers:
Companies involved and preliminary analysis of the data 40
2.4 Excess volatility tests 45
2.5 The effect of the bankruptcy of the Crédit Mobilier on
the companies with which it was involved 53
2.6 Bankruptcy effect on the companies not affiliated to
the Crédit Mobilier 65
vi
CONTENTS
2.7 Comparative analysis of the results obtained for affiliated

and non-affiliated companies 71
2.8 Bankruptcy effect on the companies either affiliated or not affiliated
to the Crédit Mobilier 73
3 CORPORATE INVESTMENT, CASH FLOW AND

FINANCIAL CONSTRAINTS OF FIRMS
The case of the Crédit Mobilier 76
3.1 Introduction 76
3.2 The value of association with the Crédit Mobilier on cash
flow, investment and market value 78
3.3 Transformation of the raw data according to ‘traditional
financial accounting’ 91
3.4 Extensions of the results: description and interpretation
of a robustness test 98
3.5 The expression of all variables in real terms 104
4 THE SUPERVISORY ROLE OF THE CRÉDIT MOBILIER

Some interpretations 109
4.1 Introduction 109

4.2 The supervisory role of the Crédit Mobilier 111
4.3 Advantages and disadvantages of an association with a
bank for the financing of productive projects 116
4.4 The Crédit Mobilier banks: a comparative analysis
between France and Germany 124
4.5 The attempt by the Crédit Mobilier’s shareholders to
prevent the bank’s bankruptcy 128
4.6 Conclusion 129
5 GENERAL CONCLUSIONS 132
APPENDIX A
Data on the Crédit Mobilier 136
A.1 The Crédit Mobilier between 1852 and 1867 136

A.2 The Crédit Mobilier between 1903 and 1914 136
vii
CONTENTS
APPENDIX B
Data relative to the general indices (GNP, share prices)
for France between 1852 and 1914 140
B.1 General price indices 140
B.2 Share price series 141
APPENDIX C
Data relative to the affiliated companies 142
C.1 The whole data 142
C.2 The Compagnie Immobilière 142
APPENDIX D
Data relative to non-affiliated companies 146
APPENDIX E
Monthly share prices for affiliated companies
between 1866 and 1868 148
APPENDIX F
Data for investment and cash flow tests 152
Notes 161
Bibliography 166
Index 174
viii
FIGURES
1.1 Moral hazard with hidden action 10

1.2 Adverse selection models 10
1.3 Signalling model 11
2.1 Prices, dividends, and perfect-foresight fundamental
sample 1870–1914 46
ix
TABLES
2.1 Ownership of the shares of the Immeubles de la Rue de

Rivoli 32
2.2 Advances to companies, 1856–66 32
2.3 Fluctuations of share prices for affiliated companies
between 1855 and 1856 34
2.4 Crédit Mobilier’s liquidation account in 1901 36
2.5 Market product prices by sector 41
2.6 Volatility tests for the French, German and American
stock markets between 1870 and 1914 49
2.7 Average asset prices for affiliated firms 54
2.8 Relative impact of the bankruptcy of the Crédit Mobilier 55
2.9 Relative impact of the bankruptcy of the Crédit Mobilier
including a dummy variable d1867 56
including a dummy variable d1867/68 57
2.11 Regression for d1867 58
2.12 Regression for d1867/68 59
2.13 Results for 1866 61
2.16 Results for 1866–8 64
2.17 Relative impact of the bankruptcy of the Crédit Mobilier 67
including a dummy variable d1867 68
including a dummy variable d1867/68 69
2.20 Results for d1867 70
2.21 Results for d1867/68 70
2.22 Cross-section results excluding the Compagnie Immobilière 73
2.23 Cross-section results including the Compagnie Immobilière 73
3.1 Companies included in the data set 80
x
TABLES
3.2 Summary statistics comparing Crédit Mobilier and non-Crédit

Mobilier firms 81
(a) Sample 1861–75 81
(b) Sample 1861–7 81
(c) Sample 1868–75 82
3.3 Investment regression, 1861–75 84
3.6 Investment regression equation for the whole sample 88
3.7 Fixed effect model for 1861–7 89
3.9 Fixed effect model for the whole sample 89
3.10 Summary statistics comparing Crédit Mobilier and
non-Crédit Mobilier firms when introducing a
depreciation rate of 5 per cent 92
3.11 Investment regression for the whole sample 1861–75 93
3.12 Investment regression over 1861–7 94
3.13 Investment regression over 1868–75 94
3.14 Results for affiliated companies as presented in Becht and
de Long’s paper 95
3.15 Results for non-affiliated companies as presented in
Becht and de Long’s paper 95
3.16 Fixed effect model for the whole sample 96
3.19 Fixed effect model for all firms 97
3.20 Investment regression equation for affiliated companies
and independent firms over the whole sample 98
(a) Case when d = 3% 98
(b) Case when d = 7% 98
and independent firms before the bankruptcy 99
and independent firms after the bankruptcy 100
3.23 Cash flow parameters 100
3.24 Sample 1861–75 robustness results for Crédit Mobilier
firms without the Compagnie Immobilière 101
3.25 Sample 1861–7 robustness results for Crédit Mobilier
firms without the Compagnie Immobilière 101
xi
TABLES
3.26 Investment regression equation for all companies over the

different samples 102
(a) Investment regression equations when d = 3% 102
(b) Investment regression equations when d = 7% 102
3.27 Cash flow parameters 103
3.28 Q parameters 103
3.29 T-statistics for sample 1861–7 as regards the cash flow
parameters 104
3.30 Summary statistics comparing Crédit Mobilier and
non-Crédit Mobilier firms when parameters are expressed
in real terms 105
3.31 Investment regression equation for the whole sample 105
3.32 Investment regression equation for 1861–7 106
3.33 Investment regression equation for 1868–75 106
3.34 Investment regression equations for all firms 106
3.35 Regression for the whole sample without the Compagnie
Immobilière 107
3.36 Comparison of cash flow and Q parameters 108
4.1 Financial participation of the Crédit Lyonnais in
‘big deals’ 122
A1 Balance sheet of the Crédit Mobilier from 1853 to 1866 137
A2 Average profit of the Crédit Mobilier from 1855 to 1865 138
A3 Share prices of the Crédit Mobilier from 1853 to 1866 138
A4 Balance sheet of the Crédit Mobilier from 1903 to 1914 139
A5 Share price series 141
A6 Data available for affiliated companies: share prices 144
A7 Revenues of the Compagnie Immobilière 145
A8 Expenses of the Compagnie Immobilière 145
A9 Share price and dividend distribution, 1855–66 145
A10 Data for non-affiliated companies 147
A11 Share prices for affiliated companies for 1866 149
A14 Data available for the railway companies 153
A15 Data available for mining industries 154
A16 Data available for the steel industries 155
A17 The Compagnie Immobilière 156
A18 New values of Q when introducing a depreciation rate of
5 per cent 157
A19 Transformation of the data when changing origin 158
A20 Data in real terms 159
xii
PREFACE
Economists seem to agree on the monitoring power of banks as regards

their clients. Their arguments are based on the ability of these
institutions to collect information on exposure to risk, and on their
flexibility to diversify investment. In order to exert this monitoring role,
a banker should always evaluate the trade-off between profitability and
efficiency.
This book examines how universal banks could help relax the credit
constraints of firms, improve the supervision of managers and stabilise share
prices. The case of the Crédit Mobilier, which went bankrupt in 1867, is a
good illustration of the weakness of the monitoring role of banks. It exhibits
the potential of enterprises to diversify their investments in order to avoid
financial destabilisation. A historical approach adds evidence to this last
assertion and casts doubt on the beneficial role of closed relationships with
powerful banks.
As far as bankruptcy is concerned, the attitude of the central bank is
an interesting factor. At that time, the Banque de France and its
governors were hostile to the founders of the Crédit Mobilier, the
Péreires; consequently, its participation in rescuing the bank was almost
non-existent. Beyond the historical context, this fact illustrates the
importance of political pressures in financial decision-making. This
debate remains very sensitive, particularly in view of the discussions
concerning the future of the European central bank. As a whole, the
questions mentioned in this essay are still valid today.
This book was written in order to present some new ideas regarding
the control and monitoring of financial institutions and, more
specifically, universal banks. While the flexibility of other structures
must be acknowledged, there are moves afoot to create a unified
European banking system in which universal banks will have a privileged
position on the financial markets. The justification for this concerns their
perceived superiority as evaluators of information and risk. The
conclusions of the empirical analysis discussed here have a direct
bearing on the power of universal banks.
xiii
ACKNOWLEDGEMENTS
This book is based on my Ph.D. dissertation at the European University

Institute in Florence, written between 1991 and 1995. I would like to
express my gratitude to all the people who supported me at that period:
many thanks to Robert Waldmann who has acted with great sensibility as a
supervisor; the History Department at the Institute, and especially Peter
Hertner, Albert Carrerras and Kirti Chaudhuri, who have always been
interested in my research; William Kennedy, who kindly provided me with
useful comments and encouragement; and last, but not least, Maurice Levy
Leboyer who, through written exchanges, enabled me to form a new opinion
on the personality and position of the Péreire brothers in nineteenth-century
financial circles.
My research led me to collect data in Paris. I am grateful to the personnel
of the Paris Archives for their contribution to this work.
Much of what follows is based on a simple idea: the Crédit Mobilier and
its founders, the Péreires, failed because they encountered difficulties that
were due more to external constraints than is usually acknowledged. In other
words, environment matters. My own environment in the course of writing
the present book has been as follows. For several years Francesc Relaño has
endured my ramblings on the course of banking policy over the last century.
My parents have provided support during the most austere period of the
writing process.
My last thought is to dedicate this book to my brother, who will never
have the opportunity to read this piece.
To Patrick.
xiv
INTRODUCTION
Bankers and investors influence industrial investment through the financial

power they exert on the managers of firms. The purpose of this research is
to study the interrelations that exist between these managers, their
shareholders and the banks which finance investment projects when defining
a debt-equity contract. The subject will be approached from two diverse
perspectives: the first one is theoretical (Chapter 1), the second empirical
(Chapters 2, 3 and 4).
Theoretical arguments can be, and have been, advanced to describe how
the conflict of interest between managers and owners of firms can be solved.
The purpose of the first chapter is to survey the literature on principal-agent
theory and the signalling approach which concentrates on these questions. In
particular the power of banks will be debated. Two questions are considered:
• Why might banks want and be able to obtain the power to influence and
monitor managers?
• Why might shareholders benefit if the bank has such power?
Agency and signalling theory are concerned with the information and
interest conflicts which can exist between the different economic actors
associated with the firm: managers, shareholders and creditors. This
theory holds that some agents (the managers) are better informed about
the quality of the enterprise than the principals (shareholders, creditors).
To resolve the conflict of interest, contracts (implicit or explicit) must be
negotiated between the lenders, the creditors, the shareholders and the
managers.
The term ‘agency’ derives from the fact that corporate decisions are
delegated to agents (e.g. managers) who act on behalf of other parties
(external investors). The relation between investment and debt is explained
by the choice between debt and equity for the financing of a project.
The signalling approach makes the assumption that there is asymmetric
information as between management and investors: the managers have
more information than the shareholders on the value of their enterprise.
1
THE ROLE OF BANKS IN MONITORING FIRMS
The problem is the accurate transmission of such information, to give

shareholders the incentive to finance productive projects.
Both approaches feature the monitoring role exerted by lenders on
borrowers: control remains one of the main factors of agency. The concept
of monitoring will presently be examined, focusing on the role of banks in
performing this control function. According to agency theory, the bank will
have less information about the project than the managers; the role of
collateral can be useful in dealing with the problem of adverse selection.
Adverse selection arises when lenders cannot distinguish managers of firms
who undertake low-risk projects from the ones who realise high-risk
projects. Then a contract that is optimal when offered to less risky
borrowers also attracts more risky borrowers and thus becomes unprofitable
to the lenders. The use of collateral constitutes one way to control this risk.
Several other tools will be discussed.
Two examples are presented to illustrate the argument: Japan and
Germany. The aim of this analysis will be to compare the conclusions
reached for these two countries with the results obtained for the Crédit
Mobilier.
Historians have argued that the influence exerted on industry by banks
such as the German Großbanken or the French Crédit Mobilier was
substantial during the nineteenth century. The role played by the
Großbanken—those banks which, like the Crédit Mobilier, took a great role
in the industrialisation of Germany—in monitoring and supervising
corporate management was an accepted part of German finance theory in the
years before the First World War.
The second step of my analysis is to gain, through empirical research, an
insight into the role of banks in the financing of enterprises during the
period 1850–1913. In particular, I study the Crédit Mobilier’s role in France
during the pre-WWI period. This bank, founded by the Péreires in 1852,
until it went bankrupt in 1867, contributed largely to French
industrialisation. It failed a second time at the beginning of 1900 and
merged in 1932 with the Banque de l’Union Parisienne.
The Crédit Mobilier has been chosen for two reasons:
• Its organisation was similar to that of the German Großbanken. The

comparison with these institutions can give some additional insight to
the results mentioned in the literature.
• As the bank went bankrupt, it will be interesting to analyse the impact
of this failure on the firms affiliated to it. If a real monitoring role were
exerted by the Crédit Mobilier, affiliated firms should have faced
difficulties when the bankruptcy occurred.
First the role played by the Crédit Mobilier on the stock market is
determined. The level of volatility will be quantified in order to assess the
2
INTRODUCTION
French general economic context between 1852 and 1913. Share price series
are then used to investigate the impact of the bankruptcy of the Crédit
Mobilier on the stock market for affiliated and non-affiliated companies.
Share prices did not fall when the bank failed. The conclusion is that the
Crédit Mobilier provided liquidity to firms, but apparently did not seem to
be supervising them, and its seal of approval was probably not worth as
much as that of J.P.Morgan1 or the Großanken.2
Some remarks must be made concerning the data used to construct the
test. During the period statistical sources such as share prices and
balance sheets were not always kept according to strict rules. In addition
the amount of statistical information available over the period was quite
limited because managers were not obliged to publish their accounts. So
the validity of our results has to be assessed in the context of a panel
data of twenty five companies which provided information on their
financial situation.
As the Crédit Mobilier seems not to have succeeded in controlling
firms, it would be interesting to measure the effect of an association with
this establishment on investment policy. The basic theoretical idea is as
follows. In their world of perfect information and no agency costs, firms
can borrow and invest optimally; liquidity has no effect on the
investment decision. In their classic paper, Franco Modigliani and
Merton Miller (1958) showed under what conditions financial structure
becomes irrelevant for real investment decisions. In a context of
asymmetric information this proposition is no longer true. The next step
in our research is to analyse the relationship between investment and
cash flows between 1860 and 1880.
The effect on investment policy of an association with this
establishment will then be measured and compared to Großbanken
policy during a similar period. This proposed analysis then tests the
proposition that the Crédit Mobilier acted like a German-style universal
bank. To that intent it examines the effects on the investment behaviour
of French firms of affiliation with the Crédit Mobilier. Investment
equations for a modified version of Tobin’s Q-model are estimated.
Two samples are considered: the first consists of corporations that were
affiliated to the Crédit Mobilier, the second includes companies which
had to resort to other sources of finance. The focus of this construction
is on contrasting the financing of the Crédit Mobilier’s affiliated
enterprises with those which, for one reason or another, did not form
part of the nexus.
The chronological limits of this study are imposed by the data sources:
after 1880 the Crédit Mobilier lost its powerful role in the banking sector
and thus no records exist after this point.
The sample includes fifteen companies: seven were affiliated with the
Crédit Mobilier, eight with the Crédit Lyonnais. As each company, in any
3
event, had a relationship with a bank, the interpretation of the results must
be in some sense cautious.
Using this evidence, the following two conclusions are reached:
1 companies that maintained close ties with the Crédit Mobilier were less
liquidity constrained after the bank’s crisis in 1867 than before;
2 unaffiliated companies experienced no more difficulty in financing
projects than those which resorted to the Crédit Mobilier.
These results strongly contradict earlier findings. The interpretation will

define the context in which they are valid.
Some empirical extension of the results is provided. In particular, we
intend to develop two different models: the first will try to re-evaluate the
book value for all companies entering the sample by introducing a fixed
depreciation rate of 5 per cent. This is because I suspect the account figures
for book value changed only rarely from year to year. The second
proposition is to divide all variables by the general level of prices. If the
conclusions are not substantially different they support the arguments
mentioned above.
An interpretative chapter discusses whether the supervision of the Crédit
Mobilier was really effective—if the firms with which it was involved had
not anticipated the fall of the bank, and diversified their sources of financing
corresondingly. Our empirical results prove that the bank did not really
control firms. The regressions suggest that the affiliated firms were more
liquidity constrained before the bankruptcy than after. This supports the
argument that the firms did not really trust the Crédit Mobilier as a banking
institution—they just used it to finance their investment projects. A large
part of the economic literature considers that a bank-firm association is
always profitable for both parties because:
• the bank prevents the firm from undertaking projects that are too risky;
• the firm signals its reputation to the market by having a regular relation
with a bank.
The Crédit Mobilier’s case gives no clear evidence to confirm this

argument. The question is then: What is the most efficient way to control a
firm: an interaction between firms and banks (which means substantially a
bank-firm association), or a clear separation of banks and firms (as
presented in our first chapter)? In other words, should banks supervise, or
just choose whether or not to loan? In particular, we can ask if the principal
(the Crédit Mobilier) was not an agent acting for firms which needed
liquidity over the period.
A comparison between France and Germany is then drawn. Other
researchers 3 have shown the efficiency of the German Großbanken
4
INTRODUCTION
(which were explicitly modelled on the Crédit Mobilier). The diverse

results obtained in the two cases enable me to say that in many countries
the Crédit Mobilier was the inspiration for a new type of banking, but
not the model. The last point will be a debate on the rationality or
irrationality of the shareholders. In particular, some explanation for their
passive attitude in the face of the bankruptcy will be advanced. The last
chapter concludes.
5
1
AGENCY THEORY
AND MONITORING
A theoretical and empirical interpretation
For French business, whether medium or ‘big’, the obsession

with independence vis à vis the lender (the bank) was constant.
It was with reference to the level of their current bank accounts
that the Schneiders of the nineteenth century, for example,
calculated their degree of independence and their room for
manoeuvre…. The domination of banks over business, if it
existed was not achieved without resistance.1
Bouvier (1988:109)
1.1 INTRODUCTION
In this chapter the basic concepts of agency and signalling theories applied
to the financing of the firm are presented. In particular the chapter will
study how a debt-equity contract is defined between the manager of a firm,
the shareholders and the bank which provides liquidity.
Agency theory, although not quite new, is a part of decision theory
which is currently characterised by a dynamic development. The well-
known basic model of decision theory is related to a situation in which
one person—the so-called decision-maker—has to make a decision as well
as to bear its consequences. This assumption is often unrealistic.
Therefore, in agency theory the decision-maker (the agent) and the
‘beneficiary’ of the decision (the principal) are all distinct. A typical
example concerns the management of the firm: the agent, the manager of
the firm, chooses an investment project. The principal—the shareholders of
the firm and (or) the bank which finances the enterprise—partly support
the risks involved in this investment.
Hence, the common element is the presence of at least two individuals.
The first (the agent) must choose an action from a number of alternative
possibilities. The action affects the welfare of both the agent and the
principal. The principal has the additional function of prescribing payoff
rules; that is, before the agent chooses the action, the principal determines a
rule that specifies the fee to be paid to the agent as a function of the
7
principal’s observation. The problem acquires interest only when there is

uncertainty at some point and, in particular, when the information available
to the participants is unequal.
As information is asymmetric between the economic agents, control is
one of the main factors of agency. The second part of this chapter will
be devoted to the definition of this control mechanism in order to
determine which of the two principals, the shareholders of the firm or
the bank with which the enterprise is involved, is able to exert
effectively this monitoring role.
1.2 PRESENTATION OF THE FINANCIAL

THEORIES
(a) Basic concepts of agency and signalling theories
The financial theory of agency focuses on the relationships between

different groups of security holders (equity and bondholders) in the
context of optimal financing of the firm. The standard economic
theory of agency considers two individuals: a principal, who provides
the capital, and an agent (the manager), who provides the effort. Both
are assumed to be utility maximisers. Principals value end-of-period
wealth which is derived from their share in the realised value of the
firm. Agents value their end-of-period wealth stemming from the
realised profit.
Agency problems arise because, under the behavioural assumption of self-
interest, agents do not invest their best efforts unless such investment is
consistent with maximising their own welfare. The agency model is basically
a formulation of the principal’s problem of choosing the ‘best’ employment
contract for the agent, where ‘best’ is defined in the context of Pareto
optimality. A Pareto optimal contract is such that no other contract can
improve the welfare of one party without reducing the welfare of the other.
Observability of the agent’s effort is really the core of the incentive
problem in this theory. While the effort level determines the level of output
of the firm (i.e. the end-of-period value or cash flows), the output is also
affected by random events that are beyond the control of the agent. An
agency problem arises when the consequences of the agent’s efforts cannot
be distinguished entirely from the consequences of other random events by
observing output alone. In the tradition of the agency theory, the output
(payoff) level is mutually observable by the principal and the agent, but the
effort level is observable only by the agent. First best contracts induce an
optimal risk-sharing between agent and principal.
Further adverse selection is often introduced between the two
parties. Consider an owner-manager who seeks to finance a project by
8
AGENCY THEORY AND MONITORING
selling securities, while the true nature of the return distribution of the
project is unknown to the outside market. Management possesses
valuable information about the project that is unavailable to the market.
If this information were revealed fully to the market, the market would
value the project at V . Otherwise, the market is unable to distinguish
a
this project from another less profitable project with a value V . This
b
asymmetry may be corrected, at a cost, through various ‘signalling
mechanisms’. In the absence of an unambiguous signal, however,
management will obtain less for securities sold than their ‘fair value’
reflected in the true nature of the project A. The difference between the
‘fair’ price and the actual price is the agency cost associated with
informational asymmetry. It exists for the issuing of debt as well as
new equity securities, provided that there is a differential probability of
bankruptcy for the two projects. In their seminal paper, Jensen and
Meckling (1976) proposed agency costs as a key tool in evaluating
designs of a principal-agent relation. Agency costs were defined as the
sum of:
1 the monitoring expenditure by the principal;

2 the bonding expenditure by the agent; and
3 the residual loss, i.e. the monetary equivalent of the reduction in welfare
experienced by the principal due to the divergence between the agent’s
decision and ‘those decisions which would maximise the welfare of the
principal’ (ibid.: 308). The agent receives compensation for his work and
supports part of the agency cost (bonding expenditure).
Adverse selection can be described through three categories of model.

The first is represented by a situation of moral hazard with hidden
action. Under moral hazard, it is easier for the principal to observe the
manager’s output than his effort. The principal offers a contract to pay
the agent based on output which depends on the agent’s effort. We can
represent this situation by a game tree. In the model, the principal (P)
offers the agent (A) a contract, which he accepts or rejects. After the
agent accepts, nature (N) adds noise to the task being performed. Figure
1.1 illustrates this situation.
If the principal knows the agent’s ability but not his information level
about the potential investment, the problem is moral hazard with hidden
action. An example is given by F.Gjesdal (1982), whose paper considers
the use of imperfect information for risk sharing and incentive purposes
between bondholders and stockholders when perfect observation of
actions and outcomes is impossible, making complete contracting
infeasible. The incentive problem consists of two parts: the choice of
information system and the design of a sharing rule based on the
information in the agency problem.
9
Figure 1.1 Moral hazard with hidden action
Figure 1.2 Adverse selection models
The second case concerns the largest part of the agency literature:
adverse selection models. Adverse selection can be represented by
Figure 1.2.
Kose and Kalay (1982) illustrate adverse selection by analysing the
conflicts between two groups (stockholders and bondholders) concerning the
riskiness of company projects. They derive an optimal set of contractual
arrangements which minimise the cost of this conflict. In that sense, these
authors come back to the idea of Myers (1977), who has argued that
stockholders who control the firm could attempt to transfer wealth from the
bondholders, even by rejecting profitable investment projects. In other words
they can choose policies which reduce the market value of the firm. Myers
has indicated that self-imposed constraints on dividend pay-outs by
stockholders are a possible solution to the problem of under-investment.
The debt contract between a bank and the manager of a firm defines the
adverse selection problem. According to the recent literature, such as
Laffont and Freixas (1988), Gale and Hellwig (1985) and Diamond (1984),
the assumption of asymmetric information leads the bank to prefer credit
rationing, to an equilibrium between supply and demand at a specific
interest rate. The agency relations between a bank and a firm prevent the
lender from distinguishing sound creditor risks from bad ones. The
requirement of a guarantee can solve such problems and enables the lender
to separate good borrowers from the bad ones. In this context Deshons and
Freixas (1987) establish two types of loan system:
• the optimal separating contracts, where good credit risks are identified;
• the rationing credit solution, where neither guarantee nor interest rate
make the selection role possible since the bank is unable to take the
surplus of its loan customers.
10
Figure 1.3 Signalling model
The latter situation, represented by Figure 1.3, is the signalling model.

The agent sends a signal to the principal.
For example, the managers of firms transmit information to the
shareholders and the creditors of the firm they administer. This activity of
signalling has two aspects:
• the manager of a good firm sends information to the external investors

to signal that the enterprise is a good firm;
• however, the manager of a bad firm might use the same signal to make
investors think that the enterprise is a good one, but it is more costly for
him to do so.2
A penalty system must be in place to prevent the manager of a bad firm

from sending false signals. The problem of signalling in the
determination of the debt level of a firm is expressed by the
diversification of a security portfolio for a manager-shareholder. If the
latter possesses a good productive investment, he will inform the market
by the composition of his portfolio. If the project is really good, he will
devote a significant part of his savings to it, to the detriment of other
projects. As he is the best informed, his less diversified portfolio is a
signal which can indicate the value of the project. Leland and Pyle
(1977) conclude that the value of the enterprise is positively correlated
to the proportion of capital held by the shareholder-manager. But the
authors emphasise that this relation is a non-causal statistical relation
between the value of the firm and its debt level. Every modification of
the manager’s portfolio induces a change in the perception of the future
liabilities by the market: it results in a new financing policy, and
ultimately another value for the firm.
In this context, the level of diversification in the assets held by a firm
constitutes the most valuable information as regard the value of the firm. As
observed by Jensen and Meckling (1976) or Grossman and Hart (1983), if a
firm does not want to be indebted, its risk of bankruptcy is limited but the
market will suppose that the maximum performance for the firm has not
been attained. In the opposite case, a loan from a bank increases its risk of
bankruptcy, which obliges the managers to be more efficient. This argument
illuminates the divergence of objectives between the different partners: the
11
shareholders, who seek dividends, the managers, who must signal the value
of their organisation by diversifying their portfolio, and the banks, which
grant credit and want to avoid losses if the firm goes bankrupt.
To conclude these first two points, it seems that the application of agency
and signalling theory to finance decisions is concerned with the influence of
conflicts between shareholders and managers over the value of the firm, and
the resolution of these conflicts by the use of debt. The latter involves an
incentive scheme for performance. The debt increases the bankruptcy risk,
which is a threat sufficient to encourage the manager to administer
efficiently. Bankruptcy will mean for the manager not only loss of his work,
but also being held responsible for the situation.
In these models, debt is defined by an optimal contract for the firm,
and the optimal structure of the enterprise is a combination of debt and
stocks. So, the financial parameters justifying the limits of debt policy
can be deduced by analysis of the shareholders’ and managers’ points
of view (e.g. level of risk of the project, expected profit, etc.).
However, these theories have a certain number of limitations. In
particular, the same financial indicators have specific roles in each
approach. The debt level constitutes a very manipulable variable for the
administration of the firm. It signals the quality of an enterprise for
Ross (1977). It is the origin of the agency cost for Galai-Masulis
(1976) and Jensen and Meckling (1976). The results of these analyses
do not seem to be very coherent: an increase in debt or in the part
attributed to the manager and shareholders in the capital induces an
increase in the market value of the firm, according to the signalling
models. But the difficulties due to agency relations then become more
and more pronounced. Myers (1977) has taken an interest in this
problem. He explains that three financial parameters must be taken into
consideration at the same time: the dividend distribution, the amount of
investment and the level of debt.
Agency costs appear there as one of the essential elements of a new
kind of signalling model. For Myers, the existence of debt for an
enterprise can generate a sub-optimal level of investment: the nature of
equity and the financial structure can be modified such that the investors
do not maximise the value of the firm but the value of the stocks, to the
prejudice of debt.
(b) Common agency
As indicated above, agency theory aims to explain the conflicts of interest and
differing information that could exist between the different partners of the firm:
managers, shareholders and lenders. Every partner determines the optimal level
of debt according to his own interest. The manager wants to find the amount of
12
funds he thinks sufficient. For the shareholders and the bank, investment is a
way to realise profit. However, the latter actors bear different costs. The
shareholders give the manager financial compensation for this work, while the
banker pays interest costs for granting the credit (Laffont and Freixas 1988).
In this section, possible areas of common interest between these three
actors in the financing of the firm will be discussed. In particular the
following question will be asked: are these different objectives sufficiently
compatible such that the common determination of debt for the firm might
be possible? Are there situations where the three partners are led to define
jointly the very price to pay, so that the firm could finance its productive
activities through an optimal level of debt?
When several economic agents must take a decision in which they are all
concerned (for example, the conclusion of a debt contract between bank and
manager, preserving the solvency of the enterprise and the wealth of the
shareholders) the economic argument is both normative (what is the optimal
price of debt for a firm?) and strategic (which price results from the
confrontation of these different interests?). In fact the shareholder is likely
to be in favour of debt which gives him more profit and decreases his level
of risk-sharing. However, if the manager invests in high-risk projects, the
risk of bankruptcy increases. Lars Stole (1990) has considered the
contracting problem facing multiple principals, each of whom wants to
contract with the same agent. This has been termed the common agency.
Common agency can either be delegated or intrinsic. Under delegated
common agency, the choice of contractual relationship is delegated to the
agent, who can choose whether to contract with both, one or none of the
principals. Alternatively, when common agency is intrinsic, the agent’s
choice is more limited: the agent can choose only between contracting with
both or with neither.
Consider a contracting environment with two principals and one agent. A
first possibility is that each principal wants to share the profit of a given
outcome which the agent produces. Each agent is represented by a type
which describes the characteristic of the projects undertaken by the manager.
A direct revelation mechanism incites the agent to announces his type to
each principal separately. If the agent contracts only with one principal we
return to the literature defining contracts between either shareholders and
manager or between bank and manager. If the agent contracts with no one,
as he has not enough money to realise the project, he cannot carry it out.
The only interesting case is where all parties contract:
• Stage 1 The principals announce the contract which they wish to sign
with the manager (agent). A contract specifies a project, a reward
scheme for each configuration of the market (principal-agent) and the
credit to be granted. We assume that offers between principals and agent
are private information.
13
• Stage 2 The manager of the firm (agent) announces which of the offered
contracts he is willing to accept. The announcement is public
information (each agent must inform the principals whether he wishes to
deal with that project exclusively or in conjunction with projects of
other firms). Public observability of the announcement is not essential,
but simplifies the description of an equilibrium.
• Stage 3 Each principal selects the manager which is willing to
undertake the project and to accept the proposed contract. At the same
time, principals announce the amount of capital they grant for its
realisation. These choices are observable.
Assumptions
First, all players are risk-neutral. While risk-neutrality on the part of the
principal is not essential, a risk-neutral agent is indispensable in what
follows. For example, as mentioned in stage 2, a manager can undertake
several projects with differing levels of profitability, which enables him
to spread risk. Second, principals are allowed to condition the terms of
contracts upon the market. This feature of our model illustrates an
important aspect of agency delegation. Since a principal’s decision to
abandon a common agent (manager) constitutes an observable change in
market environment, competitors may renegotiate their contracts in
response. 3
Stage 1 of our model provides a simple way to represent this process of
reactive renegotiation: the principal simply announces an amount of funds
and an incentive scheme for every possible market configuration. Note,
however, that, as long as a principal retains the same agent, other principals
cannot respond directly to changes in its contracts since contract offers are
private information. Consider now the model described above, but modify
the description of the agency delegation process as follows. In stage 1, each
principal makes one contract offer (credit incentive scheme) to the same
agent (the manager of the firm). In stage 2 this single agent either accepts
both offers or neither. If he rejects, principals earn zero profit and the game
ends. In this simple model, common agency is intrinsic in the sense that the
agent has no alternative to the pre-specified delegation. Bernheim and
Whinston claim that:
non co-operative behaviour induces an efficient (potentially second

best) action choice if and only if collusion among principals would
implement the first best level cost.4
In an earlier paper5 they discussed the possibility of collusion between

shareholders and bank to reach an equilibrium. This leads us to consider the
supervisory role of banks and shareholders in the agency theory.
14
Agency as control is tied to compensation in at least three different

ways. The first is a causal relation: compensation helps to define the
agency relation, much more than the reverse. Compensation’s incentive
role (giving the agent an incentive to act in the principal’s interest) is
less important than its informative aspect: agents must know what their
responsibility is (for example, as regards the solvency of their
organisation) and how it changes each time the financial structure is
modified.
Performance-related compensation for managers may be primarily a
vehicle for defining their duty. But if the bonus scheme or other form of
incentive compensation is mechanical, conveying little about the kind of
performance to be elicited, the scheme does not signal agency relations. This
first connection between compensation and agency has to do with bringing
agency into the formal organisation.
The second connection brings agency into public markets, answering the
question: in which direction should the cash flow component of
compensation flow? For example, in a financial corporation this constitutes a
way to achieve control on managers of firms, by giving them an incentive to
maintain a reputation towards their lenders.
As a third illustration of the relation between agency and
compensation, consider the question of appointment of the different
actors of the firm. The manager is the agent not only of the shareholders
but also of the bank which finances its projects. Any study of corporate
structure must therefore consider two very different interfaces: that
between the shareholders and the manager, and that between the manager
and the external investors.
1.3 THE SUPERVISORY ROLE OF THE BANK IN

AGENCY LITERATURE
The bilateral principal-agent question considers that one individual, called

the agent (e.g. the manager of a firm) performs duties on behalf of another
party, the principal (external investors: e.g. the bank and the shareholders).
This theory explains
1 why debt was relied upon as a source of capital before debt financing
offered any tax advantage relative to equity;
2 why shares are issued when financing a project.6
In this section, I study the theoretical framework concerning the role of

banks in monitoring the enterprises with which they are involved. More
generally, how do financial institutions affect the allocation of funds for
investment?
15
Some authors who have considered this question have observed that
financial systems differ significantly between countries and across periods,
and the question arises how these differences between financial systems
have affected the functioning of the different economies. In this section I am
going to concentrate on two case studies, Japan and Germany, in order to
contrast the results obtained for these two countries with those relating to
the Crédit Mobilier in Chapter 2. As a first illustration of this concept, let us
consider the paper of Mayer (1988). The author suggests there that there are
systematic differences in performance between financial systems in which
banks play a prominent role, and those financial systems in which banks are
less prominent. He explains these differences in performance in terms of
differences in the mechanisms that reduce or eliminate moral hazard
between entrepreneurs and financiers. Specifically, he suggests that the more
bank-oriented systems of Germany and Japan involve greater commitment
on the part of the firm and the bank to a long-term relationship, which
allows them to enjoy the benefits of long-term contracting.
In this section, the following issue is considered: in what sense is ability
to support long-term relationships the clue to assessing financial
institutions? The focus of the discussion will be on financial institutions
(more specifically, the role of banks in the provision of finance to firms),
rather than financial instruments.
(a) Internal finance, external finance and bank finance:

theoretical presumption in historical assessment
Theoretical framework
The manager-debtholder literature has two sub-literatures. First there is the
issue of the manager-bondholder conflict. There are numerous theoretical
discussions of the role of bond convenants in controlling the manager’s and
(or) shareholders’ incentive to increase the risk of the firms. This literature
has treated debt mostly as a homogeneous instrument creating incentives, in
direct conflict with those associated with equity financing. Second, there is
the borrower-lender literature which has developed from the simple
examination of the incentive of the borrower to default on a bank loan, to
the role of the bank as the delegated monitor of a portfolio of borrowers,
i.e., a monitoring role for the bank’s depositors.7 These latter models have
served to strengthen the argument that there is a unique information and
monitoring role of bank loans and that the phenomenon of credit rationing is
due to the nature of information asymmetry in the market.
The argument discussed here is richer than the traditional literature
briefly presented above. Essentially based on the model of Devinney and
Milde (1990), it incorporates the role of ‘inside debt’. Inside debt, as
16
defined by Fama (1985), is debt which explicitly incorporates monitoring by

giving the debtholders access to otherwise proprietary information.
The modelling framework used here is that of an asymmetric information
game. Managers are viewed as choosing amongst alternative investment
projects which require the same initial outlay but differ in terms of risk.
Greater bank monitoring affects the project chosen by the manager, while
the bank’s choice of loan contract terms affects management effort. Outside
stockholders choose the managerial incentive contract and, hence, the level
of risk-sharing between managers and stockholders.
The major assumptions of the model presented by Devinney and Milde
are the following:
1 The expected payoff of the project, µ is a function of management effort

e with the properties µe > 0 and µee < 0.
2 Both debt L and equity S are used to finance the project.
3 All debt is inside debt (bank loan) as defined by Fama (1985).
4 The equity position consists of both ‘inside’ equity and ‘outside’ equity.
Inside equity is owned by the management and is a fraction α of the
total equity outstanding. The remaining fraction (1–α) is the equity
position held by outside owners.
5 Outside shareholders determine the fraction of equity given to the
managers, α, and the managers’ fixed income component H. Managers
are prohibited from trading the equity of their corporation in the stock
market.
6 Banks determine the level of monitoring activities m, the loan size L,
and the (gross) loan rate R=(1+r)
7 Managers choose the project, i, and hence its riskiness, σ and the level
of effort, e.
8 Banks can monitor the managers’ choice of project (risk) but
monitoring is costly. Total monitoring costs are M=νm, where ν is the
per unit monitoring cost.
9 Neither outside stockholders nor banks can monitor the level of
managerial effort.
10 Stockholder and bank expectations are rational in the sense that they
can calculate the management’s response to changes in their decision
variables.
11 All decision-makers are risk-neutral.
Under these assumptions, how can the bank-stockholders management

equilibrium be determined? The manager’s responsibility is to accept or
reject the projects and choose their effort. We consider that only one project
can be undertaken. The chosen project is characterised by specific µ (its
expected payoff) and its riskiness s is observable to the bank but not outside
17
stockholders. By assumptions 6 and 8, managers cannot do anything but

follow the bank’s instructions regarding s.
The effort level choice e=e* is optimal if the marginal expected payoff to
the manager is equal to his/her marginal opportunity cost. More specifically
the relationship between e* and the decision variables determined by outside
shareholders and banks are the following:
• an increase in the fraction of internal equity increases the manager’s

effort;
• monitoring leads to higher effort levels by management;
• increasing the loan size and/or the loan rate will increase the effort until
it threatens the solvency of the firm.
Banks can monitor the manager’s choice of s and/or participate in the

decision-making process. By increasing monitoring and/or participation,
banks can enforce their interest more and more, thus eliminating the
tendency of management to undertake higher-risk projects. Then the bank’s
decision-making process leads to three components characterising a loan
contract: interest R*, loan volume L* and risk s*. The risk, s*, however, is
implicitly determined by bank monitoring, m*. This contract specifies the
second best solution and is consistent with the manager’s effort response
function.
The distinction between internal and external debt is similar to that
between internal and external equity and can explain the monitoring role of
the participants in the financing of investment projects.
Theoretical presumption in historical assessment

The economic theory stresses the notion of effort to explain the monitoring
role of banks and stockholders towards the managers of firms. The main
argument of the literature is based on the notion of ‘capital scarcity’. Capital
is regarded as scarce because:
• in the late nineteenth century, a hundred years after the beginning of the
industrial revolution, the adoption of up-to-date technology required
large capital outlays; and
• the failure of some countries to participate in the earlier stages of
industrialisation meant there was a lack of accumulated funds that could
be used to finance these outlays.
Gerschenkron (1962) studied this notion of ‘capital scarcity’ and applied it

in the context of German industrialisation from 1850 to 1914. He observes
the prominent role of German banks in providing liquidity for heavy
industry with large capital requirements.
18
Gerschenkron’s argument is as follows: over the period, the firms’

retained earning were insufficient to finance desired investment. At this
point, Gerschenkron raises two questions:
1 Why should an insufficiency of internal finance be seen as characteristic

of a situation of ‘economic backwardness’?
2 Why should the large-scale provision of external finance to industry
require the intervention of banks or government?
The first question really concerns the role of internal finance in an

industrially advanced country. It is clear that, in the second half of the
nineteenth century, certain advanced technologies required large-scale
investments. It is also clear that firms in the less developed countries had
not had time to accumulate the funds required for such investment.
However, it is less clear why firms in the industrially more advanced
country should have retained earnings to such an extent that further
industrialisation could be financed without the same level of recourse to
outside finance.
Gerschenkron’s interpretation is that the involvement of banks or
government is needed to provide outside finance to industry on a large scale.
External finance from the anonymous markets of our theoretical models is
not seen as an alternative.
Bank finance here is not necessarily just loan finance. During certain
periods, especially prior to 1873, German companies obtained substantial
amounts of equity finance from banks. The shares would be held by banks,
or by clients of banks acting on the banks’ advice, and in many respects
banks were as much involved in equity finance as in loan finance. While
share markets were organised, they were certainly not anonymous and free
for all.
The question is to what extent bank involvement in equity and loan
finance was actually necessary. According to Gerschenkron’s argument,
bank finance could do something that anonymous organised financial
markets could not. So we need to discuss what exactly the comparative
advantage of bank finance is. This is the subject of the following section.
(b) The advantage of bank finance versus market finance
The discussion of the advantage of bank finance over market finance implies
that financial systems are institutions which reduce or eliminate problems of
moral hazard or asymmetric information between firms and financiers.
Financiers typically have less information about firms than entrepreneurs or
managers. Moreover, investors are subject to various types of moral hazard:
moral hazard concerning managerial effort, moral hazard concerning the
19
riskiness of the firms strategies and moral hazard concerning reported ex

post return realisation. These problems of moral hazard and asymmetric
information cause difficulties for the provision of finance to industry.8
Intermediaries are taken to reduce these difficulties by engaging in
monitoring and control activities.
Diamond (1984) presents an explicit example in which intermediation
successfully reduces the agency costs of outside finance under moral hazard.
In his analysis, the feasibility of financial intermediation rests on two key
propositions:
1 Monitoring and control of a firm involve natural scale economies: a

single intermediary can monitor and control the firm as effectively as
thousand of shareholders—but much more cheaply.
2 If the intermediary has a well-diversified portfolio of firms that he
finances, then relations between himself and his own financiers—the
final investors—are not much affected by moral hazard and asymmetric
information because his own return is approximately riskless, so for him,
fixed interest debt finance is feasible and does not involve any moral
hazard.
On the basis of these two propositions, Diamond shows that under certain
circumstances incentive-efficient allocations cannot be made without
intermediation.
Diamond’s notion of financial intermediation as delegated monitoring (or
delegated control) is closely related to Gerchenkron’s account of bank
involvement in firms at the early stage of industrial development, although
this is not explicit in Gerschenkron’s work. As emphasised by Mayer, bank
initiative and bank participation in entrepreneurial planning may be a way to
obtain enough information and control to reduce the moral and
informational hazards of finance to a tolerable level. We may therefore look
at the imperfect information approach to financial intermediation as the
theoretical basis for Gerchenkron’s view that banks were needed to provide
outside finance during the early stages of industrialisation in Germany, when
capital was ‘scarce and diffuse’ and ‘the distrust of industrial
activities…considerable’. In Diamond’s terminology, Gerschenkron’s
assumption must have been that the sum of the monitoring costs of the
banks and direct agency costs of bank deposits was less than the agency
costs of direct finance, perhaps even that the agency costs of direct finance
were so high that this was not a genuine alternative at all.
According to this approach, supervision seems to be profitable for both
the bank and the shareholders, which stand to benefit from higher profits by
reducing their agency costs. It will be interesting to analyse which type of
contract is induced to exert this control. In other words, are long-term
commitments necessary to control firms?
20
(c) Monitoring and the definition of contracts between banks

and firms
‘Monitoring’ ought to be understood in a broad sense as any form of

information-gathering about a firm, its investment prospects and its
behaviour. This information is useful because it serves to sort out ‘bad’
projects and/or to punish ‘bad’ behaviour. The literature contains the
following justifications for monitoring:
• Monitoring of the firm’s profit returns makes it possible to conclude

contracts in which the financiers’ claims are related to the firm’s profits.9
• Monitoring of the firm’s characteristics, i.e. a creditworthiness test,
makes it possible to avoid bad loans.10
• Monitoring of the firm’s behaviour during the loan application stage
makes it possible to avoid loans to firms that are following too risky an
investment strategy.11
In all these examples, the role of monitoring is straightforward. The main

distinction is between monitoring that takes place before a contract is
agreed, and monitoring of the execution of a contract. The former serves to
reduce the proportion of bad loans, the latter serves to improve performance
under the agreed contract.
Monitoring that takes place before a contract is signed may give rise
to an interesting problem of competition for contracts. In the following
chapter, when describing the role of the Crédit Mobilier in French
industrialisation, this competition will be referred to. In particular, the
fight between the Péreires, founders of the bank, and the Rothschilds, as
regards the railway companies, illustrate this point perfectly. More
specifically, the competition between these two families discouraged
monitoring: over 1861–7 the Rothschilds’ first objective was to prevent
the Péreires from expanding their financial power. The Rothschilds’
reputation in the market won them preferential access to many
companies, including the railways. This competition prevented the
Péreires from exerting their control efficiently.
Monitoring that takes places after a contract has been concluded
stresses the length of the commitment between the firms and the bank
that provides the funds. Mayer suggests that financial intermediation may
lengthen the investment horizon of firms. He considers a model with two
investment periods. There are two types of firms, ‘good’ and ‘bad’.
Firms can choose between two investment strategies, short-term and
long-term. Both strategies require identical amounts of funds in both
investment periods. However, the long-term strategy has a relatively
higher expected payoff in the second investment period. Model
parameters are specified in such a way that ‘bad’ firms should not
21
receive funds in either period. However, ex ante, there is no way for

financiers to distinguish between ‘bad’ firms and ‘good’ firms. After the
first investment period, a partial distinction may be possible because
there are two sources of information: costless observation of first-period
returns, and costly monitoring. In the absence of monitoring, banks
interpret low first-period return realisations as an indication of poor
quality, and discontinue financing. Anticipation of such behaviour
induces firms to opt for the short-term investment strategy even though
the long-term strategy may eventually be more profitable.
This theoretical consideration can be illustrated by two examples:
France and Germany. The policy of German banks as regards investment
was essentially short-term and concerned enterprises where the expected
profit was quite high (the mining industry). However, this does not mean
that the German bank never undertook risky projects in the late
nineteenth century. According to Tilly, 12 these banks used their
information advantages and their freedom from liquidity worries to
engage in investment strategies that involved high risks as well as high
returns. The commitment of a German bank in a risky project induced a
more careful control of the performance of the company. This idea will
be developed further in the two following chapters when interpreting the
empirical work proposed for the Crédit Mobilier. This last bank was
quite different as regards the control it could exert on the companies
with which it was involved. Despite the fact that it undertook risky
projects, its supervisory role was very limited, and sometimes totally
ineffective (as in the case of the Compagnie Immobilière).
The real question is how monitoring fits into the overall relationship
between the intermediary and the firm. Monitoring as a form of information-
gathering about the firm is useful only because the information that is
collected has consequences for behaviour and resource allocation within the
relationship. The notion of financial intermediation as delegated monitoring
must encompass such uses of information as well as the act of collecting the
information.
This consideration brings us back to the idea of incentive contracts
presented in the preceding section. Let us recall the basic concept. From the
theory of optimal incentive contracts, we know that even noisy monitoring
will improve performance under an optimal incentive contract. 13 By
combining this result with the procedure of Diamond that was outlined
above, we obtain a new version of financial intermediation as delegated
monitoring, one that emphasises the incentive effects rather than the sorting
effects of monitoring.
An interesting point will be to apply it to the financial systems of the
late nineteenth century. Monitoring as an element in an incentive
contract relationship is not at all that far from the accounts of
Gerschenkron and Mayer.
22
Consider Gerschenkron’s observation that banks played a key role in

the formation of cartels in German industry. Any cartel is subject to the
moral hazard problem that the individual firm has an incentive to
under-cut and to serve more than its allotted share of the market. Banks
that are involved with several firms in the same industry have an
incentive to restrain such behaviour, in order to increase the aggregate
gross return they can earn from the industry. In the following chapter
we compare these considerations on the German case to that of the
Crédit Mobilier.
So far we have studied how the interdependence between information
provision to firms and information collection about firms affects the
functioning of financial intermediation as delegated monitoring. It will be
interesting to debate how this delegation takes place effectively.
(d) Financial intermediation as a mechanism

of commitment
In this section the advantages of a close relationship between bank and

firms will be described. Mayer (1988) considers that financial
intermediation in the form of a close relationship between the bank and
the firm constitutes a mechanism of commitment. Mayer suggests that
Japanese banks are more willing to engage in corporate rescues than
financiers elsewhere because the bank-firm relation in Japan involves a
mutual long-term commitment. Some rather striking quantitative evidence
on this point is provided by Hoshi et al.14 For a sample of Japanese firms
they found that the cost of financial distress was significantly less for
firms that had close relations to a ‘main bank’ than for firms that did not.
To measure this relationship, they evaluated the degree of correlation
between investment and cash flow for firms affiliated regularly with a
bank, and for independent firms. Formally, they prove the following: that
firms that have close banking ties appear to invest more and perform better
than firms that do not have such ties.
This proposition is crucial for the work presented here. In the next
chapter, which is devoted to the empirical study of the Crédit Mobilier, the
validity of this hypothesis will be investigated: as this bank went bankrupt in
1867, if the above affirmation were valid, the share prices of the companies
with which the Crédit Mobilier was involved should have fallen relative to
other firms; surprisingly, a formal test presented in the next chapter does not
support this hypothesis.
The question we should ask now is, to what extent should a system of
corporate finance based on intermediation through a ‘main bank’ be
regarded as internally stable? Mayer seems to believe that the superior
performance of such a system is a guarantee of its persistence over time. In
23
contrast, Gerschenkron (1962) regarded it as being transitory, with firms

depending on outside finance through banks only until they have enough
inside finance available. The empirical research reported below will test
these assertions in order to clarify the financial role played by the Crédit
Mobilier in French industrialisation.
Similar patterns can also be observed in other countries. Thus Hoshi et
al. report that in Japan, too, the larger, more profitable companies availed
themselves of the newly developing organised markets to become more
independent of their banks.
In the long term, there is a certain tendency for firms to emancipate
themselves from such a relationship, using markets, competition amongst
banks and, above all, reliance on internal rather than external finance.
The emancipation of firms from close banking relationships is certainly
not costless. Recall the observation of Hoshi et al. that the costs of financial
distress in Japan are significantly larger for firms without close banking
relations than for firms with close banking relations. For firms that are not
in financial distress, Hoshi et al. report a similar observation: hence the
emancipation of a firm from a ‘main bank’ relationship is accompanied by a
significant increase in the sensitivity of current investment to fluctuations in
current earnings and liquidity. Bank loans are less used and/or available to
smooth over fluctuations in earnings.
From the perspective of Jensen and Meckling (1976) one might argue that
internal finance has priority because the agency costs of internal finance are lower
than the agency costs of external bank or market finance. This is also the
explanation given by Myers and Majluf (1984) in the context of a model of
asymmetric information and signalling. Internal finance is taken to have no
agency costs (signalling costs, simple inefficiencies) because it represents the use
of funds available to the firm itself. External finance has no agency costs because
information asymmetries and externalities preclude the attainment of a first-best
allocation in the arrangement between the firm and its outside financiers.
From this perspective, the Hoshi et al. observation of investment sensitivity
to current earnings should be seen as evidence of the agency costs of outside
finance: investment projects that might be expected to be profitable under
internal finance are deemed to be unprofitable under outside finance when the
agency costs of outside finance are added to the mere opportunity costs of
funds. The inefficiency in the allocation of funds across the firms that results
when investment opportunities are less than perfectly correlated with earnings
is nothing but an element in the overall agency cost of outside finance.
1.4 GENERAL CONCLUSION
This first analysis describes the relationships that exist between the manager
of a firm, its shareholders and the bank which finances the investment
24
project, in the context of asymmetric information. As we have shown in the

first section, each partner wants to define the debt contract according to his
own interest. The manager is willing to find the amount of funds he thinks
acceptable to undertake his investment project. For the shareholders and the
bank, this commitment is a way to realise profit. However, the latter actors
bear different costs. The shareholders give the manager a financial
compensation for his work, while the banker supports the interest cost for
granting the credit.15
As there is asymmetric information between the economic agents, we
must determine which actor(s) is (are) able to control the manager and if
necessary punish him when he does not truthfully reveal the necessary
information about the productive project.
There are two possible candidates: the shareholders of the firm, or the
bank which finances the projects. Let us consider first the shareholders. In
order to exert this control effectively, the shareholders have to agree
unanimously, or at least on a majority basis. On the other side, the bank
disposes of a number of tools to control its affiliated companies more
efficiently.
Our analysis shows the power of banks in screening the investment
projects of firms with which they are involved. As pointed out in the survey
of the literature, some authors carried out empirical studies of the efficiency
of certain financial systems. More precisely, Gerschenkron for the nineteenth
century and Hoshi et al. for the recent period have analysed the power of
banks. They have stressed the importance of banks in controlling the
investment policy of their affiliated companies. My aim is now to apply their
methodology to the French case. As I have chosen the same period (1852–
1914), I intend to compare my results to those obtained for Germany and
Japan. In particular, I shall investigate whether the power of banks in the
French financial system was as strong as in Germany.
25
2
THE CRÉDIT MOBILIER

AND THE FRENCH STOCK
EXCHANGE 1853–1914
An empirical perspective
The fate of the Crédit Mobilier is characteristic not only of the

rivalry between financial groups, and of the dangers of
speculation on the stock exchange: it highlights an ever present
problem of industrial involvement.1
Bouvier (1988:109)
2.1 INTRODUCTION
The aim of this work is to gain, through empirical research, an insight into
the role of banks in the financing of enterprises during the period 1850–
1913 in France. Some authors (e.g. de Long (1989) and Ramirez (1992))
have already applied this analysis in the United States.
The credit which a bank gives to a company enables it to exert an
influence over the company. 2 J.Bradford de Long has illustrated the
role of investment bankers in his article ‘Did J.P.Morgan’s men add
value? A historical perspective on financial capitalism’. In the period of
which he writes, financiers possessed strong voices in corporate
management. The author illustrates his point of view by focusing on
the J.P.Morgan Company. J.P.Morgan and his partners saw themselves,
as did other participants in the pre-WWI securities industry, as
fulfilling a crucial monitoring and signalling intermediary role between
firms and investors, in a world where information about the underlying
values of firms and the quality of management was scarce. De Long
stresses that investment bankers played an important role, in that they
took great care to make sure that the firms they watched had the right
managers. This analysis fits our hypothesis that the bank can control
the management of a firm and motivate it to reveal truthfully the
information about the project it proposes to undertake.
Another line of research has been made in the German context. The
roles of four banks—the Deutsche, the Diskonte Gesellschaft, the
Dresdner, and the Darmstädter—have been studied, producing the same
26
THE CRÉDIT MOBILIER AND THE STOCK EXCHANGE
conclusion. Bradford de Long and Marco Becht (1992) have used price
and dividends series to test the volatility of the German stock market.
In particular, they examined whether the role played by the German
Großbanken in the pre-WWI stock market might have been the cause of
the low comparative volatility of the German stock indices before 1914.
These ‘Great Banks’ were founded in deliberate imitation of the French
Crédit Mobilier. There has been much discussion about the origin of
the early German credit banks, and especially to what extent the Crédit
Mobilier served as a model. The first significant joint-stock bank was
created in 1848: the Schaaffhausen’schen Bankverein. During the crisis
of the 1860s’ recession it faced a liquidity problem. Because of the
bank’s importance and the strained atmosphere of the times, normal
liquidation procedures did not occur. The Darmstädter was a direct
result of the French Crédit Mobilier. It counted among its founders
Abraham Oppenheim, who had been co-founder of the Crédit Mobilier.
Much of the initial capital of the bank was supplied by the Péreire
brothers.
These banks had access to considerable capital with which to
provide finance to corporate enterprises. They contributed to the
intensity of the boom of the 1850s. Despite the crash in 1857 and the
depression of the 1860s, most of the banks survived. The early German
banks remained permanently tied in with the companies which they
promoted, through representation on boards of directors and direct
participation in ownership. How this cautious approach can be
reconciled, at least in theory, with the role the banks wished to play in
the development of industry, is clear from the Schaaffhausen’schen
Bankverein for 1852
The management has proceeded from the principle that it is the

function of a great banking institution not so much to call great
new branches of industry into existence through large scale
participation on its own account as to induce the capitalists of the
country by the authority of its recommendations, based on
thorough investigations, to apply their idle capital to undertakings
which, properly planned, corresponding to real needs, and
equipped with expert management, offer prospects of reasonable
profits. 3
The case of the Crédit Mobilier will be analysed to see if, as in the
American and German contexts, bankers possessed strong voices in
corporate management. From 1853 to 1913 this bank took part in the
financing of firms. In fact, two ‘Crédit Mobiliers’ existed. The first, founded
by the Péreires, went bankrupt in 1867. After its bankruptcy it was
reorganised, but it went bankrupt again in 1900. It reappeared in 1903 as the
27
Crédit Mobilier Français. In 1932 it merged with the Banque de l’Union

Parisienne.
The most interesting banking model was, of course, the first Crédit
Mobilier, because of its innovative characteristics and because of the
power it had reached in financial markets and among banks. That was
certainly one reason why the Péreires faced a certain amount of
criticism.
The first criticisms arose when they envisaged the possibility of founding
a bank capable of issuing money. The Banque de France refused
categorically to allow this. Their commitments in various risky sectors (for
instance, the property business with the Compagnie Immobilière)
contributed to their failure.
However, they tried new banking techniques for financing industries.
The advances they granted to the companies with which they were
involved could be considered as innovative, matching the ideas the
Péreires had on finance. To the Crédit Mobilier, the role of a merchant
bank was to provide the necessary funds in order to contribute as fully as
possible to the industrial development. This idea was new and sometimes
unrealistic.
The aim of this chapter is to verify whether, as de Long (1989) suggests,
a close relationship with a bank is always beneficial for firms, increasing the
market value of their shares. In that respect, the place of the Crédit Mobilier
in the banking sector will be analysed. De Long maintains that an
association with J.P.Morgan facilitated the assessment of company
managers’ performance and the careful evaluation of projects. Having J.P.
Morgan act as an intermediary partially resolved the principal-agent problem
between shareholders and managers.
A similar question will be asked about the Crédit Mobilier for the period
1853–1914; namely, ‘did the Crédit Mobilier increase the ratio of market
value to book value?’ Attention will then turn to the private participation of
the Péreire brothers in the financing of firms. In particular, we will
investigate how the Crédit Mobilier may have increased the value of shares
by providing liquidity. This question will be of interest at the point when the
Crédit Mobilier went bankrupt.
Bradford de Long maintains that investment banker representation on
boards allowed bankers to monitor the performance of firms’ managers,
quickly replace managers whose performance was unsatisfactory, and
signal to ultimate investors that a company was well managed and
fundamentally sound. If this argument is to be followed, the companies
where the Crédit Mobilier was personally involved should have
experienced difficulties in 1867, the year of its bankruptcy. In the
second section of the chapter I will study the fluctuations of the share
prices of the companies involved with the Crédit Mobilier between
1866 and 1868. According to de Long’s argument, these prices should
28
have fallen. A formal test will be presented to analyse the validity of

this proposition for a sample including affiliated companies and non-
affiliated companies. The last section of this chapter will present the
results obtained for a pooled panel including both affiliated and non-
affiliated firms.
A comparison between France and Germany will be made.
Differences and the analogies will be identified. As the German
Großbanken were founded after the bankruptcy of the Crédit Mobilier,
their success can be interpreted as a learning effect from the mistakes
made by the Péreires. The Germans were more careful to screen
projects according to their risk level, to avoid situations similar to that
of the Compagnie Immobilière.
2.2 FINANCIAL CAPITALISM AND DATA ON

THE CRÉDIT MOBILIER IN FRANCE, 1853–1914
(a) Some historical background about the Crédit Mobilier,

the banking system and French industry
The interpenetration of finance and industry was intensified by the counter-

movement of industrialists-turned-bankers. André Koechlin, a scion of the
notable Alsacian family, head of Dollfus Mieg et Cie, and founder of the
machine shops that produced the first French locomotives, became one of
the most daring and successful promoters and investment bankers of the
second Empire, both in France and abroad. Emile and Isaac Péreire went
from railways to the Crédit Mobilier, and one of their principal associates,
Baron Seillière, from a Vosges textile family, also helped to finance several
large metallurgical establishments, including Le Creusot and Wendel in
France and Krupp in Germany.
In part, for a relatively small section of its total operations, the Crédit Mobilier
functioned as an ordinary banking enterprise—that is, it accepted deposits,
discounted commercial paper and made loans and advances on security. The
Crédit Mobilier’s principal function consisted of lending to merchants and
industrialists on ‘two-signature papers’ for a period of up to two years. (The Bank
of France discounted only three-signature paper with maturities not longer than
ninety days.) In an effort to obtain the sources for such loans, the bank proposed
that it should be allowed to issue what would have been, in effect, interest-bearing
bank notes (bearer bonds, interest bonds). The provision for interest (at 3.65 per
cent per year—one centime per 100 francs per day) would have ensured
circulation of bonds and if they were made receivable in payment of taxes, as the
Péreires proposed, they would have acquired virtual legal tender status. (In some
respects the proposal resembled a vast paper money scheme, in as much as the
loans granted by the bank as well as their repayment might be in bonds.) To
29
enhance confidence in the institution and in its attractiveness to potential investors

and bondholders, the Péreires also proposed that the French government should
guarantee it against losses up to twenty five million francs.
In the Péreires’ own words, ‘The new establishment should be considered as
a lending and borrowing office [“bureau de prêt et d’emprunts”] where industry
will borrow from all capitalists on the most favourable terms through the
intermediary of the richest bankers acting as guarantors, because the holders of
the bonds, who are the true stockholders of the association, will find an easier
and safer investment for their capital without the hazards of individual loans.’4
In the spring of 1853, soon after the Crédit Mobilier began to function, the
Péreires and Fould requested an agreement for a ‘Caisse Générale des Sociétés de
Crédit Mutuel’, intended as the nucleus of a network of mutual credit societies in
various localities and different lines of industry. The Caisse Générale was
intended to perform the same or similar functions for small firms and handicrafts
that the Crédit Mobilier did for large-scale industry; thus the four-fold programme
implicit in their abortive Saint Simonian scheme of 1830 would be complete:
commercial credit (Comptoir d’Escompte), industrial credit (Crédit Mobilier),
mortgage credit (Crédit Foncier), and mutual credit for small entrepreneurs.
However, the Conseil d’Etat, reasserting its prerogatives in such matters,
quietly shelved the project. Shortly afterwards, the Corps Legislatif passed a
law to permit departments and communes to convert and consolidate their
floating debts. The Crédit Mobilier requested permission to exchange its
own bonds for the debts of these local government bodies, to have them
stamped by an official of the government, made payable through the
receivers-general of taxes, and accepted for purchase by the Caisse des
Depots et de Consignations (the official government loan office). Such
measures would, in effect, have advanced the Mobilier’s bonds to the status
of fiduciary money. This was never realised in practice.
The power of the Crédit Mobilier had an important effect on financial
institutions at that time (P.Dupont Ferrier (1925)). Jean Bouvier, François
Furet and Marcel Gillet (1965)5 drew up a balance sheet of the Crédit
Mobilier from its foundation until its fall, with the average profit calculated
over five years (see Appendix A), which supports this argument.
(b) The financing of enterprises during the period
During the period the number of financial intermediaries providing short-

term credit6 varied. Probably there were around 400 institutions on average.
It is difficult to gauge the extent of their activity, as only a limited number
published their balance sheets. Only the report by the General Assembly
together with the balance sheets of those companies with which the banks
had relations provide evidence of the amount of credit devoted to the
financing of production.
30
Statistics on the issue of shares and the interest rate on the financial
market give an idea of the fluctuations on an aggregate level. 7 These
statistics serve as a reference point for the analysis of the relations between
the Crédit Mobilier and the companies with which it was in contact.
Nevertheless, operating only with its own capital and deposits, the Crédit
Mobilier performed prodigious feats of industrial promotion and financial
manipulation. In its first major operation, it subscribed to 40 million francs
in the Crédit Foncier’s bonds, with which it maintained close ties in the
early years.
In its first full year of operation the Crédit Mobilier financed, by
making advances and underwriting bond issues, the Midi railway as
well as the newly chartered Grand Central and the French Eastern,
Mulhousen and Strasburg-Basle railways. The Crédit Mobilier also
secured the merger of three short railways in the Loire coal basin with
the Grand Central, reorganised the Loire coal industry and promoted
the Darmstädter Bank. It played a leading role in the mergers which
took place in 1855 in the French Western (railway), and promoted or
financed the railways of the Pyrenees, Dauphine, the Ardennes and
several shorter lines. It created, with generous government subsidies,
the Compagnie Générale Transatlantique (French Line), the Compagnie
Générale des Omnibus (Paris), and the Société de l’Hôtel et des
Immeubles de la Rue de Rivoli (subsequently the Compagnie
Immobilière).
The Crédit Mobilier founded by the Péreires played a great role in French
industrialisation. In every company financed by this bank, a member of the
Crédit Mobilier’s board of directors was represented in the shareholders’
assembly. The annual report of April 1860 is very explicit on this subject:
Our company has always considered as a principle of its high

commercial morality never to open a subscription, nor recommend a
firm, without first having a large proportion of interest in it and its
administrators having become associated with it.8
A representative from the Crédit Mobilier was present on the board of

directors of each of the companies asking for funds. Table 2.1 gives
details of the ownership of the shares of the Immeubles de la Rue de
Rivoli in 1854 (subsequently the Compagnie Immobilière) (Aycard,
1867). The par value of capital of the company was 24 million francs,
divided in 240,000 shares. The three latter members of this list were
subscribers of the Crédit Mobilier. Table 2.1 shows clearly that the
affiliated companies belonged to the Crédit Mobilier, unlike the German
case or J.P.Morgan company.
Let us refer to the question J.Bradford de Long and Marco Becht (1992)
have asked in their paper: ‘Did association with the Crédit Mobilier reduce
31
Table 2.1 Ownership of the shares of the Immeubles de la Rue de Rivoli
Sources: Archives d’Entreprises Privées (General shareholders’

assembly and newspapers) of the Société Immobilière or Immeubles
de la Rue de Rivoli, 1857–67
Table 2.2 Advances to companies, 1856–66
the variance of the ratio price to dividends?’ This question is an important

one for the Crédit Mobilier. To be more precise in answering it, let us
analyse how the Crédit Mobilier used to loan companies money.
The Crédit Mobilier granted credit to the companies with which it was
involved. Most of the time it gave advances to enable them to undertake
their investment. Table 2.2 shows the aggregate advances made to companies
between 1856 and 1866.9
Examples of advances being made without any special guarantee include
the following:
• In 1856 and 1857, 6 million fr, to the Compagnie Parisienne du Gaz.

• In 1857, 13 million fr. to the Compagnie Transatlantique.
• In 1855, 1856 and 1857, 29 million fr. to the Compagnie des Chemins
de Fer du Midi.
• In 1865, out of 53,971,339 francs advanced, 52,080,709 francs went to
the Compagnie Immobilière.
32
If we refer precisely to the balance sheet figures (see Appendix A), we

observe that the advances are part of the investment account. These
advances represent at least one third and sometimes half of the total amount
of the account. This demonstrates the risky behaviour of the bank, since it
did not ask for collateral.
Collateral can be useful in dealing with the problem of adverse selection,
as discussed in the preceding chapter. As mentioned earlier, the collateral
in the debt contract has a dual role: in addition to providing financial cover
in the event of bankruptcy of the firm, it acts as a signal transmitted between
the parties.10 Any manager is ready to offer collateral if he thinks that the
probability of bankruptcy is low. The Crédit Mobilier did not benefit from
this information because it never asked for guarantees from its affiliated
companies.
Several hypotheses can be formulated to explain this practice. The
first is that the Crédit Mobilier wanted to attract clients by offering
them optimal conditions to finance their projects. Another argument
refers to the situation of French industry over the period. Large
numbers of companies were new. Many of the firms involved with the
bank were in heavy industry (the railways, for example). They needed
huge amounts of credit. The Crédit Mobilier had made the choice to
invest in these sectors, thinking it could be profitable in the long term.
In any case, if the French credit policy is compared to the German
policy over the same period, we can see that German banks used
similar practice as regards the financing of the productive projects of
their companies.
As the Crédit Mobilier often owned part of the companies’ equity with
which it was involved, the bank was not only an intermediary but an active
partner. The question is: did the Crédit Mobilier reliably screen and
supervise companies? In other words, as the Crédit Mobilier took a great
part in the enterprises where they granted credit, was it possible for the bank
to correctly evaluate the risk of its investment?
As the Crédit Mobilier gave credit without asking for any collateral, it
increased the risk of bankruptcy (see the next section, concerning the fall of
the Crédit Mobilier). It could be thought that, for affiliated companies, an
association with the Crédit Mobilier did not always reduce the variance of
share price towards dividends.11
The scale of the advances to companies emphasises the power of the Crédit
Mobilier. This power was important among banks until 1865 (see Appendix
A). After this date, the ‘Banque de Savoie affair’ provoked hostility among
financial circles and at the Banque de France. Prominent investors such as the
Rothschilds wanted to force the Crédit Mobilier out of the banking sector.
However, it was the Crédit Mobilier’s position on the financial market and the
Péreires’ investments in risky sectors (e.g. the Compagnie Immobilière) that
eventually brought about the failure of the establishment. Before 1864 the
33
Table 2.3 Fluctuations of share prices for affiliated companies between 1855 and
1856
Sources: Archives d’Entreprises Privées (General shareholders’ assembly and

newspapers)
Crédit Mobilier had been a victim of strong speculative movements: for

investors, the profit realised by the institution made investors believe in its
stability. To illustrate this effect, Table 2.3 sets out the values of the shares of
some companies involved with the Crédit Mobilier in 1856.
These last remarks lead us to ask the same question as de Long and Becht:
Did association with the Crédit Mobilier stabilise the share price compared to
book value? The speculative movements indicated here cast doubt on this. In
the next sections of this chapter, a formal test will be presented to analyse the
impact of the Crédit Mobilier bankruptcy on the stability of assets of the
companies with which it was involved. The aim is to see if the Crédit
Mobilier ever managed to supervise firms properly and thus to stabilise share
prices. Table 2.3 suggests that the bank never really achieved this.
(c) The fall of the Crédit Mobilier
The fall of the Crédit Mobilier was essentially caused by the attempt to
salvage the Société Immobilière. This company’s eagerness to participate
fully in the extension of Paris led its managers to undertake huge
investments with poor potential.
From 1864–5 the Crédit Mobilier had to confront serious difficulties
arising from the Banque de France’s decision to increase interest rates. The
Crédit Mobilier decided to double its capital, to support advances for the
Compagnie Immobilère. But in June 1866 its situation became worst and the
fall of the share prices seemed to be the consequences of the depreciation of
its portofolio. The Péreires redoubled their efforts to obtain cash. It was said
at the time that a loan would be granted to the Crédit Mobilier by the city of
Paris, but the deal was never realised. By the end of 1866, the Péreires were
contemplating the possibility of a merger between the Mobilier, the
Mobiliario Espanol and the Immobilière, with a loan from the Caisse des
Consignations (an agency of the government), but they did not succeed. In
addition to these difficulties, the Péreires faced a stockholders’ dispute
34
concerning the increase of the bank’s capital. The Crédit Mobilier’s true
situation could no longer be concealed. In March 1867, for the first time in
its existence, the prices of its shares fell below par.
The Péreires were forced to resign from both the Crédit Mobilier and the
Immobilière. Between 1867 and 1870, little information was available. The
Crédit Mobilier went bankrupt, Emperor Napoleon III, who had supported
the Péreires, fell and France had to face a war with Germany. In 1871,
Baron Haussmann made a deal with the bank for the repayment of its debt.
This involved formally dissolving the old company, and immediately
forming a new one. The new bank, called the Société de Crédit Mobilier
Français, was to take over the assets of the old Crédit Mobilier.12 The impact
of this new bank on the financing of French industry was so small that it
disappeared from the economic literature. The bankruptcy and loss of
resources meant it played only a secondary role.
(d) The reconstitution of the Crédit Mobilier and its

final bankruptcy
After the fall of the Péreires in 1871, Baron Haussmann, an old member of
the board of directors, became President of the new Crédit Mobilier.
However, as mentioned previously, the bank had lost its earlier pre-eminent
role in the financing of French industry. The first aim of the new bank was
to solve the difficulties of the Compagnie Immobilière. The capital of the
company was divided in two parts, between Paris and Marseille. Part of this
credit came from the Crédit Foncier, which had contributed to the financing
of the Compagnie Immobilière: 55 million francs came from the credit of
the Paris division and 32 million from Marseille. Floating assets were
liquidated. The Crédit Mobilier reduced its holdings of the Compagnie
Immobilière’s debt to 51 million francs by selling part of its bonds and
issued new assets for an amount of 76 million francs. Half of these assets
were bonds with fixed revenue (5 per cent) and half were bonds with
variable revenue to the amount of 5 per cent of the capital. The liquidity
problems of the Compagnie Immobilière were solved by the partial
liquidation of its debt. The Crédit Mobilier became a firm with a par value
of only 40 millions francs, undertaking more projects abroad than in France.
In 1875, Emile d’Erlanger, another member of the old board of directors
(of the first Crédit Mobilier) became President. During this period the bank
had relations with the Bank of Brussels and the Spanish government. No
balance sheet was published in the reports to the shareholders’ assembly,
despite the fact that they had been specifically requested by the
shareholders.
In August 1876, Baron Emile d’Erlanger submitted his resignation and
Charles Wallut, the bank’s Vice-President, became President. The Crédit
35
Mobilier continued to start new ventures with foreign companies, but little
information was published relating to its activities within France. The bank
participated in the Italian Ice House Society, as well as contributing to an
Algerian government loan.
In June 1881, Eugène Péreire, founder of the first Crédit Mobilier, was
nominated Administrator of the bank. He proposed to add a bank service to
the Compagnie Transatlantique. The Banque Transatlantique was founded
with a capital of 50–60 million francs, of which one quarter was to be given
immediately to the Company Transatlantique. Twenty thousand Comptoirs
Maritimes shares were absorbed by the new bank. The Crédit Mobilier took
10,000 shares in this new society. In August 1881, M.Péreire was appointed
to represent the Crédit Mobilier on the syndicate which was formed to issue
the Banque Transatlantique shares. The Crédit Mobilier, with the Romanian
government, undertook the constitution of a Romanian Crédit Mobilier.
Economic relations were developed with the United States and Spain for
railway construction.
On 4 November 1882, M.Péreire raised the issue of the participation of
the Crédit Mobilier in the profits of the Compagnie Transatlantique. He
thought that the idea of a combination of capitalisation or repurchase of the
right to the profit sharing should be examined. On 3 February 1883, Péreire
submitted his resignation. This was followed, on 29 June that year, by the
Crédit Mobilier asserting its rights to the benefits of the Compagnie
Transatlantique. It was believed that the Crédit Mobilier should have 25 per
cent of these profits and the shareholders 8 per cent. Subsequent court
proceedings resulted in the Compagnie Immobilière being ordered to pay
367,000 francs to the Crédit Mobilier for 1881 and 1882.
In 1886, the Crédit Mobilier was a company with 30 million francs of
Table 2.4 Crédit Mobilier’s liquidation account in 1901
Sources: Crédit Mobilier: Procès Verbaux des Assemblées Générales; Procès Verbaux
du Conseil d’Administration
36
capital. Its financial situation got worse and worse, until it became
completely bankrupt in 1902. One consequence was that, in 1898, funds
were cut to two main borrowers, causing the Puerto Rican railway to stop
and the Rhone railway line to be liquidated. In 1901 the Crédit Mobilier’s
liquidation account was as shown in Table 2.4.
(e) The new Crédit Mobilier
After the first Crédit Mobilier’s bankruptcy and subsequent

reorganisation, the Crédit Mobilier Français had to confront the
scepticism of its clients. The strategy that was adopted was ‘to make the
Crédit Mobilier Français become an intermediary between the medium-
sized business and the public’ (Rapport sur l’exercice, 1903–4). Its
capital of 7,525,000 francs was constituted of subscriptions in cash, all
the capital of the Office des Rentiers, together with all the Crédit
Mobilier capital, namely:
• rights to the Compagnie Generale Transatlantique;

• the Société Publique Financière capital;
• the Société Civile Franco-Mexicaine capital;
• the Puerto Rico business.13
The Crédit Mobilier was anxious to prove that its portfolio was
performing and it took part in business which was small and easy to realise.
Between 1903 and 1904 the Crédit Mobilier began to develop: its capital
reached 10 million francs and, the year after, 25 million francs. During this
period the bank had business relations with the large financial house of
Thalmann and Co. and had just concluded a South American business deal.
The bank was then capable of entering, either alone or with the
participation of other companies, more suitable financial operations (e.g. the
Serbian national debt). The Crédit Mobilier maintained relationships with
provincial banks to guarantee full realisation of share transactions. The
report of 1911 states:
We have sent more correspondents abroad and the portfolio fluctuation

has increased. In this aim we have taken interest in the banks’
institutions…If this policy has your approval, we propose to improve
our effort in this direction and to create a greater base of operations.14
From 2 million francs in 1909, the discount portfolio reached 20 million

francs in 1910 and 39 million francs the year after. Financial advances to
industry reached 12 million francs in 1910. At this point, the Crédit
Mobilier succeeded in concluding a contract with the Banque de Paris and
37
entered the bank consortium. By 1910, after an increase in capital to 60

million francs, the Crédit Mobilier had become a great financial
establishment. In 1921 its capital was 100 million francs.
The balance sheet, now officially audited, shows the first important
variations: few elements of comparison exist between the preceding system
and the new one. To illustrate this idea, let us analyse the balance sheet of
the new Crédit Mobilier (see Appendix A).
The credit side is constituted by monetary funds available in cash and
funds resulting from liquidation on the stock exchange. Their use
corresponds to financial operations such as bills in portfolio, reports and
advances. The deficit account of investment banks consists, in the major
part, of lending to companies in a direct relationship with the Crédit
Mobilier, in order to guarantee their development. Credit for the assets-
portfolio and financial participation during the period 1900–13 amounted to
a maximum of 43.4% and a minimum of 10.3%, with the average value
24.6%.
For the liabilities, with the exception of the reserves which have already
been discussed, the most interesting aspect is the deficit current accounts.
The Crédit Mobilier made a distinction between the debit accounts and the
deposit account, but this distinction is not very useful from the eligibility
point of view.
The merchant banks had expanded considerably and as a consequence
they experienced an increase in their balance sheets. The Crédit Mobilier’s
balance sheet totalled 248 million francs in 1913 against 40 million francs
in 1903. The average profit over the period 1904 to 1908 was about 5.3
per cent, and between 1909 to 1913 approximately 12.3 per cent. The
distribution of the dividends was stable and from 1910–13 represented 7
per cent.
This survey of the Crédit Mobilier demonstrates that the three
different establishments of that name were not at all comparable. If we
consider the first Crédit Mobilier and the second one, few points of
comparison exist, except the fact the formal organisation of the firm was
identical and a large number of the shareholders in the Péreires’ bank
maintained their participation in the financing of firm’s projects, despite
the bankruptcy of 1867.
Concerning the relationship between the second Crédit Mobilier and the
third the differences are much stronger as the beginning of the twentieth
century heralded a new role for the merchant banks which was implied by
the new role of the Banque de France. This latter exerted an increasing
monitoring role over banks. However, in regard to the bankruptcy of the
Crédit Mobilier, we can ask whether the central bank in France really
fulfilled its interventionist role as the lender of last resort.
Henry Thornton (1802) and Walter Bagehot (1873) developed the key
elements of the classical doctrine of the lender of last resort in England.
38
They argued that the central bank should intervene for illiquid but solvent
banks. Otherwise supporting failed companies would damage healthy
competitors and weaken the discipline of a market economy.15
For Thornton, the Central bank’s responsibility in times of panic was
to act as lender of last resort, providing liquidity to the market and
discounting freely the paper of all solvent banks, but denying aid to
insolvent banks no matter how large or important (Humphrey 1975,
1989). In other words, the lender of last resort must prevent illiquid but
solvent banks from failing. 16
At the other extreme, Charles Goodhart (1985, 1987) advocates
temporary central bank assistance to insolvent banks. He argues that the
distinction between illiquidity and insolvency is a myth, since banks
requiring lender of last resort support because of ‘illiquidity will in most
cases already be under suspicion about…solvency.’ Furthermore ‘because
of the difficulty of valuing [the distressed bank’s] assets, a Central Bank
will usually have to take a decision on last resort support to meet an
immediate liquidity problem when it knows that there is a doubt about
solvency, but does not know just how bad the latter position actually is’
(1985:35).
Solow (1982) is also sympathetic to assisting insolvent banks. He argues
that any bank failure, especially a large one, reduces confidence in the
whole system. To prevent a loss of confidence caused by a major bank
failure from spreading to the rest of the banking system, the central bank
should provide assistance to insolvent banks. However, such a policy
creates a moral hazard, as banks respond with greater risk-taking and the
public loses its incentive to monitor them. In the event, the Banque de
France did not rescue the Crédit Mobilier. Its reasoning followed
Thornton’s argument, according to which:
It is by no means intended to imply that it would become the Bank

of England to relieve every distress which the rashness of country
banks may bring upon them: the bank, by doing this, might
encourage their improvidence. There seems to be a medium at which
a public bank should aim in granting aid to inferior establishments,
and which it must often find it difficult to be observed. The relief
should neither be so prompt and liberal as to exempt those who
misconduct their business from all the natural consequences of their
fault, nor as scanty and slow as deeply to involve the general
interests. These interests, nevertheless, are sure to be pleaded by
every distressed person whose affairs are large, however indifferent or
even ruinous may be their state.
(Thornton 1802:50)
39
Hence the non-intervention of the Banque de France was profitable for the
whole system, because an insolvent bank disappeared.
From an historical and human point of view, a simple study of the main
shareholders of the Banque de France sheds further light on this non-
intervention. During the period, the Rothschild family, who disliked the
Péreires for both personal and professional reasons (and may have been
jealous of the power the Crédit Mobilier had obtained in the financial
market) had contributed to the decision of the ‘Banque de France’ not to
salvage the Crédit Mobilier. The only intervention that had occurred was to
prevent a too sudden and huge fall of share prices on the stock market,
which would have provoked a financial crisis for the whole economy. As
regards this argument, it will be interesting to analyse the impact of the
Crédit Mobilier’s bankruptcy on the share prices for both affiliated and non-
affiliated companies.
2.3 DATA AVAILABLE OVER THE PERIOD OF

THE TWO CRÉDIT MOBILIERS:
COMPANIES INVOLVED AND PRELIMINARY
ANALYSIS OF THE DATA
This section presents and discusses all the data sets that will be used in the
empirical tests explained below. Three points are developed: the first aims to
elicit the general context of the French economy and to analyse in more
detail the situation of the financial markets; this will constitute the basis of
our excess volatility test.
The second and third parts describe the data available for companies that
were affiliated with the Crédit Mobilier and independent ones; this will
support our statistical research relative to the examination of the impact of the
Crédit Mobilier’s bankruptcy for both sub-samples and for the whole sample.
(a) General context of the French economy during

the period
As Bradford de Long and Marco Becht (1992) did for the United States
and Germany, I intend to analyse the excess volatility of the French
stock market from 1852 until 1914. In this respect some general ideas
about the French economy from 1853 to 1914 are given. 17 More
specifically, general figures on share prices, market prices, interest rate
and dividend are provided.
The aggregate data are relatively simple. The money stock was stable.
The huge saving rate compared to a reasonable demand maintained
interest rates at a moderate level over the period: between 3 per cent and
40
4.5 per cent (Loutchitch 1930). The rate of discounting assets with fixed
revenue varied between a minimum of 3.22 per cent in 1897 and 1898
and a maximum of 3.73 per cent in 1913. The official discount rate
decreased to 2 per cent in 1896 and 1897. Its value was around 3 per
cent from 1900 to 1910. This low return on credit gave the banks an
incentive to invest abroad. Some explanations are required to explain
how the indices are calculated. To construct the industrial, transport and
trade series, fifty four separate assets have been taken into consideration
for the years 1895 to 1913 and twenty four assets for the years 1870 to
1895. Amongst these were some issued by companies involved with the
Crédit Mobilier. From 1895 to 1913 these were:
• Electrometallurgie de Dives;
• Chemin de Fer PLM, Nord, Orleans, Midi;
• Compagnie Générale Transatlantique;
• Société Parisienne d’Eclairage.
From 1870 to 1895 they were:
• Chemin de Fer PLM, Nord;

• Compagnie Générale Transatlantique;
• Compagnie d’Eclairage par le Gaz.
From this series, the impact of the Crédit Mobilier’s bankruptcy on the
level of production can be analysed. If the Crédit Mobilier had been a
‘Great Bank’, the impact of its bankruptcy on affiliated companies could
have induced a relative effect for the sectors to which these firms belonged.
As, for a large number of sectors (e.g. the railways), the Crédit Mobilier was
an active partner in the financing of companies, it will be interesting to
present some figures to support the argument.
Table 2.5 compares market product prices by sector on an aggregate
level. The sectors I have chosen are those in which the Crédit Mobilier
Table 2.5 Market product prices by sector
Source: Lenoir (1919:87–8) basis average of the market prices for 1901–10
41
was interested, such as railways and mining. A simple examination of

these figures shows that the situation was very stable between 1867 and
1870. No big change can be observed when the Crédit Mobilier went
bankrupt. The bank’s failure does not seem to have had a great impact on
the companies.
As regards the financial point of view, some general comments on these
series can be made. From 1870 to 1872 wars (the Franco-Prussian, the
Commune of Paris and a civil war) led to an increase in the interest rate.
The rate of 3 per cent increased from 4.17 per cent in 1869 to 5.51 per cent
in 1871 and 5.47 per cent in 1872. The yield on bonds went from 4.73 per
cent to 5.72 per cent between 1869 and 1872. The dividend yield on stock
passed from 5.07 per cent in 1870 to 7.28 per cent in 1872. Prices followed
this trend between 1869 and 1872. The year 1873 was one of general crisis,
particularly in Germany, Great Britain, and in the United States. France
experienced economic difficulties until approximately 1879. Between 1882
and 1884, the interest rate increased, but a new crisis in 1890 led to a
remarkable trend towards a reduced discount rate proposed by banks (see
Appendix B: 2.18 per cent in 1878, 3 per cent in 1883, 2 per cent in 1896).
During this period, the dividend yield decreased regularly, remaining one
point below the interest rate on long-term bonds. A slow and almost regular
increase of the interest rate characterised the period from 1897 to 1913.
Shares did not experience any general marked movement. Goods price
fluctuations were small,18 where the author studies the evolution of the
general price level between 1809 and 1925 (see Appendix B).
(b) List of the companies affiliated to the Crédit Mobilier
Here an analysis similar to that of Bradford de Long and Marco Becht

(1992) will be employed. Companies affiliated and not affiliated to the
Crédit Mobilier must be taken into consideration in order to answer the two
following questions:
• Did the Crédit Mobilier stabilise the ratio of market value to dividends
of affiliated firms?
• Did the Crédit Mobilier increase the market value compared to book
value of affiliated firms?
In addition, I want to study the power of the Crédit Mobilier among the
banking sector, its policy towards investment and the importance of an
affiliation with the bank as regards credit policy.
P.Dupont Ferrier in his book Le Marché Financier sous le Second Empire
provides a list of the societies founded by the Péreires between 1853 and
1867. These enterprises were:
42
• The railway enterprises: Compagnie de l’Est, PLM, Compagnie du

Midi, Compagnie d’Orleans and Compagnie de l’Ouest.
• Insurance companies: l’Union.
• La Compagnie Transatlantique.
• La Compagnie Immobilière de Paris. This company was very important
for the Crédit Mobilier, Appendix C gives an indication of its financial
situation in 1867, the year of the bankruptcy of the bank, and describes
the market price of shares and dividends of the Compagnie Immobilière
between 1855 and 1866.
• La Compagnie des Omnibus.
• La Société Parisienne d’Éclairage et de Chauffage par le Gaz.
The bonds issued by the firm were guaranteed by its capital. The
received capital during the year could not be greater than the capital
realised (Cameron 1991:142). Plenge (1903) gives the data on the
Crédit Mobilier’s shares throughout its existence, their value in 1867
(the year when it went bankrupt) and the distribution of dividends (see
Appendix A).
The new Crédit Mobilier (1903–14) was initially involved with the
Société métallurgique de Montbard-Aulnaye, in which the bank was
largely represented, and with the Société d’Electro-métallurgie de Dives.
The Crédit Mobilier also founded the Société Française des Machines-
Outils. In varying degrees, the Crédit Mobilier participated in the
Compagnie Parisienne de Distribution d’Électricité and in the
reorganisation of the Compagnie Générale des Omnibus de Paris. It
contributed to the development of the Société de Raffineries Say and to
the expansion of the Société de l’Éclairage Électrique. The bank also
concluded contracts abroad. In 1909–10 it participated in an Argentinian
government loan whose interest rate was 4.5 per cent and in the Santa Fe
business affair (interest rate 6 per cent).
In summary, the firms in which the Péreires were interested were the
railway companies, l’Union, the Compagnie Transatlantique and the
Compagnie Immobilière de Paris. Two societies were involved with both the
Péreires and with the Crédit Mobilier Français: the Société des Omnibus de
Paris and the Compagnie d’Eclairage et de Chauffage par le Gaz. Firms
involved with the new Crédit Mobilier included the Société des Machines
Outils, the Société Montbard-Aulnoye, the Electro-metallurgie de Dives and
the Raffineries Say.
In contrast to the first Crédit Mobilier founded by the Péreires, the
second one was not always represented on the boards of firms with which
it was involved. In the Raffineries Say, however, the Crédit Mobilier was a
financial partner of the company, while with the Société des Machines
Outils, the Société Montbard-Aulnoye and the Electro-Metallurgie de
Dives the Crédit Mobilier was strongly represented on the boards of
43
directors, and sometimes also the principal shareholder, as with the Société
des Machines Outils.
However, the new Crédit Mobilier’s functioning was not very different
from that of the Péreires, particularly when it was well represented on
boards of directors. The close relations that the Crédit Mobilier
maintained with the societies with which it was involved enabled it to be
not only a financing establishment but also an active partner in the
realisation of projects.
As advances were granted to the societies where the Crédit Mobilier was
officially represented, the risk of bankruptcy was increased. In this case we
can ask the following questions:
• Did association with the Crédit Mobilier increase the value of shares
compared to book value?
• Did association with the Crédit Mobilier reduce the variance of the
price dividend ratios?
• Did association with the Crédit Mobilier always facilitate loans and
advances to the companies? If so, in what measure?
More specifically, it will be interesting to study share fluctuations when the

Crédit Mobilier went bankrupt. As in de Long (1989) I would like to extend
this analysis to companies that were not affiliated to the Crédit Mobilier.
(c) List of companies not affiliated to the Crédit Mobilier
This data set supports the hypothesis that companies not affiliated to a bank
tend to be liquidity constrained (and face a larger gap between external and
internal costs of funds) than those which are. Bertrand Gille (1968) provides
us with a list of companies which were affiliated with the Crédit Lyonnais
for the financing of their projects. These companies were
• Alais;
• Schneider;
• Marine;
• L’home;
• Terrenoire;
• Denain;
• Fourchambault;
• Maubeuge;
• Givors;
• Firminy;
• Franche Comté;
• Chatillon.
44
Lenoir (1919) enables us to complete this list by considering the case of

the Houillères du Nord and the textile sector, which was not affiliated to the
Crédit Mobilier on the period. Appendix D presents the data available for
these companies. The comparison of the results of the test for the two
groups of companies will allow us to evaluate the effectiveness of an
association with the Crédit Mobilier as opposed to an association with the
Crédit Lyonnais (some of the firms, e.g. Marine, Fourchambourg, Firminy
and Chatillon, were involved with the bank). Unfortunately, no data were
available on firms which were affiliated with neither bank.
From the historical data collected here, a test can be constructed to
show if:
• association with the Crédit Mobilier increased the value of shares

compared to book value;
• association with the Crédit Mobilier always facilitated loans and
advances to the companies. As advances (without collateral) were
granted to the societies where the Crédit Mobilier was officially
represented, the risk of bankruptcy increased.
2.4 EXCESS VOLATILITY TESTS
As in de Long and Becht’s paper, an analysis can be undertaken to test for

excess volatility. More specifically, we can evaluate volatility of the prices
relative to trend. Let us recall the idea of such of a test. Shiller’s19 first key
insight was that the level of the stock market should be a forecast of the ex
post perfect foresight fundamental. An investor who buys and holds, and
pays less than the ex post fundamental, receives a supernormal return.
Arbitrage, therefore, should push prices to an efficient forecast of the perfect
foresight fundamental. If prices are too volatile to trend, investors can make
better forecasts of ex post fundamentals by taking as their forecast some
linear combination of the market price and the trend, and betting that returns
will be low whenever the market price is above the trend. How can such a
test be expressed more formally?
(a) Presentation of the test
Following de Long and Becht’s argument, let us call p* the fundamental

price, which can be expressed as: where p is the stock index divided by the
(2.1)
45
Figure 2.1 Prices, dividends, and perfect-foresight fundamental sample 1870–1914

Legend Lnrp* represents the natural log of the perfect foresight fundamental, Lnrp is
the log of the real stock price and Lnrd for the real dividend
price level, d represents the dividends divided by the price level and r is
chosen to represent either the an arbitrary real required rate of return or a
risk premium paid on assets plus the risk-free rate minus the inflation rate.
Here r is similar to the value taken by Becht and de Long (8 per cent) to
facilitate the interpretation of the result. Other values for r between 5 per
cent and 10 per cent would not have changed the shape of the graph in
Figure 2.1.
In the French case I have calculated the fundamental price using
Dessirier’s series for the share prices and the dividend and Lenoir’s series
for the general level of prices. All the series constructed are deflated by
the GNP deflator from Lenoir. The same methodology used by Bradford
de Long and Becht (1992) will be followed. As the general price series,
the share price series and the dividends series do not refer to the basic
year in the construction of the indices, normalisation of all these prices is
necessary to avoid any bias problems in the test. Moreover, the dividend
series presented in Dessirier’s paper was not used as such. We evaluated
the ratio as dividend.100/share price. This series corresponds to a constant
term times the ratio of dividend index to price index. After having
normalised the dividend, a calculation of the volatility ratio, that is the
price dividend ratio, is made in order to construct the test. Figure 2.1
provides individual surveys of the behaviour of prices, dividends, and
perfect foresight fundamentals which are obtained by computing equation
46
(2.1). They plot for the period the logs levels of prices and the log ex post
perfect-foresight fundamental.
Shiller’s test says that:
pt = E(p*t | It) (2.2)
var pt < var p*t. (2.3)
This formulation was the subject of much controversy because the

definition of p* is said to be impossible. If so, investors will take
t
advantage of this to buy when the price is low, and sell when it is high.
The basic idea of solving this problem is to consider a naive forecast
noted p equal to a constant times dividends. The problem can be
N
reformulated and expressed by:
var (ln pt – ln pN) < var (ln p*t – ln pN) (2.4)
where pN=αd and α is a constant.

Dessirier’s series and Levy-Leboyer’s series enable us to construct such a
test. The question is then: are the conclusions obtained for Germany
comparable with those we will obtain for France? To answer this question,
let us consider the result represented by Figure 2.1 and that of de Long and
Becht.20
Before discussing the similarities and the differences between these
two graphs some explanations about the underlined data are worth
making. The share prices series is taken from Dessirier (1928b) and
presents some odd elements. The first remark is that the figures reported
in the article are flat and the financial terms used to explain the data are
confusing. For example, the values in column 2 of the share prices series
retrace the dividend for assets with variable revenue. The French term
used by Dessirier is ‘revenue des actifs à revenu variable’. For the
description given in text it is clear that the two terms (‘revenu des actifs
à revenu variable’ and ‘dividend’) reflect the same parameters. What is
more ambiguous is on what basis the author built this series. Therefore
the interpretation of the whole test is relative to the underlying data.
The second argument is related to the figures themselves. When reading
Dessirier’s data the symbol ‘is found in place of numbers. The problem is
whether this a representation of unavailable data, or did the author choose
not to report it as it was identical to the preceding value? Dessirier wrote
four papers in three years and proceeded the same way each time. When a
number was non available, he clearly stated this and when, for two
subsequent years, numbers were identical, he always used the same symbol’.
This clarification being made, arguments can be advanced to explain the
existence or the absence of volatility in France over the period.
47
As in the German case the fundamentals follow a trend. From 1890 to

1913 the German and the French results are quite comparable: dividends,
prices and perfect foresight fundamentals follow a similar increase. Hence a
test on excess volatility in this period would probably give similar results to
those obtained by de Long and Becht. The second remark about the graph is
that for France, between 1870 and 1890, the price/dividend ratio was
increasing.
How can we explain these differences? In 1882 France suffered a
banking crash due to the failure of the Banque de l’Union Générale.
All industrial activity decreased between 1881–4 and 1885–6, with
some variations according to the sectors and the firms. There was then
a slight increase, which was interrupted during 1887–9 due to the
bankruptcies of the Compagnie du Canal de Panama and the Comptoir
de l’Escompte.
What is more interesting to analyse is the behaviour of the Banque de
l’Union Parisienne just before the crash, that is at the end of the summer of
1881 and afterwards. Prior to this date, a revision of all their policy towards
industry took place. They suspended part of their participation in the
creation of enterprises. They decreased the supply of credit in the short and
medium term. The crisis, which the bank anticipated, resulted in a decrease
in their profit, which was at the origin of their defensive attitude. This
situation became more pronounced during the second quarter of 1881. The
price of money increased and most of the important investors and lenders
followed a deflationary policy.
According to Bertrand Gille in La Banque et le Crédit en France de 1815
à 1848, the disparity that was essentially responsible for the crises was an
‘excess of investment’21 inducing a relative gap in the ‘circulating capital’22
and hence the capital of firms. ‘The investment chains provoke the excess
and the break in the equilibrium’.23 The crisis happens when ‘the investment
ceases’ and this issue is unavoidable:
The investments cease because the accumulation of capital is destroyed

and there is no longer any liquidity; they cease because the scarcity of
circulating money may increase its rate of interest; because certain
projects are revealed to be bad or speculative.24
Certain bankers of the first half of the nineteenth century (and more later
also such as Laffite, or, more importantly, the Péreires) thought and said that
the progress in the banking system, in feeding payment and credit, enabled
growth and avoided ‘commercial crises’. The facts contradicted their
optimism. The promotion of banking and the modern form of credit did not
prevent cycles. They furnished, on the contrary, a source of new turmoil: the
rise of credit expansion and the confirmation of depression due to the
increase of banking operations.
48
The Großbanken and the Crédit Mobilier were different. Several

arguments can be adduced to explain these differences:
• The first one concerns the effective power of each bank. In particular
the Crédit Mobilier and the Großbanken did not have similar powers
with regard to screening companies’ projects. The German Great Banks
were more cautious and evaluated the level of risk for any potential
investment project.
• The second deals with the ability of the Crédit Mobilier to supervise the
managers of the firms which whom it was involved. The differences
described in the text between 1870–95 on the one side and 1895–1914
on the other side give a hint of the different policy which had been
exerted by the German banks and the French Crédit Mobilier.
• The third refers to the economic situation in France during the
period and, more specifically, the difficulties faced by the banking
system because of its innovative character as evidenced by Saint
Simon’s theory.
(b) Volatility ratios
Presentation of the results

Table 2.6 presents summary statistics on the volatility of the French
stock market. For comparative purposes, it also reports similar statistics
for the United States and Germany. If actual prices were rational
estimates of fundamentals, they should exhibit less volatility relative to
some ‘naive’ forecast than do the ex post perfect. As the values are
Table 2.6 Volatility tests for the French, German and American stock markets
between 1870 and 1914
49
extremely flat before 1895 (some values are identical for subsequent
years) and reported yearly after 1895, the ratios are evaluated for the
whole sample and for 1895–1914.
The first line of Table 2.6 shows the volatility of log prices (p) around
the log of the perfect foresight fundamental (p*–p) calculated using an 8
per cent per year real discount. The second line shows the volatility of
the log price-dividend ratio (p–d). The third line shows the volatility of
the log ex post perfect foresight fundamental to dividend ratio (p*–d). If
prices are more volatile relative to ‘naive’ forecasts than perfect
foresight fundamentals are, relative to the naive forecast, investors could
and should have constructed a better forecast: a weighted average of the
market price and the naive forecast would have generated smaller
forecast errors. Thus the values in the second line of the table, the
volatility of the price-dividend ratio, should under the efficient market
hypothesis be smaller than the values in the third line, the volatility of ex
post perfect foresight fundamental. For France it is true after 1895 as for
Germany. Between 1870 and 1914 the situation is more similar to the
United States. Line four reports this volatility ratio. Line five of the table
calculates another volatility ratio. The volatility of the perfect foresight
fundamental about the actual price should be less than the volatility of
the perfect foresight fundamental about the naive forecast. For both
periods the price and the dividend are almost exactly equally good
forecasts of the perfect foresight fundamentals.
A final implication of the efficient markets hypothesis is that the two
ratios of lines four and five—the sum reported in line six—should add
up to one. If not (as is the case of France for the whole sample, and the
United States) the log difference between price and the perfect foresight
fundamental (p–p*) is correlated with the log price-dividend ratio (p–d).
Profits could have been earned by trading on this correlation of the
price-dividend ratio and value relative to price. Surprisingly, between
1895 and 1914 the situation of France is similar to that of Germany. The
analysis of the results obtained for France leads us to the following
conclusions:
1 Prices are volatile relative to dividend over the period 1870–1914. Prices
are much more volatile than perfect foresight fundamentals relative to
the naive forecast. Thus, tests based on market volatility ratios shows
traces of excess volatility in the French stock market before the First
World War. However, between 1895 and 1914 volatility tests present
strong similarities between France and Germany
2 The situation described by the French stock market is much more similar
to that of the United States for the whole sample than that of Germany.
The symbol ‘ used in Dessirier’s data should lead to a bias which gives
opposite results for the whole sample and the period 1895–1914.
50
Interpretation of the results

How can we explain the similarity and the differences between France,
Germany and the United States? We propose to give an interpretation of
these results by explicating the monitoring role played by the banks and
their relations with the stock market.
The monitoring role of banks in France,

Germany and the United States
According to Becht and de Long, ‘the key position of the Großbanken in the
German economy is linked to the absence of excess volatility. The
Großbanken dominated pre-World War I financial markets in Germany.’25
The system involved well-developed analytical abilities, strong views
regarding the proper levels of stock prices and an incentive to moderate
fluctuations away from fundamentals. Vincent P.Carosso and Richard Sylla,
when examining the role of J.P.Morgan said:
Bankers generally undertook non financial responsibilities out of

necessity—to protect their own interest and those of the investors to
whom they sold the companies’ securities.26
Their role in the affairs of the corporations they sponsored sometimes

required them to assume entrepreneurial responsibilities they generally
would have preferred to avoid. The fact that they were capable and stood
ready to undertake these tasks when the occasion demanded it gave the
private bankers immense influence far beyond the sums of money they
themselves commanded.
In the United States these private bankers were largely responsible for
facilitating the rise of giant enterprises, the consolidation of industries and
the organisation of the earliest multinational companies, but did not monitor
as thoroughly as the Großbanken.
This strong monitoring role in the case of Germany explains the
absence of volatility on the German stock market: the fluctuations of share
price were reduced over the period. However, the stock market price seems
to imply that this role was less efficient in the French case from 1870–95.
The failure of the Crédit Mobilier and the bankruptcy of the Banque de
l’Union Parisienne in 1880 can partly explain the high volatility ratios
over the period.
The banks and the Stock Market

According to Becht and de Long, the Großbanken, as well as trying to
create value for the investors to whom they sold newly issued securities,
51
sought also to support the prices of the securities they issued. For Riesser,
this support was not an attempt to produce an artificial market or to swindle
the public. These authors wonder if the Großbanken had sufficient resources
to keep asset prices close to their own internal estimates of fundamentals. As
the Großbanken were by far the largest actors in the German economy, an
affirmative answer can be given to this question. In the United States the
situation was similar to that in Germany: the banks occupied a powerful
position. To illustrate this argument, let us just quote Karl Erich Born
(1983:93):
What probably explains their success is that they had collaborated in

the formation of big trusts at the end of 19th and the beginning of the
20th century thereby not only gaining opportunities for control and
influence in industry and transport but also making considerable profit.
Vincent P.Carosso and Richard Sylla (1991) added:
Their chief contribution in the industrial development which was the

ultimate basis of their strength and influence, was their access to the
world’s principal source of capital.
In France such a situation did not exist except for the ‘Haute Banque’
and more specifically the Rothschilds, whose aims was to fight the Péreires.
The Rothschilds found the Péreires’ investment policy too risky to consider
them as real bankers. On the other hand, they saw them as competitors they
had to force out of the market. Evidence of this lack of supervision will be
provided in the following section by analysing the impact of the Crédit
Mobilier’s bankruptcy on the financial market.
(c) Conclusion
It appears that the Crédit Mobilier benefited from a powerful position up

until 1862–3. Even before the bankruptcy, however, it faced liquidity
difficulties. Its place on the financial market was delicate, as most of the
bankers of the period were hostile to the Péreires. The Péreires wanted to be
innovative as regards credit policy: in particular they presented themselves
as promotors of firms, whose principal aim was to provide liquidity to
industry. But the control they exerted on firms was limited. When comparing
the Crédit Mobilier with the German Großbanken, Engberg (1981:35) states
that:
The Crédit Mobilier never became a ‘universal’ bank as did the

German banks soon after their organisation. The Crédit Mobilier
52
represented a bold and imaginative idea; it was designed to effect

revolutionary changes on a grand scale of entire railway, industrial and
credit systems. The German credit banks were called into existence to
meet new and enormous demands for ‘finance’ in general. The Crédit
Mobilier was created to implement change; the early German credit
banks were created in response to changes of the society that were
taking place.
2.5 THE EFFECT OF THE BANKRUPTCY OF

THE CRÉDIT MOBILIER ON THE COMPANIES WITH
WHICH IT WAS INVOLVED
E.Nouette-Delorme (1865) pointed out the evolution of the prices of

shares and bonds of these societies just before the bankruptcy of the
bank. I have chosen to add to this analysis the evolution of the share
prices in the year immediately following the fall of the Crédit Mobilier
to answer the following question: what happened to the companies with
which the Crédit Mobilier was involved between 1867 and 1869?
Because of the bankruptcy of the Crédit Mobilier, the share prices of
these companies should have fallen relatively to the share prices of other
firms. Table 2.7, however, which reproduces the average asset prices for
affiliated firms, shows the opposite.
Table 2.7 suggests that the railways suffered least from the crisis of
the Crédit Mobilier. The Rothschilds were a partner of one of the
railway companies and came to the assistance of the railway companies
during this difficult period. The importance of railways in the French
economy, and the low level of competition that existed between the
companies in the period, probably provided stability for these
companies. Moreover, monthly share prices for affiliated companies
between 1866 and 1868 show that the companies did not underperform
in the market.
In what follows, the impact of the bankruptcy will be analysed in three
ways:
1 the effect of price level and dividend on share prices for the Crédit
Mobilier’s firms;
2 the correlation between the failure of the bank and the fluctuation of
share prices for the affiliated companies;
3 an evaluation of the association with the Crédit Mobilier between 1866
and 1868.
53
Table 2.7 Average asset prices for affiliated firms
Sources: Archives d’Entreprises Privées (General shareholders’ assembly and

newspapers); Société Immobilière or Immeubles de la Rue de Rivoli, 1857–67;
Compagnie Générale Transatlantique, 1854–67; Compagnie Générale des Omnibus de
Paris 1855–1914; Compagnie Parisienne d’Eclairage et de Chauffage par le Gaz, 1855–
1914.
Notes
a
When the share price increased this is indicated by a minus sign (i.e.
negative loss).
*
This must be considered as the loss per share before the reorganisation of
the Crédit Mobilier and the issue of the new shares. After the fall of the
Crédit Mobilier the Cie Immobilière went bankrupt: so the same comments
as those made for the Crédit Mobilier apply in that context.
**
After the bankruptcy of the Crédit Mobilier the Compagnie Transatlantique
chose to reduce its partnership with this bank. For example, in 1864, the
amount due to the Crédit Mobilier was 10,126,461.08 francs. The company
was supposed to give 25 francs of dividends deductable from a reserve fund
in case of insufficient profit over the year. In 1867 the dividend was only 15
francs per share and the Crédit Mobilier became a debtor of the company
for an amount of 4,514,610.78 francs. In 1868, M.Dollfuss became President
of the Compagnie Transatlantique; M.Péreire had to resign.
(a) The general influence of the

Crédit Mobilier’s bankruptcy
The aim of this section is to see if there was any marked effect on price
level and dividends due to the bank’s failure. This effect will be examined
by measuring changes in share prices for affiliated companies, correlated
with price level and dividend. The ordinary least square regression is
expressed by:
ln(pt / pt–1) = constant + α ln(pyt / pyt–1) + β (divt / divt–1)
54
Table 2.8 Relative impact of the bankruptcy of the Crédit Mobilier
where py stands for the general level of prices for the whole economy, divt
stands for general dividend series and pt represents the share prices for each
of the affiliated companies. The sample considered is 1861–75.
Table 2.8 does not reveal any significant figures regarding any influence
of the failure that are distinct from the effects of price level or dividends.
Dummy variables are introduced to concentrate the analysis on this specific
aspect. The table provides us with very low coefficients for both the GNP
deflator and dividends. If the figures produces by the second regression are
much higher, they remain insignificant as most of them are close to 0.
The regression equations which are going to be considered are:
ln(pt / pt–1) = constant + α ln(pyt / pyt–1) + β ln(divt / divt–1) + γd18

67
where d1867=1 in 1867 and 0 otherwise.

This test is carried out for each of the companies separately and over the
same sample, that is 1861–75. The results are given in Table 2.9 (standard
errors are in brackets).
The statistical shown in Table 2.9 provide no more explanatory power than the preceding
ones, which is quite surprising as the focus of the regression is the bankruptcy. The
coefficients are not significant except for two companies: the Compagnie Immobilière
and the Compagnie Transatlantique. Both these firms were confronted with financial
difficulties, for different reasons. For the former, its simultaneous failure with that of the
Crédit Mobilier (and causing it) explains the correlation between the different parameters.
In the Compagnie Transatlantique’s case, the reason behind the failure is found in the
struggle between the Compagnie, which wanted to sever all relations with the bank, and
the Péreires, who wanted to maintain their position on the board of directors.
55
Table 2.9 Relative impact of the bankruptcy of the Crédit Mobilier including a
dummy variable d1867
Table 2.10 Relative impact of the bankruptcy of the Crédit Mobilier including a dummy variable d1867/68
To complete the argument the same regression can be carried out for d1867/
68
considering that d1867/68=1 in 1867 and 1868, 0 otherwise. The results are
given in Table 2.10.
Table 2.10 confirms the results obtained in the preceding test. Except for a
relatively small number of firms such as the Compagnie Immobilière and the
Compagnie Transatlantique, the bankruptcy did not lead to any great perturbation
that could not be explained by the price level (e.g. the GNP deflator coefficient)
or the financial market (e.g. the dividend coefficient). In what follows, these two
parameters will be avoided: we concentrate on the correlation between the
fluctuations of the share prices for affiliated companies and the bankruptcy, and
we evaluate the validity of an association with the Crédit Mobilier.
(b) The impact on shares for affiliated companies
To analyse the magnitude of the fluctuations of the share prices, we analyse

two sub-samples: the first includes the railways companies, which suffered
less from the failure (in total six companies); the second includes all other
firms which were confronted with financial problems because of the policy
of the Crédit Mobilier toward credit.
Let us consider the six railway companies. An ordinary least square
regression can be run by calculating ln(p /p ) for each of the companies
t t–1
between 1862 and 1867. The equation of the regression will be then:
ln (pt / pt–1) = c + βd1867

The same regression can be done for d considering that d =1 in
1867/68 1867/68
1867 and 1868, 0 otherwise. The dependent variables are the share prices of
the Compagnie de l’Est.
Table 2.11 Regression for d1867
58
Table 2.12 Regression for d1867/68
The aim of the test is to analyse the impact of the bankruptcy of the
Crédit Mobilier on the share fluctuations.
Specifically we intend to test the null hypothesis:
H0 β=0: ‘The bankruptcy of the Crédit Mobilier has no influence on the share
prices’ against H1 β<0
The results of the test are given in Tables 2.11 and 2.12.
These tables lead us to the following comments.
• The coefficient β in both regressions is close to zero and positive. A

negative sign was expected.
• Except in two cases in the first regression (North and West), and three
cases in the second one (East, North and West), the significance level is
greater than 0.60.
• All t-statistics are insignificant.
Some particular comments can been made as regards the companies

themselves. For the Compagnie Immobilière and the Compagnie
Transatlantique, the arguments are quite different as both firms tried to sever
links with the Crédit Mobilier after its bankruptcy. In these two cases only,
the bankruptcy of the Crédit Mobilier induced a decrease in share prices. If
we examine the coefficient in the regression, we see that the Compagnie
Immobilière has a negative coefficient because of its simultaneous
bankruptcy with that of the Crédit Mobilier. For the Compagnie
Transatlantique the coefficient is positive, but greater than that obtained for
the railway companies, as this firm was liberated by the bankruptcy of the
Crédit Mobilier. The T-statistic is larger compared to the railway companies
but remains insignificant. The results for the Omnibus Compagnie and the
59
two insurance companies are very similar to the those obtained for the
railways. The only comment that can be made concerns the size of the
regression coefficients. For the Omnibus Compagnie and for l’Union Life it
is of the same order as the railways. But as they are all positive the
conclusion is identical: H0 is not rejected by the data and there is no
evidence that the bankruptcy of the Crédit Mobilier had any influence on the
share prices.
(c) The evaluation of the association with

the Crédit Mobilier between 1866 and 1868
The last point of our analysis will be to analyse the correlation between the
fluctuations of the share prices for affiliated companies and the variations in
asset prices for the Crédit Mobilier. With that aim in mind, monthly data
from 1866–8 are used to run two regressions. The first is expressed as:
Log(pt) = constant + β Log(pt CM).
The second has as its objective the evaluation of the interaction of the
change in asset price of the Crédit Mobilier on shares of affiliated
companies. The regression is given by:
Log(pt / pt–1) = constant + β Log(pt CM / pt–1 CM).
All these tests are made separately for each year. A final regression will
consider the sample as whole for 1866–8. The results are presented as a set
of two tables representing the figures obtained for each of the two
regressions each. Standard errors are in brackets.
Some comments are worth making on these regressions. For the year
1866 the sign of the constant term for the Compagnie Immobilière changes
between the two regressions: in the first regression it is positive and in the
second one it is negative. What interpretation can we give to this
phenomenon? In the upper tables the t-statistics are uninterpretable as, under
the null hypothesis, the disturbances are log(p ) and are not independently
t
identically distributed or even probably stationary.
This set of results as a whole leads to the following interpretation:
except for the Compagnie Immobilière (where the β coefficient is strongly
positive, as shown by the first table of each year) and the Compagnie
Transatlantique, nothing serious happened when the Crédit Mobilier went
bankrupt. What can be said when considering the sample as a whole? The
regression figures are presented in Table 2.16 (standard errors are in
brackets).
Tables 2.13 to 2.16 reinforce the preceding arguments.
60
Table 2.13 Results for 1866
Table 2.16 Results for 1866–8
(d) Summary and general interpretation
The preceding econometric results, as regards the impact of the bankruptcy

for affiliated firms, show that the Crédit Mobilier provided liquidity to firms
but did not seem to be supervising them (as witnessed by the bankruptcy of
the Compagnie Immobilière). Since the Crédit Mobilier itself went bankrupt,
its seal of approval was probably not worth as much as that of J.P.Morgan &
Co or the Großbanken.
According to Bradford de Long and Becht’s paper, ‘the Großbanken had
the power to carry out whatever threats they might make to or conditions they
might lay down for corporation managers. Their organisations controlled a
large part of voting stock—and a still larger part of the stock voted’. In spite
of the fact that the Crédit Mobilier often owned the biggest part of the shares
of firms, in which it was involved, such threats were not applied.
The authors add: ‘The key to the profitability of the Großbanken lay in
[their] ability to securitize [their] credit commitments, and to place the
bonds and shares of the corporations [they] sponsored at high prices…. The
Großbanken thus sought to enhance their issue credit,…as their reputation as
honest brokers selling sound securities, who would monitor corporations to
protect both their own investment and the investments of those who had
bought on the bank’s recommendation’.
Thus, the formal tests prove that de Long’s proposition does not apply in
the case of the Crédit Mobilier. According to this author ‘J.P.Morgan and
Co’s approval of an issue had become…a large factor which inspires
confidence in the investor and leads him to purchase’. Such was not the case
with the Crédit Mobilier, whose active participation of this bank in a
company did not ‘add value’ to the share prices. The inverse proposition
seemed to be closer to the reality in this case.
An analysis of the place of the Crédit Mobilier on the financial market
enables us to give an interpretation of this result. From an economic point of
view, this bank was at the height of its power on the Stock Exchange at the
beginning of the 1860s. Around 1863, the innovative component of its
investments made it lose part of its force in financial matters. In particular,
the Crédit Mobilier was subject to speculative movements which contributed
partly to its new fragility. However, it was its significant participation in the
financing of firms such as the Compagnie Immobilière that was the origin of
the fall of the Crédit Mobilier.
2.6 BANKRUPTCY EFFECT ON THE COMPANIES

NOT AFFILIATED TO THE CRÉDIT MOBILIER
To complete the analysis, it will be interesting to extend the study to the firms
that were independent of the Crédit Mobilier. As in the preceding section,
65
formal tests can be constructed to analyse the fluctuations of the share prices for
the companies which were not affiliated to the Crédit Mobilier. Let us consider,
then, the twelve steel companies and the Lenoir series to study the investment
constraint, if any, of these companies. A methodology identical to that followed
in the preceding section is used for the non-affiliated sample. In what follows,
the impact of the bankruptcy will be analysed, taking into account the general
effect of the bankruptcy on the financial market and the correlation between the
failure of the bank and the fluctuation of share prices for affiliated companies.
(a) The general influence of the Crédit Mobilier’s

bankruptcy for non-affiliated firms
The aim of this section is to see if there was any marked influence on price
level and on the financial market due to the failure of the bank. This effect
will be examined in two stages:
• first, in the absolute, by evaluating the effect of price level and dividend
on share price for non-affiliated firms;
• second, by measuring the provoked modification in share prices for
non-affiliated companies in correlation with the change in price level
and dividend.
The ordinary least square regression is expressed by:
ln(pt / pt–1) = constant + α ln(pyt / pyt–1) + β (divt / divt–1)
where py stands for the general level of prices for the whole economy, pt
represents the share prices for each of the affiliated companies and div is the
general dividend series. The sample considered is 1861–75.
The results are given by Table 2.17 (standard errors are in brackets).
As previously, Table 2.17 does not produce any significant figures
regarding any influence of the failure distinct from the effects of price level
and dividends (the coefficient obtained for the logarithm of py and dividends
being close to 0 or negative) dummy variables are introduced to concentrate
the analysis on this specific aspect. The regression equations which are
going to be considered are:
ln(pt / pt–1) = constant + α ln py + β ln(div) + d18

67
ln(pt / pt–1) = constant + α ln(pyt / pyt–1) + β ln(divt / divt–1) + γd1867
66
Table 2.17 Relative impact of the bankruptcy of the Crédit Mobilier
This test is done for each company separately and over the same sample,
that is 1861–75. The results are given by Table 2.18 (standard errors are in
brackets).
The statistical results given by Table 2.18 do not add any information to
the preceding tables. To complete the argument the same regression can be
done for d1867/68 considering that d1867/68=1 in 1867 and 1868, 0 otherwise.
The results are given by Table 2.19.
This comparison shows that there was no general increase in share prices in
1867 or 1868 in response to the Crédit Mobilier’s bankruptcy. For more than
half of the sample the coefficients for d1867 and d1867/68 are significant and
positive, while for the rest of the sample the same coefficients are slightly
negative and thus not really significant. On the other hand the argument
coefficients (ln(py) and ln(div) for the first regression, ln(pyt/pyt–1) and ln(divt/
div t–1 ) for the second one) are small and not significant. This shows
(unsurprisingly, given the results for affiliated companies) that the Crédit
Mobilier’s bankruptcy did not cause a general panic. In what follows, these
two parameters will be avoided: we concentrate on the correlation between the
fluctuations of the share prices for affiliated companies and the bankruptcy,
and we evaluate the validity of an association with the Crédit Mobilier.
(b) Share price fluctuations while

the Crédit Mobilier went bankrupt
To analyse the magnitude of the share price fluctuations, an ordinary

least square regression can be run by calculating ln(p t /p t–1 ) for
67
Table 2.18 Relative impact of the bankruptcy of the Crédit Mobilier including a dummy variable d1867
Table 2.19 Relative impact of the bankruptcy of the Crédit Mobilier including a dummy variable d1867/68
each company between 1862 and 1867. The equation of the regression
will be then:
ln (p p ) = c + βd
t / t–1 1867

The same regression can be done for d1867/1868 considering that d1867/68 =1
in 1867 and 1868 and 0 otherwise.
The results are presented in Tables 2.20 and 2.21. These tables lead us to
the following comments.
Table 2.20 Results for d1867
Table 2.21 Results for d1867/68
70
The coefficient β is close to zero and positive except in the cases of

Chatillon, Denain and Allais. These results must be put in perspective since
these three companies had encountered difficulties during the recession of the
steel industry at the beginning of the 1860s and around 1870. So the size and
the sign for these three companies can be explained by factors other than a
lack of association with the bank. In general, a negative sign would be
expected if the failure of the Crédit Mobilier caused financial disruption. In
particular, if the Crédit Mobilier were truly a great bank, as it is presented in
the historical and economic literature, its bankruptcy should have had an
influence on the other banks (like the Crédit Lyonnais, which was the
financial institution of the companies quoted above) and subsequently on the
companies affiliated to these banks. The results of the test give no support to
this hypothesis. All t-statistics are low, and none of these coefficients are
significantly different from zero.
The conclusion is then that H is not rejected by the data and there is no
0
evidence that the bankruptcy of the Crédit Mobilier had any influence on the
share prices for companies not involved with the bank.
(c) Conclusion
The Crédit Mobilier did not cause general distress (unsurprisingly, given the
results for affiliated firms). There is only weak evidence of a general boom
which could mash the effect of the Crédit Mobilier’s bankruptcy. The results
for affiliated and non-affiliated firms are very similar. This suggests that de
Long’s proposition does not apply to the Crédit Mobilier.
2.7 COMPARATIVE ANALYSIS OF THE

RESULTS OBTAINED FOR AFFILIATED AND
NON-AFFILIATED COMPANIES
In the preceding sections, we present regressions run for two samples:

affiliated and non affiliated companies. The most interesting results are
given by studying the fluctuations of share prices when the Crédit Mobilier
went bankrupt. The aim of this section is to compare the results obtained for
the two samples. Let us first recall the equations of regression:
ln(pt / pt–1) = c + βd1867
or ln(pt / pt–1) = c + βd1867/68
where d 1867=1 in 1867 and 0 otherwise, d1867/68=1 in 1867 and 1868, 0

otherwise.
71
The hypothesis tested is H0 β=0: ‘The bankruptcy of the Crédit Mobilier

has no influence on the share prices’ against H1 β different from 0.
Let us first compare the results obtained in the two cases for d1867 and
d1867/68 as regards the coefficient β. Let us recall these values for d1867.
For the railways companies we obtain
0.018<β<0.080 and all are positive.
For all the affiliated companies except the railways the values are given by:
–0.044<β<0.322 and the magnitude of the coefficients is more important.
For the non-affiliated companies (excluding Allais) we have:
–0.1125<β 0.097.
For d1867/68 the railways coefficients ß are described by
0.014<β<0.092 and are all positive.
For all the affiliated companies except the railways the values are given by:
–0.09<β<0.14 and the magnitude of the coefficients is more important.
For the non-affiliated companies (excluding Allais) b is included in:
–0.1125<β<0.097.
As regards the standard error, affiliated companies as a whole show for d1867
a value smaller than the ones obtained for d1867/68 and close to 0. The same
consideration can be held for the non-affiliated companies.
What kind of interpretation can we give to these results? The railways
were best placed to resist bankruptcy for the following reasons:
• the Rothschilds were a financial partner;

• they benefited from government subsidies when facing difficulties;
• they were able to diversify their investment which enabled them to resist
the crisis.
72
As a whole, the affiliated companies did not suffer from the bankruptcy of
the Crédit Mobilier. The results obtained for the affiliated and non-affiliated
companies are very similar (coefficients are slightly higher for d1867/68, but
with a magnitude too small to be significant).
In conclusion, we can affirm that the bankruptcy of the bank had no
influence on either affiliated or non-affiliated companies. An association
with the bank does not seem to have increased the values of shares. This
expresses the lack of supervision of the Crédit Mobilier.
2.8 BANKRUPTCY EFFECT ON THE COMPANIES

EITHER AFFILIATED OR NOT AFFILIATED TO
To extend the preceding results, a cross-section test is used, defined as:
ln(pt / pt–1) = c + β(CM)
where CM=1 if the companies were affiliated to the Crédit Mobilier, CM=0
otherwise.
We run the regressions for the years 1866, 1867, 1868 and evaluate the
ratios:
ln p68 / p67
ln p68 / p66
The results, not significantly different for either ratio, are presented in
Tables 2.22 and 2.23.
Several comments can be made on Tables 2.22 and 2.23.
Table 2.22 Cross-section results excluding the Compagnie Immobilière
Table 2.23 Cross-section results including the Compagnie Immobilière
73
• The figures obtained for the cross-section are not very different from
those resulting from the preceding OLS.
• The coefficient β in both regressions is close to zero. The only negative
sign appears when the Compagnie Immobilière is included.
• All t-statistics are insignificant.
Except when the Compagnie Immobilière is included, coefficients provide

no evidence that the bankruptcy had any effect on the share prices of
affiliated companies. Even when a negative sign is obtained, the magnitude
of coefficients is too small to be significant.
H0 is not rejected by the data.
The preceding tests prove the irrelevance of de Long’s proposition to the

Crédit Mobilier case. According to de Long, ‘J.P.Morgan and Co’s approval
of an issue had become…a large factor which inspires confidence in the
investor and leads him to purchase’. The active participation of the Crédit
Mobilier in a company did not ‘add value’ for the share prices.
More specifically, our argument concerning the period 1861–70 proves
that association with the Crédit Mobilier had no great influence on the
share prices of the companies, as the failure of the Crédit Mobilier had
little effect.
The analysis of the situation of companies that were not affiliated to the
Crédit Mobilier reinforces this hypothesis. Its bankruptcy did not cause a
big enough financial perturbation to induce share price fluctuations on the
important business concerns over the period.
Between 1852 and 1870 the Crédit Mobilier does not appear to have been
a bank capable of supervising properly the firms to which it granted credit.
It essentially provided liquidity but without screening the firms’ managers.
Moreover, the communication it entertained with other banks did not reflect
the power of the Crédit Mobilier on the financial market. Numbers of banks
over the period (like the Banque de France, the Société Générale and the
Crédit Lyonnais) had serious doubts about the innovative ideas of the Crédit
Mobilier concerning the financing of business projects. The results of the
test as regards companies that were not affiliated to the Crédit Mobilier
confirm this idea.
All these findings lead us to question the concept of ‘control’: can a bank
effectively control a firm through the credit it grants? No unequivocal answer
can be given to this question. Edmond Baldy in his book Les Banques
d’Affaires en France depuis 1900 bases his argument concerning effective
control on the percentage of participation a bank has in a firm. Strictly
speaking, the required participation for absolute control is 50 per cent plus
74
one, but this is, of course, subject to the ownership of the assets between the
shareholders. In particular, if one shareholder holds a huge amount of assets
of the company, the bank, in order to exert a supervisory role, must acquire a
majority, strictly larger than the external group. As regards the Crédit
Mobilier, if the participation in the firms was always major, nevertheless it did
not, typically, represent the largest proportion of the assets issued in the
companies. Hence the monitoring role of the bank was diminished, according
to this argument. However the strongest explanation of this lack of control
remains the bank’s risk-prone investment policy (e.g. the Compagnie
Immobilière).
Beyond this economic interpretation, a historical explanation can be
given. Around 1865 the Crédit Mobilier had lost some of its power. As early
as 1863, the Péreires requested permission to double the capital of the
Crédit Mobilier, but the government refused. With more and more of its
assets tied up in advances to the Société Immobilière, in 1866 the Mobilier
eventually persuaded the government to allow a doubling of its capital, in
return for modifications in its statutes restricting still further its freedom of
action. Naturally, the Péreires did not reveal to their stockholders the true
situation of the bank, but spoke vaguely and grandly of the need for new
resources to capitalise on the many opportunities for profitable operations
which daily came its way. In fact, of course, the Mobilier undertook no new
enterprises, but devoted all of the new capital to an attempt to salvage the
Société Immobilière.
Added to these difficulties, the Péreires faced a stockholders’ lawsuit
against the increase of the bank’s capital. In 1867, the Crédit Mobilier went
bankrupt. The first reactions of the governors of the Banque de France were
outrage and indignation. But when the financial and political implications
that the sudden failure of the Mobilier could provoke were pointed out, the
governors agreed, in effect, to permit a more gradual liquidation of the
enterprise. This constitutes one additional possible explanation of the lack of
fluctuations for the share prices of the companies with which the Crédit
Mobilier was involved.
75
3
CORPORATE INVESTMENT,
CASH FLOW AND FINANCIAL
CONSTRAINTS OF FIRMS
French industry with its tradition of self finance had few direct
links with the banking system…Whether or not this was so,
and if so whether it reflected the strength or weakness of
industrial enterprises, is still a subject for debate.
Cameron and Bovykin (1991:15)
3.1 INTRODUCTION
As shown in the preceding chapter, the Crédit Mobilier played a crucial

role in the industrialisation of French industry over the period 1852–70.
But contrary to the conclusions reached by other researchers studying
other banks in other countries (e.g. de Long’s assertions about J.P.
Morgan & Co), it does not appear to have been effective in facilitating
the assessment of firms’ managerial performance and in the evaluation of
projects.
The Crédit Mobilier provided liquidity to firms but does not seem
to have been supervising them. Since the Crédit Mobilier went
bankrupt itself, the value of its seal of approval was correspondingly
diminished.
Moreover, the formal tests we have constructed for the year of
bankruptcy of the Crédit Mobilier cast further doubt on the assertion that its
seal of approval might have added value to French corporations. According
to de Long, ‘J.P.Morgan and Co’s approval of an issue had become …a
large factor which inspires confidence in the investor and leads him to
purchase’. The evidence suggests that this was not true of the Crédit
Mobilier. The active participation of this bank in a company did not ‘add
value’ to the share prices. If anything, the inverse proposition seems to be
closer to the reality.
To analyse more deeply the role of the Crédit Mobilier in the financing
of industrial companies, it would be interesting to see if an association with
76
CORPORATE INVESTMENT, CASH FLOW AND FINANCE
the Crédit Mobilier reduced such companies’ costs of external borrowing

and consequently made their investment less sensitive to cash flows.
In a world of perfect information and no agency costs, firms can
borrow and invest optimally; liquidity has no effect on the investment
decision. In a classic paper, Franco Modigliani and Merton Miller (1958)
demonstrated the conditions under which financial structure becomes
irrelevant for the real investment decisions. Jorgenson (1992) has
provided an extensive survey of the theory of investment behaviour,
distinguishing between classical, neo-classical accelerator and Q theories;
Modigliani-Miller (M-M) might apply to all of them. Today, however,
there are many models of corporate finance that have moved away from
the assumption of perfect information, stipulating some kind of
information problem.
The assertion that imperfect information invalidates M-M has many
consequences:1 all these models predict that liquidity matters. Thus, our
strategy is to see whether liquidity is a more important determinant of
investment for independent, unaffiliated firms than for firms with close
banking ties.
Recent articles have explored the empirical relationship between
corporate financial structure and investment.2 Fazzari, Hubbard and Petersen
(1988) classify firms according to their a priori beliefs about whether a firm
faces information problems in the capital market, and then test whether the
information problems are presumed to be severe. Their basis of comparison
is corporate dividend policy, and they argue that firms that retain more of
their earnings are more likely to be liquidity constrained. Indeed, they find
that investment is more sensitive to liquidity for firms that consistently
retain a larger fraction of their earnings.
Hoshi, Kashyap and Scharfstein (1989, 1991), on the other hand, show
that investment expenditures of Japanese companies affiliated to a bank
(e.g. a member of a zaibatsu) display almost no sensitivity to cash flows
in contrast to the investment expenditures of firms which are not so
affiliated. This finding supports the hypothesis that companies that are
not affiliated to a bank tend to be more liquidity constrained (and face a
wider gap between external and internal costs of funding) than those
which are. This work follows their approach and investigates the
assertion that the Crédit Mobilier played a role in the financing of
French industry that was similar to that of J.P.Morgan in the United
States3 and of the Großbanken in Germany. 4
In order to construct the empirical case study, two samples are
considered: the first consists of corporations that were affiliated to the
Crédit Mobilier, the second set of companies had to rely on other sources
of finance. However, all firms were affiliated with a bank: independent
companies used the Crédit Lyonnais to provide liquidity. This does not
prevent us from assessing the impact of the bankruptcy on investment for
77
affiliated and non-affiliated firms. The focus of this paper is thus on

contrasting the financing of the Crédit Mobilier’s affiliated enterprises
with those which, for one reason or another, did not form part of the
nexus. In particular, the study of its investment policy before and after the
bankruptcy will indicate the nature of the supervisory role of the Crédit
Mobilier. To determine the appropriate procedure for deciding if firms
were affiliated with the Crédit Mobilier, a short discourse into the general
nature of the business relationship between banks and industry, as was
suggested by contemporary sources like Jeidels (1905) and Riesser (1911),
might be in order.
3.2 THE VALUE OF ASSOCIATION WITH

THE CRÉDIT MOBILIER ON CASH FLOW,
INVESTMENT AND MARKET VALUE
There are several ways in which a bank and an industrial corporation can
interact. A company might conduct its daily business through a current
account with one or more banks. This was not the most common practice
with the Crédit Mobilier. In the case of a private company, a bank might
help to issue new equity or debt, for example bonds. A bank might take a
direct stake in the company, or it appoint representatives to sit on the
company’s board of the directors.
This last arrangement was the most frequent for the Péreires’
establishment. In every company financed by the Crédit Mobilier, a member
of its board of directors was represented in the shareholders’ assembly.5 The
question of the advances to companies which were granted by the Péreires
emphasises the direct participation of the bank.
Lending, issuing shares, direct investment and supervisory board
members helped the bank to monitor and in many cases influence decision-
making in the industrial enterprise. Nevertheless, Jeidels (1905) carefully
distinguishes between monitoring, in the sense of information-gathering, and
direct influence on a corporation’s business decisions. Although collecting
information seems to have been likely in the Crédit Mobilier’s case, any
kind of direct influence over the decisions of management seems improbable
(cf. the empirical results in Chapter 2). However, generally speaking, placing
bank directors or members of the bank onto the board of corporations was
common practice for the Péreire brothers and could have been the most
effective means of influencing managerial decisions of affiliated
corporations. In Chapter 2 we tried to analyse whether this supervision was
effective in the Crédit Mobilier’s case. We intend to reinforce this argument
by examining the investment policy of the bank.
The aim of this section is then to analyse the relationship between
investment and cash flows between 1860 and 1880. Two samples are
78
considered: the first consists of corporations that were affiliated to the

Crédit Mobilier, the second includes companies which relied on other
sources of finance. The focus of this construction is on contrasting the
financing of the Crédit Mobilier’s affiliated enterprises with those which,
for one reason or another, did not form part of the nexus. The
chronological limits of this study are imposed by the data sources: after
1880 the Crédit Mobilier lost its powerful role in the banking sector and
thus no records exist for this period. The sample is constituted of fifteen
companies: seven were affiliated with the Crédit Mobilier, eight with the
Crédit Lyonnais. As each company had a relationship with a bank the
interpretation of the results must be in some sense cautious. Conscious of
these limits, we discuss the validity of the actual research.
(a) Data selection
The principal source for data underlying the statistical analysis comes
from the General Assembly’s reports and some articles and books such
as those by Jean Denuc (1934) and Bertrand Gille (1968). In his book,
Bertrand Gille has reproduced a great number of company balance
sheets, particularly steel companies; these figures make it possible to
deduce cash flows, investment and book values. Jean Denuc provides
values for other companies; the railways companies were reported
separately. The different indices Denuc gives for all series enable us to
deduce the book value and the cash flows for the different companies.
The investment factor was given by an article issued by the Institut
National de la Statistique which recapitulated the economic situation of
France from 1850 until 1914. Here, we want to evaluate the correlation
between cash flows CF, common equity Q and investment I. The
regression specification was taken directly from Hoshi, Kashyap, and
Scharfstein (1989, 1991) and was also used in Ramirez (1992). Table
3.1 lists the companies included in the sample taken into consideration
and records their affiliation with the Crédit Mobilier. As can be seen,
railways companies represent the major part of the Crédit Mobilier’s
investments.
We can note also that all companies were affiliated with either the Crédit
Mobilier or the Crédit Lyonnais, both as powerful banks over the period.
When the banks offered credit for business firms at this time, industrial
enterprises were rarely involved because they had been warned by
experience. Significant amounts of capital were provided by merchant banks
especially to railways, shipping companies and urban transport projects, but
industrial enterprises received little. As far as we are concerned with public
utilities, the bank’s affiliation was omnipresent. This suggests that it may be
difficult to deduce the effect of affiliation with a big bank by comparing
79
Table 3.1 Companies included in the data set
Crédit Mobilier’s affiliated companies to non-affiliated companies. However,

we should note that the degree of affiliation was totally different between
the two banks. In the Crédit Mobilier’s case, the bank was substantially the
principal shareholder of the client firms, whereas for the Crédit Lyonnais,
affiliation only meant that the bank was a privileged financial partner of the
firm involved. A useful approach is to contrast the effect of the Crédit
Mobilier’s bankruptcy on those companies with which it was affiliated, and
on Crédit Lyonnais affiliated companies.
Hoshi, Kashyap and Scharfstein’s methodology used four parameters to
construct a regression of the gross invesment I, the gross value of share K,
the free cash flow CF, and Q measured as the market value of common
equity divided by its book value. K constitutes a problem in the French
context. It corresponds to the gross value of assets—in French, ‘le capital
brut’. In the balance sheets of the companies taken into consideration in the
sample, this factor seems to have be particularly stable over the period. The
question arises: does this variable mean the same thing in the work
presented by Hoshi et al. and in the French situation? It is hard to answer
this question.
Entrepreneurs over the period 1860–80 did not use sophisticated
accounting and management techniques: what counted was the overall state
of receipts and expenses seen in a concrete manner in the cash flow of the
business. The next step of sophistication was the establishment of an annual
budget, which was important because it allowed an assessment of net profit.
But accounting rules were not imposed. Every entrepreneur had his own
accounting conventions, to some degree, which makes it difficult to interpret
80
and compare these balance sheets. In order to avoid any ambiguity, this
parameter will be called Cb, which should represent for the French case the
gross value of shares.
All variables, the market prices of common equity, the cash flows CF and
the investment I, are normalised by the stock of depreciable assets at the
beginning of the period. First, some summary statistics comparing Crédit
Mobilier firms and non-Crédit Mobilier firms are presented. As the Crédit
Mobilier went bankrupt in 1867, the mean is calculated in constant terms in
regression, taking three samples:
• 1861–67—the situation before the bankruptcy;

• 1868–75—the situation after the failure of the bank;
• 1861–75—for general information.
The Compagnie Immobilière is either excluded or included in the sample,

as it went bankrupt with the Crédit Mobilier.
Table 3.2(a) shows the differences in the means for Tobin’s Q, investment/
share value and cash flows/share value. Although some means are different, the
standard errors would seem to suggest that no difference existed. In addition, the
figures obtained for Q are quite low. An examination of the accounting rules over
the period seems necessary. Financial reporting at the beginning of the nineteenth
Table 3.2 Summary statistics comparing Crédit Mobilier and non-Crédit Mobilier
firms*
81
Table 3.2 Continued
Note: Averages are calculated for all firms and all years. Standard errors are in
parentheses. Capital (Cb) and investment (I) are for depreciable assets. Sources and
definitions of variables are described in the text. The sample for the Crédit Mobilier
firms is constituted of seven firms, for non-affiliated firms eight.
century was characterised by the odd, scrappy profit calculation and balance sheet
which contained assets valued in a wide variety of ways. In fact the main aim of
the published balance sheet was to provide creditors and shareholders with a
statement of company solvency and to reassure them that dividends had not been
paid out of capital. There was no requirement for the publication of a profit
statement, nor were there any minimum disclosure rules regarding the content of
the balance sheet. During the nineteenth century the calculation of profit became
the first stage in the preparation of the final accounts. The new approach to profit
was the use of profits as the basis for dividend declarations. This adoption of the
matching concept reduced the scope for management to manipulate profit; with
valuations, there is a need not only to make estimates, but also to decide on the
appropriate method of valuation from the various options available, including
present value, selling price and replacement cost. This last parameter is of such
importance that we should refer to James Tobin’s Q theory of investment.6
This can be summarised as follows. The rate of investment—the speed at
which investors wish to increase the capital stock—should be related, if to
anything, to Q the value of capital relative to its replacement cost.
Over the period, replacement accounting delayed the recognition of
capital consumption until expenditures were made for renewals.
Furthermore, since periodical renewals would require substantial sums, there
was a bias against renewals as costs because replacement reserves were not
generally maintained. On the one hand, a firm might not have been able to
charge substantial renewals to expenses without creating a deficit, while on
the other hand, without sufficient internal funds available, it would be
necessary to finance replacement expenditures with sources external to the
firm. This discussion suggests that the application of replacement accounting
tends to create serious liquidity problems and that renewals might have been
deferred or treated as additions in many cases.
82
There was considerable scope for suspicion between bankers and

industrialists over these years: bankers were cautious when dealing with
industrialists because the former feared long-term commitments which were
not thought profitable on a short term basis, while the industrialists
considered the bankers as simply rapacious. Hence debt took the form of
advances on capital, and was not reported as such in the balance sheets. On
this basis it is hard, if not impossible, to approximate debt as such.
(b) The investment regression
The selected regression specification takes the following form:
It/Kt = α0 + α1 CFt/Cbt + α2 Q
where:
I = gross investment at time t, measured simply as the change in the gross

value of depreciable assets (plant and equipment);
Cb = gross value of assets;
CF = free cash flow, i.e. free surplus plus nominal allowances (mostly
depreciation);
Q = the market value of common equity stock divided by its book value.
Different sub-samples were chosen to run this regression:
• the whole sample;

• the period before the bankruptcy, 1861–7;
• the years following the bankruptcy.
The main results are presented in Tables 3.3, 3.4 and 3.5. The evidence
presented in these three tables suggests that free cash flow was more
important in both magnitude and statistical significance among independent
firms. That means that free cash flow was an important determinant of
investment expenditures among the independent mining and steel companies.
This is also evidenced by the fact that the overall pattern of the regression
for affiliated companies is significantly different from that of the non-
affiliated companies. That is, we cannot prove that the explanatory variables
are explained by different investment expenditures among independent and
Included in these tables is an interaction dummy which takes the value 1
times cash flow when the corporation is affiliated to the Crédit Mobilier.
This variable is introduced to estimate the difference and statistical
importance of the cash flow coefficient between affiliated and non-affiliated
companies.
83
Table 3.3 Investment regression, 1861–75
Note: *CF.CM represents CF times Crédit Mobilier affiliation: affiliated firms 7,

independent companies 8, all firms 15.
Also included in the tables is a regression for the whole sample,

excluding the Compagnie Immobilière, and a regression for the sample
taking into consideration this firm, which went bankrupt at the same time as
the Crédit Mobilier. This last case produces different results depending on
whether or not we consider the situation of the Compagnie Immobilière.
This can be explained by the fact that the Crédit Mobilier invested a huge
amount of capital in the firm which was irrelevant at the end and provoked
its own bankruptcy.
The different coefficients entering the investment equation were evaluated
by identifying only whether or not a firm was affiliated to the Crédit
Mobilier. The results obtained describe the general situation as regards the
investment policy of the Crédit Mobilier. No change in investment is taken
into consideration this time. The aim is to analyse the relationship between
common equity and cash flow on investment for affiliated and non-affiliated
companies.
The question is then whether the structure of ownership and the behaviour of
large shareholders differs between firms that were affiliated with a bank, and
those that were independent of such groups. The empirical results compiled in
order to answer this question are presented in Tables 3.3, 3.4 and 3.5.
Table 3.3 enables us to identify the difference between affiliated and non-
affiliated firms and also to stress the importance of the Compagnie
Immobilière in the investment policy of the Crédit Mobilier. If independent
84
firms were liquidity constrained by having to rely on cash flows to finance

their projects, the coefficient CF/Cb must be much higher for these firms
than for affiliated firms.
Let us consider the interaction effect CF.CM. Everything else being
constant, a unit increase in cash flows contributes about 0.02 units increase
in investment expenditures among the Crédit Mobilier’s firms. Unlike the
separate regressions, this suggests that affiliation with the Crédit Mobilier
reduced liquidity. The results vary, as other coefficients must be the same
for all firms as reported in columns four and five. The coefficient is then
insignificant.
An analysis of the impact of the bankruptcy of the Crédit Mobilier is now
required. Two sub-samples of the initial sample (before and after the
bankruptcy) are considered. A dummy variable reflects the affiliation of a
company to the bank.
For the sub-sample 1861–7, we suppress the Compagnie Immobilière: the
absence of too many values prevents us from obtaining the relevant results.
The results are presented in Table 3.4.
Let us comment on the interaction coefficient, CF.CM. If we compare
the result obtained for the whole sample and the one we have for the years
preceding the bankruptcy, we note that the value is much larger for 1861–
7 than for the whole sample. That means that an association with the
Crédit Mobilier was relevant during the period 1861–7. Remarkably the
Note: *CF.CM represents CF times Crédit Mobilier affiliation: affiliated firms 7,

independent companies 8, all firms 15.
85
coefficient is significantly positive which suggests that association with the

Crédit Mobilier reduced liquidity and thus increased the effect of the cash
flow on investment. As regards the separate regressions, note that the cash
flows are very similar for affiliated companies compared to the
independent firms.
What happened after the bankruptcy? Surprisingly, the coefficient on CF/
Cb for affiliated companies is tiny and insignificant. It is as if the failure of
the Crédit Mobilier removed a liquidity constraint, the opposite of what
theory would predict. The coefficient on the interaction CF.CM is negative,
as if companies once affiliated to the new bank are less liquidity constrained
than ever in the past. Note that for 1861–7 this coefficient was positive
(Table 3.4 provides us with a coefficient equal to 0.101 for all companies
including the Compagnie Immobilière and equal to 0.131 for the whole
sample excluding the Compagnie Immobilière).
The results, taken together, are the opposite of those predicted by the
theory. It is as if affiliation with the Crédit Mobilier created liquidity
problems.
Conclusion
The results obtained for France in the Crédit Mobilier case are very different
from those presented for the United States and Japan by de Long and
Ramirez, or for Germany by Marco Becht. Substantially, de Long (1989)
asserts that the monitoring role of the Morgan Corporation was, in general,
successfully achieved. The Pujo Committee of 1913 agreed that the House
of Morgan had sufficient resources and influence in the financial world to
render it the most powerful and active of the few investment houses. The
econometric work of Ramirez (1992) suggests that association with
J.P.Morgan relieved a credit constraint.
The evaluation of Q and its fluctuations is not completely different from
the results obtained by de Long and Ramirez. However, the participation of
the bank in the firms’ projects was so non-uniform that a direct association
with the Crédit Mobilier did not always contribute to an increase of Q for
the affiliated firms (Table 3.5). It seems on the contrary that the bankruptcy
of the Crédit Mobilier removed a liquidity constraint for affiliated firms. The
comparison of the results obtained for 1861–7 on the one side and for
1868–75 on the other side indicates that, after the failure, the affiliated firms
did not suffer from liquidity constraint as they had before.
As a whole, our results seem to indicate that association with the Crédit
Mobilier created liquidity problems. The next section intends thus to isolate
the bankruptcy effect in order to evaluate more precisely the profitability of
an association with the Crédit Mobilier (if any) on Q.
86
(c) The fixed effect model
The regression equations are evaluated by introducing year dummy and firm
dummy variables. In other words, we determine investment for every year
except for one, separately, and for every firm except one, separately. As in
the preceding test we consider three samples:
• the whole sample 1860–75;

• the sub-sample 1860–7;
• the sub-sample 1868–75.
We run the regressions for affiliated companies, non-affiliated companies

and all firms.
The selected equation is modified as
I/Cb = α0 + α1 Q + γ.CF/Cb
where:
I = gross investment at time t, measured simply as the change in the gross

value of depreciable assets (plant and equipment);
CF = free cash flow, i.e. free surplus plus nominal allowances (mostly
depreciation);
Q = common equity.
87
Table 3.6 Investment regression equation for the whole sample*
Note: *The dependent variable is investment relative to capital stock at the beginning
of the period. The regressions include year dummies and firm dummies. Standard errors
in brackets: affiliated firms with the Compagnie Immobilière 7, independent companies
8, all firms 15.
The regression equations are evaluated by introducing year dummy and firm
dummy variables. In other words, we determine investment for every year
except for one separately and for every firm except one separately.
Let us consider first the Crédit Mobilier firms and non-affiliated
companies. Tables 3.6, 3.7 and 3.8 present the main results.
Some general remarks can be made regarding these tables:
• the coefficient of Q obtained for affiliated companies is quite high,

whatever the sample considered;
• however, for these same companies, the coefficient on cash flows is low;
• generally speaking, the standard errors are quite large.
The evidence presented in these tables suggests that the coefficient on

cash flow gains importance in magnitude among the non-Crédit Mobilier
companies.
Table 3.9 presents the results of the regression specification over the
entire sample. The regression includes an interaction dummy which takes
the value 1 times the cash flow when the corporation is part of the Crédit
Mobilier sample. This variable is introduced to estimate the difference and
statistical importance of the cash flow coefficient among affiliated and non-
The findings are not very robust: except for the sample 1861–75, the
interaction coefficient is negative and the standard errors are quite large.
Crédit Mobilier firms relied less on cash flow to finance their investment
expenditure over the first period than on other sources of credit. For the two
latter samples, no evidence is shown by the table.
Previous explanations of the importance of cash flow for non-affiliated
companies from Japan and the United States are not really applicable to the
88
Table 3.7 Fixed effect model for 1861–7*
8, all firms 15.
8, all firms 14.
Table 3.9 Fixed effect model for the whole sample*
Note: *The dependent variable is investment relative to capital stock at the beginning of
the period. The regressions include year dummies and firm dummies. Standard errors in
brackets. The number of companies is 15 for 1861–67 and 1861–75, and 14 for 1868–75.
89
French case. The Crédit Mobilier did not act as a universal bank throughout
the whole period. Because of asymmetric information, non-affiliated
companies face much higher costs of external financing than did affiliated
ones. However, this cost was partly compensated by the risky behaviour of
the bank towards investment.
How can we explain these differences between France and Germany?
A series of eminent economic historians 7 had held that the major
innovation in industrial financing in the nineteenth century was the
Crédit Mobilier. Gerschenkron’s theory was elaborated into a model, a
backwardness, in which banks, depending on how backward a country
was, substituted liquidity for entrepreneurship. This had been the driving
force in the industrial revolution in the United Kingdom, for example.
But evolution proceeded rather differently in other countries. German
banks took the example of the Crédit Mobilier and developed it to the
utmost. Apart from Hamburg and Frankfurt, which clung longer to
commercial banking, the German banks went in for close relations with
industrial and mining firms.
Theoretically, the Crédit Mobilier’s case illustrates perfectly the following
dilemma. Adherents of rational expectations and efficient markets tend to
believe that markets always work. Institutionalists, of course, take the
opposite view, that institutions determine historical outcomes. Between the
two positions, there is room for an eclectic view that markets mostly work
but occasionally break down, and that sometimes institutions adapt to
underlying changes in demand and supply conditions and sometimes they do
not. The Crédit Mobilier was not capable of adapting to the change.
(d) Conclusion
The main premise of this section is that, unlike the Morgan company, 8
the Crédit Mobilier was not fulfilling an effective monitoring role but
only provided liquidity to firms whenever necessary. This bank does not
seem to have alleviated the impact of the imperfections of the capital
market.
To strengthen the argument, we consider two extensions of this work in
the following chapter. The first step will be to include in the book value a
depreciation rate δ, equal to 5 per cent, in order to conform more to the Q
theory. All details for the description of the parameters will be given in
Chapter 4. Some additional value for δ will be considered to evaluate the
robustness of the test. The next step will be to regress investment, cash flow
and Q on the basis of their real value, that is by dividing the parameters by
the general level of prices in France.
90
3.3 TRANSFORMATION OF THE RAW DATA

ACCORDING TO ‘TRADITIONAL FINANCIAL
ACCOUNTING’
As previously mentioned, the Q values are quite low because of the

absence of replacement cost. The first transformation will be to introduce
a depreciation rate in the evaluation of Cb and therefore Q in order to be
consistent with James Tobin’s Q theory. This theory of investment can be
summarised as follows. The rate of investment—the speed at which
investors wish to increase the capital stock—should be related, if to
anything, to Q, the value of capital relative to its replacement cost (Tobin
1971:330). Wilfried Feldenkirchen (1982:269) explains that over the
period this replacement cost was not always reported. For German
companies a depreciation rate was only introduced when the firms were
making a profit. The same procedure occurred in France. The appendix
describes the new evaluation of Q. The second proposition is to regress
investment, cash flow and Q on the basis of their real value, that is by
dividing the parameters by the general level of price in France.
(a) The introduction of a depreciation rate
This section includes a depreciation rate in the evaluation of the parameter

Cb in order to modify the factor Q in the investment regression. Why
implement such a procedure?
The purpose of the depreciation charge today is to spread the net cost of
a fixed asset (original cost minus sales proceeds) over its estimated useful
life. The aim is to ensure that the revenue arising during each accounting
period bears a fair share of the total costs incurred.
During the nineteenth century, depreciation involved no specific outlay of
funds and could be ignored in the short term; even when recorded, many
accountants apparently regarded depreciation as a segregation of profits rather
than an expense. Depreciation reserves, a subject that occasioned considerable
confusion in the nineteenth century, were viewed as a kind of surplus account.
In early accounting texts, the term depreciation was often used to describe
the difference between fixed asset valuations at two different dates.
Depreciation, when charged, was viewed principally as a means of earmarking
for retention resources which could be set aside to finance replacement, rather
than as a bona fide cost of production. The data we have for Cb fits these
parameters. Moreover, as early in the development of railways as 1852, it is
evident that very few people had a good conception of the relationship between
depreciation and net income. The methods used over the period for reflecting
depreciation in the accounts seemed to receive little support in railway circles.
An annual revaluation of properties was one of these; setting aside an annuity
91
Table 3.10 Summary statistics comparing Crédit Mobilier and non-Crédit

Mobilier firms when introducing a depreciation rate of 5 per cent*
Note: *Averages are calculated for all firms and all years. Standard errors are in
parentheses. The sample for the Crédit Mobilier firms is constituted of 7 firms over
1861–7 and 1861–75, and 6 firms over 1868–75. The number of affiliated firms is 8.
which would accumulate to the desired sum by the time replacement became
necessary was another. This explains the flat series for Cb. How can we
transform it, and what is the definition of the new parameter?
Let Cbk1861 represent the first element of the series which is defined as:
Cbk1861 = I1861/δ
where δ is constant and is chosen so that the series is slightly increasing.

The whole series is given by: Cbkt+1 = Cbkt (1 – δ) + It .
We call this new variable Cb1. Appendix F describes in detail the
construction of the series and the new values obtained for Q called Q1. Table
3.10 presents some descriptive statistics to analyse the modifications
produced by this alteration of the data.
If we compare these numbers with those obtained in the preceding section,
we note that they remain low. The problem then is to examine if this decrease
is realistic or not. To complete the analysis presented here, can we explore the
modification implied in the data set if we consider depreciation rates d both
greater and smaller than 5 per cent (see Appendix F).
(b) The investment regressions
The selected equation is modified as:
I/Cb = a + a Q + ?.CF/Cb
1 0 1 1 1
where:
I = gross investment at time t, measured simply as the change in the

gross value of depreciable assets (plant and equipment);
92
Table 3.11 Investment regression for the whole sample 1861–75
Note: Standard errors are in brackets. *CF.CM represents CF times Crédit Mobilier
affiliation: affiliated firms 7, independent companies 8, all firms 15.

CF1 = free cash flow, i.e. free surplus plus nominal allowances (mostly
depreciation);
Q = common equity including the depreciation rate given by Appendix F.
1
In the following section we present the results obtained and compare them
to those of the preceding section. As before, we consider three samples:
the whole sample 1860–75, the sample prior to the bankruptcy i.e. 1860–7,
and the years after the bankruptcy. The results are given in Tables 3.11,
3.12 and 3.13.
If we compare the results obtained here with those of the preceding
section, we can make the following comments. Let us first consider the
whole sample 1861–75. The coefficient for CF/Cb and Q are not very
1 1
different for affiliated and independent firms. The association with the
Crédit Mobilier, even if it was effective (largest value for the dummy
coefficient), was not important enough to cause a difference between the
two sets of companies.
As regards the sample 1861–7, a comparison with the results obtained by
Becht and Ramirez (1992:28) will be interesting. Let us recall them briefly.
Tables 3.14 and 3.15 only give us the coefficients obtained by running a
regression similar to the one presented in this chapter. We deliberately
choose not to report the standard errors, the aim being to analyse the
analogies and the differences between the two countries.
93
Table 3.12 Investment regression over 1861–7
affiliation: affiliated firms 7, independent companies 8, all firms 15.
Table 3.13 Investment regression over 1868–75
affiliation: affiliated firms 6, independent firms 8.
If we compare these figures with the ones we obtain, we remark that the
coefficients of the cash flows and the common equity are in exactly the
reverse order for the Crédit Mobilier:
• CF/Cb is close to 1 for the affiliated companies and close to zero for
1
independent firms, which means that the affiliated companies were
94
Table 3.14 Results for affiliated companies as presented in Becht and de Long’s
paper
Table 3.15 Results for non-affiliated companies as presented in Becht and
de Long’s Paper
liquidity constrained before the bankruptcy (Becht and Ramirez

conclude from the test they build that an association with a universal
bank limited the constraint towards liquidity);
• Q is very small (between 0.150 and 0.250), which does not represent a
1
significant parameter in the investment regression.
What kind of interpretation can we give to these results? The founders of

the Crédit Mobilier cannot be consider as real ‘bankers’, but rather as
promoters of enterprises. Their aim was the realisation of the highest profit
possible and the construction of a solid reputation for the financiers who
selected carefully the projects to which they granted credit.
If we consider the situation after the bankruptcy (sample 1868–75) and we
once again compare our results to those obtained by Ramirez and Becht, we note
a strong analogy between the two cases. After the failure, the results are what one
would expect if the Crédit Mobilier began to act as a big bank even though it had
lost its powerful position among French financiers. The coefficients of the cash
flows are negative for the affiliated companies, and Q conforms more closely to
1
the theory for the independent firms. We can conclude that the bankruptcy was the
origin of a complete change in the investment policy: the Crédit Mobilier began to
monitor effectively the firms with which it was involved. It will be interesting to
see if these conclusions are confirmed by the fixed effect model.
(c) The fixed effect model
I/Cb1 = α0 + β1Q1 + ß2CF/Cb1 + γt + ai + εit
95
Table 3.16 Fixed effect model for the whole sample*
are in brackets. Affiliated firms with the Compagnie Immobilière 7, independent
companies 8, all firms 15.
are in brackets. Affiliated firms with the Compagnie Immobilière 7, independent
where α0=intercept, αi=(N–1) firms specific fixed effects, and γt=(T–1) year
dummies and e is a residual term.
The results from these estimations are given by Tables 3.16, 3.17, 3.18
and 3.19.
If we analyse the coefficient we obtain for the cash flows, we notice that
the Crédit Mobilier firms were more constrained before the bankruptcy (CF/
Cb =0.112 for Crédit Mobilier firm over 1861–7) than after (where CF/
1
Cb =–0.103). This is the opposite of what is predicted by theory. The
1
collapse of the Crédit Mobilier should have created liquidity problems for
affiliated firms. The most remarkable result of Tables 3.16 and 3.17 is the
96
of the period. The regressions include yearly dummies and firm dummies. Standard
errors are in brackets. Affiliated firms 6, independent companies 8.
Table 3.19 Fixed effect model for all firms*
are in brackets. All firms with the Compagnie Immobilière 15.
difference that exists between a sample with the Compagnie Immobilière

and one without it. Let us consider, for example, those samples, 1861–75 and
1861–7, as regards the coefficient on CF/Cb . It seems all affiliated companies
1
were liquidity constrained but that the Compagnie Immobilière was much more
constrained than the others.
The results presented for all companies corroborate this analysis, except that
the difference between the set including the Compagnie Immobilière and the set
excluding it is not so strongly marked. As a whole, the introduction of a
depreciation rate in the data contributes to an increase in the value of CF dummy,
which represents the effect of an association with the Crédit Mobilier on liquidity
constraints. The pattern ‘affiliation causes liquidity constraints’ is strengthened.
97
(d) Conclusion
This econometric work provides some additional evidence suggesting that

the Crédit Mobilier seems not to have functioned like a Großbank. The
affiliated companies were more liquidity constrained before the bankruptcy
than after.
3.4 EXTENSIONS OF THE RESULTS:

DESCRIPTION AND INTERPRETATION
OF A ROBUSTNESS TEST
All the empirical work in this section is based on the assumption that the
depreciation rate is equal to 5 per cent. This choice can be seen as arbitrary,
but leads to a slightly increasing series for Q. To complete the analysis
presented here we introduce the modification implied in the data set if we
consider a depreciation rate d greater and smaller than 5 per cent. Appendix F
presents the figures obtained for the coefficient Q when introducing a
Table 3.20 Investment regression equation for affiliated companies and independent
firms over the whole sample*
are in brackets. Affiliated companies with the Compagnie Immobilière 7, independent
firms 8.
98
firms before the bankruptcy*
are in brackets. Affiliated companies with the Compagnie Immobilière 7, independent
firms 8.
depreciation rate of 3 per cent and 7 per cent. In what follows, the new
variables for Q and Cb are noted Q2 and Cb2 when δ=3 per cent, Q3 and Cb3
when δ=7 per cent.
The aim of this section is then to measure the modification induced by a
small difference in the depreciation when analysing the correlation between
investment, cash flows and common equity. For that purpose we concentrate
the econometrical analysis on the fixed effect model. The previous tests
presented in section 3.2(a) are not going to be considered, in order to
alleviate a fastidious examination of a large number of tables. The
regression we run considers the set of affiliated companies on the one side
and the set of independent companies on the other side over three samples:
the whole sample 1861–75, the situation before the bankruptcy 1861–7, the
event after the bankruptcy. We then evaluate the correlation between
investment and cash flows for the whole set of companies over the same
samples. The results presented in Tables 3.20, 3.21 and 3.22 successively
report the figures obtained for the two values of depreciation rates in order
to facilitate their interpretation.
99
firms after the bankruptcy*
are in brackets. Affiliated companies 6, independent firms 8.
Table 3.23 Cash flow parameters
Note: *CF/Cb1 stands for cash flows including a depreciation rate of 5%; CF/Cb2
stands for cash flows including a depreciation rate of 3%; CF/Cb 3 stands for cash
flows including a depreciation rate of 7%.
Some general comments can be made on this first set of results. In order
to summarise the arguments we construct a small table (Table 3.23)
reporting all the figures we obtain for the coefficient on cash flows.
100
As can be noted, the coefficients for the cash flows are very similar for
affiliated companies when the Compagnie Immobilière is excluded. whatever
level of deprecation rate is considered. The same argument can hold for the
independent companies.
As regards the coefficient on Q, the preceding tables exhibit very strong
similarities for every set of companies (affiliated companies with or without
the Compagnie Immobilière, as well as independent companies). The
constant term always has the same sign and the same magnitude. The same
conclusions follow for R2.
Some additional comments can be made on the fluctuations of the t-statistics
as regards the robustness of this test. In what follows we concentrate on the
whole sample (1861–75) and on the years before the bankruptcy (1861–7)
because the rest of our results exhibit no big change for the t-statistics. More
specifically we consider the set of companies excluding the Compagnie
Immobilière. The analysis is summarised by two tables (Tables 3.24 and 3.25)
which reproduce the coefficient on CF/Cb in order to appreciate the degree of
i
liquidity constraints for affiliated companies over these two period.
First, it is remarkable that the introduction or the exclusion of the Compagnie
Immobilière should have an influence on the interpretation of the liquidity
constraints that the affiliated firms had to suffer during these two periods. The
Table 3.24 Sample 1861–75 robustness results for Crédit Mobilier firms without the
Compagnie Immobilière
Table 3.25 Sample 1861–7 robustness results for Crédit Mobilier firms without the
Compagnie Immobilière
101
Table 3.26 Investment regression equation for all companies over the different
samples
of the period. The regressions include yearly dummies and firm dummies. Standard
errors are in brackets. Affiliated firms with the Compagnie Immobilière 7, independent
figures reported in these tables show that the coefficient on cash flows is more or
less significant according to the level of depreciation rate included in the
regression. As this situation does not represent the norm over the entire statistical
work presented here, we can say the results are robust: the sign and the amplitude
of the point estimates have not changed when the depreciation rate varies.
Let us now consider the whole set of firms for each sample. The results
are presented in Table 3.26. In order to compare the figures obtained for the
three values of d we construct two tables (Tables 3.27 and 3.28) which
summarise the results obtained for Q and CF/Cb . Some general comments
i i
will be made for the dummy coefficient and the R2 parameter. Let us first
consider the coefficient obtained for CF/Cb .
i
The values reported in Table 3.27 exhibit strong similarities for the three
values of d, except in one case, CF/Cb for all companies 1861–7 when the
3
102
Table 3.27 Cash flow parameters
Compagnie Immobilière is excluded. To complete the analysis we intend to

Table 3.28 Q parameters
study the parameter Q and to examine if the same particularity happens over
the same sample.
As regards Table 3.28 it seems that the similarities for this coefficient are
even stronger. If we now consider the dummy variable, we see that it always
has the same sign and the same magnitude. The same line of argument can
be made for the constant term.
For all firms, R2 is included between:
0.647 ≤ R2 ≤ 0.995
which means that the correlation between the variables is quite high.
As before, we add some indication as regards the variation of the t-
statistics, concentrating on the only sensible sample, 1861–7. Two
coefficients are taken into consideration: the cash flow and the dummy
variable, reported in Table 3.29.
The most surprising element of this table is the small value for the t-
statistics obtained for the dummy variable when the depreciation rate is
equal to 5 per cent. If this number is compared to the whole set of figures
reported in the table, it seems that this situation represents a marginal case
and does not alter the validity of our results.
103
Table 3.29 T-statistics for sample 1861–7 as regards the cash flow parameters
(a) Conclusion
The robustness test constructed here reinforces the preceding conclusion, in

the sense that no evidence is given that the Crédit Mobilier behaved over the
period like a Großbank. As before it seems that affiliated companies were
more liquidity constrained before the bankruptcy than after. This empirical
work enables us, however, to consider that the chosen value for d does not
have a greater influence over the results obtained in the test. In the statistical
analysis presented below we then concentrate on a medium value for d equal
to 5 per cent.
3.5 THE EXPRESSION OF ALL VARIABLES IN REAL

TERMS
As the figures presented above remain surprisingly low, an additional

transformation is done in order to express all the factors in real terms. This
means that we divide It by Pt which gives Bkt+1 in real terms. Q is then
computed as: value of shares/(Pt.Bkt). The new series can be described as:
Cbkt = Cbkt–1 (1–δ) Pt/Pt–1 + It.Pt/Pt–1.
The appendix describes in detail the series retained for Pt and the method
used to compute Q. Let us present some descriptive statistics to analyse the
modifications produced by this alteration of the data.
If we compare these figures with those obtained previously we note that
they do not conform to what could be expected.9 Taking them as correct the
point is then to evaluate the impact of this modification on the investment
regression.
I/Cb2 = α0 + β1Q2 + β2CF/Cb2 + γt + αi + εit
104
Table 3.30 Summary statistics comparing Crédit Mobilier and non-Crédit Mobilier
firms when parameters are expressed in real terms*
Note: *Averages are calculated for all firms and all year. Standard errors are in
parentheses. Affiliated firms with the Compagnie Immobilière 7, independent firms 8.
Table 3.31 Investment regression equation for the whole sample*
are in brackets. Affiliated companies with the Compagnie Immobliere 7, independent
firms 8.
where: α0=intercept, αi=(N–1) firms specific fixed effects and γt=(T–1) year
dummies.
The results from these estimations are given by Tables 3.31, 3.32, 3.33 and 3.34.
Some comments can be made on these four tables. Let us first consider
what happens over the whole sample and for 1861–7. The first observation
is that the coefficient on Q is larger for affiliated companies and the second
2
that the results obtained when the Compagnie Immobilière is part of the
sample are much different from the ones where it is excluded.
As regards the independent companies, the coefficients are not well
estimated. This difference appears more pronounced than in the preceding
test. One reasonable explanation could be the lack of data: missing values
are introduced for 1861–4.
105
Table 3.32 Investment regression equation for 1861–7*
the period. The regressions include year dummies and firm dummies. Standard errors are
in brackets. Affiliated companies with the Compagnie Immobliere 7, independent firms 8.
Table 3.33 Investment regression equation for 1868–75*
in brackets. Affiliated companies 6, independent firms 8.
Table 3.34 Investment regression equations for all firms
in brackets. All firms with the Compagnie Immobilière 15.
106
If we consider the period after the bankruptcy, the results do not

provide us with new information regarding the correlation among the
parameters.
What seems to be more interesting is the figures relative to all
companies. If the whole sample is compared with the years prior to the
bankruptcy, it will be noted that the Q coefficient is much smaller when
2
the Compagnie Immobilière is excluded over 1861–7 than when it is part
of the sample. It will be of interest to analyse the influence of the
bankruptcy over 1861–75.
For that purpose we introduce a new dummy variable we call
CF.AFFI.67 which measures the influence of an affiliation with the Crédit
Mobilier after the year of bankruptcy. With the interaction dummy which
takes the value 1 times cash flow when the corporation is affiliated to the
Crédit Mobilier, we multiply another dummy which takes the value 1 after
the bankruptcy, 0 otherwise. We run the regression for the whole sample
without the Compagnie Immobilière. The results are given in Table 3.35.
Let us compare the figures obtained here with the ones we have in the
preceding regressions (Table 3.6).
When we exclude the Compagnie Immobilière from the sample and
analyse the impact of the bankruptcy as regards the investment policy, we
see that, even if there is an effect (see the last column of the table) the
difference is either small (a decrease of 10.6 per cent in the coefficient on
cash flow can be noted for a bank whose assets were largely liquidated) or
non-existent (Q is increasing).
2
Let us consider the affiliated dummy itself. In the second regression the
sign of this variable becomes negative, reflecting the failure of the Crédit
Table 3.35 Regression for the whole sample without the

Compagnie Immobilière*
Note: *The dependent variable is investment relative to capital stock

at the beginning of the period. The regressions include year dummies
and firm dummies. CF.AFFI.67 represents CF times Crédit Mobilier
affiliation times dummy variable, 0 before 67, 1 afterwards. Standard
errors are in brackets.
107
Table 3.36 Comparison of cash flow and Q parameters
Mobilier. This again suggests that the failure of the Crédit Mobilier relaxed
liquidity constraints for affiliated firms.
All these figures prove that the Crédit Mobilier’s firms were more liquidity
constrained before the bankruptcy than after. The bank never managed to
supervise firms. As long as stock values were appreciating, the Crédit
Mobilier found it easy to participate in the stock market while diversifying
its activities. The bank participated in railway buildings in Europe,
international banking and French real estate. This multifaceted activity was
in part due to the Péreires’ very broad view of the goals of the Crédit
Mobilier, but it was also a response to the booming international capital
market that occurred after mid-1850.
The extraordinary degree of activity of the Crédit Mobilier meant that it
operated with few reserves, especially given that it distributed a substantial
part of its profit for nearly a decade. The absence of reserves made it
impossible for the bank to survive the economic slowdown that followed
1865 and relieved credit constraints for affiliated firms.
The next step of our argument will be to discuss the place of the Crédit
Mobilier in the banking system, in order to explain the minor effect on
companies when the Crédit Mobilier collapsed. We pay attention in our
analysis to the differences between France, Germany and the United States
over the same period.
108
4
THE SUPERVISORY ROLE OF

Some interpretations
The Crédit Mobilier has been, if I may be forgiven the term, a

promoter of business, working on the premise that present
interest almost invariably dominates future interest.1
Bonnet (1865:115)
La revue des deux Mondes, November 1865
4.1 Introduction
In this chapter we intend to use the different econometric results we have

obtained to highlight the theoretical questions they suggest. In Chapter 2, we
showed that over the period, prices were more volatile than perfect foresight
fundamentals relative to naive forecasts would have suggested. Thus, tests
based on market volatility ratios show traces of excess volatility in the
French stock market in the period before the First World War.
Concerning the fluctuations of the share prices of affiliated and non-
affiliated companies, the test constructed for the sample 1867–70
demonstrates that share prices for affiliated companies did not fall. My
tentative conclusion concerning the period 1867–75 was that an association
with the Crédit Mobilier did not have a great influence on the share prices
of the companies with which the bank was involved.
I then produced a series of statistical analyses concerning investment I,
cash flows CF and common equity Q for affiliated and non-affiliated
companies. As regards Q, we can say that before the bankruptcy, an
association with the Crédit Mobilier was correlated with high value. But the
figures we obtained in the Crédit Mobilier’s case prove that the participation
did not add as much value as in the Morgan Company’s case. The standard
interpretation suggests that the unaffiliated companies were not credit
constrained compared to the affiliated ones, and that affiliated firms were
more liquidity constrained before the bankruptcy than after.
These results give rise to four different questions. The first concerns the
supervisory role that the Crédit Mobilier had exerted over the firms with
which it was involved. In regard to this, we study two main points:
109
• the possibility of a financial crisis due to the Crédit Mobilier’s fall;

• the idea of a banking panic.
More specifically, if this supervisory role were effective and necessary,

as the Crédit Mobilier went bankrupt, a financial crisis would have
followed for the affiliated companies. Except for the bankruptcy of the
Compagnie Immobilière, however, nothing spectacular happened. To
support our argument we compare the example of the Ohio Life
Insurance and Trust Co. in the United States to the statistical results
concerning the fluctuations of the share prices of affiliated companies
in the Crédit Mobilier’s case. We will then draw conclusions on the
qualification of financial crisis in the Crédit Mobilier’s case.
The second issue concerns the consequences of the bankruptcy for
the banking sector: as the Crédit Mobilier played a great part in the
industrialisation of France (providing a large amount of liquidity) we
can ask why its failure did not provoke a banking panic. In regard to
this, two definitions of banking panic are presented and applied to the
Crédit Mobilier’s case to see if this crisis can be defined as a liquidity
crisis. The empirical research concerning investment, common equity
and cash flows will inform our argument.
The question then asked is this: what is the most efficient way to
control a firm—an interaction between firms and banks (which means
substantially a bank-firm association) or a separation of banks and
firms (as presented in Chapter 1)? In other words, should banks
supervise, or just choose whether to loan or not? In the latter case,
supervision is exerted by the credit policy of the bank toward
enterprises. The power is derived from the firms’ need for funds to
realise their productive investments. This will constitute the second
point of our discussion.
The third point of our argument will consist of a comparison
between France and Germany. More specifically, some German
Großbanken (in particular the Darmstädter) were modelled on the
Crédit Mobilier. The research that Marco Becht and Carlos Ramirez
(1993) have carried out on these banks demonstrates the efficiency of
the system in contrast to the French experience. It will then be
interesting to compare the two banking systems in order to explain the
differences in the results obtained.
The last point consists of a discussion about the role played by the
shareholders (if any) when the Crédit Mobilier went bankrupt. In particular,
we want to assess the rationality or irrationality of their behaviour when the
firm was facing financial difficulties.
110
THE SUPERVISORY ROLE OF THE CRÉDIT MOBILIER
4.2 THE SUPERVISORY ROLE OF THE

CRÉDIT MOBILIER
If the Crédit Mobilier were truly a great bank, its bankruptcy would have
provoked financial difficulties not only for the enterprises with which it was
involved but also for the other banks on the market.
(a) Did the Crédit Mobilier’s bankruptcy create a

financial crisis?
According to Kindleberger (1984) and Miskin (1991), a financial crisis

involves either sharp declines in asset prices, or failures of large
financial and non-financial firms. To support this argument, Miskin in his
article entitled ‘Asymmetric information and financial crises’, takes the
example of the Ohio Life and Trust Company, a major financial
institution with substantial investments in western land and railroads as
well as in commodity futures, which failed on 24 August 1857. The most
interesting part of this analysis is that the timing of events in the panics
of 1857 seems to fit an asymmetric information interpretation of the
financial crisis as in our preceding study of the relationship between
bank, manager and shareholders. Rather than starting with the bank panic
of October 1857, the disturbance to the financial markets seems to have
arisen several months earlier with the rise in interest rates, the stock
market decline, major failures of financial firms and the widening of the
interest rate spread.
The asymmetric information analysis provides an explanation of how
the financial crisis could have led to a severe economic downturn. The
rise in the interest rate and the stock market decline, along with the
failure of Ohio Life and Trust Co. which increased uncertainty, would
magnify the adverse selection and agency problems in the credit markets.
Indeed, the stock market crash might have been linked to the general rise
of interest rates, which would have lowered the present discounted value
of future income streams. In this case, the panic of 1857 can be viewed
as a financial crisis. The net result of the increase in the adverse
selection and agency problems is that investment activity and aggregate
economic activity would decline, causing low expectations of further
economic contracting and business.
This example is also interesting because the period during which it
happened is not so far away from that of the Crédit Mobilier’s bankruptcy.
If we compare the results given by the volatility ratios and by the test on
the fluctuations of share prices for affiliated firms, we find that there is no
evidence that the economic situation described above can apply to the
French case. If the price-dividend ratio decreased during 1870–80 (just
111
after the bankruptcy of the Crédit Mobilier), it is impossible to tell if the

failure of the bank has anything to do with this movement. The Franco-
Prussian War, the Commune of Paris and the Revolution of 1870 caused
considerable economic disturbances over this period. However, the general
crisis of 1866 that occurred in England, Italy and Spain was not a major
event in France and thus had no influence on the Crédit Mobilier’s
bankruptcy.
Some authors (e.g. Friedman and Schwartz (1963)) have established the
close connection of financial crisis (especially the bank failures) with
changes in real output. Such a phenomenon did not happen in 1867. In that
sense, we can affirm that the Crédit Mobilier’s bankruptcy did not create a
financial crisis.
The second strong argument concerns the relation between banks,
managers and shareholders when financing a productive project.
According to Stiglitz (1985), raising capital through banks results in
more effective control over capital than raising it through an equity
market.2 Because both adverse selection and moral hazard are inevitable,
at least in realistic loan markets, credit will often be rationed, implying
that firms will typically face borrowing constraints which really do limit
their investment opportunities from time to time. Moreover, the tightness
of these borrowing constraints for any particular firm will clearly depend
on the amount of collateral that the owner-manager is able and willing to
offer, and not just on the present discounted value of what he can afford
to repay without surrendering collateral in the worst possible future his
firm can face.
Stiglitz points out the interdependence of the different partners in the
firm:3 the shareholders and the bondholders all benefit from the control
of the banks. In fact, the intention of banks to grant credit to a firm is an
effective positive signal. The actions of firms which try to find funds on
the capital markets because they cannot obtain them from their bank
could be interpreted in an adverse way. According to this scenario, when
the Crédit Mobilier went bankrupt, the share prices of the affiliated
companies should have fallen: this would have proved that the Crédit
Mobilier added some value to the shares for affiliated companies. But
nothing like this happened to the share prices of these companies in the
year of the bankruptcy. It could be argued that the investors had
anticipated the fall of the bank and had already diversified their
investments and discounted the effect it would have on affiliated firms.
In his book La Naissance d’une Banque: le Crédit Lyonnais, Bouvier
remarks that a great number of firms which were financed by the Crédit
Mobilier before the bankruptcy turned to the Crédit Lyonnais for funds
afterwards. In particular the Compagnie Transatlantique, which was one
of the largest companies of the Crédit Mobilier, requested funds as early
as 1865. Another explanation for the stability of the share prices is the
112
intervention of the Banque de France which prevented a crisis from

developing; this encouraged other banks to lend funds to the companies
affiliated with the Crédit Mobilier, except for the Compagnie
Immobilière. This reinforces the idea that the failure of the Crédit
Mobilier did not really induce a financial crisis.
The third argument concerns the role of financial institutions in the crisis.
Was the Crédit Mobilier’s bankruptcy in 1867 an isolated case, or was it a
reflection of the fragility of capitalism over the period?
The first remarkable element is that the bankruptcy of the Crédit Mobilier
had nothing to do with the general crisis of 1866 in England and Italy. As
Kindleberger pointed out in his book entitled Manias, Panics and Crashes,
the crisis occurred in France in 1864 and was solved by an extension of the
maturities of bills. Moreover, the plots of share prices of ten affiliated
companies presented in the second chapter of that book prove that no
particular event could be noted for this year. Thus the Crédit Mobilier crisis
seems to have been an isolated event.
Secondly, financial institutions stand between firms and households. To
a large extent, the liabilities of firms are owned by financial intermediaries
and the assets of households are largely liabilities of financial
intermediaries. These intermediaries (here banks) are self-, or profit-,
seeking institutions. Of these profit-seeking organisations, one set plays an
exceptionally delicate role in capitalist economies. This set consists of the
investment bankers who either as brokers—who bring buyers and sellers
together—or dealers—who take financial liabilities into their own
accounts—act as midwives to young companies and finance their
continuing operations. Essentially, these operators have superior
knowledge about those customers who need financing (they have a need
for funds) and those customers who have a need for outlets in which
money can be placed. They turn this private knowledge of the conditions
under which funds are desired and the conditions under which funds are
available to their own advantage, even as they perform the social function
of selecting the investments that the economy makes. In the Crédit
Mobilier’s case the problem was that the bank had strongly focused on
short-term profit-seeking and neglected its selection role. Its desire to
create a new style of credit banking prevented the Péreires from carefully
studying the projects in which they participated.
These financial intermediaries are of critical importance in determining
the values attached to collections of capital assets held by firms. In a
balance sheet, the difference between the sum of the values entered for
capital and financial assets and the value of debts on the liability side is the
book value of the owner’s interest in the firm. Dividing the book value of
the owner’s equity by the number of outstanding shares yields the book
value of a share. However, for the main companies in a large economy there
is a large market for equity shares and this market value may be less than,
113
equal to or greater than the book value. A major consideration in decisions

to invest is that the market evaluation of the capital assets needs to exceed
the supply price of the required investment goods.
One consequence of the introduction into the markets of these layers
of profit-seeking organisations, which determine the value of financial
instruments, is that the value imputed to capital assets can and does vary
independently of the cost of investment goods. Furthermore, the extent to
which internal funds are available to finance investment depends upon
the excess of anticipated cash flows over the amount needed to service
liabilities that were issued to finance such acquisitions in the past. This
scenario describes perfectly the situation of the Crédit Mobilier at its
foundation. The power it had on the financial market influenced the
value of assets and was at the origin of the speculative movements
beginning in 1864.
Thus capitalist techniques of valuing capital assets, the market determination
of liability structures, and the possibility of sharp increases and decreases in the
market price of capital assets and financial instruments lead to systematic
increases and decreases in the price of assets relative to the price level of current
output. Once current profits fall to the extent that cash flows to highly indebted
operations are insufficient to meet commitments or liabilities, then the pressure
of the need to validate debts and to meet withdrawals by depository institutions
leads to a proliferation of attempts to make positions by selling out. The result
can be a sharp fall in asset values. These considerations cannot be applied to the
Crédit Mobilier’s case, as was shown in Chapter 2.
The literature relative to investment constraint also suggests that a related
transmission mechanism may operate through the banking sector: a
reduction in bank liquidity makes it difficult for firms to obtain capital. As
Bernanke (1983) has shown, the large fall in bank liquidity may help to
explain the depth and the persistence of the Great Depression. This,
however, does not apply to the Crédit Mobilier’s case. The fall of the bank
did not provoke any financial crisis. This leads us to analyse the position of
the Crédit Mobilier in the banking sector in France.
(b) Did the bankruptcy of the Crédit Mobilier provoke a

banking panic?
The term ‘banking panic’ is often used somewhat ambiguously and, in

many cases, synonymously with events in which banks fail, such as
recessions and stock market crashes. Historically, bank debt has
consisted largely of liabilities which circulate as a medium of exchange,
such as bank notes and demand deposits. The contract defining this debt
allowed the debtholder the right to redeem the debt (in hard currency) on
demand at par.
114
We define a banking panic as follows. A banking panic occurs when bank

debtholders as a whole, or many banks in the banking system, suddenly
demand that banks convert their debt claims into cash (at par) to such an
extent that the banks suspend convertibility of their debt into cash.
One theory of banking panics is based on identifying the conditions
under which bank depositors would rationally change their beliefs about
the riskiness of their bank. The core of the theory is that banking panics
serve a positive function in monitoring banks in an environment where
there is asymmetric information about bank performance. In an
environment with asymmetric information, a panic can occur as follows.
Bank depositors may receive information leading them to revise their
assessment of the risk of banks, but they do not know which individual
banks are most likely to be affected. Since depositors are unable to
distinguish individual bank risks, they may withdraw a large volume of
deposits from all banks in response to a signal. Banks then suspend
convertibility and a period follows during which the banks themselves
sort out which banks among them are insolvent. Indeed, it is possible to
view panics as a means for depositors to force banks to resolve
asymmetries of information through collective action (i.e. monitoring and
closure) The efficiency of this mechanism derives from a supposed
comparative advantage that banks possess.
This last analysis explicitly utilises the notion of liquidity to define the
banking panic. This parameter has been studied when analysing the
relationship between investment and cash flow. In particular, we can ask the
question: was there an increased sensitivity of investment to cash flow by
the Crédit Mobilier’s affiliated companies in the period preceding and
following the bankruptcy of the bank, related to the changes in the financing
patterns that occurred at the same time?
Answering this last point will enable us to give a positive answer to the
question as to whether firms had anticipated the difficulties the bank was to
face later. Is the following hypothesis as stated in Hoshi, Kashyap and
Scharfstein (1991), true in the Crédit Mobilier’s case: ‘Bank relationships
relax liquidity constraints’?
The answer will be ambiguous, for different reasons invoked in the paper.
As Q is greater for affiliated firms than for non-affiliated ones, the bank-
firm relationship seems to be profitable. However, as nothing really
important happened during the year of the bankruptcy, we can add the
following question: ‘Do firms care who provides their financing?’ If the
manager of a firm has to choose between banks and shareholders to finance
a project, the answer will be positive, because the origin of the funds is
different (the information cost in particular is different between bank and
shareholders). If the manager has to choose between different banks the
answer is a function of the debt policy and of the power of these banks to
generate a good reputation for the firm. This second situation is illustrated
115
by the Crédit Mobilier case. In particular, despite the facility to obtain

liquidity from the Crédit Mobilier, some firms may have preferred more
control to more flexibility. They may have chosen another bank to signal
their honesty.
(c) Conclusion
The results for the Crédit Mobilier’s case are different from the
German Goßbanken or the J.P.Morgan Company. Becht and Ramirez
conclude that ‘mining and steel companies affiliated with one of
Imperial Germany’s “Great Banks” were not liquidity constrained’.
Their paper provides evidence suggesting that this limit can be very
high. The power of the Großbanken seems to have helped the
development of the mining industry by alleviating the impact of capital
market imperfections.
The Crédit Mobilier’s case, however, proves the fragility of banks’
power; it shows that the supervisory role can be very limited and
dispensable. The Crédit Mobilier was vulnerable to speculative
movements. It is possible that the hostility of famous financiers like the
Rothschilds contributed to its continuously decreasing power in financial
circles.
4.3 ADVANTAGES AND DISADVANTAGES OF AN

ASSOCIATION WITH A BANK FOR THE
FINANCING OF PRODUCTIVE PROJECTS
The bank-firm association can be debated especially as regards the results

concerning the correlation of investment cash flows. More particularly, the
fixed effect investment model proves that the interaction coefficient (which
takes the value one times the cash flow when the corporation is part of the
Crédit Mobilier sample) is very small and the standard errors are large
except for the years 1861–7. The extensions of the basic model exhibit a
liquidity constraint that is larger for affiliated companies than for non-
affiliated ones over 1861–7.
The efficiency of a bank-firm association is, then, the object of
discussion. We proceed here in two ways:
1 a general analysis of the bank-firm association;

2 as in our sample the non-affiliated companies were in a relationship with
the Crédit Lyonnais, a comparison of the behaviour of the two banks is
drawn in order to explain the puzzling effects on investment policy of
firms affiliated with the Crédit Mobilier’s case.
116
(a) The (dis)advantages of an association between a bank

and firms: a theoretical analysis
To begin our thesis, let us give some of the background to the banking
sector in the nineteenth century. During the period of this study there were a
number of sounder financial institutions than the Crédit Mobilier, which had
an enormously swollen industrial portfolio greatly exceeding its capital, and
which was dependent on favourable developments on the stock exchange to
support the continuation of its activities. Universal banks such as J.P.Morgan
and the Großbanken established the closest possible relations with their
client industrial enterprises. Universal banks accompanied an industrial
enterprise from the cradle to the grave, from establishment to liquidation
through all the vicissitudes of their existence. Through the device of
formally short-term, but in reality long-term current account credits and
from the development of the institutions of the supervisory boards to the
position of the most powerful organs within corporate organisations, the
banks acquired a formidable degree of power over industrial enterprises,
which extended far beyond the sphere of financial control into that of
entrepreneurial and managerial decisions. This argument is strongly
supported by evidence presented by de Long and Becht, Ramirez and Becht
and Ramirez.
Association with the Morgan company or with one of the Großbanken
appears to have added value to enterprises. These were banks that took an
active part in the decisions related to choosing a project, in order to defend
their reputation in the financial markets. In the Crédit Mobilier’s case this
argument is problematic, for one essential reason: as noted above, the bank
extended its role to that of a shareholder (and often the principal
shareholders of the firm). Therefore it suffered the same risks as an ordinary
shareholder.
The answer to the question: ‘Is it preferable for banks and firms to be
associated or not?’ is neither always yes nor always no. For Aoki (1988) the
bank can be a monitoring agent. Two essential reasons can be given to
support this proposition:
1 The firm can raise investment funds by borrowing from the bank at a
low cost, and does not have to rely on equity financing. Therefore, its
strategic business decisions (corporate policy) can be made in the
interest of individual stockholders.
2 Individual stockholders of the firm are underprivileged in the sharing of
economic returns to the firm, as they receive only a very small fraction
of profit as dividend.
Hence the main bank is in a position to be briefed about the company’s

general business and affairs in the capacity of a major stockholder and is
117
able to scrutinise the company’s strategic plan in the capacity of a major

lender. It often sends its representative to the company’s board of directors.
As point out by Wrück (1990), highly concentrated bank borrowings are
also associated with superior performance. For Japanese firms, close bank
lending relationships and equity ownership go together. On average, the
largest lender holds 23 per cent of the firm’s bank debt and 4 per cent of its
equity. The implications of the results in Hoshi et al. for US firms are
difficult to determine because the financial structures associated with
superior performance by Japanese firms in financial distress are illegal in the
US, where the Glass-Steagall Act prevents banks from holding large equity
positions in firms, including the firms that borrow from them.
Thus only the main bank-cum-major stockholder has sufficient ability to
closely monitor its customer companies. This situation corresponds to the
Crédit Mobilier’s Case. However, did it exercise its expertise and
informational advantage only for its own gain?
One possible scenario is as follows: in the normal course of affairs and
business of the portfolio company, the individual stockholder, obviously
disadvantaged in his/her capacity to gather information and monitor the
management of the company, may feel more secure if the main bank closely
and responsibly monitors management of the company to reduce the risk of
bankruptcy. One may thus say that the individual stockholder is willing to
delegate the monitoring function to the main bank-cum-major stockholder.
The deviation from share price maximisation by the company at the expense
of individual stockholders and in the interests of the bank may, in part, be
regarded as an ‘agency fee’ paid by the individual stockholders to the bank
for that service.
If the above scenario is correct, then it is vital for the main bank to
maintain its reputation as a competent and responsible monitor. This may
explain an aspect of the rescue operation put forth by a main bank when it
discovers serious financial difficulties in a company. The problem in the
Crédit Mobilier’s case is that this argument was pushed to the extreme.
When the Compagnie Immobilière faced difficulties, the bank tried
everything to rescue the firm, eventually provoking the bankruptcy of the
bank itself.
This proves that the bank-firm association contains some risks and
hence disadvantages. However, the Crédit Mobilier presents some
peculiarites: as soon as the bank chose to rescue the Compagnie
Immobilière whatever the risk, it stopped acting as a bank. This company
had played a special role in the Crédit Mobilier’s story. This vulnerable
young firm needed big loans. In anticipation of high returns, the Péreires
chose to invest freely. This attitude was not totally unreasonable: other
banks in the same period granted risky loans to support the financing of
new companies. The problem as regards the Crédit Mobilier was that the
profit took longer to be realised than had been foreseen.
118
The Crédit Mobilier had attempted a new style of banking, but never
managed to apply it correctly: it still stuck to the old style, which consisted
of holding 51 per cent of the assets of a company in order to participate
actively on its financing. The companies affiliated to the Crédit Mobilier
were joint stock companies managed like limited partnerships, which
corresponded to the French tradition.
Concerning the advantage for a firm of an association with a specific
bank, Mayer (1988) suggests that there exist systematic differences in
performance between the financial systems where banks play a primary
role and those where the banks are not active in the same way. These
differences must be explained by the mechanisms that each system
provides to suppress the moral hazard in relations between entrepreneurs
and financiers. More specifically, Mayer affirms that the systems
proposed by the German or Japanese banks enable the firms and the
banks to sustain long-term contracts which in turn enable them to extract
the surplus associated with these contracts and to reduce moral hazard.
The answer to the question of the (dis)advantage of bank-firm
association will substantially depend on:
• the proportion of assets held by the bank: are the decisions taken by
the shareholders’ assembly the expression of the majority or, as in the
Péreires’ case, the will of the founders? The results of our tests prove
that the second alternative is more realistic in the Crédit Mobilier’s
case. In fact this argument leads us to analyse the position of
shareholders when a bank holds a large amount of assets in a firm.
Considering the Crédit Mobilier’s case, it seems that the role they
played was limited. We will discuss this idea in the next section when
comparing the French case to the German one;
• the expectations of financiers on the possible profitability of the firm
(e.g. confidence on the potential revenue for the bank in the future); this
reflects essentially the bankers’ point of view, whose aim is to make
profit;
• the power the bank exerts on the financial market (e.g. capacity to solve
the capital market imperfections). This last point corresponds essentially
to the situation of the German Großbanken, in which case an association
was profitable on average: for example adding value to the enterprise’s
shares.
Conclusion
It is relevant to point to the stress laid by Saint Simon and his followers
upon industrialisation and the great task they had assigned to banks as an
instrument for organisation and development of the economy. This, no
doubt, appealed greatly to the creators of the Crédit Mobilier, who liked to
119
think of their institution as a ‘bank to a higher power’ and of themselves as

‘missionaries’ rather than bankers.
Despite the collapse in 1867, the Péreires profoundly influenced the
history of banking in Europe from the second half of the last century
onwards. As Gerschenkron (1962:13) noted, the number of banks in
various countries created in the image of the Péreires’ bank was
considerable. However, more important than their imitation was the
creative adaptation of the basic idea of the Péreires and its incorporation
in the new type of bank, the universal bank, which in Germany, along
with most other countries on the Continent, became the dominant form
of banking. The difference between banks of the Crédit Mobilier type
and commercial banks in the most advanced industrial country of the
time (England) was very great. The German banks, which may be taken
as a paradigm of the type of the universal bank, successfully combined
the basic idea of the Crédit Mobilier with the short-term activities of
commercial banks.
Over the period, the Crédit Mobilier had acted more as an investor than a
banker which supervises and controls the firms with which it was involved.
This justifies the idea that the Crédit Mobilier essentially provided liquidity
to the enterprises and did not supervise them. A bank-enterprise association
is then profitable for both parties, so long as both play an appropriate role.
The bank must screen companies according to their expected returns; a firm
can expect additional value from a bank in terms of the power it has to
solve capital market imperfections.
(b) A comparative analysis of the Crédit Mobilier and the

Crédit Lyonnais over the period
When presenting the methodology of the investment regression we have

stressed the importance for all companies of their sources of financing. On
the one side, the affiliated companies were obviously involved with the
Crédit Mobilier, on the other side the -affiliated ones were financially
related to the Crédit Lyonnais. The statistical analysis proves that the non-
affiliated companies were not dramatically credit-constrained. As both
samples were affiliated with a bank, it will be interesting to compare the
finance policy in order to interpret the results of Chapter 2 and 3. With this
aim, we ask four questions:
• Were both of the banks merchant banks?

• Were the specific activities of the two banks similar?
• What were the characteristics of the credit policies of the two banks?
• Were the dividend policies similar for the two establishments?
120
The definition of the two banks

The argument developed here is based on a definition taken from the
Encyclopaedia Universalis (1985). The way the Crédit Lyonnais was
created was very different from the Crédit Mobilier. The Crédit Lyonnais
was a limited liability company 4 founded for a period of thirty years
from 1 July 1863 and whose main office was fixed at Lyon. It was
founded first to receive only deposits and to use them to grant, in the
very short term, credit to industrial establishments. Unlike the Crédit
Mobilier, no long-term loans or equity holdings were contemplated. The
objective of the Crédit Lyonnais is set out as follows by its founder,
Henri Germain:
The cash currently held in our bank was in great part hidden in the
drawers, without any interest for those who kept it, without any utility
for French society.5
The aim of the Crédit Lyonnais was quite different from the Crédit Mobilier.
In 1854, Isaac Péreire had declared that the Crédit Mobilier was ‘a limited
partnership’ for industries, 6 a lending bank and a bank of issue. The
ambitions of this bank were wider than those of the Crédit Lyonnais. This
diversification made the Crédit Mobilier powerful, but was also at the origin
of its difficulties. Specification of its different activities will enable us to
understand the functioning of the establishment.
The specific activities of the two establishments

As we have shown with the statistical tests presented in the two preceding
chapters, the Crédit Mobilier’s main activity was to provide liquidity to
firms in order to finance productive projects during the period of French
industrialisation. For the Crédit Lyonnais, two types of activity could be
distinguished:
• participation in the productive investment (like the Crédit Mobilier)

which represents the ordinary transactions; in other words the ordinary
affairs including overdrafts on current accounts;
• financial participation such as government loans, advances to foreign
government and huge investments in large companies which necessitated
a huge amount of liquidity. These were quite limited.
Table 4.1, after Bouvier (1968), illustrates this latter point. Bouvier asserts
that during its early years, the Crédit Lyonnais was a kind of readjusted
Crédit Mobilier, less ambitious, more prudent and more solid.
If we study Table 4.1, we note that the ‘big deals’ only exceptionally
121
Table 4.1 Financial participation of the Crédit Lyonnais in ‘big deals’
represented more than half of the total transactions of the bank. This
enabled the bank to avoid liquidity problems during the French financial
crises of 1870 and 1875. The Crédit Mobilier, on the other hand, was
closer to what Sayous (1907) calls a financial firm. Talking of these
firms he said:
But they (the financial firms) generally play a more active role: they
hold a block of shares, prepare the market or wait for any favourable
circumstance and pass the assets to the public; they themselves
transform the private enterprises into limited companies from which
they later issue certificates. They go as far as to create industries or
to organise business using an idea which was suggested to them.
They are then the brokers, the traders of this modern merchandise
called stocks: they facilitate and even organise the contact between
savings and those who need it.7
This definition describes the Crédit Mobilier’s range of activities. Sayous

adds that generally these financial institutions took a speculative attitude
to the stock market. Large participations in firms and risk-taking were
the order of the day. This can partly explain the empirical results
obtained in Chapters 2 and 3: the affiliated companies were liquidity
constrained before the bankruptcy because of a defensive attitude
towards the Crédit Mobilier’s credit policy. Such an argument will not be
valid in the Crédit Lyonnais’s case. This is the argument we intend to
discuss now.
122
The credit policy of the Crédit Lyonnais

This point is the area of most significant difference from the Crédit
Mobilier. As we have already pointed out, the Crédit Mobilier granted credit
to companies with which it was involved without asking for any collateral.
This might seem to have been totally unreasonable when we consider the
traditional economic view that the provision of collateral is a good signal for
the bank as regards the quality of a project and the reliability of the
enterprise which offers it. My hypothesis is that the Crédit Mobilier’s
practice has to be considered in the context of a country which was at the
beginning of its industrialisation and where the notion of risk was totally
different. Investors and bankers had to accept risks in the short term in order
to make profits in the long term.
The policy of the Crédit Lyonnais was totally different to this. The most
significant example is the establishment as early as 1870 (that is, only seven
years after the foundation of the establishment) of a department in charge of
the financial study of the different customers of the bank. Its duty was to
establish secure information, brought regularly up to date, concerning the
financial circumstances of the most important companies in France and
abroad. However, it is perhaps significant that this department was
established in 1870, after the bankruptcy and liquidation of the Crédit
Mobilier. Probably the Crédit Lyonnais was responding to the experience of
this failure by adopting a more prudent attitude.
The policy towards the distribution of dividend

The Crédit Mobilier was in the habit of distributing large dividends to its
shareholders. I will support this claim in more detail below. The point here
is to stress the difference between the Crédit Mobilier and the Crédit
Lyonnais towards dividend policy. Henri Germain applied the theory of non-
distribution of the exceptional benefits. In fact, he introduced to the bank a
technique which was close to industrial self-financing, distinguishing
between normal benefits arising from the short-term transactions (the
classical banking services), non-distributable benefits (statutory reserves,
increase of capital, etc.), and exceptional dividends. Of the total profit
realised by the bank, only a small percentage was distributed to the
shareholders: around 15 per cent. This illustrates the determination of the
Crédit Lyonnais to maintain its solvency.
(c) Conclusion
The points outlined above have a bearing on the validity of the econometric
results set out earlier. All the statistical analysis has to be interpreted in the
123
context of the Crédit Mobilier’s failure. The analysis relates to movements

in investment, share prices and cash flows for affiliated and non-affiliated
companies. Recent articles have explored the empirical relationship between
corporate financial structure and investment.8 Fazzari, Hubbard and Petersen
(1988) discussed the idea according to which the supply of investment
finance is not perfectly elastic for firms that face asymmetric information
problems in financial markets. They find that investment is more sensitive to
cash flows for firms that consistently retain a larger fraction of their
earnings.
Hoshi, Kashyap and Scharfstein, on the other hand, show that investment
expenditures of Japanese companies affiliated to a bank (i.e. a member of a
keiretsu) display almost no sensitivity to cash flows, in contrast to the
investment expenditures of firms which are not so affiliated. This finding
supports the hypothesis that companies that are not affiliated to a bank tend
to be more liquidity constrained (and face a larger gap between external and
internal costs of funds) than those which are. The absence of any
fluctuations for affiliated companies is a quite surprising fact that we have
already discussed in the preceding section.
As the non-affiliated Crédit Mobilier’s companies were involved with the
Crédit Lyonnais, the preceding explanation of the attitude of the bank as
regards credit and dividend helps to explain why the non-affiliated
companies were not credit constrained.
So, despite the small number of companies surveyed, we can advance the
hypothesis that the analysis of a bank failure illustrates the capacity of firms
to diversify their financing to avoid liquidity constraint. J.Bouvier provides
evidence that supports this argument: after the bankruptcy the Crédit
Lyonnais bought the insurance companies. The Rothschilds, who were
investors in two railways companies (Nord and Quest) contacted the Société
Générale to provide liquidity for further investment.
4.4 THE CRÉDIT MOBILIER BANKS: A COMPARATIVE

ANALYSIS BETWEEN FRANCE AND GERMANY
The next point we are going to consider is the divergence of the results
obtained in the French and the German cases. As noted at the beginning of
this chapter, the results obtained for Germany and France are very different.
The aim here is to give some interpretative reasons for this difference. More
specifically, we examine if the type of participation of the Großbanken in
firms followed the same model as the Crédit Mobilier.
The financial commitment of every ‘mixed bank’ consists in the
promotion and financing of industrial firms (short- and long-term financing:
the long-term corresponding to the renewal of short-term credit and
advances). The German financiers were real entrepreneurs, facilitating access
124
of firms to the stock exchange. As Tilly (1986) mentions: ‘the German

banks represented and reinforced not only the supply but also the demand
for the investment funds’. The development of large firms and cartels was
linked closely to the changing role of the banks in the German industrial
system. The large corporate banks, which had appeared in the third quarter
of the century, had first played a central role in the financing of industry in
the boom of the 1850s and in the years around 1870. However, from the
beginning of the 1890s, the co-operation between banks and industrial
concerns increased steadily.9 Long-term credit became the main basis of
banks’ relations with industry, to the extent that the expansions, mergers and
conversions from private to joint-stock companies became more important
and more frequent than the foundation of new companies.10 The issuing of
shares and bonds, which was performed by the large banks for industrial
concerns, was usually only the second step after co-operation had been
established on a long-term basis. The banks sought to monopolise the
financial arrangement of their industrial concerns and to service them in
various way by developing a comprehensive policy towards them and being
prepared to accept short-term financial losses in order to secure their long-
term co-operation.
How can we explain the difference between France and Germany? First,
as mentioned in the above review of the relations between banks and firms,
the main period of development occurred in Germany around 1870 and in
1890, that is after the Crédit Mobilier’s experience and after the crises that
had occurred between 1850 and 1870.
Secondly, the role of the central bank was stronger in Germany than in
France over the period. In particular, the Imperial Bank imposed strict rules
concerning the issue of bank notes and the emission of bonds on the
financial markets. On the other hand, the opening of discount window as a
role of a lender of last resort and a favourable policy as regards discount
factors avoided liquidity problems for banks (Tilly 1986a, b).
Thirdly, German banks had less liquidity than the French banks (Sayous
1899). In a normal context, such a situation represents an advantage because
speculative movements have less power. But, as stressed by the author, this
placed the German banks in a position of suspension of payment if any
crisis were to occur.
Fourthly, the participation of the German banks in the financing of
enterprises was totally different from that of the Crédit Mobilier. As the
German capital market was less developed than the French one, a credit had
a stronger implication for firms than in a French context. In most cases, the
Crédit Mobilier held the major part of the debts. In the German case, the
situation was different. In particular, the supervisory board (Aufsichtrat)
elected by the annual general meeting of the shareholders (or their
representatives) was obligatory by law from 1870 and was strengthened by a
law of 1884. Its functions were not limited to the appointment and
125
supervision of the executive board (Vorstand), it could also make or

influence the most important strategic decisions (especially investment
decisions) at its quarterly or monthly meeting. However, the banks could
threaten to withdraw credit, without being requested on the supervisory
board. In addition, to hold a major part of the equity was a bad signal for a
Großbanken. In such a case, the shareholders’ annual general meeting
played a relatively peripheral role.
To illustate the efforts of Großbanken to control their clients, let us quote
Jeidels (1929:126) when he reported a letter which the Dresdner Bank wrote
to the Executive of the North West German Cement Syndicate on 19
November 1900:
According to the notice published by your Company in the

Reichsanzeiger of the 18th, we have to reckon with the possibility
that the decisions will be taken at the General Meeting held on the
30th which are likely to introduce changes in the scope of your
business of a kind not agreeable for us. On this ground we are
regretfully obliged herewith to withdraw the credit granted to you;
accordingly we ask you to make no further drafts on us and at the
same time politely request that you will repay the balance due to
us, at the latest by the end of this month. Should, however, no
decisions of this kind be taken at the General Meeting in question
and safeguards be offered us in this respect for the future, we
declare ourselves very ready to enter negotiations with you with a
view to the granting of a new credit.
If all these arguments seem to stress the prudent and rational attitudes
as regards credit policy, one point should be mentioned which concerns
the question of advances to companies. Sayous (1899) affirms that the
economic development of Germany, in establishing a strict distinction
between issuing banks and ordinary banks had, as a consequence,
facilitated the credit and financial operations; the relationship between
banks and industry became stronger and stronger as the activity
increased. However, the politics as regards advances was not different
at all. Simple loans based on certificates and assets did not play a
crucial role under the classical form in the relations between trade and
industry; more important were the credits granted on current accounts.
If the name differed in Germany, there is no doubt that the results were
identical. Both are advances granted to industries without any special
guarantee.
The success of the Großbanken may simply have been due to the fact that
they developed around 1870, twenty years after the mistake of the first
Crédit Mobilier.
126
(a) Conclusion
As regards the principal-agent theory, the German banks really acted as

principals which supervised the firms with which they were involved. In the
Crédit Mobilier’s case it seems that the firms had used the bank to finance
their productive projects and managed to avoid any kind of control. The
bank, which should have acted as a principal, had become the agent of the
societies to which it provided liquidity. An analysis according to the degree
of affiliation with the Crédit Mobilier reinforces this argument. In the
sample used for affiliated companies to construct the tests concerning the
share prices fluctuations and the investment, two subsamples must be
considered:
• companies where the Crédit Mobilier was an official shareholder (and

sometimes the main shareholder) and where the Péreires were the
directors of the companies;
• companies where the Crédit Mobilier was only a financial partner.
The eformer situation, where the bank held the main part of the assets in
company, would have been impossible in the German case. For the
Großbanken this would have signified difficulties for the companies in
negotiating their titles on the equity market. Such a ‘signal’ would have
induced non-participation for a German bank (Whale 1930). This distinction
is very useful with regard to the power relationships between the firms and
the Crédit Mobilier.
For the first subsample, power was derived from the leading position the
Crédit Mobilier held by virtue of the number of shares belonging to the
Péreires. However, one surprising point is that the other shareholders were
strong enough to force the Péreires to resign after 1867. How can this
conundrum be explained? For this subsample (excluding the Compagnie
Immobilière), the Crédit Mobilier served the firms by providing sufficient
liquidity for the firms to grow during the period of French industrialisation.
In terms of the principal-agent theory, it seemed that the bank—in
appearance the principal—in reality was an agent which served the whole
firm (that is, the other block of shareholders). This perhaps helps to explain
the powerful situation the firms were in when the bank failed.
For the second subsample, the roles were well defined: the bank (the
principal) chose to grant credit to the firms because it expected thereby
to earn some profit. However, something is odd in this case: the firms
did not suffer from the bankruptcy. If an association with the Crédit
Mobilier had been a signal of reliability, the shares of firms should have
declined in value when the bank went bankrupty and thus lost its
reputation for sound financial judgement. This suggests that the Crédit
Mobilier never had such a reputation.
127
The Crédit Mobilier specialised in company promotion and provided

general financial services for the enterprises it patronised. In this regard,
credit was granted on the basis of future profit, which prevented the bank
from evaluating properly the risk of the project presented by the firms. The
exceptionally high growth experienced during the French industrialisation
probably constituted one of the reasons why the Crédit Mobilier lost control
over the firms’ managers. Up to that point, the firms had enjoyed an
‘exploitative’ attitude toward the bank.
4.5 THE ATTEMPT BY THE CRÉDIT

MOBILIER’S SHAREHOLDERS TO PREVENT
THE BANK’S BANKRUPTCY
The shareholders are generally sheep; sometimes tigers.11

Henri Germain
The purpose of this section is to see which of these two categories of

shareholders best described those of the Crédit Mobilier.
The behaviour of the shareholders seems to have been passive in the
Crédit Mobilier’s case. We can discuss the power of the shareholders and
debate if their behaviour was rational or irrational when the bank went
bankrupt.
We may wonder why the Crédit Mobilier’s shareholders reacted so
imperceptibly when the bank began to face its first liquidity problems in
1864. Several arguments can be made to explain this attitude:
• the power of the Péreires was so great in the shareholder’s annual

general meeting that it prevented anyone from influencing their
investment policy;
• most of the shareholders of the Crédit Mobilier were big bankers; the
difficulties that began in 1864 were a sign of future troubles for the
bank. They deliberately chose not to react in order to divide the spoils
between themselves after the bankruptcy. This constituted, in fact, a way
to control the Péreires, whose attitude towards risk was judged to be
reckless;
• during the period when share values increased, instead of diversifying
their portfolio, the Péreires had distributed their capital gain to the
shareholders. Some 68.2 million francs of dividends were thus wasted in
the four years between 1855–6 and 1862–3. In 1864, despite a new
increase in the dividend, the Crédit Mobilier faced difficulties and
managed to hold firm by the deposits of current accounts. As long as
they received dividends, the shareholders adopted a passive attitude
128
towards the policy of the bank. We can discuss whether this high
dividend policy did not correspond to a partial liquidation of the
establishment. Perhaps, since shareholders received a high enough
dividend for four years, they considered that, even if the worst
happened, they would still gain.
To refer again to Henri Germain’s definition, the shareholders were

consciously ‘sheep’. This argument is reinforced by the discourse of K.
Paris, who states:
The shareholders under the magic of language, of promises and

revenues, were not worried about the sudden shocks and relevant signs
that reached the asset prices issued by the firm. What was the
importance of an unexpected decrease? When the Crédit Mobilier fell
down, it was like a balloon to rebound and go up higher! And the
frenzy of speculation was maintained as tumultuous, as intense, as
disorganised as before.12
This quotation points out the enthusiasm of the shareholders for a firm
which had given them huge dividends and managed in the past to endure all
speculative movements.
According to the traditional theory of the firm, this attitude was
irrational. The firm must, generally speaking, be considered as a profit-
maximising entity, that is:
• profit maximisation is in the best interest of shareholders of the firm;

• shareholders can and do create incentive schemes for managers that
force managers to do what is best for the shareholders. So we conclude
that the managers maximise profits in order to give the shareholder the
highest dividend possible and to maintain the solvency of the firm.
The case of the Crédit Mobilier contradicts these two propositions. The
Péreires did not act in the shareholders’ interests, but in their own interests.
As they held the major part of the capital of the bank, the power of the
other shareholders was very limited. Hence the control policy inside the
bank was not effective. This constitutes a good example of the difficulties in
supervising companies when financiers are at the same time decision-makers
in the productive process.
4.6 CONCLUSION
This interpretative chapter essentially has discussed relations between banks

and industry. The subject of mutual relations between banks and industrial
129
enterprises in either furthering or hindering industrialisation, in the process

of concentration, in business cycles and in economic growth generally has
been the cause of speculation and debate since the turn of the century.
Renewed interest among economic historians was engendered by Alexander
Gerschenkron’s essay on relative economic backwardness in historical
perspective in the early 1860s. Recently, the rediscovery of Rudolf
Hilferding’s Finanzkapital has once again brought vividly to our attention
the issue of the interactions of banking and industry.
In my comments the (dis)advantages of close relationship between
industrial companies and banks are presented. In particular the bankruptcy
of the Crédit Mobilier raises two issues:
• the supervisory role of a bank over the companies with which it is

involved;
• the association between bank and firms.
To supervise firms, the bank could have:
• made the borrower’s actions part of the agreement and allowed

repayment to depend on the outcome of the borrower’s project;
• written a simple agreement (payment of such amount to be made on
such a date), then made the loan if it believed that the borrower was
likely to repay;
• used a collateral.
In the Crédit Mobilier’s case nothing like this was done. The only task the
Péreires accomplished was to provide liquidity and to try to realise an
immediate profit, without really taking care of the solvency of their own
establishment.
Another issue concerns whether the bank-firm relationship consists of a
domination by the banker, or a more flexible contract. As regards the Crédit
Mobilier, the policy of placing a member of the bank on the board of
directors in each company demonstrates the intention of the banker to place
himself in a leadership position. But the results that followed do not bear
out this assumption.
Close relationships between banks and industry developed everywhere
basically on the same lines during the nineteenth century. However, the evolution
of these relationships proceeded rather differently in different countries.
Geographically, my illustrative evidence is limited to Germany and the United
States. The whole set of results exhibited by the Crédit Mobilier’s case contrasts
with those obtained for the German Großbanken or the J.P.Morgan company.
Considering then the existing theory concerning the investment policy of
firms regularly associated with a bank, it is interesting to establish their
predictive power. Our results have two specifics:
130
• they concern a bank, the Crédit Mobilier, which went bankrupt;

• the set of non-affiliated companies was closely linked with the Crédit
Lyonnais, another bank.
They prove anyway that, contrary to the suggestion of Bernanke and Gertler
(1989), capital market imperfections do not always contribute to excessive
output fluctuations. For the pre-WWI period both assumptions, the scarcity
of capital (illustrated in France by an underdeveloped capital market
substituted by powerful financiers) and the decisive role of banks in
furthering industrialisation were tested. The main findings with regard to the
Crédit Mobilier serve to intensify the debate on the assumption of the
banks’ supremacy over industry.
131
5
GENERAL CONCLUSIONS
This book has considered the supervisory role between 1852 and 1880 of
the Crédit Mobilier, a bank which went bankrupt in 1867 and was
reorganised subsequently.
Economic historians have often studied the prominent role played by the
Crédit Mobilier in industrial financing in the nineteenth century.
Gerschenkron elaborated a model of backwardness, in which first banks,
then government, depending on how backward a country was, substituted
themselves for the entrepreneurship that had been the driving force in the
industrialisation of the United Kingdom, called ‘the first industrial nation’
by Peter Mathias. Cameron explained in detail how the Crédit Mobilier and
its imitators were responsible for the economic development not only of
France but of Continental Europe as a whole.
In order to evaluate the impact of monitoring on this bank, two
approaches have been adopted. First, we have studied the impact of the
Crédit Mobilier’s bankruptcy on share prices for affiliated and non-affiliated
companies to ascertain if—as in de Long’s study of J.P.Morgan —an
association with the Crédit Mobilier added value to the firms with which the
bank was involved. The sample we have constructed rejects such a
hypothesis. No significant effect of the bank’s bankruptcy could be detected
either on the economy as a whole or on the particular sample of twenty five
firms, half of which were involved with the Péreire brothers, the others
being totally independent of this bank. Our conclusion is therefore that the
Crédit Mobilier does not seem to have been supervising them properly
(hence the bankruptcy of the Compagnie Immobilière). The Crédit
Mobilier’s ‘seal of approval’ was probably not worth as much as that of
J.P.Morgan & Co or the Großbanken. De Long in his paper mentioned that
‘J.P.Morgan and Co’s approval of an issue had become…a large factor
which inspires confidence in the investor and leads him to purchase.’ In our
case, the active participation of the bank in a company did not ‘add value’
to the share prices.
Secondly, we evaluated the monitoring role of the bank as regards its
investment policy. This study of corporate finance in France concentrates on
132
GENERAL CONCLUSIONS
three samples: companies affiliated with the Crédit Mobilier, non-affiliated

companies and both. We also ran the regressions for three periods: the entire
period 1860–75 and two sub-samples 1860–7 and 1868–75 to compare
investment policy before and after the bankruptcy.
The econometric work presented here is based on the theory of
investment behaviour, and more specifically, the Q models (models of
investment demands). In particular, we investigate the assertion that the role
that the Crédit Mobilier played in the financing of French industry was not
dissimilar to that of the House of Morgan in the United States1 or the post-
war keiretsu and the pre-war zaibatsu in Japan.2 In other words, the Crédit
Mobilier should have supervised firms and so relaxed the credit constraint
for affiliated companies. The investment equation we use fits the Q models
with possible liquidity effects. If the Crédit Mobilier had adopted the same
attitude as the German Großbanken, the statistical results should have shown
a lower correlation between cash flow and investment for affiliated
companies.
The data set has not provided any evidence that the Crédit Mobilier
helped the development of French industry by alleviating the impact of
capital imperfections. No significant differences as regards Q and CF/K
coefficients can be identified between affiliated and non-affiliated
companies: neither category was liquidity constrained. Moreover, the
affiliated companies were more liquidity constrained before the bankruptcy
than after. Hence, the results do not fit in the Hoshi, Kashyap and
Scharfstein approach.
Thus, despite the small sample, we reveal a situation where a bank was
clearly not capable of controlling firms. This constitutes probably the most
interesting point of the research. The Crédit Mobilier is a good illustration
of the difficulties a bank can encounter when supervising a firm. Even if the
financial literature insists on the fact that a bank can also put pressure on
firms to enforce them to behave safely, these techniques are not efficient
when the institution stops behaving like a real bank. Historians have been
indulgent to the Péreires. They were not real bankers, but promoters of
enterprises: they aimed to expand the industrial sector and to gain prestige
and power through their ever-increasing role in financing firms.
The Péreires’ attitude towards risk was inadequate. Even if the expected
profit in the short term was sometimes high, they should have been able to
evaluate the risk properly in order to assure continuity in their policy. If we
admit the idea that the Péreire brothers were something other than bankers,
we can question whether their establishment was really a merchant bank and
discuss if, as mentioned in most of the literature, the Crédit Mobilier really
was ‘a great bank’ which significantly contributed to French
industrialisation.
As regards the econometric work done in this research and the existing
literature, it seems that the definition of the Crédit Mobilier has been
133
unclear. When the bank was founded, Courtois fils, a specialist on financial
questions and author of a Traité élémentaire des operations de bourse et de
change3 wrote:
The credit establishments which contribute to the conversion of fixed

capital in stocks, are generally said to be Crédit Mobilier.4
This definition does not make clear the role and the limitation of such
establishments. The merchant banks are distinct from other banks in the
sense that they are, at least in France, the only institutions to offer the
companies some help in their creation and organisation. The French
merchant bank must evaluate the worth of the potential project that the
firm wishes to undertake. It founds and reorganises the company around
this idea. Its aim is not to keep control permanently: once the enterprise is
well established in its industrial sector, the merchant bank will publicise
its assets, mostly by its introduction on the stock market. The merchant
bank does not expect any further profit from its participations: it will
maintain a very small holding in the company and a representation in the
shareholders’ assembly. This was not how the Crédit Mobilier saw its role:
it was not a merchant bank but an investment bank: participation for the
Péreire brothers signified a source of profit to be exploited in both the
short and long term.
The most serious restrictions to our conclusions are:
• the short period over which the results are valid;

• the sample is constituted of firms which were affiliated either to the
Crédit Mobilier or the Crédit Lyonnais. However this disadvantage (both
groups of companies had relations with a powerful bank that helped
French industrialisation) allowed us to compare the two institutions as
regards their investment policy.
The Péreire brothers, with their initiative and their ambitions were
themselves the cause of the liquidity difficulties. The search for immediate
profit led them to make huge investments (e.g. the Compagnie Immobilière).
Henri Germain of the Crédit Lyonnais was more careful:
The Crédit Lyonnais was a long-term work and twenty years of trial
and error were needed by its founder to make its foundations solid.
More lucky than the Péreires, it had the chance and the ability to last. 5
Secondly, as Levy-Leboyer (1976) mentioned, the Crédit Mobilier’s failure

has to be understood within the context of stabilisation of the market, with
the Crédit Foncier between 1867 and 1869 willing to consolidate its
liabilities especially towards the City of Paris and to reduce its short-term
134
GENERAL CONCLUSIONS
debt. This establishment had hastened the crisis by imposing a veto, on 3

August 1867, on new advances to the ‘Compagnie Immobilière’. The
balance sheet of this society had, up to then, been maintained in equilibrium
by an overvaluation of its stocks. Moreover the Banque de France chose not
to fulfil its ‘lender of last resort’ role properly. Thus it seemed that the
Crédit Mobilier was the victim of these establishments: both institutions
considered the financial situation of the bank to be so catastrophic that they
adopted a laissez-faire policy.
In some sense, the Péreires were naive to believe that they could bring
about radical change to French banking policy. Even if they intended to
manage their firms as ‘limited liabilities’, they pursued the old techniques
(limited partnership companies) which consisted in holding the majority of
assets in a firm. They succeeded in most cases (e.g. the railways). However,
one firm, the Compagnie Immobilière, provoked their downfall. This proves
that being a banker requires not only stability and ingenuity, but also the
ability to adapt to the rules of financial society.
135
APPENDIX A
Data on the Crédit Mobilier
A.1 THE CRÉDIT MOBILIER BETWEEN

1852 AND 1867
The following data are presented in order to understand better the place of
the bank in the financial market of the time. The balance sheet of the Crédit
Mobilier is presented in order to discuss the policy regarding investment and
dividends in the operations of the Crédit Mobilier (cf. Table A1).
The investment account and the cash/balance account show us the power the
bank had over this first period: these two accounts reflect the risky attitude of
the founders. This is reinforced by the dividend distribution (described by the
two last columns and the data below). The dividends declared were considerably
higher than those distributed by other merchant banks during the same period.
Further information on the bank is also disclose by a careful examination
of the share prices fluctuations over the period. Let us first consider the
prices reported in Table A3. These figures reflect the apparent power of the
institution from 1852 to 1867.
Some comments can be made on these prices. In 1856 and 1857 the
fluctuations of the share prices are large; at that time the bank issued
240,000 new shares at a price of 500 francs. In 1862 the share price went
from 705 francs in January to 1,285 francs in October. The general increase
in liquidity in France was probably one reason why lots of speculators
bought shares in financial markets in this period. In 1864 the increase of
liquidity of the Banque de France caused the same phenomenon.
A.2 THE CRÉDIT MOBILIER

BETWEEN 1903 AND 1914
Table A4 reflects the improvement concerning the presentation of the private

accounts. The Crédit Mobilier had become, over the period, a normal
merchant bank, less powerful than the one of the Péreires.
136
Table A1 Balance sheet of the Crédit Mobilier from 1853 to 1866 (in millions of francs)
Source: Dupont Ferrier (1925:234).

Table A2 Average profit of the Crédit Mobilier from 1855 to 1865
Source: Jean Bouvier et al (1965:449).
Table A3 Share prices of the Crédit Mobilier from 1853 to 1866
Source: (1903:111).
138
Table A4 Balance sheet of the Crédit Mobilier from 1903 to 1914 (in millions of francs)
Source: Baldy (1922:234).

APPENDIX B
Data relative to the general indices
(GNP, share prices) for France between
1852 and 1914
The aim is to present the data necessary to construct the excess

volatility test.
B.1 GENERAL PRICE INDICES
The sources of data on prices are Chadeau (1988:216) and Levy Leboyer
(1970:79–119). The general price index A is composed of the two following
series: B agricultural price index and C industrial index. The repartition
between the two sectors is implied by the weight attached to agriculture and
industry in the GNP. For a detailed presentation see Levy Leboyer’s article
pp. 115–17.
B.2 SHARE PRICE SERIES
Table A5 Share price series
140
APPENDIX B
Source: Dessirier (1928b:167).

Notes: (a)According to the average annual market price of the rent 3% (average between
the maximum and the minimum price until 1918). (b)Average rate: 1870–82 rent 3%,
bonds of the railways 3% and industrial bonds 3%, 1883–92 the three latter and ground
rent 3%, 1893–1913 in addition industrial bonds 4%. (c)According to the average of the
discounted rate taken from the monthly balance sheets. (d)Average rate of eight groups
of assets with fixed revenue: 3.84 (for the five quoted in note (b)and in addition ground
rent 4% in 1913, railways bonds 4% and industrial bonds 5%).
141
APPENDIX C
Data relative to the affiliated companies
C.1 THE WHOLE DATA
The aim of this appendix is to present the data underlying the analysis
presented in Chapter 2. The share prices between 1852 and 1867 presented
here constitute an insight of the whole data set used to run the regression
and will be provided in order to give an idea of the impact of the Crédit
Mobilier’s bankruptcy on the share prices of companies.
As regards Table A6 no fluctuations in the share prices can be noted for
the set of affliated companies.
C.2 THE COMPAGNIE IMMOBILIÈRE
The Compagnie Immobilière was constituted from three societies. The

first, founded in December 1854, was the Société de l’Hotel et des
Immeubles de la rue de Rivoli. Its capital of 24 million francs was
divided in 240,000 shares of 100 francs. E. and I.Péreire possessed at its
foundation 42,220 shares; the Crédit Mobilier as a whole more than
182,195. E. Péreire was its President. In 1858, after a modification of its
statute, the company was renamed the Compagnie Immobilière de Paris.
In 1859, it merged with the Société des ports de Marseille whose capital
of 1,500,000 francs was divided into 30,000 shares of 500 francs. In
June 1863, the society reached an agreement with the Société de la rue
Impériale de Marseille.
A distribution of 125 francs of dividends was made to the shareholders
of the Crédit Mobilier. The bankruptcy of the first Crédit Mobilier was
largely caused by that of the Compagnie Immobilière. It is interesting to
examine the situation of this company from 1870 up to 1874. As we have
already pointed out, the Crédit Mobilier had granted this company many
advances and loans, using the Crédit Foncier as intermediary. In 1870, the
net profits that the shareholders could expect after having paid all the
interests on loans would be:
142
APPENDIX C
in 1871, 985,031.73 francs;

in 1872, 2,419,811.03 francs;
in 1873, 3,605,013.19 francs;
in 1874, 4,142,288.97 francs.
The reorganisation of the Crédit Mobilier after the fall of the Péreires
obliged the society to liquidate part of the claims it had with its partners,
that is the Crédit Foncier.
The revenues of the company during the period were as shown in Table
A7. The expenses were as shown in Table A8.
In March 1881, the Crédit Mobilier reached an agreement with the
liquidators of the Compagnie Immobilière to rebuy some property not yet
liquidated. From this period on, all the claims with the enterprise were
resolved.
Table A9 shows the share price and the dividend distribution between
1855 and 1866.
143
Table A6 Data available for affiliated companies: share prices (francs)
Source: Annales de la Statistique de Paris: Numéro Récapitulatif (1954) pp. 180 et seq. for the railways; private account 65AQ
Archives Nationales de Paris for the insurance company l’Union, the Compagnie Transatlantique, the Compagnie Immobilière, the
Compagnie Parisienne d’Eclairage.
Note: na stands for not available.
APPENDIX C
Table A7 Revenues of the Compagnie Immobilière (francs)
Source: Private accounts on the company 65 AQ Archives Nationales de Paris.
Table A8 Expenses of the Compagnie Immobilière (francs)
Table A9 Share price and dividend distribution, 1855–66

Note: *For 1864 no data are available.
145
APPENDIX D
Data relative to non-affiliated companies
146
Table A10 Data for non-affiliated companies
Source: Chadeau (1988:120), Gille (1968:233–79).

APPENDIX E
Monthly share prices for affiliated
companies between 1866 and 1868
The most remarkable observation concerning these figures is the absence of

fluctuations for the affiliated companies except for the Compagnie
Immobilière and the Compagnie Transatlantique. The data show share prices
remaining stable despite the failure of the bank.
148
Table A11 Share prices for affiliated companies for 1866 (francs)
Source: Le temps 1866.

Note: The table reports monthly share prices evaluated the first day of each month.


APPENDIX F
Data for investment and cash flow tests
In this appendix, the raw data and its transformation in order to construct
the test will be described. To run the regressions, three parameters must be
available for all companies in the sample: investment, cash flow, and Q. To
estimate this last parameter the book values taken from the balance sheets
have been used. Before reporting any data, some explanations and
clarifications might be made. The investment variable is clearly specified in
the balance sheets and exhibits few changes as regards the actual accounting
.
If we now consider the existing literature on Tobin’s Q, it appears that
most authors take into consideration Q which is the ratio of the valuation of
corporate physical capital in the stock markets to its estimated cost of
reproduction. In the statistical work presented here we do not use this
definition of Q but rather common equity Q. Because of insufficient data on
the market value of debt, we could only approximate ‘Tobin’s Q’ with what
we should be calling strictly speaking ‘common equity Q’. As regards the
definition of the variable K it corresponds to the ‘capital brut’ (gross capital)
which over the period did not include a systematic depreciation rate. In
order to avoid any confusion it will be called Cb.
Table A14, A15 and A16 describe the values of shares that we call p,
investment, gross values of assets Cb and cash flows respectively for the
railway companies, the mining industry and the steel firms. The sample
cover the years 1860–75. As can be seen, some values are missing for
certain years. For the unavailable data, missing values have been substituted.
The situation of the Compagnie Immobilière must be considered
separately. As this company went bankrupt at the same time as the Crédit
Mobilier, little information was available over the period (from 1861 until
1869). However, the introduction of this element in the sample is relevant as
it produces many elements of explanation and interpretation on the bank’s
policy toward investment (the analysis of the results proves that the Crédit
Mobilier did not behave like a Großbanken before the bankruptcy, but
changed policy after 1868). The sources of our information in this special
case are the balance sheets collected in the Archives Nationales de Paris and
152
Table A14 Data available for the railway companies
Sources: Balance sheets collected in the Archives Nationales de Paris

Legend: p stands for value of share, Cb represents the gross value of assets, cf are the cash flows, i is the gross investment.
Table A15 Data available for mining industries
Sources: Balance sheets collected in the Archives Nationales de Paris.

Table A16 Data available for the steel industries

also some financial newspapers over the period. The data are reported in
Table A17.
All figures represent real values in millions of francs.
Some general comments can be made on Table A17. In comparison to the
first three, we remark that the fluctuations of the parameters are larger and
seem to be difficult to justify. They reflect in some sense the ‘irrationality’
of the Péreires as regards their investment policy. Up to the end (1868
represents a significant date), the bank continued to believe in the
profitability of the company and provided a large amount of liquidity (e.g.
the magnitude of investment i). The test presented for this company alone
will support this argument.
The last variable, Q, is simply the ratio of the market value to book
value. This procedure is virtually identical to the one de Long used in his
work.
As the raw data do not conform to traditional accounting rules, some
transformation has been made. First a depreciation rate δ is introduced for
the book value. The first value noted Cbk1861 is chosen as I1861/δ where δ=5
per cent. Some words of explanation are necessary to justify the choice of
departure for the calculation. If we refer to Jean Bouvier et al. (1965:165–
91) we can note that there was a lack of information for the preceding
period. According to the data available for this work we then choose to
begin in 1860.
The series is then evaluated by:
Cbkt+1 = Cbkt(1 – δ) + It
where d is a constant and chosen such that the series is slightly increasing.
This transformation of the data leads to an increase of Cb.
Table A17 The Compagnie Immobilière

Legend: p stands for value of share, Cb represents the gross value of assets, cf are the
cash flows and i is the gross investment.
156
Table A18 New values of Q when introducing a depreciation rate of 5 per cent
Table A18 exhibits the new value of Q. Some remarks can be made on
these figures. If the series exhibits numbers that are increasing slightly over
the period, the figures obtained for Q are quite different from those obtained
previously. In order to analyse this difference, we propose to see what
would have been the data had another origin been chosen.
Let us consider that the first value is Cbk1862=I1862/δ. The following value
are then obtained by
The whole series is then given by:
Cbkt+1 = Cbkt(1 – δ) + It.
As the set of data as a whole has the same characteristic (slightly increasing) over
the period we concentrate our argument on the first values for a few companies
(three railway companies, for example). The figures are reported in Table A19.
It appears that the first values have an influence on the data: if we
compare the third and the fourth columns, we see that the figures for Q
are increasing. The same phenomenon is to be noted for the other cases.
However the biggest change comes from the introduction of a
depreciation rate.
The second step of our argument will be to choose other values for d. In
order to see if our choice of d is valuable or not, we evaluate two other
series by considering small variations for d. The idea is then to discuss the
modification of data when choosing a depreciation rate smaller then 5 per
cent and to see the fluctuations of the same parameter when d is larger. The
figures obtained for Q remain low, but fit the theory (i.e. they are smaller
and drecreasing when d decreases.
The last step is to divide all variables by the general price index. Levy Leboyer
(1970:79–119) provides us with a series. As all the figures are expressed in basis
1880, we transpose them in real terms by applying the formula:
Table A19 Transformation of the data when changing origin
158
Table A20 Data in real terms
As before we evaluate Cbk by Cbk1861=I1861/δ. The whole series is given by:
Cbkt+1 = Cbkt(1 – δ). Pt+1/Pt + It. Pt+1/Pt.
The new values of Q are given in Table A20.
160
NOTES
INTRODUCTION
1 Cf. de Long (1989).

2 Cf. de Long and Becht (1992).
3 Cf. Becht and Ramirez (1993).
1 AGENCY THEORY AND MONITORING
1 Translation from the original:
Chez l’industriel français qu’il soit ‘moyen’ ou ‘grand’ la hantise de

l’indépendance vis à vis du créancier (la banque) fut constante. C’est au niveau
plus ou moins élévé de ses comptes courents vis à vis des banques que les
Schneider du XIXème siècle par exemple mesurent leur degré d’indépendance
et leur capacité de mouvement…. La domination des banques sur les
entreprises, si elle existait ne s’est pas faite sans résistance.
2 See in this respect Ross (1977) and (1978).

3 A similar issue arises in the literature on insurance provision. Rothschild and
Stiglitz (1976) and Miyazaki (1977) have proposed equilibrium concepts in
which firms are allowed to react to changes in contracts offered by other firms.
4 Bernheim and Whinston (1986:923).
5 Bernheim and Whinston (1985).
6 Cf. Jensen and Meckling (1976).
7 Cf. Diamond (1984).
8 Cf. Jensen and Meckling (1976).
9 Cf. Diamond (1984) and Gale and Hellwig (1985).
10 Cf. Broccklcr (1990).
11 Cf. Diamond (1989), who points out that the requirement of a guarantee can
address this problem and enables the lender to distinguish sound creditors from
bad ones.
12 This author has devoted a large part of his research to the study of German banks.
More specifically he discusses the characteristics of these banks as regards investment
policy and therefore justifies the control exerted by the banking sector on industry. As
an illustration, let us quote two basic sources: Tilly (1986a) and Tilly (1986b).
161
13 See Grossmann and Hart (1983).

14 Cf. Hoshi, Kashyap and Scharfstein (1991) and (1990).
15 See in this respect Laffont and Freixas (1987).
2 THE CRÉDIT MOBILIER AND THE FRENCH

STOCK EXCHANGE 1853–1914: AN EMPIRICAL
PERSPECTIVE
1 Translated from the original: Le sort du Crédit Mobilier n’est pas seulement
caractéristique des rivalités entre groupes financiers et des dangers de la
spéculation boursière: il rend sensible le problèms permanent de l’engagement
industriel.
2 Cf. Baldy (1922). The author provides several examples, among which are the
Crédit Lyonnais and the Banque de l’Union Parisienne. Some explanations are
necessary on what could be understood by ‘control’. The literature on
asymmetric information and incomplete contract presented in the first section
indicates that debt does not provide control. Instead it provides access to
privileged information, which the bank is then able to use in the process of
monitoring and enforcing the contract. Control, however, stays with the
managers of firms. As the Crédit Mobilier provided both debt and equity to its
customers, it can be said that this bank clearly exerted control over the firms
with which it was affiliated.
3 Huth (1918:11).
4 The quotation is taken from a memoir addressed to a group of bankers whose
support the Péreires hoped to obtain.
5 The whole of Bouvier, Furet and Gillet’s Le mouvement du Profit en France au
19ème Siècle provides general information about the fluctuations of share prices
over the period. It gives an indication of the increasing importance of the
financial market over the period.
6 By ‘short-term’ we mean three to six months.
7 See Bouvier, Furet and Gillet (1965).
8 Translation from the original: ‘Notre société a toujours considéré comme un
principe de haute moralité commerciale de ne jamais ouvrir une souscription, de
ne jamais recommander une entreprise sans s’y interesser elle même dans une
forte proportion, et ses administrateurs se sont fait un devoir de s’y associer.’
9 In the balance sheets where these advances were disclosed, the Administrators of
the Crédit Mobilier never mentioned that these advances to companies were not
based on any collateral. ‘Shares and bonds on collateral’, to cover additional
loans, had to be taken into consideration independently of the advances to
companies.
10 Cf. Leland and Pyle (1977).
11 See de Long and Becht (1992).
12 Statutes of the Société de Crédit Mobilier, 8 November 1871. The new bank was
capitalised at 80 million francs with 160,000 shares of 500 francs each; holders
of the old stock obtained two new shares for five old and could subscribe at par
for 48,000 additional shares. The founders of the new bank (including only two
members from the original bank, Charles Mallet and Frederic Grieninger) took
the remainder.
13 Numbers of events happened over the period, including the Spanish-American
war and the US annexation of Puerto Rico. In 1900 the Crédit Mobilier had
162
NOTES
invested funds for the construction of around 70 miles of railways in Puerto

Rico.
14 Translation from the original: ‘nous nous sommes appliqués à accroître le
nombre de nos correspondants à l’étranger et le mouvement de notre portefeuille
de banque. C’est notamment dans ce but que nous avons pris des intérêts dans
les institutions de Banques…Si cette politique a votre approbation, nous nous
proposons d’étendre encore nos efforts dans cette voie et de nous créer de plus
en plus d’appui.’
15 Bageholt has been criticised for not stating clearly when the central bank should
intervene, for not giving specific guidelines to distinguish between sound and
unsound banks, and for not realising that provision of the lender of last resort
facility to individual banks would encourage them to take greater risks than
otherwise. See in that respect Thornton (1802).
16 In part, Humphrey’s summary of the classical position is as follows:
The lender of last resort’s responsibility is to the entire financial system and
not to specific institutions…The lender of last resort exists not to prevent the
occurrence but rather to neutralise the impact of financial shocks…[Its] duty is
a twofold one consisting first, of lending without stint during actual panics and
second, of acknowledging beforehand its duty to lend freely in all future
panics…[It] should be willing to advance indiscriminately to any and all sound
borrowing on all sound assets no matter what the type…In no case should the
central bank accommodate unsound borrowers. The lender’s duty lay in
preventing panics from spreading to the sound institutions, and not in rescuing
unsound ones…. All accommodations would occur at a penalty rate, i.e., the
central bank should rely on price rather than non-price mechanisms to ration
use of its last resort lending facility…The overriding objective of the lender of
last resort was to prevent panic-induced declines in the money stock.
(Humphrey 1975:9)
17 Cf. Dessirier (1928a), (1928b) and Lenoir (1919).

18 Cf. Levy-Leboyer (1970).
19 Cf. Shiller (1981) and (1989).
20 See de Long and Becht (1992:12).
21 Gille (1959:373).
22 Ibid. (p. 374).
23 Ibid.
24 Translation from the original: ‘Les investissements cessent parce que
l’accumulation du capital est détruite et qu’il n’y a plus de disponibilités; ils
cessent parce que la rareté de l’argent circulant en fait remonter le taux; parce
que certaines affaires se sont révélées mauvaises ou spéculatives.’
25 Becht and de Long (1992:22). The Großbanken are still present in the German
economy today, although not nearly as important as they were before the First
World War. Rudolf Hildering could say that all that was needed in order to attain
socialism was the nationalisation of Germany’s six largest banks.
26 Cameron (1991:65).
163
3 CORPORATE INVESTMENT, CASH FLOW AND

FINANCIAL CONSTRAINTS OF FIRMS: THE
CASE OF THE CRÉDIT MOBILIER
1 See Chapter 1 with reference to credit rationing literature, and Stiglitz (1981).
2 The debate as regards the advantage of a strong relationship between banks and
industry is one of the oldest in financial economics. The originality of Hoshi,
Kashyap and Scharfstein (1989, 1990, 1991) was to use an econometric
methodology in order to support the argument. Hoshi, Kashyap and Scharfstein
(1989) is a basic reading which constitutes the foundation of the statistical work
presented in this paper.
3 De Long (1989) and Ramirez (1992).
4 Hoshi, Kashyap and Scharfstein (1989, 1991).
5 Annual report: April 1860.
Our company has always considered as a principle of its high commercial

morality never to open a subscription, nor recommend a firm, without first
having a large proportion of interest in it and its administrators having become
associated with it.
6 Cf. Tobin (1971).

7 E.g. Gerschendron (1962).
8 Cf. Ramirez (1992).
9 As we have already pointed out throughout this chapter:.
Economic logic indicates that a normal equilibrium value for Q is 1 for

reproducible assets which are in fact being reproduced and less than 1 for
others. Values of Q above 1 should stimulate investment, in excess of
requirements for replacement and normal growth, and values of Q below 1
discourage investment.
(Tobin and Brainard 1977:238)
4 THE SUPERVISORY ROLE OF THE CRÉDIT

MOBILIER: SOME INTERPRETATIONS
1 The original:
Le Crédit Mobilier a été, qu’on me pardonne l’expression, un lanceur

d’affaires. Pour lui l’intéret du présent domine presque toujours celui de
l’avenir.
2 These are not the only alleged control mechanisms in a capitalist economy. For
instance, there are evolutionary arguments which suggest that firms which
manage their resources efficiently will survive, while those that do not, will not.
For some criticisms of this argument, see, for instance, Diamond and Stiglitz
(1974), Stiglitz (1982).
3 In this article Stiglitz suggests that the firm should be considered as a problem
of multiple agents and multiple principals.
4 In French, ‘Société anonyme à responsabilité limitée’.
164
NOTES
L’argent aujourd’hui déposé notre caisse était en grande partie enfoui dans les
tiroirs, sans interêt pour ceux qui le gardaient, sans utilité pour la société.
6 In French, ‘Société Commanditaire pour l’industrie’.

Mais elles (les sociétés financières) jouent généralement un rôle plus actif:
elles ‘prennent ferme’ un certain ‘paquet’ de rentes ou d’obliga-tions,
‘préparent’ le marché ou attendent quelque circonstance favor-able et font
‘infiltrer’ les titres dans le public; elles ‘transforment’ elles-mêmes les
enterprises privées en sociétés anonymes dont elles émettent ensuite les
actions. Elles vont jusqu’à ‘créer’ des industries ou ‘organi-ser’ un commerce,
utilisant une idée qui leur fut insufflée. Ce sont done les courtiers, les
commissionnaires, les négociants de cette marchandise moderne que l’on
appelle valeurs mobilières: elles facilitent et même organisent le contact entre
l’épargne et ceux qui en ont besoin.
8 See Hoshi Kashyap and Scharstein (1990, 1991).

9 See Becht and Ramirez (1993).
10 Cf. Weber (1903:337): of eighteen industrial corporations in Rheinland and
Westphalia founded in the years 1896–1900 and quoted on the Berlin stock
exchange in 1901, sixteen emerged from the conversion of previously existing
private companies; only two were new foundations.
11 Translation from the original: Les actionnaires sont en général des moutons,
parfois des tigres.
Les actionnaires pris à la magie du langage, des promesse et des revenus ne

s’inquiètaient même plus des brusques soubresauts et des signes révèlateurs qui
atteignaient le cours des titres emis par la Société. Que leur importail une
chute inatlendue? Quand le Crédit Mobilier tombait, c’était comme le ballon
pour rebondir et s’élever plus haut! Et la frénésie de la speculation se
maintenaient aussi tumultueuse, aussi intense, aussi désordonnée.
‘Le Crédit Mobilier et ses actionnaires’ (1867:50).
5 GENERAL CONCLUSIONS
1 See de Long (1989) and Ramirez (1992).

2 See Hoshi, Kashyap and Scharfstein (1990, 1991).
3 Courtois (1889:13).
4 ‘Les établissements de crédit qui aident à la conversion des capitaux stables en
capitaux mobiliers, sont dit, d’une manière générale, Crédit Mobilier.’
5 ‘Le Crédit Lyonnais fut une oeuvre de longue haleine et il fallut à son fondateur
vingt ans de tâtonnement pour l’asseoir sur des bases solides. Plus heureux que
les Péreires, il eut la chance et l’habileté de durer.’
Quotation extracted from Bouvier (1968:150).
165
BIBLIOGRAPHY
Aghion, P. and P.Bolton, (1986) ‘An “Incomplete Contract” Approach of

Bankruptcy’, Mimeo, MIT Harvard University, March, 484.
Akerlof, G. (1970) ‘The market for lemons: qualitative uncertainty and the market
mechanism’, Quarterly Journal of Economics, 84: 488–500.
Aoki, M. (1984) The cooperative game theory of the firm, Oxford: Clarendon.
—— (1988) Information, Incentive and Bargaining in the Japanese Economy,
Cambridge University Press.
Aoki, K. and J.Kyle (1992) Corporate Investment in Japan: a Foreign Investor’s
Guide to Making Strategic Alliances, London: Woodhead Faulkner.
Autin, J. (1884) Les Frères Péreire: le Bonheur d’Entreprendre, Paris.
Aycard, M. (1867) Histoire du Crédit Mobilier 1852–1867, Paris.
Bagehot, W. (1873) Lombard Street: A description of the Money Market, London:
H.S.King.
Baldy, E. (1922) Les Banques d’Affaire en France depuis 1900, Paris.
Barham, V. and O.Lefevre (1990) ‘Bank loans securitization with asymmetric
information: a principal-agent solution’, Working Paper Core Louvain La Neuve,
November, 9072.
Baron, D. and B.Holmstrom, (1980) ‘The investment banking contract for new issues
under asymmetric information delegation and the incentives problem’, Journal of
Finance, December, 4: 1115–1138.
Batchelor, R.A. (1986) ‘The avoidance of catastrophe: two nineteenth century crises’,
in F.Capie and E.Wood Geoffrey (eds) Financial crises and the world banking
system, New York: Saint Martin’s Press.
Becht, M. and C.Ramirez (1993) ‘Finance capitalism, Großbanken and the German
mining and steel industry 1906–1912’, Working Paper EUI, Italy, 93/26.
Becker, E.B. and C.A.Olson, (1989) ‘Unionization and shareholders’ interests’,
Industrial and Labor Relations Review, 42, 2: 246–261.
Benston, G.J. et al. (1986) Perspectives on Safe and Sound Banking: Past, Present
and Future, Cambridge: MIT Press.
Bernanke, B. (1983) ‘Nonmonetary effects of financial crisis in the propagation of
the great depression’, American Economic Review, LXXIII, 2: 257–276.
166
BIBLIOGRAPHY
Bernanke, B. and M.Gertler (1989) ‘Agency costs, net worth and business
fluctuations’. American Economic Review, LXXIX, 1: 14–31.
Bernheim, D. and M.Whinston (1985) ‘Common marketing agency as a device for
facilitating collusion’, Rand Journal of Economics, 16, 2: 269–281.
—— (1986) ‘Common agency’, Econometrica, 54, 4: 923–943.
Besanko, D. and A.V.Thakor (1987) ‘Collateral and rationing: sorting equilibria and
competitive markets’, International Economic Review 28, 3: 671–689.
Bigo, H. (1947) Les Banques Françaises au cours du XIXème Siècle, Paris.
Blee, E. (1882) ‘La Vérité sur la Crise Financière, Paris.
Bonin, H. (1988a) Histoire Economique de la France depuis 1880, Masson.
—— (1988b) ‘La Grande Depression Française de la Fin du Dix Neuvième Siècle:
Reflexion sur sa datation et sur sa fonction’, Histoire, Economie et Société 6, 4:
509–533.
Bonnet, V. (1865) La Revue des deux Mondes, Paris.
Born, K.E. (1983) International Banking in the 19th and 20th century, Leamington
Spa: Berg Publishers.
Bouvier, J. (1960) ‘Les Péreire et l’affaire de la Banque de Savoie 1860–64’, Cahiers
d’Histoire 3: 383–410.
—— (1968) Naissance d’une banque: le Credit Lyonnais, Paris: Flammarion.
—— (1972) ‘Systèmes bancaires et entreprises industrielles dans la croissance
Européenne du XIXe siècle’, Annales ESC 1: 46–70.
—— (1988) Histoire économique et histoire sociale, Geneva: Librairie Droz.
Bouvier J., Furet F. and Gillet M. (1965) Le Mouvement du Profit en France au
19ème Siècle, Industrie et Artisanat, Mouton and Co.
Brief, R. (1981) ‘Nineteenth century accounting error’, in J.R.Edwards (ed.)
Company Legislation and Changing Patterns Disclosure in British Company
Accounts 1850–1914, London.
Brown, R. (1905) A History of Accounting and Accountants, Edinburgh: T.C. &
E.C.Jack.
Calomiris C.W. and Gorton G. (1991) ‘The origins of banking panics: models, facts
and bank regulation’, in G.Hubbard (ed.) Financial Markets and Financial Crises,
University of Chicago Press / National Bureau of Economic Research.
Cameron, R.E. (1953) ‘The Crédit Mobilier and the economic development of
Europe’, Journal of Political Economy, December, 4: 461–488.
—— (1967) Banking in the Early Stage of Industrialization: A Study in Comparative
Economic History, New York: Oxford University Press.
Cameron, R.E. and V.I.Bovykin (1991) International Banking 1870–1914, New York:
Oxford University Press.
Caron, F. (1973) Histoire de l’Exploitation d’un Grand Réseau de Chemin de Fer: La
Compagnie du Chemin de Fer du Nord 1846–1937, Paris: Mouton.
Carosso, V.P. and Sylla, R. (1991) ‘US banks in international finance’, in R.Cameron
and V.I.Bovykin (eds) International Banking, 1870–1914, New York and Oxford:
Oxford University Press, 48–71.
Cassis, Y. (1992) Finance et Financiers in European History 1880–1960, Cambridge
University Press.
Chadeau, E. (1988) ‘L’Economie Nationale aux XIXème et XXème siècles’,
Annuaire Statistique de l’Economie Française aux XIXème et XXème siècles,
Presses de l’Ecole Normale Supérieure, (1, 3).
167
Chastenet, E. (1861) Le Crédit Mobilier et la Bourse, Dentu.

Conant C.A. (1969) A History of Modern Banks of Issue, (Reprints of Economic
Classics), New York: Auguste M.Kelley.
Corrigan, E.G. (1987) Financial Market Structure: A Longer View, New York: Federal
Reserve Bank of New York.
Courtois, P. (1889) Traité élémentaire des Opérations de Bourse et de Change, 10th
edn, Paris.
Darrough, M.N. and N.M.Stoughton (1986) ‘Moral Hazard and Adverse Selection:
the Question of Financial Structure’, Journal of Finance,, XLI, 2: 501–513.
de Long B. (1989) ‘Did J.P.Morgan’s men add value? A Historical Perspective on
Financial Capitalism’, Working Paper, Harvard University and NBER, 3426.
de Long B. and M.Becht (1992) ‘“Excess volatility” and the German stock market
1876–1990’, Working Paper, Harvard University and NBER, 82.
de Long B., A.Schleifer, L.H.Summers and R.Waldmann (1990) ‘Noise trader risk in
financial markets’, Journal of Political Economy, 98, 4: 703–729.
Denuc, J. (1934) ‘Dividendes, valeur boursière et taux de capitalisation’, Bulletin de
la Statistique, XXIII, IV: 691–767.
Deshons, M. and X.Freixas (1987) ‘Le Role de la Garantie dans le Contrat de Prêt
Bancaire’, Revue Finance, 2: 7–32.
Dessirier, J. (1928a) ‘Les Indices des Valeurs Mobilières en France et en Divers
Pays’, Revue d’Economie Politique, 42: 202–231.
—— (1928b) ‘La Prévision Statistique des Mouvements des Valeurs de Bourse’,
Journal de la Société Statistique de Paris, 153–194.
Devinney, T.M. and H.Milde (1990) ‘Agency Contracting and Inside Debt’,
Schweizerische Zeitschrift für Volkswirtschaft und Statistik 1 2: 97–111.
Diamond, D. (1984) ‘Financial Intermediation and Delegate Monitoring’, Review of
Economic Studies 51: 393–414.
Diamond, P.A. and J.E.Stiglitz (1974) ‘Increases in Risk and in Risk Aversion’,
Journal of Economic Theory, 8(3), July, 337–360.
Dupont Ferrier, P. (1925) Le Marché Financier sous le Second Empire, Paris.
Edwards, J.R. (1981) Company Legislation and Changing Patterns of Disclosure in
British Company Accounts 1850–1914, London.
—— (1937) A survey of French Contributions to the Study of Cost Accounting
During the 19th Century, London.
Engberg, H.L. (1981) Mixed Banking and Economic Growth in Germany 1850– 1931,
New York: ARNO Press.
Fama, E.F. (1985) ‘What’s different about banks?’, Journal of Monetary Economics,
15, 29–39.
Fazzari S., R.G.Hubbard a nd B.Petersen (1988) ‘Financing constraints and corporate
investment’, Brookings Papers on Economic Activity, n.1.
Feldenkirchen, W. (1982) Die Eisen- und Stahlindustrie des Ruhrgebiets 1879–1914:
Wachstum, Finanzierung und Struktur ihrer Großunternehmen, Steiner Verlag
GmbH.
—— (1991) ‘Banking and economic growth: banks and industry in the nineteenth
century and their changing relationship during industrialisation’, in W.R.Lee (ed.)
German Industry and German Industrialisation, London: Routledge.
Fohlen, C. (1981) Entrepreneurship and Management in France in the Nineteenth
Century, Paris.
168
BIBLIOGRAPHY
Freixas, X. (1986) ‘Contrats de dettes en information imparfaite’, Working Paper,

GREMAQ, Toulouse, Juillet.
Friedman, M. and Schwartz, A. (1963) Monetary History of the United States 1867–
1960, Princeton University Press.
Gale, D. and M.Hellwig (1985) ‘Incentive compatible debt contracts: the one period
problem’, Review of Economic Studies, LII: 647–663.
Gerschenkron, A. (1962) Economic Backwardness in Historical Perspective,
Cambridge, Harvard University Press.
Gille, B. (1959) La Banque et le Crédit en France de 1815 à 1848, PUF.
—— (1968) La Sidérurgie Française au XIXème Siècle, Geneva: Librairie Droz.
—— (1970) La Banque en France au XIXème siècle: Recherche Historique, Geneva.
Gjcsdal, F. (1982) ‘Information and incentives: the agency information problem’,
Review of Economic Studies, 49, 3: 373–90.
Goodhart, C.A.E. (1985) The Evolution of Central Banks, Cambridge, MIT Press.
—— (1987) ‘Why do banks need a central bank?’, Oxford Economic Papers (March)
39, 1: 75–89.
Grossman, S.J. and O.D.Hart (1983) ‘An analysis of principal-agent problem’,
Econometrica, 51, 1: 7–46.
Guesnerie, R. and J.J.Laffont (1984) ‘A complete solution to a class of principal-
agent problems with an application to the control of a self-managed firm’, Journal
of Public Economics, 25, 2: 329–369.
Hammond, J.P. (1993) ‘Financial distortions to the incentives of managers, owners
and workers’, Lecture prepared for the Conference on the Structure of Capitalism
and the Firm in Contemporary Society, Milan, 18–20 March.
Holmstrom, B. (1979) ‘Moral hazard and observability’, Bell Journal of Economics,
10: 74–91.
Holzer, P.H. and W.Rogers (1990) ‘The origins and the developments of French
costing systems (as reflected in the literature)’, The Accounting Historians
Journal, 17, 2: 57–71.
Hoshi T., A.Kashyap and D.Scharfstein (1989) ‘The role of banks in reducing the
costs of financial distress in Japan’, Massachussetts Institute of Technology, 3435.
—— (1990) ‘Bank monitoring and investment: evidence from the changing structure
of Japanese corporate banking relationships’, in R.Glenn Hubbard (ed.)
Asymmetric Information, Corporate Finance, and Investment, University of
Chicago Press / National Bureau of Economic Research.
—— (1991) ‘Corporate structure, liquidity, and investment: evidence from Japanese
industrial groups’, Quarterly Journal of Economics, 1: 33–60.
Hughes P.J. (1988) ‘Risk sharing and valuation under moral hazard’, in Gerald A.
Feltham, Amin H.Amershi and William T.Zemba (eds) Economic of Information
and Contracts, Kluwer Academic Publishers.
Humphrey, T. (1975) ‘The classical concept of the lender of last resort’, Federal
Reserve Bank of Richmond Economic Review 61, January: 2–9.
—— (1989) ‘Lender of last resort: the concept in history’, Federal Reserve Bank of
Richmond Economic Review, 75, March/April: 8–16.
Huth, W. (1918) Die Entwicklung der Deutschen und Französischen Großbanken im
Zusammenhange mit der Entwicklung der Nationalwirtschaft, Berlin: Franz
Siemenroth.
169
Innes, R.D. (1990) ‘Limited liability and Incentive contracting with ex ante action
choices’, Journal of Economic Theory, 52: 45–67.
Jacquillat, B. and M.Levasseur (1984) ‘Signaux, mandats et gestions financière: une
synthèse de la littérature’, Finance, 5: 31–63.
Jeidels, D. (1905) ‘Das Verhältnis der deutschen Großbanken zur Industrie mit
besondere Berücksichtung der Eisenindustrie’, in Staats- und
Sozialwissenschaftliche Forschungen, Band XXIV, Heft II: , 1–271, Leipzig:
Duncker & Humblot.
Jensen, M.C. and W.C.Meckling (1976) ‘The theory of the firm: managerial behavior,
agency costs and ownership structure’, Journal of Financial Economics, 3: 305–
360.
Johnston H.T. and R.S.Kaplan (1987) Relevance Lost: the Rise and Fall of
Management Accounting, Boston, Massachusetts: Harvard Business School Press.
Jorgenson, D. (1971) ‘Econometric studies of behavior: a review’, Journal of
Economic Literature, IX: 1111–47.
Kaufmann P. (1914) La Banque en France, Paris.
Keely, M.C. (1990) ‘Deposit insurance, risk about market power in banking’,
American Economic Review, 4: 1183–1200.
Kindleberger, C.P. (1984) ‘Banking and industry between the two wars: an
international comparison’, Journal of European Economic History, 39, 2: 7–26.
Kocka, J. (1978) ‘Entrepreneurs and Managers in German Industrialization’, in
Mathias P. and M.Postan (eds) The Cambridge Economic History of Europe, Vol.
VIIThe Industrial Economics: Capital Labour and Enterprise Part I, Cambridge:
Cambridge University Press.
Kose, J. and A.Kalay (1982) ‘Costly contracting and optimal payout constraints’,
Journal of Finance, XXXVII, 2: 373–390.
Krause, D. (1991) ‘Zur Geschichle der Mitteldeutschen Creditbank’, Bankhislorisches
Archiv, 17: 17–41.
Kreps, D. and R.Wilson (1982) ‘Reputation and imperfect information’, Journal of
Economic Theory, 27: 253–279.
Kreps, D., P.Milgrom, J.Roberts and R.Wilson (1982) ‘Rational cooperation in the
finitely-repeated prisoner’s dilemna’, Journal of Economic Theory, 27: 245– 252.
Laffont, J.J. (1987) ‘Le risque moral dans la relation de mandat’, Revue Economique,
41, 1: 5–23.
Laffont, J.J. and X.Freixas (1988) ‘Contrats Bancaires Optimaux’, Toulouse:
GREMAQ.
Laffont, J.J. and E.Maskin (1986) ‘The theory of incentives: an overview’, in
W.Hildenbrand (ed.) Advances in Economic Theory, Cambridge University Press.
Lambert, R. (1983) ‘Long term contracts and moral hazard’, Bell Journal of
Economics, 14, 2: 441–452.
Landes, D.S. (1956) ‘Vieille banque et banque nouvelle: la révolution financière du
XIXème siècle’, Revue d’Histoire Moderne et Contemporaine, 3: 204–220.
Leland, H.E. and D.H.Pyle (1977) ‘Informational asymmetries, financial structure and
financial intermediation’. Journal of Finance, 32, 2: 381–387.
Lenoir, M. (1919) ‘Le mouvement des cours des valeurs mobilières françaises depuis
1856’, Bulletin de la Statistique Générale de la France, Octobre, 65–92.
—— (1922) ‘Les valeurs mobilières et la crise de 1920 (indices économiques
compares en France et divers pays)’, Bulletin de la Statistique Générale de la
170
BIBLIOGRAPHY
France, Janvier, 171–221.

Levy Leboyer, M. (1970) ‘L’héritage de Simiand: prix, profits et termes de l’echange
en France au XIXème siècle’, Revue Historique, 493: 79–119.
—— (1976a) ‘Le crédit et la monnaie: l’apprentissage du marché’, in Fernand
Braudel and Ernest Labrousse Histoire Economique de la France Vol. III:
‘L’Avénement de l’Ere Industrielle’, Paris: Presse Universitaire de France.
—— (1976b) ‘La spécialisation des établissements bancaires’, pp. 431–467 in
Fernand Braudel and Ernest Labrousse Histoire Economique de la France Vol. III:
‘L’Avenèment de l’Ere Industrielle’, Paris: Presse Universitaire de France.
Lindenberg, E. and S.Ross (1981) ‘Tobin’s q ratio and industrial organisation’,
Journal of Business, January, 54: 1–32.
Littleton, A.C. (1960) ‘Accounting Evolution to 1900, New York: Russell.
Luce, R.D. and H.Raiffa (1957) Games and Decisions: Introduction and Critical
Survey, New York: Wiley.
Mackie-Mason, K. (1989) ‘Do firms care who provides their financing?’ Working
Paper NBER, 3039.
Mandelker, G. and A.Raviv (1977) ‘Investment banking: an economics of optimal
underwriting contracts’, Journal of Finance, June, 2: 61–75.
Mankiw, G. and M.Shapiro (1985) ‘Do we reject too often? Small sample properties
of tests of rational expectations models’, Economics Letters, 20: 139–145.
Mankiw G., Romer D. and M.Shapiro (1985) ‘An unbiased reexamination of stock
market volatility’, Journal of Finance, , XL, 3: 677–689.
—— (1991) ‘Stock market forecastability and volatility: a statistical appraisal’,
Review of Economic Studies, 58: 455–477.
Marriner, S. et al. (1978) Business and Businessmen: Studies in Business, Economic
and Accounting History, Liverpool University Press.
Mayer, C. (1988) ‘New issues in corporate finance’, European Economic Review,, 32:
1167–1188.
Milgrom, P. and J.Roberts (1982) ‘Limit pricing and entry under incomplete
information’, Econometrica, 50: 443–460.
Miskin, F. (1991) ‘Asymmetric information and financial crises: a historical
perspective’, in G.Hubbard (ed.) Financial Markets and Financial Crises,
University of Chicago Press / National Bureau of Economic Research.
Moussa, P. (1980) ‘A quoi sert une banque d’affaires?’, Banque, 394 Avril, 429– 433.
Murphy K.M., A.Schleifer and R.W.Vishny (1989) ‘Industrialization and the big
push’, Journal of Political Economy, 97, 5: 1003–26.
Myers, S.C. (1977) ‘Determinants of corporate borrowing’, Journal of Financial
Economics, 49: 993–1007.
Myers, S. and N.Majlhuf (1984) ‘Corporate financing and investment decisions when
firms have information that investors do not have’, Journal of Financial
Economics, 77: 187–221.
Nelson, J. (1990) ‘Principals and agents: does the assignment of information
matter?’, Economic Letters, 23: 223–227.
Neuberger, D. and M.Neumann (1991) ‘Banking and antitrust: limiting industrial
ownership by bank?’, Journal of Institutional and Theoretical Economics, 147:
188–199.
Nouette-Delorme, E. (1865a) Les sociétés de speculation: le Crédit Mobilier et ses
valeurs, Paris: Castel.
171
—— (1865b) Le Crédit Mobilier et ses Valeurs: Histoire de 12 ans, Paris: Castel.

Page, F.H. Jr. (1989) ‘Incentive compatible strategies for the general Stackelberg
games with incomplete information’, International Journal of Game Theory, 3:
409–421.
Paris, K. (1867) Le Crédit Mobilier et ses Actions, Paris.
Péreire, E. and I. (1879) La question des Chemins de Fer, Paris: Imprimerie de C.
Motteroz.
—— (1912–1920) Oeuvres, Paris: Alcan.
Plenge, J. (1903) Gründung und Geschichte des Crédit Mobilier, Tubingen: These.
Plessis, A. (1985) La Politique de la Banque de France de 1851 à 1870, Paris: Droz.
—— (1993) ‘Bankers in the French society 1860s–1960s’, in François Crouzet (ed.)
The Economic Development of France since 1870, Elgar Reference Collection Vol.
II.
Prowse, S.D. (1992) ‘The structure of corporate ownership in Japan’, Journal of
Finance XLVII, 3: 1121–1140.
Pujo Committee, US Congress, House, Committee on Banking and Industry (1913a)
Report of the Committee Appointed Pursuant to House Resolutions 429 and 504 to
Investigate the Concentration of Control of Money and Credit, Washington, D.C.:
G.P.O.
—— (1913b) Money Trust Investigation. Hearings Before the Committes Appointed
Pursuant to House Resolutions 429 and 504 to Investigate the Concentration of
Control of Money and Credit, Washington, DC: G.P.O.
Ramirez, C. (1992) ‘Did J.P.Morgan’s men add liquidity? Corporate investment, Cash
flow and financial structure at the turn of the twentieth century’. Working Paper,
Harvard University.
Rasmusen, E. (1989) Games and Information: An Introduction to Game Theory, Basil
Blackwell, Oxford and Cambridge.
Riesser, J. (1911) The Great German Banks and their Concentration, Washington
D.C.: US Government Printing Office.
Ross, S. (1973) ‘The economic theory of agency: the principal’s problem’, American
Economic Review, May, 3: 363–367.
—— (1977) ‘The determination of financial structure: the incentive signalling
approach’, Bell Journal of Economics, 8, 2: 23–40.
—— (1978) ‘Some notes on financial incentive signalling models, activity choice and
risk preferences’, Journal of Finance, XXXIII, 43: 777–794.
Rothschild, M. and J.Stiglitz (1976) ‘Equilibrium in competitive insurance market: an
essay in the economics of incomplete information’, Quarterly Journal of
Economics, 90: 331–359.
Salinger, M. (1984) ‘Tobin’s q, unionization, and the concentration-profits
relationship’, Rand Journal of Economics, 15: 159–170.
Sayous, A.E. (1899) ‘Les banques allemandes en cas de crise ou de guerre’, Revue
d’Economie Politique, 142–166.
—— (1907) Lex Banques de Dépôts, les Banques de Crédit et les sociétés
Financières, Paris.
Scott, K.E. (1991) ‘Banking and antitrust: limiting industrial ownership by bank?
Comment’, Journal of Institutional and Theoretical Economics, 147, 1: 202–206.
Shavell, S. (1979) ‘Risk sharing and incentives in the principal and agent
relationship’, Bell Journal of Economics, 10: 55–73.
172
BIBLIOGRAPHY
Shiller, R. (1981) ‘Do stock prices move too much to be justified by subsequent
changes in dividends?’, American Economic Review, 71, 2: 421–36.
—— (1989) Market Volatility, Cambridge, MIT.
Smirlock, M., Gilligan T. and W.Marshall (1984) ‘Tobin’s q and the structure
performance relationship’, American Economic Review, 74, 4: 1051–1060.
Solow, R.M. (1982) ‘On the lender of last resort’, in C.P.Kindleberger and J.P.
Laffarge (eds) Financial Crises: Theory, History and Policy.’ Cambridge
University Press.
Stiglitz, J.E. (1985) ‘Credit markets and the control of capital’, Journal of Money,
Credit and Banking, 17: 133–152.
—— (1982) ‘The inefficiency of the stock market equilibrium’, Review of Economic
Studies, 49(2), April, 241–261.
Stiglitz, J.E. and A.Weiss (1981) ‘Credit rationing in markets with imperfect
information’, American Economic Review, 71: 393–410.
Stole, L.A. (1990) ‘Mechanism design under common agency’, Mimeo, MIT,
November (unpublished).
Teichova, A. (1988) ‘Rivals and partners: banking and industry in Europe in the first
decades of the twentieth century’, Uppsala Working Papers in Economic History,
Report from the Vienna Banking Industry Symposium.
Thornton, H. (1802) An Enquiry into the Nature and Effects of the Paper Credit of
Great Britain, edited by F.A.Hayek, Fairfield Augustus: M.Kelley.
Tilly, R. (1966) Financial Institutions and Industrialization in the Rhineland 1815–
70, London: University of Wisconsin Press.
—— (1986a) ‘German banking 1850–1914: development assistance for the strong’,
Journal of European Economic History, 15, 1: 113–152.
—— (1986b) ‘Germany 1815–1870’, in R.Cameron, et al. (eds) Banking in Early
Stages of Industrialisation: a study in comparative economic history New York,
London, Toronto: Oxford University Press.
Tobin, J. and W.R.Brainard (1977) ‘Asset markets and the cost of the capital’, in
Balassa B. and Nelson R. (eds) Economic Progress, Private Values and Public
Policy: Essays in Honor of William Fellner, North Holland.
Webb, D. (1987) ‘Conflict of interest, bond rationale and the value maximisation
rule’, Economica, 54, 3: 455–463.
Weber, A. (1903) Die rheinisch-westfälischen Provinzialbanken und die Krisis,
Leipzig.
Weston, J.F. and T.E.Coppeland (1987) Financial theory and corporate policy, 3rd
edn, Addison Wesley Publishing Company.
Whale, B.P. (1930) Joint Stock Banking in Germany: a study of the German Credit
Banks before and after the War, London: Macmillan and Co.
Wilson, C. (1977) ‘A model of insurance markets with incomplete information’,
Journal of Economic Theory, 16: 167–207.
Wrück, K.H. (1990) ‘Financial distress, reorganization and organizational efficiency’,
Journal of Financial Economics, 27, 2: 419–444
Yoshikawa, H. (1980) ‘On the q theory of investment’, American Economic Review,
3: 739–743.
173
INDEX
adverse selection 9–10 banks: association with firms, analysis

affiliated companies 42–4; data on 117–20; financial intermediation by
142–5; depreciation rate and see financial intermediation;
investment 92–5, 98–104; effect of monitoring by see monitoring by
bankruptcy on 53–65; investment and banks; and stock market 51–2;
cash flow 80–90; and non-affiliated supervisory role of see supervisory
companies, comparisons 71–3; share role of banks. See also central banks
price fluctuations 60–4; share prices, Banque de France 29, 38; and
monthly (1866–1868) 148–51; bankruptcy of Crédit Mobilier 112
summary statistics on 105 Banque de l’Union Parisienne 28, 48,
agency costs 9, 12 51
agency theory 1, 7; banks, supervisory Banque Transatlantique 36
role of 15–24; basic concepts 8–12 Becht, M. 27, 31, 34, 40, 42, 45–6, 51,
Allais (company) 44, 67; bankruptcy, 86, 93, 95, 110, 116–17
impact of 68–70; data on 147 Bernanke, B. 114
Aoki, M. 117 Bonnet, V. 109
asset diversification 11–12 Born, K.E. 52
asymmetric information see information Bouvier, J. 7, 30, 112, 121, 124
asymmetry Bovykin, V.I. 76
Bagehot, W. 38 Cameron, R.E. 43, 76, 132

bank finance 16–19; and market finance capital scarcity 18–19
19–20; and supervisory role of banks capital stock and investment 82, 84–6;
16–19 and depreciation rate 93–5
bank-firm relations: analysis of 117–20; Carosso, V.P. 51–2
freedom from 24; Japan 23–4 cartels 23
banking panic and bankruptcy of Crédit cash flow and investment expenditure
Mobilier 114–16 77; and Crédit Mobilier 78–90;
bankruptcy of Crédit Mobilier 3, 35–7; dataon 153–61; and depreciation 103;
banking panic in 114–16; and in Japan 124; regression on 83–7
collateral 33; effect on affiliated central banks: as lender of last resort
companies 53–65, 73–4; effect on 38–9; role in Germany 125
non-affiliated companies 65–71, Chatillon (company) 44, 45, 67;
73–4; financial crisis, creating 111– bankruptcy, impact of 68–70; data on
14; general influence of 54–8; high- 148; investment and cash flow
risk projects 13; shareholders’ analysis 80–90
attempts to prevent 128 9; and stock collateral 33
market prices 41–2 commitment by banks 23–4
174
INDEX
common agency 12–15; assumptions 14 143; share price fluctuations 60–4

15; compensation 15; contract terms share prices, monthly (1866–1868)
14; risk-neutrality in 14 149–51
Compagnie des Chemins de Fer de l’Est Compagnie Générale Transatlantique 31,
43; asset prices 54; bankruptcy of 41
Crédit Mobilier, impact of 55–9; Compagnie Immobilière 31, 32, 35, 43,
data on 144; investment and cash 97; asset prices 54; and bankruptcy
flow 80–90; share price fluctuations of Crédit Mobilier 113, 118;
60–4; share prices, monthly (1866–1868) bankruptcy of Crédit Mobilier,
149–51 impact of 55–9; data on 142, 143,
Compagnie des Chemins de Fer du 144; investment and cash flow 80–
Midi 32, 41, 43; asset prices 54; 90; investment regression 98–101;
bankruptcy of Crédit Mobilier, share price fluctuations 60–4; share
impact of 55–9; data on 143; prices, monthly (1866–1868) 149–51
investment and cash flow 80–90; Compagnie Parisienne de Distribution
share price fluctuations 60–4; share d’Électricité 43
prices, monthly (1866–1868) 149–51 Compagnie Parisienne du Gaz 32
Compagnie des Chemins de Fer du Compagnie Transatlantique 32, 36, 43;
Nord 41; asset prices 54; bankruptcy asset prices 54; and bankruptcy of
of Crédit Mobilier, impact of 55–9; Crédit Mobilier 112; bankruptcy of
data on 143; investment and cash Crédit Mobilier, impact of 55–9;
flow 80–90; share price fluctuations data on 143; share price fluctuations
60–4; share prices, monthly (1866– 60–4; share prices, monthly (1866–
1868) 149–51 1868) 149–51
Compagnie des Chemins de Fer compensation in common agency 15
d’Orleans 41, 43; asset prices 54; contracts: and banks, supervisory role of
bankruptcy of Crédit Mobilier, 21–3; competition for 21; terms in
impact of 55–9; data on 143; common agency 14
investment and cash flow 80–90; Crédit Lyonnais 120–3; credit policy of
share price fluctuations 60–4; share 123; definition 121; dividends,
prices, monthly (1866–1868) 149–51 distribution of 123; financial
Compagnie des Chemins de Fer de participation 122; specific activities
l’Ouest 43; asset prices 54; 121–2
bankruptcy of Crédit Mobilier, Crédit Mobilier: advances to companies
impact of 55–9; data on 143; (1856–1866) 32–3; advantages/
investment and cash flow 80–90; disadvantages of association with
share price fluctuations 60–4; share 116–24; affiliated companies 42–4;
prices, monthly (1866–1868) 149–51 bankruptcy of see bankruptcy;
Compagnie des Chemins de Fer PLM collateral, not asked for 33; control
41, 43; asset prices 54; bankruptcy of firms 3; credit policy 123;
of Crédit Mobilier, impact of 55–9; criticisms of 28; data (1852–1867)
data on 143; investment and cash 136, 137–8; data (1903–1914) 136,
flow 80–90; share price fluctuations 139; definition 121; dividends,
60–4; share prices, monthly distribution of 123; evaluation of
(1866–1868) 149–51 association with 60–4; excess
Compagnie d’Eclairage par le Gaz 41; volatility tests on stock market
asset prices 54; bankruptcy of Crédit 45–53; fall of 34–5; historical
Mobilier, impact of 55–9; data on background 29–30; liquidation
143 account of 36; monthly share prices
Compagnie Générale des Omnibus (1866–1868) 149–51; not-
(Paris) 31, 43; bankruptcy of Crédit affiliatedcompanies 44–5; projects
Mobilier, impact of 55–9; data on 31; reconstitution of 35–7;
175
representation on boards 31; specific firms: banks’ association with, analysis

activities 121–2; supervisory role, 117–20; control of by Crédit
limited 22; supervisory role of Mobilier 3; investment performance
111–16 of 23. See also bank-firm relations
Crédit Mobilier Français 28, 37–40 fixed effect model of investment 87–90;
credit policy: Crédit Mobilier and Crédit and depreciation rate 95–7
Lyonnais 123; in supervisory role of Fourchambault (company) 44, 45, 67;
banks 110 bankruptcy, impact of 68–70; data on
credit rationing 10 148; investment and cash flow
analysis 80–90
Darmstädter Bank 27, 31, 110 France: banking crash (1882) 48;
De Long, J.B. 26–7, 28, 31, 34, 40, 42, economy (1853–1914) 40–2; interest
44–6, 51, 65, 74, 76, 86, 117, 132 rate rises (1870–1872) 42;
debt: defined 12; inside 16–17 monitoring role of banks 51; share
Denain (company) 44, 67; bankruptcy, price indices 140–2; stock markets,
impact of 68–70; data on 147 volatility tests on 49–50; supervisory
Denuc, J. 79 role of banks, comparisons with
depreciation rate and investment 91–2; Germany 124–8
regression on 92–5, 98–104 Franche Comté (company) 44, 67;
d’Erlanger, E. 35 bankruptcy, impact of 68–70; data on
Deshons, M. 10 148
Dessirier, J. 46, 47 Freixas, X. 10, 13
Deutsche Bank 26 Friedman, M. 112
Devinney, T.M. 16, 17 Furet, F. 30
Diamond, D. 10, 20
Diskonte Gesellschaft Bank 26 Galai, D. and Masulis, R. 12
Dollfus Mieg et Cie 29 Gale, D. 10
Dresdner Bank 26, 126 Germain, H. 121, 123, 128, 129, 134
Dupont Ferrier, P. 30, 42 Germany: banks, caution of 49; banks,
short-term policy 22; capital scarcity
Electrometallurgie de Dives 41, 43 in 18–19; financial participation
Engberg, H.L. 52 125–6; investment policy 125;
enterprises, financing of 30–4 monitoring role of banks 51; stock
excess volatility tests on stock market markets, volatility tests on 49–50;
45–53 supervisory role of banks,
external finance and banks, supervisory comparisons with France 124–8
role of 16–19 Gerschenkron, A. 18–19, 20, 22, 23,
120, 130, 132
Fama, E.F. 17 Gille, B. 44, 48, 79
Fazzari, S. 77, 124 Gillet, M. 30
Feldenkirchen, W. 91 Givors (company) 44, 67; bankruptcy,
Filature (company) 67; bankruptcy, impact of 68–70; data on 147
impact of 68–70; data on 147 Gjesdal, F. 9
financial crises: and bankruptcy of Goodhart, C. 39
Crédit Mobilier 111–14; and real Grossman, S.J. 11
output 112
financial intermediation by banks 20; Hart, O.D. 11
and commitment 23–4; and Haussmann, Baron 35
investment 21; role of 113 Hellwig, M. 10
Firminy (company) 44, 45, 67; Hilferding, R. 130
bankruptcy, impact of 68–70; data on L’Home (company) 44, 67; data on 148
148; investment and cash flow 80–90 Hoshi, T. 23, 24, 77, 79, 80, 115, 133
176
INDEX
Houillière (company) 67; bankruptcy, Majluf, N. 24

impact of 68–70; data on 147 Marine (company) 44, 45, 67;
Hubbard, R.G. 77, 124 bankruptcy, impact of 68–70; data on
Humphrey, T. 39 148; investment and cash flow
analysis 80–90
information asymmetry 9, 10; and market finance and bank finance 19–20
banking panics 115; and bankruptcy Mathias, P. 132
of Crédit Mobilier 111; and Maubeuge (company) 44, 67;
investment 77; and role of banks 17 bankruptcy, impact of 68–70; data on
inside debt 16–17 148
internal finance: and agency costs 24; Mayer, C. 16, 21, 22, 23, 119
and banks, supervisory role of 16–19 Meckling, W.C. 9, 11, 12, 24
investment: and capital stock, regression Milde, H. 16, 17
on 82, 84–6; and cash flow, Miller, M. 3, 77
regression on 83–7, 152–60; and Mine Firminy 80
Crédit Mobilier 78–90; and Mine Saint Etienne 80
depreciation rate, regression on 91–5, Mines Grand Combes 80
98–104; and financial intermediation Mines Montrambert la Beraudière 80
by banks 21; fixed effect model Miskin, F. 111
87–90; and information asymmetry Modigliani, F. 3, 77
77; performance of firms 23 monitoring by banks: and contracts
21–3; as financial intermediation 20;
Japan: bank lending relationships 118; model of 17
bank-firm relations 23–4; investment moral hazard 9–10
and cash flow 124; investment Myers, S. 24
expenditure and cash flow 77 Myers, S.C. 10, 12
Jeidels, D. 78, 126
Jensen, M.C. 9, 11, 12, 24 Napoleon III 35
joint-stock banks 27
non-affiliated companies 44–5; and
Jorgenson, H.T. 77
affiliated companies, comparisons
J.P.Morgan Company 26, 28, 51, 77
71–3; data on 146–7; depreciation
rate and investment 92–5, 98–104;
Kalay, A. 10 effect of bankruptcy on 65–71;
Kashyap, A. 77, 79, 80, 115, 133 investment and cash flow 80–90;
Kindleberger, C.P. 111, 113 share price fluctuations 67–71;
Koechlin, A. 29
summary statistics on 105
Kose, J. 10
Nouette-Delorme, E. 53
Laffont, J.J. 10, 13 Ohio Life and Trust Company 111
Leland, H.E. 11 Oppenheim, A. 27
lender of last resort, central banks as
38–9
Paris, K. 129
Lenoir, M. 45, 46
Péreire, Emile 29
Levy-Leboyer, M. 47, 134
Péreire, Eugène 36
L’Home (company): bankruptcy, impact
Péreire, Isaac 29, 121
of 68–70
liquidation account, Crédit Mobilier 36 performance-related compensation in
liquidity: central banks, providing 39; common agency 15
and Crédit Mobilier 86; Germany Petersen, B. 77, 124
and France, comparisons 125; and Plenge, J. 43
investment 77 Pujo Committee (1913) 86
Loutchitch 41 Pyle, D.H. 11
177
Ramirez, C. 79, 86, 93, 95, 110, stock market and banks 51–2
116–17 Stole, L.A. 13
Riesser, J. 78 supervisory role of banks 15–24, 109–31;
risk-neutrality in common agency 14 association with, advantages and
Ross, S. 12 disadvantages 116–24; bank finance
and market finance 19–20; banking
Saint Simon, Henri, comte de 119 panic in bankruptcy 114–16; and
Sayous, A.E. 122, 125, 126 cartels 23; and contracts 21–3; and
Schaaffhausen’schen Bankverein 27 Crédit Lyonnais, comparisons 120–3;
Scharfstein, D. 77, 79, 80, 115, 133 credit policy in 110; finance 16–19;
Schartz, A. 112 financial crisis in bankruptcy 111–14;
Schneider (company) 44, 67; France and Germany, comparisons
bankruptcy, impact of 68–70; data on 124–8; intermediation and
147 commitment 23–4; and investment
Seillière, Baron 29 performance of firms 23; model of
share prices: affiliated companies 60–4; 17–18
and bankruptcy of Crédit Mobilier Sylla, R. 51–2
112; fluctuations 34; indices, France
140–1; monthly, affiliated companies Terrenoire (company) 44, 67;
148–51; non-affiliated companies bankruptcy, impact of 68–70; data on
67–71 147
shareholders on bankruptcy of Crédit Thalmann and Co. 37
Mobilier 128–9 Thornton, H. 38–9
Shiller, R. 45, 47 Tilly, R. 22, 125
signalling theory 1; basic concepts Tobin, J. 82, 91
8–12; model of 11
Société de l’Hôtel et des Immeubles de Union Fire 55–9; data on 143
la Rue de Rivoli 31, 32 l’Union (insurance company) 43;
Société d’Eclairage Électrique 43 bankruptcy of Crédit Mobilier,
Société des Machines Outils 44 impact of 55–9; data on 143
Société des Omnibus de Paris 43 United States: monitoring role of banks
Société Immobilière 34 51; stock markets, volatility tests on
Société Metallurgique de Montbard- 49–50
Aulnaye 43
Société Parisienne d’Eclairage 41 volatility ratios 49–52
Société Parisienne d’Eclairage et de
Chauffage par le Gaz 43
Société Raffineries Say 43 Wallut, C. 36
Solow, R.M. 39 Whale, B.P. 127
Stiglitz, J.E. 112 Wrück, K.H. 118
178

The Role of Banks in Monitoring

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

The Role of Banks in Monitoring

Încărcat de

Drepturi de autor:

Formate disponibile

THE ROLE OF BANKS IN

Does the Anglo-American approach to the relationship between banks and

Elisabeth Paulet is Research Assistant at the European University Institute

1. ECONOMIC IDEAS AND GOVERNMENT POLICY

London and New York

John Maynard Keynes

1 AGENCY THEORY AND MONITORING

2 THE CRÉDIT MOBILIER AND THE FRENCH STOCK

2.7 Comparative analysis of the results obtained for affiliated

3 CORPORATE INVESTMENT, CASH FLOW AND

4 THE SUPERVISORY ROLE OF THE CRÉDIT MOBILIER

4.1 Introduction 109

5 GENERAL CONCLUSIONS 132

A.1 The Crédit Mobilier between 1852 and 1867 136

1.1 Moral hazard with hidden action 10

2.1 Ownership of the shares of the Immeubles de la Rue de

3.2 Summary statistics comparing Crédit Mobilier and non-Crédit

3.26 Investment regression equation for all companies over the

Economists seem to agree on the monitoring power of banks as regards

This book is based on my Ph.D. dissertation at the European University

Bankers and investors influence industrial investment through the financial

The problem is the accurate transmission of such information, to give

• Its organisation was similar to that of the German Großbanken. The

These results strongly contradict earlier findings. The interpretation will

The Crédit Mobilier’s case gives no clear evidence to confirm this

(which were explicitly modelled on the Crédit Mobilier). The diverse

For French business, whether medium or ‘big’, the obsession

principal’s observation. The problem acquires interest only when there is

1.2 PRESENTATION OF THE FINANCIAL

(a) Basic concepts of agency and signalling theories

The financial theory of agency focuses on the relationships between

1 the monitoring expenditure by the principal;

Adverse selection can be described through three categories of model.

Figure 1.1 Moral hazard with hidden action

Figure 1.2 Adverse selection models

Figure 1.3 Signalling model

The latter situation, represented by Figure 1.3, is the signalling model.

• the manager of a good firm sends information to the external investors

A penalty system must be in place to prevent the manager of a bad firm

(b) Common agency

non co-operative behaviour induces an efficient (potentially second

In an earlier paper5 they discussed the possibility of collusion between

Agency as control is tied to compensation in at least three different

1.3 THE SUPERVISORY ROLE OF THE BANK IN

The bilateral principal-agent question considers that one individual, called

In this section, I study the theoretical framework concerning the role of

(a) Internal finance, external finance and bank finance:

defined by Fama (1985), is debt which explicitly incorporates monitoring by

1 The expected payoff of the project, µ is a function of management effort

Under these assumptions, how can the bank-stockholders management

stockholders. By assumptions 6 and 8, managers cannot do anything but

• an increase in the fraction of internal equity increases the manager’s

Banks can monitor the manager’s choice of s and/or participate in the

Theoretical presumption in historical assessment

Gerschenkron (1962) studied this notion of ‘capital scarcity’ and applied it

Gerschenkron’s argument is as follows: over the period, the firms’

1 Why should an insufficiency of internal finance be seen as characteristic

The first question really concerns the role of internal finance in an

(b) The advantage of bank finance versus market finance

riskiness of the firms strategies and moral hazard concerning reported ex

1 Monitoring and control of a firm involve natural scale economies: a

(c) Monitoring and the definition of contracts between banks

‘Monitoring’ ought to be understood in a broad sense as any form of

• Monitoring of the firm’s profit returns makes it possible to conclude