The Power and the Money: The Mexican Financial System, 1876-1932 9781503619678

Citation preview

THE POWER AND THE MONEY

SOCIAL SCIENCE HISTORY

Edited by Stephen Haber and David Brady David Brady and Mathew D. McCubbins, Party, Process, and Political Change in Congress Jeffrey Bortz and Stephen Haber, The Mexican Economy, 1870-1930 Edward Beatty, Institutions and Investment Jeremy Baskes, Indians, Merchants, and Markets

THE POWER AND THE MONEY The Mexican Financial System, 1876-1932

NOEL MAURER

STANFORD UNIVERSITY PRESS

Stanford, California

Stanford University Press Stanford, California

© 2002 by the Board of Trustees of the Leland Stanford Junior University

IT.:Im Published with the assistance oflnstituto Technol6gico Aut6nomo de Mexico.

Library of Congress Cataloging-in-Publication Data Maurer, Noel. The power and the money : the Mexican financial system, 1876-1932 I Noel Maurer. p. cm.-(Social science history) Includes bibliographical references and index. ISBN 0-8047-4285-5 (cloth: alk. paper) 1. Finance-Mexico-History. 2. Financial institutions-Mexico-History. 3. Banks and banking-Mexico-History. I. Title. II. Series. HG185.M6 M38 2002 332' .09034-dc21

2002003103

Original Printing 2002 Last figure below indicates year of this printing: 11 10 09 08 07 06 OS 04 03 02 Typeset by G&S Typesetters in 10.5/13 Bembo

For my parents, Leon and Femia, the best people in the world

CONTENTS

Figures and Tables Acknowledgments

lX Xlll

Introduction 1.

Political Instability and Financial Concentration

13

2.

Structuring the Credit Market

34

3.

Central Banking in Porfirian Mexico

48

4.

The Consequences of Credibility

70

5.

Financial Capitalism in the Porfiriato

93

6.

Banks and Industrial Concentration

115

7.

Financing a Revolution

134

8.

Reconstructing Confidence

160

Conclusion

195

Notes Bibliography Index

203 237 245

LIST OF FIGURES AND TABLES

Figures 1.1.

Federal Budget Deficits in Early Porfrrian Mexico

17

1.2.

Short-term Gross Federal Borrowing

17

1.3. Banamex Specie Reserves During the Financial "Coup-d'etat" of1885

26

2.1. Banking Market Shares

45

2.2. Herfindahl Index for Porfirian Banking

46

3.1. Banknote Circulation

55

3.2. Banknotes as a Percentage of All State Bank Liquid Reserves

57

3.3. Banknote Holdings

58

3.4. Specie Reserve Ratios

61

3.5.

Ratio of State Banknote Issues to Specie Reserves

63

3.6.

Ratio of State Banknote Issues to Other Liabilities

64

3.7. Banco Central Specie Reserve Ratios

65

4.1. Risk-return on Assets

72

4.2.

73

Risk-return on Equity

4.3. Risk-return on Equity, Close-up

74

4.4. Actual and Predicted State Bank Liquidity Ratios

86

4.5. Banamex Actual and Predicted Liquidity Ratios

86

4.6. Banco de Londres y Mexico Actual and Predicted Liquidity Ratios

87

4.7. Banamex Share ofDeposit Market

88

4.8. Banamex Loan/Deposit Ratios

89 ix

List if Figures and Tables

X

4.9.

Mexico City Nominal Interest Rates

4.10. Gross Bank Lending as a Percentage ofGDP

90 91

5.1.

Loan-asset Ratios for BNL and BMM, 1893-1907

109

5.2.

Porfirian Bank Liability Structure

112

6.1.

Distribution of Firm-types in Mexico

121

6.2.

Share of Total Textile Output

121

6.3.

Average Value of Output, Logs

122

6.4.

Average Spindlage per Firm

122

6.5.

Output per Worker, by Type of Firm

125

6.6.

Output per Spindle, by Type of Firm

125

6.7.

Concentration in Mexican Textiles

132

7.1.

Changes in the Banco de Guanajuato's Balance Sheet, March-June 1912

140

Monthly Rate oflncrease in the Peso/Dollar Exchange Rate, in Annualized Terms, 1913-16

150

8.1.

Real Banking Stock Index

164

8.2.

Total Size of Banking System

182

8.3.

Banking Market Herfindahl Index

186

8.4.

Risk-return on Equity

190

8.5.

Capital and Liquidity Ratios

191

7.2.

Tables 3.1.

Banamex Note Discounts

54

3.2.

Banamex Advances to Other Banks

59

4.1.

Yields on Banking Common Stock

76

4.2.

Tobin's q Estimates for Porfirian Banks

78

4.3.

Liquidity Preference Regression Results

84

5.1.

Average Inflation-adjusted Rates ofReturn on Porfirian Railways

105

6.1.

Which Firms Grew Fastest?

123

6.2.

Spindlage Growth Regressions

124

List of Figures and Tables

xi

6.3. Which Firms Were More Productive?

126

6.4. Were Bank-connected Firms Different?

127

6.5. Dividend Payments, Current Pesos

128

6.6. Liquidity Sensitivity Test Results

130

6.7. Survival Analysis, 1893-1913

131

7.1. Forced Loans from Revolutionary Governments

158

8.1. Assets Denominated in Foreign Currency, 1925-31

178

8.2. Survival Regressions

184

8.3. Banking Profits

188

ACKNOWLEDGMENTS

Much of the research for this book depended on the extensive resources of the Archivo Hist6rico del Banco Nacional de Mexico. I would like to particularly thank Jose Antonio Batiz for his help and forbearance, making special materials available and allowing me to continue working past the archive's scheduled hours. In addition, I would like to thank Carlos Contreras and Armando Rojas at the Archivo General de la Naci6n, for helping me find newly catalogued records and lending me their research materials. Without their aid, this work would not have been possible. Funding for the initial research for this study came from the Graduate Research Office and the Department of Latin American Studies at Stanford University. Support from the Instituto Tecn6logico Aut6nomo de Mexico and the Social Science History Institute of Stanford University was crucial for the completion of the manuscript. I am deeply indebted to both ITAM and SSHI. I am also deeply indebted to my advisor, colleague, and good friend Steve Haber, who encouraged and cultivated my interest in Mexican history, counseled me on research design, and read the entire manuscript several times. Much of Chapter 7 is based on joint work with Steve, and he generously shared painstakingly assembled data sets with me. I am also indebted to Carlos Marichal, who read the entire manuscript and provided multiple concrete suggestions for its improvement. I wish to thank Avner Grief and Gavin Wright, who as doctoral thesis committee members provided irreplaceable and essential criticisms of my initial interpretations and suggestions about where to go next. Aurora Gomez generously shared her research with me. Without her aid this book would not have been possible, and I owe her a debt of gratitude. Ted Beatty also read the entire manuscript and provided invaluable research and writing advice. It gives me great pleasure to acknowledge a long list of people whose comments and suggestions have greatly aided this endeavor. Different parts of the book have benefited from participants at economic history workshops at Stanford University, ITAM, the University oflndiana, UCLA, Columbia Xlll

xiv

Acknowledgments

University, the Colegio de Mexico, CIDE, the Universidad Torcuato di Tella, and the Banco Central de Colombia. I would like to emphasize the importance of my colleagues in the Department of Economics and Centro de Investigaci6n Econ6mica at ITAM. The active surroundings acted as a hothouse for the latter stages of my research. I benefited repeatedly from discussions with my colleagues: Alexander Elbittar, Alejandro Hernandez, G6nzalo Hernandez, Silvia Hernandez, Andrei Gomberg, David Kaplan, Todd Keister, Fritz Laux, Cesar Martinelli, Francese Obiols, Sangeeta Pratap, German Rojas, Joyce Sadka, Julio Santaella, and especially Tridib Sharma, whose theoretical work provided many of the key insights for Chapter 6. One could not ask for a better or more productive academic environment. Numerous other people merit thanks for helpful comments, suggestions, and discussions at various stages of the work: Gustavo del Angel, Arturo Grunstein, Paul Ingram, Sandra Kuntz, Naomi Lamoreaux, Juliette Levy, Leonor Ludlow, Elisa Mariscal, Chiaki Moriguchi, Petra Moser, Aldo Musacchio, Armando Razo, Ken Sokoloff, Bill Summerhill, and Joachim Voth. This is a much better work for their help and input. Any and all errors are, of course, wholly my own.

THE POWER AND THE MONEY

Introduction Public and private credit are closely allied, if not inseparable. -Alexander Hamilton

In 1909, on the eve of the Revolution, Mexico possessed forty-one formally incorporated financial institutions of various kinds. 1 Two of those institutions, the Banco Nacional de Mexico (Banamex) and the Banco de Londres y Mexico, together controlled more than half the assets in the banking system. In the same year, the United States enjoyed the services of18,723 banks and trust companies. 2 High concentration has continued to characterize Mexican banking down to the present day. In fact, Mexico currently has fewer financial institutions than before the Revolution of 1910, and the two biggest banks continue to collectively control roughly half the assets in the banking system. Banamex was still the nation's single largest bank as recently as two years ago. Why did such a concentrated fmancial system emerge, why did it persist for so long, and what effects did it have on the Mexican economy? This book argues that financial concentration was the product of the way Mexican governments made credible commitments to respect private property rights. Regardless of their stated ideologies, Mexico's rulers recognized that they needed to come to an accommodation with potential creditors. First, they understood that they needed a source of credit in order to restore political order. Second, they understood that without a functioning fmancial system, there could be no economic growth, and without economic growth, there would be no tax revenues. In short, they realized that restoring the financial system was crucial to their own political survival. 1

2

Introduction

Mexico after independence experienced more than fifty years of political chaos until Porfirio Diaz assumed power in 1876. Thirty-five years later, it entered another prolonged period of instability (1910-29), punctuated by one of the great social revolutions of the modern world (1913 -17). There is no economic sector as sensitive to political instability as the financial system. Under instability, governments and factions aspiring to be governments have strong incentives to steal bank cash reserves, force financial institutions to make them loans, engage in the unrestrained printing of currency (thereby setting off an inflation that will essentially be a tax on holding cash), and change the rules that regulate banking and the securities markets so as to maximize the government's access to funds. The story of Mexico's financial system is therefore fundamentally political. How did governments facing armed resistance and a real chance ofbeing violently overthrown credibly promise to repay their debts? Porfirio Diaz's strategy in the 1880s was to create a bank with a legal monopoly over lending to the federal government. Diaz's regime also selectively enforced property rights to give elites tied to powerful local strongmen-cum-politicians a stake in the political system. The threat of revolt by these strongmen assured politically connected local elites that the federal government would not prey on their wealth. The result was an extremely concentrated banking system, which interacted with the politicized nature of property rights in Porfirian Mexico in such a way as to produce an extremely concentrated industrial structure. The Mexican Revolution (which began in 191 0) violently ended the Porfirian regime, but it failed to alter the basic political calculus. The new rulers came to power during a period of extreme political instability. Between 1910 and 1929, the country underwent a revolution, a counterrevolution, a counter-counterrevolution, three civil wars, and four violent coups or attempted coups. Just as Diaz had done, the leaders who attained power in 1920 selectively enforced property rights and allowed the bankers themselves to write the laws regarding entry into banking. In addition, the government created a hostage: a government-owned commercial bank (the Banco de Mexico, or Banxico). As a result, the domestic banking system recovered rapidly during the 1920s, despite ongoing violence and political instability. The result was the same as the Porfiriato: a continuing high level of financial and industrial concentration. This is not to say the Revolution had no effects -but in the long run, the Revolution failed to sever the links between banks and politics. If anything, it strengthened them. In more general terms, this book's argument runs as follows. Governments establish politically connected private institutions with a monopoly

Introduction

3

over lending to them in order to establish credibility. By giving one institution a monopoly, a sovereign government enables lenders to coordinate their actions in the case of nonpayment. This thereby increases the government's credit limit by raising the penalty attached to defaulting. Political connections reinforce the government's credibility, by transforming bankers from residual to priority claimants on the success of the state. The more unstable the political situation, the more important these connections become. The strategy involves granting the government's bankers special privileges, both as a compensation for risk and a result of the bankers' political connections. These bankers will therefore be well positioned to engage in rentseeking. In the absence of secure property rights, this can retard the development of the financial system, limit access to capital, encourage concentration, and slow economic growth. Why are financial systems vital for the functioning of a country's economy? In essence, because they match savers to investors. Financial systems consist of the interconnected network of banks, brokers, and exchanges that raise, securitize, distribute, trade, and continually value assets. They turn illiquid physical assets into liquid contracts that can be traded, sold, or collateralized. At the simplest level, these contracts are banknotes, which are nothing more than a promise by a bank to redeem a piece of paper for gold or silver. At a more complex level, these contracts can be shares in corporations, which entitle the holder to a percentage of the profits of that corporation for the duration of the corporation's existence. Even more complex financial contracts can be written, providing different rights contingent on different external events. Regardless of the complexity of the contract, the underlying function of the financial system is twofold. It eliminates the need for parties in a financial transaction to have to have direct knowledge of one another, and it makes financial contracts easy to value and exchange. Families, of course, can partially fill in for the lack of a financial system. Family members can share resources with each other. A father, for example, may lease land to a son, or an uncle might lend equipment to his nephew. Private individuals can also lend or transfer wealth among each other, based on their personal relationships and reputations. Unfortunately, families and family networks are only so big. An economy that relies solely on them for capital greatly limits its potential for future growth. Financial systems match savers to investors through two types of intermediaries: banks and securities markets. Banks take in funds from depositors, offering interest in return for the use of the funds. Banks also raise capital by selling equity in themselves (shares of stock) or by long-term borrowing (bonds). These funds, regardless of their source, are then lent at interest.

4

Introduction

Thus, banks serve to connect savers and borrowers-without the need for these individuals to actually know one another. In addition, the underlying contracts that allow banks to take in and lend out funds makes the wealth represented by those contracts highly liquid. Thus, to cite an obvious example, savers can transfer bank deposits in order to make payments-this is what happens when bills are paid by a check. The crucial advantage is that the underlying projects that those deposits finance do not have to be liquidated in order to make the payment. Securities markets, and the brokers who underwrite and trade on those markets, allow firms to sell claims on their future earnings for immediate cash. Endless numbers of derivative contracts have been created in recent years, but there are basically two sets of contracts that brokers and securities markets create and trade: equity participation in a firm (that is, shares of stock); or long-term loans (that is, bonds). The crucial feature ofboth sets of contracts is that they are liquid. Stocks and bonds can be sold without liquidating the underlying investment, and buyers and sellers do not need to have direct knowledge of one another. In theory, banks and securities markets are substitutes for one another. As a practical matter, however, banks and securities markets are closely interconnected and their activities are interwoven. To cite an obvious example, banks might raise capital by selling equity shares in themselves. Without the existence of a securities market to make the bank's shares liquid, few investors would purchase them. Similarly, banks may serve as underwriters of new shares in other companies, essentially purchasing the shares from the firm and then reselling them on a secondary market. Banks may also discount commercial paper (essentially a promissory note of short duration issued by a firm) that can then be traded on a market. Indeed, one of the primary functions of banks in the nineteenth century was precisely this kind of discounting. A firm would write a promissory note to a supplier that it would pay a debt in a specified number of days. The holder of the note would then sell the contract to a bank, which would give him or her immediate cash, but at a discount off of the face value of the note. These notes could also be traded on a securities market, where they might be bought and sold by banks as well as by private individuals. The functioning of the organizations and markets that make up the financial system are crucial to private individuals, firms, and governments. The financial system allows private individuals to smooth their consumption and make purchases that would require them to save for an inordinately long time from their current income. Similarly, the financial system allows firms

Introduction

5

to expand faster than would be possible had they to rely solely on their own reinvested profits. The financial system also allows firms to borrow their way through downturns in the business cycle, thereby allowing them to hold more of their assets as productive tangible capital (buildings, machines, stocks of raw materials, and inventories kept on hand during the production process), and less as unproductive cash (hoarded to protect the firm in a downturn). Finally, the financial system allows governments to raise revenues in excess of those it can obtain through taxation. Essentially, financial systems allow governments to raise funds by writing a contract that allows them to obtain immediate cash in exchange for a claim on future tax revenues. Without an efficient financial system, the cost of capital faced by firms, governments, and individuals will rise. Inefficient financial intermediation will hamper new investment because investments will be less liquid, and the time horizons of investors will shorten. Without a stable payments system, transactions costs will rise dramatically, shortening both time horizons and the ability to sell goods and services at a distance. In short, the functioning of modern economies is intimately tied to the functioning of their financial systems. Governments have a special role in the formation and evolution of financial systems. First, the public sector influences the development of financial intermediaries and markets through its own demand for credit. Second, the government establishes the rules that govern the functioning of financial intermediaries. Financial systems are intended precisely to allow claims on real -and often immobile and illiquid-resources to be represented by relatively liquid contracts. Rules laid down by the government determine the security of these claims, and how easily they may be traded or transferred. In fact, the government not only deterniines the security of financial contracts but also who can create them and the circumstances under which they can be created. It is the government, not the market, that determines which group of financiers will receive a bank charter, the level of reserves that the bank must hold, and the composition of those reserves. In short, the government is a participant in the financial system and simultaneously governs the financial system. There is, in short, a holdup problem between government and the financial system. Governments need access to credit from the financial system. The financial system needs the government to enforce property rights, without which economic agents cannot write or enforce financial contracts. The privileged role of the government, however, means that it can behave opportunistically. Banks and private individuals, foreseeing this, will decide not

6

Introduction

to invest at all. Unless the government can tie its own hands, the financial system will not develop and the government will be denied the very resources it needs for its own survival. Porfirian Mexico solved the holdup problem through the creation of a specially privileged bank. The newly established regime desperately needed money in order to carry out its basic functions, but it lacked the administrative capacity to raise internal taxes or the reputation to contract international loans. Many other countries have at some point in their history solved a similar dilemma in a similar manner. The Bank of England, for example, became England's premier financial institution by swapping its shares for debt issued by the war-pressured English monarchy in 1694. The monarchy benefited by concentrating its debt in a single institution, which could be induced to offer lower interest rates and easier repayment schedules. In return, the Bank ofEngland became the sole repository for the government's tax and loan receipts. It earned fees from transferring government funds overseas and facilitating interest payments on England's various debt issues, and enjoyed insider knowledge about future government plans. 3 Like England, the Porfirian regime needed to create a single bank large enough for its financial needs. Without such a bank, the government would have been unable to carry out its most basic functions. This would have ensured additional decades of political instability. Declaring that Mexico did not need a bank with special privileges would have been the equivalent of declaring that Mexico did not need an effective national government. Chapter 1 discusses the basic credit problem the Porfirian government faced: How can a sovereign debtor credibly promise that it will repay its debts? President Porfirio Diaz needed to establish a bank large enough to provide his government with credit and make a credible commitment to that bank's owners that his government would repay its debts. In 1884 the Diaz government therefore engineered the merger of two preexisting banks into a semiofficial superbank, the Banco Nacional de Mexico (Banamex). The charter that created this superbank granted it a monopoly on all federal lending and control over all federal spending-all tax collection and federal payments passed through its hands. Since all loans to the government now went through a single intermediator, the government could no longer strategically default on some creditors while continuing to borrow from others. The Porfirian government had in fact tried to play this game in the early 1880s, which provoked a series of financial crises. The temptation to circumvent Banamex's monopoly meant that, in practice, credibly granting a private bank a monopoly over lending was not as simple as theory might imply. The government had to make it impossible to borrow without Bana-

Introduction

7

mex's knowledge or permission. The result was that by 1885 a private institution had been given effective control over the implementation of all public spending. In short, creating Banamex solved the coordination problem among lenders and insured that the government would face a credible credit boycott should it default. This was particularly important in reestablishing the government's credibility with foreign creditors. The Porfirian government, however, had to satisfY two not entirely consistent goals. First, it had to secure a source of credit for itsel£ Second, it had to give powerful local strongmen a stake in the national political system. Chapter 2 describes how the Porfiriato reconciled these goals. The long-run solution required that a deal be brokered among all of the players: the stockholders of Banamex, the stockholders in the country's second-largest bank (the Banco de Londres y Mexico), the stockholders in smaller banks, and the powerful state governors. The resulting arrangement shared many (although not all) ofBanamex's special privileges with the Banco de Londres y Mexico. The state banks were given partial local monopolies, and the state governors were enabled to award federal concessions to their cronies. Holding the arrangement together was the fact that the federal government reserved for itself the right to charter banks. Thus, competition among states for bank business could not ratchet downward the legal barriers to entry into banking. No state could unilaterally liberalize, because this would damage the interests of the other state governors, who would support the federal government in disciplining the liberalizing state. Nor could the federal government behave opportunistically by unilaterally changing the rules, because that might spark a revolt among the state governors who benefited from the arrangement. Politicians and bankers justified the advantages granted Banamex by claiming that it fulftlled the functions of a central bank of rediscount, providing liquidity to the rest of the banking system in times of need, calming panics, and supplying a uniform national currency. Chapter 3 analyzes these claims. It musters quantitative and qualitative evidence to show that Banamex carried out none of the above functions. In fact, despite claims by its own officials, the Porfi.rian government recognized that the banking system lacked an effective lender of last resort. IfBanamex was little more than a commercial bank with special privileges, then what were the costs of those privileges? In other words, did Banamex succeed in taking advantage of its rent-seeking opportunities? The answer is yes. Chapter 4 examines the public costs of the Porfirian financial system. It uses a variety of quantitative tools to demonstrate that after risk is properly taken into account, Banamex earned returns in excess of those available in a

8

Introduction

competitive market. These estimates, and other data reported by the banks to the Finance Secretariat, are then used to test the hypothesis that the two national banks exercised market power over Porfirian lending. The results indicate that they did, to the detriment of the overall efficiency of the financial system. How did this banking structure affect the rest of the economy? In order to answer this, we need to know how the protected banks ofPorfirian Mexico made their investment decisions. In other words, how did these banks go about their business? How did they match savers to investors? Chapter 5 focuses on the individual banks and their commercial strategies, using bank minutes and private credit reports to demonstrate that Porfirian banks overcame information asymmetries by making long-term loans and investments to individuals and firms connected to their directors. This practice is known as insider lending. Elite entrepreneurial families chartered banks in order to channel capital into their enterprises. They raised funds first by selling stock in the bank on the Mexico City stock market. Investors purchased these stocks knowing they were, in effect, investing in the network of enterprises controlled by the bank's directors. Chapter 5 argues that insider lending was not deleterious in and of itsel£ Rather, it allowed Mexican banks to serve as "engines of economic development" and financed a great deal of Mexico's economic growth. The interaction of widespread insider lending and insecure property rights, however, contributed to a high level of industrial concentration. Chapter 6 discusses a simple model oflending when property rights are uncertain. It then uses evidence from the Mexican textile industry to test the model's predictions. The results indicate that firms linked to banks grew far faster than their competitors. Bank-linked firms were no more productive or profitable than their competitors-but they had access to outside capital and credit while their competitors did not. Insecure property rights meant only the largest companies could circumvent the restrictions on banking by going directly to the securities markets for capital. In other words, the only interests capable of getting around the regulations were those the regulations were designed to benefit. This and other, real and perceived, imbalances in Mexico's growth processes set off the Mexican Revolution of1910-20, an extremely violent and disruptive interregnum, in which more than one million Mexicans died or fled to the United States. The Porfirian commitment mechanisms collapsed in the face of armed revolt. Chapter 7 describes the confiscations, hyperinflation, sequestrations, and general uncertainty of the period. It details the

Introduction

9

ever-shifting but increasingly hostile relationship between the bankers and the various revolutionary factions. Political stability did not magically return after the last successful violent change of government in 1920. Attempted coups, warlordism, and sporadic outbreaks of civil war continued until 1929. The new government faced the problem of reestablishing credibility under conditions not dissimilar to those faced by Porfirio Diaz in 1876. Chapter 8 examines how postrevolutionary governments, under the financial leadership of Alberto J. Pani and Manuel Gomez Morin, succeeded in reestablishing a credible commitment to refrain from confiscation and hyperinflation and to repay new domestic debts. The essence of the strategy was the same as Don Porfirio's. The bankers and other potential creditors were invited to write the rules governing their own activities. In addition, the government used its access to petroleum tax revenues-a source of income not available in the 1880s-to proffer a "hostage" to its creditors in the form of the state-owned Banco de Mexico, or Banxico. The government put up 50 million pesos in hard currency, which was then lent out to the bankers or used to cheaply finance the "entrepreneurial" activities of selected politicians. Altering the rules of the game afterward would therefore lead to the loss of this capital and damage the interests of politically powerful individuals. In this way, an undemocratic regime with a record of predation succeeded in committing to protect an important group of private property holders. This commitment did not lose credibility until the 1970s. The new regime reinstitutionalized old barriers to entry and created new opportunities for rent-seeking. It also tied public confidence in the banks almost completely to confidence in the ruling party and further politicized the provision of credit in Mexico. The book draws on a multiplicity of sources to draw its conclusions. The primary sources for information about the internal functioning of Porfirian banks came from the minutes of Banamex's board of directors, available in the Archivo Hist6rico del Banco Nacional de Mexico (AHBNM), and those of the Banco Mercantil de Veracruz, found in Galeria 2 of the Archivo General de la Naci6n (AGN). Additional data about bank lending patterns in the Porfiriato came from the minutes of the directors of the state-run Caja de Prestamos and the credit books published by the R. G. Dunn credit agency's Mexican branch. The Caja's minutes were found in Galeria 2 of the AGN, and the AHBNM contained the R. G. Dunn volumes. These were supplemented by the memorias published by the Finance Sec-

10

Introduction

retariat, and available in the Bancroft library at the University of California's Berkeley campus, and various articles published in the contemporary financial press. Issues of both the Economista Mexicano and the Mexican Herald are available on microfilm at Stanford University in Palo Alto, California, and copies of the Bolet{n Financiero y Minero are available at the Hemeroteca Nacional, located on the main campus of the Universidad Nacional Aut6noma de Mexico (UNAM) in Mexico City. Quantitative data was gathered mainly from monthly reports submitted to the Finance Secretariat and published in the financial press, or from the market pages of the Economista Mexicano and Bolet{n Financiero y Minero. Postrevolutionary information came from Finance Secretariat reports, the Boletin Financiero y Minero, and the reports of the Comisi6n Nacional Bancaria, available at the Cosio Villegas library of the Colegio de Mexico in Mexico City and the Hemeroteca Nacional. Other data came from the annual reports of the Banco de Mexico. What are the implications of this study for the historiography of Mexico and an understanding of economic growth and development? First, undemocratic dictatorships can sustain partial credible commitments to respect property rights. Mexico sustained just such a commitment on the basis of specific promises to a subset of asset holders-the holders of federal bank charters. Commitments, therefore, are not simply credible or incredible. Limited commitments can sustain economic activity, at least for a time. There are, however, economic costs to being governed by a dictatorship-cronyism is a second-best solution compared to the rule oflaw. Second, history matters. The outcomes of short-term political processes can shape and constrain events decades later. Banks may everywhere and at all times face certain problems, and exist to carry out certain economic functions, but they do not inevitably emerge out of economic theory. Rather, they emerge from contingent factors that are often political and driven by the instrumental needs of the government. These factors then shape banks' structure and mode of operation long after the government's immediate instrumental needs have passed. Third, individual banks in Mexico operated much like their counterparts in the early-nineteenth-century United States, lending to their own directors or people closely associated with them and serving as conduits for attracting impersonal capital from across the region (via deposits), the nation (via equity sold on the Mexico City stock exchange), and overseas (via equity sold elsewhere and wealthy immigrants to Mexico). Far from being pernicious or fraudulent, this practice allowed banks to overcome the scarcity of good financial information about outside credit risks. Social linkages com-

Introduction

11

pensated for poorly enforced property rights. When combined with the general insecurity of property rights, however, insider lending contributed to an extremely concentrated industrial structure, which persisted until the late 1990s. Only the firms with social and political links to the distributors of banking charters could access outside capital. The firms that grew the fastest, therefore, were the best-connected firms, not the best firms. This was the cost of dictatorship. Perhaps, over time, a decentralized, democratic, and stable political system could have emerged from Mexico's unstable nineteenth century. That, however, is not what occurred. Moreover, it failed to occur twice, after the violence of the nineteenth century and after the violence of the Revolution. In both cases, historical circumstances and the government's immediate needs meant that the second-best theoretical option was the first-best available option. Implementing the rule oflaw is not always historically feasible. The real alternative in the 1880s and again in the 1920s was not limited democratic government, but rather continuing chaos. Only in the past few years has a pluralistic democracy emerged in Mexico and with it the hope that the constraints of the past might be overcome.

Chapter 1

Political Instability and Financial Concentration The Executive believes that in the present situation a multiplicity of credit institutions might be subject to grave upheavals and the danger of bankruptcies, which would drag down fortunes great and small, compromising the dearest interests of society. -Official declaration of Mexico's Finance Ministry, 1884

How did Banamex and the Banco de Londres y Mexico become Mexico's dominant commercial banks? Why did extreme concentration characterize Porfirian banking at its inception? This concentration was not the result of impersonal economic forces, but contingent upon Mexico's political and economic circumstances in the 1880s. Mexico's weak federal government desperately needed resources in the 1880s. Therefore, in 1881 it created a bank. In theory, a bank with a monopoly over lending to the government would improve the government's ability to borrow. In practice, the government continued to try to evade the monopoly, borrowing indiscriminately from various and sundry sources. This provoked a severe financial crisis in 1883-85. To avoid a repeat of the experience, the federal government gave a private bank (Banamex) almost complete control over federal finances. In addition, Banamex received a number of extremely lucrative additional privileges, including a monopoly over the issuance ofbanknotes. The profits from these privileges could crosssubsidize lending to the federal government. The deal ran into political opposition centered around a branch of a British bank, the Banco de Londres y Mexico (BLM). Prominent politicians were tied to BLM. The result was that Banamex lost its monopoly over note issue. Many of Banamex's other privileges were extended to the BLM. The two banks formed a cozy near-duopoly over Mexican finance until the 1890s.

13

14

Political Instability and Financial Concentration

Banamex's privileges were diluted, but not dissolved. It controlled more than half of the assets in the banking system through the end of the century. Concentration, therefore, was a result of the federal government's penury. It lacked the administrative capability to finance its needs through taxation. Mexico's history of debt defaults locked it out of European markets. It overused thin local debt markets when they emerged. The result was that the government could not afford to take a more hands-off approach to the banking system. It needed a bank capable of financing its needs. The bank needed special privileges to make credible the government's commitment to borrow only from it. A more open financial system would have only ensured that Mexico lacked a stable national government for decades more.

Historical Background General Porfirio Diaz seized power in 1876. He remained in office until 1911, with only a brief interregnum in 1880-84. Diaz so dominated the country that a derivation ofhis first name, the "Porfiriato," came to be used for the entire period. Before the Porfiriato, a seemingly endless series of caudillos passed through the Presidential Palace. For every constitutional president between 1821 and 1876, there were four interim, provisional, or irregular presidencies, and the lifespan of the average administration scarcely surpassed eight months. 1 Unstable governments failed to maintain effective control over the national territory. The nations that currently make up Central America declared their independence from Mexico in 1823. In 1834 Texas seceded. In 1848 Mexico lost California, Arizona, and New Mexico to the United States. Yucatan declared independence and maintained it for more than a decade, returning only when faced by a devastating revolt by the Maya peasantry. Secessionist movements plagued Mexico's northern territories. Only the American threat kept these provinces from seceding after 1848. 2 Mexico's states were ran as virtually independent fiefdoms. They levied tariffs (alcabalas) whenever goods crossed a state line. This provided them their own revenue sources and fragmented Mexico into a collection of small and detached local economies. Mexican states, in fact, themselves regularly fragmented or invaded each other during the nineteenth century. Campeche, for example, violently seceded from Yucatan in 1857. 3 Governor Santiago Vidaurri of Nuevo Leon annexed Coahuila and installed himself as governor ofTamaulipas. His empire later violently collapsed. 4 Mexico State hived off three other "sovereign" states in the course of the nineteenth century. Jalisco and Zacatecas similarly balkanized. 5

Political Instability and Financial Concentration

15

The states failed to maintain order within their own territories. Banditry and brigandage became ubiquitous. At one point, Mexico's national Congress failed to meet because the representatives feared the trip from their constituencies to the Federal District. Historical records indicate that most travelers expected to be robbed at least once on a long journey. On one trip in 1861, the Mexico City-Puebla stagecoach was robbed three times. In another two-week period in 1865 a robbery occurred along the Mexico CityOrizaba highway every four days. 6 As if interstate tariff barriers and widespread highway robbery were not sufficient obstacles to commerce, Mexico's road network fell into ruin by the 1870s. Since Mexico lacked navigable waterways and its major population and economic centers were located far inland, most traffic had to move by road. Unfortunately, state governments regularly tore up the roadways running through their territories in order to protect their autonomy. 7 The French-backed Imperial regime began to repair major roads in 1862-67, but subsequent governments failed to continue the program for lack of funds. By 1876, half of all federal roads were unsuitable for wheeled traffic. 8 Railroads could have solved the transport problem, but there was no way to raise the capital needed to construct them. The federal government lacked the funds to repay its own debt (or repair existing roads), let alone engage in ambitious construction projects. In fact, the government was too broke to consistently pay its army. Destitute soldiers acted little better than the bandits they were notionally repressing. Military leaders routinely rebelled for the opportunity to pillage. 9 For obvious reasons, private investors were highly reluctant to risk their capital in this environment. Mexico's first railroad did not open until 1873, and in 1877 a country larger than the United States east of the Mississippi possessed only 400 miles of track, of which 18 percent used mules instead of steam engines to pull the trains. 10 Porfirio Diaz seized control in 1876, believing that a strong and centralized state would bring prosperity. But prosperity required stability, and stability required resources to invest in tangible capital (like railroads), pay the army, and buy off opponents of the new regime. How were these resources to be raised?

Mexican Money Markets in 1880 and the Cost of Political Stability Insuring peace and domestic tranquility in Mexico required a great deal of investment in troops, guns, prisons, and the means to support them. Unfortunately, Porfirian Mexico had no convenient source of credit to which it

16

Political Instability and Financial Concentration

could turn. The country had been anathema in international capital markets since 1867, when the French-backed Empire collapsed and the Restored Republic refused to recognize the Empire's debts. That meant the Porfiriato had to finance its expenses from tax revenues or domestic sources of credit. Taxes were no solution. The economy was tiny. Even had Mexico been richer, the government lacked the necessary administrative capacity. It needed to pay tax collectors in order to collect taxes. That left the domestic credit market. Before 1859, a money market of sorts had existed in Mexico City. It primarily traded government securities, particularly "customs certificates" used for the payment of import and export duties. 11 Mexican governments also issued a wide variety of short-term promissory notes, at interest rates varying from 30 to 200 percent. A thin stock market existed around trading shares of the Banco de Avio, the privatized tobacco monopoly, and a motley collection of mining companies. 12 Unfortunately, by 1876 the Mexico City capital market was moribund. The major enterprises of the first half of the nineteenth century, including the Banco de Avio and the tobacco monopoly, had gone bankrupt. The government's debt remained, but only on paper. In 1876, private Mexican financiers held federal debts with a face value of Mx$45 million, but since no one expected them to be repaid, they traded for much less. 13 The Restored Republic, for example, bought back a portion of the debt in 1868 at 30.6 percent of its face value. An 1870 report to Congress indicated that bonds issued in 1850 traded at a mere 8 to 9 percent of their face value, while more recent debts could be purchased for 15 percent. 14 Diaz faced an uphill battle in his efforts to reestablish the government's credit.

Reestablishing the Government's Credit Since borrowing was impossible, the Porfiriato ran balanced budgets during its early years. It also tried to establish a good reputation by steadily increasing the amounts dedicated to amortizing outstanding debt. These rose from 256,000 pesos in the 1876-77 fiscal year to Mx$1.1 million in 1879-80. 15 Diaz's efforts restored some confidence. The price that it paid to amortize its old debt rose from 15 to 24.9 percent ofpar. 16 Foreign confidence also increased. Luis Tellez Kuenzler isolated the size of the political risk premium English investors placed on Mexican bonds. His estimates indicate that this premium rose from 1. 7 percent to 2.5 percent after Diaz took power in 1876, but then fell to 1.4 percent by 1881 (see Figure 1.1)_17 The Porfirian government resorted to Mexico's primitive money markets for short-term credit far more than its predecessors. Figure 1.2 details the gross

$80,000 $70,000

- - Budgeted expenduures -e-· Budgeted revenues - - etsurplu

$60,000 $50,000

0 $40,000 ~

"'-

'-

$30,000

0

c:0

o . $20,000 $10,000

S$(10,000) $(20,000) r-r--

0'-

r--

00

r-"' ;:s

00

0

I

I

00

00

....I

00 00

;:s

;:s

00

U')

N 00

r--

0'-

00

"'

00 00

I

0'"" I

"' ;:s

;:s

00

~

00 00

0

I

I

;:s

U')

"'....I

N 0'-

;:s "'

Federal Budget Deficits in Early Porfirian Mexico

Figure 1. 1. SOURCE:

r--

00 "" I

I

Derived from the Cuadros de Informacion Hacendaria, 1825-1970.

$130,000 $120,000 $110,000 $100,000 $90,000

0 $80,000 ~

"'- $70,000

'0

c: 0

-

$60,000

o.

$50,000 $40,000 $30,000 $20,000 $10,000 $r-"" I N

r--

00

U')

r--

I .... r--

00

r-r--

r-"' I

00 I

""I

r-"' ;:s

00

0

00 00

N 00

;:s

I

r--

00

00

U')

00

....I 00

;:s

r-I

"'I

"'

;:s

00 00 00

""I

00

~

0'-

00 00

0

N

I

"';:s

0'-

00

U')

"' ....I "' ;:s

Figure 1.2. Short-term Gross Federal Borrowing Derived from the Cuadros de biformaci6n Hacendaria, 1825-1970. The anomalous figure for 1880 - 81 appears to be a misprint. Multiplying it by a factor of ten removes the discrepancy.

SOURCE : N OT E:

Political Instability and Financial Concentration

18

annual amount of short-term borrowing during the late Restored Republic and early Porfiriato. Note the rise after 1876-77. Now, it should be kept in mind that this was indeed short-term borrowing. Most of the funds were paid back in less than a year. Net borrowing (consisting of taking on more shortterm obligations than were paid back in any given year) was zero until 1882. The Mexico City money market faced some built-in limitations to its ability to finance the federal government. Since the market was remarkably disorganized, investors knew that the federal government could repudiate some of its old debt while continuing to contract new credits. This possibility made them extremely reluctant to lend large amounts or for long periods. Unfortunately, rising expenditures on the army and railroad construction made it increasingly difficult to stick to a balanced budget. Since the Mexico City money market could not meet its needs, the federal government therefore began to encourage the development of a banking system.

Origins

of Formal Banking

In 1876 only two banks existed in all of Mexico. The Banco de Santa Eulalia had been founded in 1875 by an American, Francis MacManus, under a charter issued by the border state of Chihuahua. The Banco de Londres, Mexico, y Sud-America (BLM) was a branch of a British bank that had operated in Mexico since 1864. 18 A third credit institution, the Nacional Monte de Piedad (Montepio) was officially a state-owned pawnshop rather than a bank, although it had charged interest on loans since 1782 and controlled 1 million pesos in assets by 1877. 19 In 1879 the governor of the Federal District, Jose Ceballos, proposed to President Diaz that the Montepio be permitted to discount commercial paper and issue "certificates" that would be exchangeable for specie on demand. President Diaz agreed to the proposal. 20 Prominent politicians appeared on the Montepio's junta superior, including Governor Ceballos, Manuel GonzaJ.ez Cosio (who succeeded Diaz as president of Mexico from 1880 to 1884), and Romero Rubio, who later served as interior minister. These men had independent political bases. Not surprisingly, as soon as Gonzalez became president the federal government agreed to accept the Montepio's "certificates" for tax payments. 21 Two other banks opened their doors in 1881. The Banco Nacional Mexicano was a product ofastute international diplomacy on the part of the Mexican government. Emilio Velasco, a Tamaulipan lawyer acting in the employ of the federal government, used his contacts with Gustavo G. Godowa, the publisher of the Mexican publication Le Nouveau Monde, to open negotia-

Political Instability and Financial Concentration

19

tions with the Banque Franco-Egyptienne. 22 After a year of discussions, an agreement was signed on August 11th, 1881, between Edouard Noetzlin, the Banque Franco-Egyptienne's representative in Mexico, and the Finance Secretariat. The agreement made the Banco Nacional Mexicano the federal government's chief financial agent and stipulated that its banknotes would be the only ones the federal government would recognize. 23 That the government's arrangement with the Montepio violated this agreement apparently escaped everyone's attention. Scarcely days later, a group of merchants announced their intention to create the Banco Mercantil Mexicano, claiming that it was their "patriotic duty" to invest in a "free" bank in order to stop Mexican capital from leaving Mexico "to increase other nations' prosperity." 24 Ironically, most of the merchants who subscribed to the Banco Mercantil's stock were Spanish immigrants. Two Spaniards led the bank: Manuel Romano and Jose Gargollo. Joining them were additional members of the sizable Spanish mercantile community in Mexico: Francisco M. de la Prida, Pedro Martin, and Juan Martinez Zorrilla, as well as the French-born Luis Lavie. Antonio Escandon, a promoter of the Ferrocarril Mexico-Veracruz, was one of only two Mexicans to serve on the board. The other was the bank's lawyer, Indalecio Sanchez Gavito. 25 Sanchez Gavito also served on the Montepio's board. 26 The Banco Mercantil claimed that competition between banks would benefit everybody. 27 By the beginning of1882, therefore, four banks functioned within Mexico City. Within six months they had been joined by two more, the Banco Internacional Hipotecario and the Banco de Empleados. 28 The Chihuahua banking cluster also expanded. By 1883 the state government had issued eight more banking charters. 29 The Banco Nacional granted the federal government a credit line of 2 million pesos, and the government used Banco Nacional deposit accounts to hold and distribute its tariff, stamp tax, and lottery revenues. In addition, the Banco Nacional took charge of all customs certificate sales. Discounts on commercial paper dropped from 12 to 8-9 percent in 1882. The rates at which the federal government borrowed dropped even farther, from 20 percent to 6 percent. 30 This easing of the credit situation, however, proved too much of a temptation for the penurious Porfirian regime to resist.

Making Commitments Credible Why did the government agree to concentrate its lending with the Banco Nacional Mexicano? Addressing this question requires a more abstract dis-

20

Political Instability and Financial Concentration

cussion of the problem facing the Mexican government and its lenders in the 1880s, in order to explain several mysteries. Sovereign lending faces a peculiar problem. If the sovereign reneges, there is no third party that can enforce repayment. Of course, "sovereign" governments are not always all that sovereign. The great powers have been known to occupy other nations to collect debts. In fact, President Benito Juarez's refusal to honor the debts of the previous Conservative government in 1861 prompted the French Intervention and the establishment of the second Mexican Empire. 31 That particular adventure, however, collapsed after the end of the American Civil War. The only power that might have been capable of forcing the Mexican government to repay its debts after the French withdrawal in 1866 was the United States, but there is no evidence that the American government had any interest in doing so. Therefore, the only recourse creditors had to induce repayment was the threat of denying future credit. 32 The fact that creditors' only means of punishment is to deny future loans implies that sovereign debtors face near-impermeable credit ceilings. This is because lenders will not lend more than the value of the largest credible penalty that they can impose on the borrower. In the case of sovereign borrowers, the value of the penalty is the present value of any future borrowing. That is to say, if the present value of a loan is more than the value of any future borrowing, then the sovereign will default on the loan. This is why college professors with excellent credit ratings are rarely given loans of millions of dollars with no collateral. The benefit from skipping off to Uruguay is worth far more than the punishment of never receiving future credit. If"there is no penalty, one observes no lending rather than a rash ofloancontract violations." 33 The shorter a regime's life expectancy, the less the value of future access to credit. Iflenders expect a regime to be short-lived, then they won't lend, because the present value of future borrowings will be very close to zero. Short-lived regimes have little incentive to pay back their debts. The Porfiriato tried to amortize outstanding debt in order to convince investors that it expected to have a considerable lifespan. Unfortunately, the signal did not have the desired effect. This should not be surprising. With a multiplicity oflenders, the government can default on some while continuing to borrow from others. In fact, the government has no incentive not to default on any single lender. Since the marginal value of the last loan the government takes at the prevailing interest rate is presumably zero, so is the cost of defaulting on that particular creditor. 34 Lenders are presumably alert to this sort of thing and watch for defaults

Political Instability and Fitzancial Concentration

21

on debts owed their compatriots as well as themselves. They can therefore declare a boycott. Unfortunately, the government can then offer extraordinarily good terms to anyone who breaks the boycott. The more restrictive the boycott, the greater the government's assumed need for loans, and the more credible an offer of good terms for violators. This means that multiple lenders cannot punish the government with a complete credit boycott. This reduces the value of the punishment they can levy. The government's credit limit will therefore be equally low. 35 A solution to the above problem exists. If the government can credibly concentrate its borrowing in a single institution, then it can raise the amount of pain caused by a default, since a single lender will have no problem imposing a credit boycott. With a single lender, defaulting on low-value marginal loans will cost the government all its future credit, including credit needed to finance high-value activities-like paying the army. The pain of default will rise, and so will the government's credit limit. 36 The model explains the logic behind President Gonzalez's decision to concentrate federal borrowing in the Banco Nacional Mexicano. Unfortunately, the rule of law was not particularly strong in Mexico. Gonzalez had every incentive to convince the Banco Nacional Mexicano that it had a monopoly and then continue borrowing elsewhere. Additionally, President Gonzalez knew that General Porfirio Diaz remained a potent political force and intended to reacquire the presidency as soon as possible. Therefore, Gonzalez's time horizon was quite limited. Once an opportunity opened up to borrow without the Banco Nacional Mexicano's knowledge, President Gonzalez gleefully seized it. Of course, once Gonzalez's borrowing became generally known, the government's credit rating suddenly collapsed. A severe financial crisis was the result. By then Gonzalez was already on his way out of office, he and his cronies having prospered mightily. 37

The Crisis of 1883 Federal spending began to outpace revenues after 1882 (see Figure 1.1, above). The federal government began to borrow against future revenues by issuing customs certificates and other short-term paper. 38 The sale of these certificates, of course, represented a particularly dispersed form of borrowing. It was not immediately clear from the government's published books that it was issuing new customs certificates faster than it was redeeming old ones. The government sold as many customs certificates as it could. By 1884 the government had mortgaged 60 percent of the port of Campeche's customs revenues, 84 percent ofVeracruz's, and 90 percent of Matamoros's and

22

Political Instability and Financial Concentration

Tampico's. 39 Promises of future railway subsidies grew even faster than direct outlays. 40 In other words, foreign railroad companies financed construction on the promise of future subsidies, and the army was paid with the promise of future tariff revenues. The government also raised funds from the befuddled Banco Nacional Mexicano. It borrowed to the limit of its credit line, and then requested additional advances. In May 1883 the bank lent federal authorities 150,000 pesos, followed in November by an additional loan of Mx$700,000 from the Banco Nacional, in exchange for Mx$1,000,000 in customs certificates due six months later.41 The Banco Nacional's management was already becoming aware of the government's financial antics-it charged a usurious effective nominal interest rate of 104 percent on the loan. 42 In addition, foreign railway investors slowly became aware of the vast scope of the government's subsidy promises. They began to curtail construction. Imports of manufactured iron and steel products used in railroad construction (Mexico's domestic steel industry would not be substantial until 1900) provide a rough-and-ready measure of the drop in foreign investment. After rising from US$1.3 million in 1880 to US$4.2m in 1882, imports of iron and steel products slid to US$3.8m in 1883, plunged to US$2.4m in 1884, US$1.2m in 1885, and bottomed out at US$904,554 in 1886. 43 The drop in investment occurred at the worst possible time. Mexico was hit by a serious crop failure in 1884, which required the importation oflarge quantities of corn. Specie flowed out to buy corn just as foreign capital inflows dried up, while the federal government gobbled up domestic credit. This provoked a severe liquidity crunch. Plantation owners could not borrow their way through the bad year, and rural bankruptcies multiplied. In mid-1883, the Banco Nacional Mexicano severely limited new lending. In fact, it informed clients that loans would not be renewed for more than four months. 44 The other banks did likewise. By the end of the year, credit had dried up in both Mexico City and the regions dependent on its capital markets. The crisis brought down the Montepio, which had invested heavily in long-term mortgages. A run occurred on the Montepio's banknotes. In April 1883 alone over half of the Montepio's circulating banknotes were redeemed in its offices. 45 The federal government tried to help by mandating that 20 percent of all federal tax payments be made in Montepio banknotes. 46 The Banco Nacional and Banco Mercantil also agreed, under strong pressure from the federal government, to accept Montepio banknotes as payments. 47 This proved too little, too late. The Montepio could not liquidate mortgages fast enough. Despite a healthy long-term balance sheet, it shut its doors. 48

Political Instability and Financial Concentration

23

In short, the expansion of the federal budget deficit provoked a serious crisis. If earlier federal participation in the credit markets had laid the foundation for a healthy capital market, the loss of control over federal finances in the early 1880s was nothing short of a disaster.

The Birth

cif Banamex

At the beginning of1884, the government requested a loan ofS million pesos from the Banco Nacional Mexicano, just enough to cover the leftover federal deficit for the 1883-84 fiscal year. The Banco Nacional refused. 49 Lack of credit, however, could bring down the regime. The solution was to engineer the merger of the Banco Nacional and the Banco Mercantil into a new bank, the Banco Nacional de Mexico, or Banamex. Banamex would then receive much more airtight control over federal finances than its predecessors, in order to prevent a repeat of the Gonzalez administration's credit blowout. It would also receive a monopoly over banknote issue-in short, seignorage revenues from the issuance of paper money would be retained by Banamex. Edouard Noetzlin arrived from France in February 1884 to begin the negotiations over the merger, and an agreement was signed on May 31. Antonio Mier y Celis, Felix Cuevas, Leon Stein, and Edouard Noetzlin made up the Banco Nacional's representatives to the merger negotiations. NicoLis de Teresa, Manuel Ibanez, and Rafael Ortiz represented the Banco Mercantil Mexicano. The new merged bank would have a monopoly over all lending to the federal government. In order to prevent future cheating, the bank would handle all the federal government's financial responsibilities, from the collection of tax payments to the dispersal of federal funds. 50 The government had two mutually reinforcing goals. One was to increase its own access to credit. The new bank extended the government an additional credit line of 8 million pesos, at an interest rate of 6 percent. 51 The other was to create a central bank that could act as a lender-of-last-resort, able "to face successfully any emergency," and regulate the issue of paper currency. Monopoly served both aims. "The Executive believes that in the present situation a multiplicity of credit institutions might be subject to grave upheavals and the danger of bankruptcies, which would drag down fortunes great and small, compromising the dearest interests of society." 52 Banamex's new concession banned new banks from issuing banknotes. 53 Existing banks would face a 5 percent tax on their banknote issues. In addition, Banamex was exempted from all federal taxes, except the stamp tax. Banamex also received the right to issue banknotes up to three times the

24

Political Instability and Financial Concentration

value of its reserves, which included federal government bonds in addition to specie. 54 Banamex moved immediately to secure its monopoly. In June 1884 it gave away shares to a group of commercial houses in Guadalajara,Jalisco, that wanted to form a bank, persuading them to become a Banamex branch instead. 55 In November they established a branch in Chihuahua and began talks with the local banks over giving up their note issues. The Chihuahuan bankers appeared amenable to the idea, ifBanamex would provide them with liquidity in times of need. 56 In short, the government justified Banamex's monopoly in terms of the country's financial stability, and not just its own increased access to credit. In that light, how did the newly created "National Bank" respond to the economic crisis confronting it at its creation?

Banamex's Response to the Crisis

cif 1885

The federal government's voracious appetite for credit was facilitated, not ended, by Banamex's creation. Banamex lent the government 4 million pesos when it opened its doors in February. In June 1884 the bank agreed to advance the government future stamp tax payments. 57 By the beginning of July the federal government owed Mx$5,686,559. 58 Banamex officials conferred worriedly among themselves, noting in ominous terms that despite the fact that the Progreso, San Blas, and Mazatlan customhouses sent their revenues directly to the bank, the debt was being liquidated at a minuscule rate. 59 Something had to be done to prevent a replay of 1883. First, Banamex flooded the market with liquidity. Interest rates dropped from 12 to 8 percent during the first three quarters of 1884. 60 Second, the government announced vicious cost-cutting measures, with the aim of balancing the budget. Third, the government used Banamex's connections with French financiers to try to negotiate a new foreign loan at lower interest rates, which it would use to refinance its debt and reduce its interest burden. President Gonzalez commissioned Edouard Noetzlin to return to Europe and assemble a financing package as quickly as possible. Noetzlin quickly completed his remit. The plan would transform Mexico's outstanding debt, some dating back to independence, into £6,000,000 of sterlingdenominated bonds at much lower interest rates. Unfortunately, the "Noetzlin contract" ignited a political firestorm when the finance secretary publicized it in October 1884, arriving as it did in the supercharged atmosphere accompanying what were to be the last days of

Political Instability and Financial Concentration

25

Gonzalez's presidency and Diaz's reaccession to power. Several congressional deputies bitterly opposed the contract, objecting to the massive £1.3 million commission charged by Banamex. As Carlos Marichal has pointed out, these commissions were probably intended to cover the government's debt to Banamex, but this did litde to douse the opposition. 61 In fact, before Congress could vote on the contract, violent demonstrations closed the legislature. Noetzlin, on hearing of this, renounced the commissions-but the loan was dead. 62 Once Diaz took back the presidential sash from Gonzalez, he formally suspended the loan contract. 63 Not surprisingly, the failure of the Noetzlin contract caused interest rates to skyrocket. Rates charged by private lenders in the Mexico City market shot back up to 12 percent by the end of December. 64 Banamex charged the government an effective interest rate of 39.4 percent on a 300,000 peso advance in January 1885. 65 By March a desperate government had borrowed an additional Mx$1,094,201 from Banamex by not paying money orders drawn on New York. 66 A month later the bank's board approved a 300,000 peso "emergency" loan to the government. 67 The government was rapidly approaching its credit limit. On June 2, 1885, the bank's chieflegal representative, Indalecio Sanchez Gavito, angrily insisted that the government begin reducing its debts. 68 Two weeks of discussions resulted in an agreement on an amortization schedule. 69 Unfortunately, on June 22, 1885, two days after the discussions ended, Finance Secretary Dublin announced what the contemporary financial press called a "coup-d'etat": the suspension of all interest payments on shortterm debt. 70 The suspension prompted a run on Banamex offices in Mexico City and Puebla. Figure 1.3 gives an idea of the run's seriousness. After paying out Mx$1.15 million in less than a week the panic was stilled. 71 Banamex survived. The run had a cost, however. Banamex halted practically all ofits commercial operations until October, and the other Mexico City banks followed suit. 72 The government's credit dried up completely. Under pressure from Banamex, Secretary Dublin instituted a rigorous budget-cutting program that slashed government salaries by as much as 50 percent. He succeeded in balancing the budget for the 1885-86 fiscal year. Dublin and Diaz knew, however, that they still needed access to credit. They made it clear that the suspension was a temporary measure and made no attempt to circumvent the limitations on borrowing without Banamex's approval. In October 1885 the federal government arranged the repayment of its debt to Banamex, including missed interest payments. 73 Fresh lending to the government soon resumed.

26

Political Instability and Financial Concentration

$7,000

~

.--------------------------::::::-r 25.0%

$6,000 -~.--------­

23.0')1,

$5,000

21.0%

" $4,000 ~

19.0'){,

g

17.0%

$3,000

o.

$2,000

15.0%

$1 ,000

13.0%

$- +=~==~~~~~~--~~~--~-+ 11.0% "'I "'I "'I "'I "'I "'I "'I "'I "'I "'I "',_;I "'I ...; ..; >v ciJ ci.. ..ci "3 c.. ... -, "' -~, " -, 0 u.. ::E " < ~ Q" ~

00

00

00

00

00

00

00

"

Figure 1.3.

00

00

00

00

Vl

Banamex Specie Reserves During the Financial "Coup-d'etat"

of1885 SOURCE:

El Ewnomista Mexicano.

In short, Banamex insured that the 1885 crisis would be a temporary obstacle on the road to financial stability, rather than a regime-destroying crisis. The threat to cut off credit was credible. The government did not and could not borrow in the subsequent fiscal year. In addition, Banamex proved its capacity to weather a crisis. It survived the May 1885 bank run. It paid 640,000 pesos in dividends in 1885, despite the suspension, and earned a positive 3.3 percent return on its assets.74 It appeared that Banamex's position was now secure. The benefits of a central bank appeared manifest. Why then was Banamex unable to retain its monopoly over note issue?

The Battle over Banamex's Monopoly on Note Issue The government wanted Banamex to function along the lines of the Bank of England. A monopoly over note issue would help the bank accomplish two goals. First, it would allow the bank-and, indirectly, the government - to benefit from seigniorage. Second, it would give Banamex an additional source ofliquidity, which it could use to ease credit conditions during financial crises. The government had to tread delicately, however, because there already existed several institutions in the country with the right to issue their own paper currency.

Political Instability and Financial Concentration

27

The government issued the Commerce Code of1884 on April20, 1884. The code banned all banks without federal charters from issuing notes and limited the note issues of already established banks to the lesser of their paidin capital or three times their specie reserves minus deposits. The code also authorized the federal government to tax banknotes. 75 Congress unanimously approved Banamex's charter under the code on May 30, 1884.76 In Article 8, section IV of the charter, the government promised to impose a 5 percent tax on the paper currency of all existing banks of issue and to prohibit the establishment of new banks of issue for fifty years. 77 The tax was sufficient to insure that other banks would withdraw their note issues. Most charters permitted the existing banks of issue to issue notes up to three times their specie reserves. 78 This meant that for every silver peso in their vaults, banks could make three pesos in loans. Borrowers would receive paper banknotes. The banks would earn the prevailing interest rate on these loans. 79 If that rate was, say, 10 percent, then the return to note issue would be the following: Equation 1.1: Returns to note issue at 10 percent discount rate 80 ($3 X 0.10) - ($1 X 0.10) - ($3 X 0.005)

= 0.30 -

0.116

= 0.184

The first term represents the profits earned from the three pesos ofloans made in paper money. The second term is the opportunity cost of holding a one-peso coin in the vaults. The third is the average cost per peso-note issued of printing the money times the number of pesos issued. 81 The end result is an average return of18.4 percent on note issue. That is to say, for each paper peso issued, the bank would earn 18.4 centavos in profit. A 5 percent note tax, however, changes the equation: Equation 1.2: Returns to note issue with a 5 percent tax ($3 X 0.10)- ($1 X 0.10)- ($3 X 0.055)

= 0.30-

0.265

= 0.035

The cost of note issue per peso-note faced by existing banks is now 5.5 centavos per peso. This is enough to reduce the return on note issue to 3.5 percent. For those banks that could legally issue currency up to only twice their reserves, the math was worse. They would lose money on their note issues, as equation 1.3 demonstrates. 82 Equation 1.3: Returns to note issue with a 5 percent tax and a 2 : 1 limit on issue ($2 X 0.10) - ($1 X 0.10) - ($2 X 0.055)

= 0.20-

0.21

=

-O.Q1

28

Political Instability and Financial Concentration

Although the accounting logic of the note tax was clear, the political logic was not. In the long-term, there existed real, if debatable, reasons for granting Banamex a monopoly over note issues. Unfortunately, in the aftermath of the crisis of1883, the sudden imposition of a tax on the banks' primary means of financing their operations made little sense, as contemporary commentators realized. One observer, in a letter to colleagues in England, noted in early May, before the results of the negotiations with Noetzlin over Banamex were made public: "The rumors of a 5 percent tax on bank circulation cannot be verified, but if such a law is enforced the general impression is that the limit will be reached and the crisis will culminate in numerous failures." 83 Once the rumors were verified, the political outcry was such that the government had to suspend its plans to impose the 5 percent tax. 84 The change of government in 1884 did not help Banamex's case. President Diaz returned to office. His new secretary of finance, Manuel Dublin, had publicly opposed the idea of a banknote tax as long ago as 1882. His primary objection had been that it gave too many advantages to existing banks over potential entrants. 85 Now he was finance secretary. He did not want to drastically upset the banks' operations in the midst of a severe crisis and in pursuit of a policy he considered ill-advised to begin with. Dublin also faced an important political limitation-the Commerce Code offered nothing to the powerful state governors. The governors were autonomous within their fiefs and often possessed militias larger than the local contingents of the federal government. In fact, the law not only gave most governors no reason to support it but also trampled on interests in the powerful border state of Chihuahua. 86 On June 20, 1884, the day the code was scheduled to come into effect, Dublin gave the Banco de Londres y Mexico a six-month extension. Two months later, in August, he granted the Chihuahuan banks an eighteenmonth exemption from the code's provisions and deliberately looked the other way when an American-Mexican partnership in Guanajuato began issuing its own banknotes. The government ignored Banamex's complaints. 87 A law passed by a unanimous vote of Congress was simply ignored when not convenient. For example, when the Banco de Londres y Mexico's exemption expired on January 20, 1885, Banamex appointed a commission of notables-Leo Stein, Manuel Ibanez, and Francisco M. de la Prida-to pressure the government into enforcing the law. Pablo Macedo, Banamex's lawyer, drew up a solicitation. Banamex's commission presented it to the federal authorities the very next day. The Finance Secretariat publicly responded that it agreed

Political Instability and Financial Concentration

29

with Banamex's position and ordered the Banco de Londres y Mexico to close. The order never took effect. On January 26, 1885, a federal judge placed an injunction over the entire Commerce Code on behalf of the Banco de Londres y Mexico. One judge, who owed his office to President Diaz, threw out all of Mexico's law governing its internal commerce for the sake of political convenience. 88 The legal battle was a sideshow. In fact, the Supreme Court never ruled on the issue. That said, the rhetoric was often quite colorful. The lawyer for the Banco de Londres y Mexico, Rafael Donde, wrote that although the Bank of England had been a positive asset for the United Kingdom, Mexico is an "essentially democratic and liberal country," unsuited to institutions developed in the "lands of monarchy." 89 Donde's colleague at Banco de Londres y Mexico, Luis Labastida, pointed out that the Bank ofEngland had been so successful because it was "identified with the State," and not private interests. 90 Eventually the war of words got so vicious that Banco de Londres y Mexico's director, Edmund Anson, and its chief lawyer, Rafael Donde, appealed directly to Banamex's board of directors to tone down the debate. 91 Legal arguments did not matter. What mattered was the balance of competing political interests, as well as the degree to which the federal government depended on them for support. In early 1886 Secretary Dublan told Banamex that it would support it in its negotiations with the Banco de Londres y Mexico's London administration, "on the condition that the new modifications also include the Chihuahua banks." 92 Dublan could not risk a revolt in the state of Chihuahua, where the governor, Luis Terrazas, was himself a banker. The American bankers and entrepreneurs who had settled in the state were extremely well-connected to Terrazas. Anything that put their businesses at risk also put Luis Terrazas's prosperity at risk-and that could prompt a revolt that could ultimately bring down the fragile national government. 93 Negotiations began in London and Chihuahua City in January 1886.94 In March, the representative of the four Chihuahua banks, Carlos Pacheco, proposed the following deal to the American manager ofBanamex's branch in Chihuahua CityY5 Mter reminding Banamex that these banks were duly charted by the State of Chihuahua, and that "businesses created in this form deserve respect," he proposed that for the next twenty-five years the Chihuahuan banks be allowed to issue a total of Mx$900,000 in notes backed by copper small change, redeemable in silver at a fixed discount of 8 percent below their face value. In addition, they should be allowed to issue Mx$700,000 in silver-backed notes, redeemable at par. At the end of this pe-

30

Political Instability and Financial Concentration

riod, all notes would have to be redeemed. The Chihuahuan banks would retain the right to unilaterally suspend convertibility in emergencies, but any banks that closed their doors to the public for more than six months would lose their right to issue notes. In return for these limits, the Chihuahua banks would receive the same federal tax exemptions granted Banamex, and Banamex would agree to extend credit to them in emergencies. Banamex happily agreed to the proposal. 96 By June the negotiations in London finally began to bear fruit, or so it appeared. The agreement hammered out stipulated that for the next five years, Banamex would open the Banco de Londres y Mexico two credit lines of up to half-a-million pesos, one at the current market discount rate of7 percent, and the other at a variable rate of 3 percent or a fixed rate of 4.5 percent. In return, the Banco de Londres y Mexico would retire its note issue. 97 There were some ominous warnings from Noetzlin that the Banco de Londres y Mexico was looking into buying the concession of the Banco de Empleados, a small, practically defunct bank with the right to issue banknotes. Banamex's board of directors responded, "We are convinced that after the experience with the Banco Hipotecario it [the Banco de Londres y Mexico] is no longer interested in any Mexican bank.... We believe that it would be imprudent to agitate over this question, which would demonstrate fear and weakness on our part." 9 H The agreement went before the board of directors on June 15, 1886. It provoked a heated debate between supporters and opponents. Felix Cuevas began in defense of the proposal: It concludes the fight that we have to continually sustain with the Banco de Londres y Mexico, and in which the government has to take part, perhaps placing our concession into question. It is certain that we will have to give up one million pesos at an excessively low interest rate ... which may seem prejudicial at first glance, but is more than compensated for by the advantages .... Our bills will grow and the bills of the Banco de Londres y Mexico be retired. And the credits extended them are guaranteed by excellent credit lines or first-class commercial discounts. 99 Martinez Zorrilla then took the floor against. "Before discussing the motives for this arrangement, I want to ask if accepting the proposal ... will not mean the Board members' renunciation of the obligations they have taken on to defend our concession and statutes?" He feared that the plan meant that the Banco de Londres y Mexico only had to renounce issuing paper currency for five years.

Political Instability and Financial Concentration

31

F. M. de Prida corrected Mr. Zorrilla's misapprehensions, explaining the five-year timeframe applied only to the credit line. "It is understood that after five years the Banco de Londres y Mexico would not be able to enjoy the right to resume issuing banknotes and would be considered a new bank of issue, against which the Commerce Code and our concession is very clear." Vice President Nicolas de Teresa then asked Pablo Macedo, the lawyer, if he believed that a five-year contract would violate the Commerce Code's six-month limit on loan and discount operations. Macedo responded that there were no difficulties in carrying out operations with terr.1s longer than six months. "This has already been done with various other operations made by the bank .... The bank could open a current account for six months with the promise to renew it for nine additional semesters." Antonio Escandon, however, remained opposed, believing that all the advantages went to the Banco de Londres y Mexico. "We would have great difficulty reimbursing other funds .... If a crisis occurred it would be very difficult to insist upon the payment of the loan." Prida countered that the increase in Banamex's circulation would be a huge advantage. The agreement passed, but only by one vote, 8-7. Soon afterward, Banamex opened credit lines to BLM. 100 Banamex did not know it, but Secretary Dublin was already working to subvert its aspirations to become Mexico's monopoly bank of issue. Too many important interests depended on the Banco de Londres y Mexico and the Chihuahua banks. He was aided by a well-placed Hacienda functionary, Jose Ives Limantour, whose wealthy family also happened be BLM shareholders. Dublin agreed to grant the Banco de Empleados, a small bank intended to serve federal officials, a concession allowing it to issue banknotes. Then, at Limantour's suggestion, the Banco de Londres y Mexico purchased the Banco de Empleados-and the right to issue banknotes. 101 On July 12 Banamex received a hint of this plan in a communication from the manager of the Banco de Empleados, Miranda y Iturbe: I have the honor to present to [you] a proposal that I believe very advantageous to [your] institution. The Banco de Empleados, of which I have always been manager, has obtained a concession ... allowing it to change its name to the Banco Comercial and extending the circle of its operations. You ought to easily understand that in order to refrain from using the right of issuing banknotes that our concession grants the Banco Comercial, it is necessary to double our effective capital, and for the same reasons necessary to rediscount our obligations at a rate sig-

32

Political Instability and Financial Concentration

nificantly lower than the market rate, for otherwise there would be no profit in the operations the bank will practice. 102 Miranda y lturbe demanded that Banamex agree to rediscount the Banco Comercial's commercial obligations at an annual rate of 5 percent, which was not acceptable to Banamex. Banamex did not, however, want this new bank to issue paper money, and so Pablo Macedo opened discussions. Miranda y Iturbe actually increased his demands. He now insisted that Banamex open a credit line of at least 3 million pesos to his bank at a rate no higher than 5 percent, and that Banamex renounce the right of issuing "guarantee bonds"-essentially commercial paper backed by real inventories of goods for sale-for ten years, except through the Banco Comercial. Banamex authorized Gustavo Struck to make Miranda y Iturbe a generous counteroffer. It would rediscount up to Mx$500,000 of Banco Comercial paper each year for ten years, and the Banco Comercial would renounce its ability to issue banknotes for only the duration of the contract. 103 Talks proceeded, but a bombshell broke over Banamex's board on August 18, 1886: the Banco Comercial was negotiating with Banco de Londres y Mexico over the possible sale of its concession. 104 On August 24, Struck reported that the negotiating committee "has had various meetings with Sr. Miranda y Iturbe, but unfortunately nothing has been settled." 105 Three days later the Finance Secretariat approved the concession's transfer. 106 The Banco de Londres y Mexico was reorganized as a fully domestic Mexican bank, with full rights of issue. The Chihuahua banks abandoned their agreements with Banamex. In 1888 the federal government recognized them as legal banks of issue. The following year Secretary Dublin permanently suspended the parts of the Commerce Code dealing with the banks altogether.

Conclusion Banamex's creation and early dominance over the financial system was not determined, but rather contingent on the political and economic circumstances of the early Porfiriato. Had Mexico enjoyed access to other sources of capital, or been able to raise domestic taxes, its creation would not have been necessary. Attempts to give Banamex a monopoly over note issue collapsed in political acrimony and legal infighting over Banamex's monopoly and ended the possibility that Banamex might follow the path of the Bank of England, Banque de France, and other European central banks with origins as private institutions. After 1888, Banamex was simply a large, privileged private bank. Only Banamex could open branches across state lines,

Political Instability and Financial Concentration

33

only its notes were acceptable in federal offices, only it enjoyed the flow of federal tax and tariff receipts through its coffers, and only it benefited from the business oflending to the federal government. Banamex needed to preserve these privileges. This was by no means certain. Both the Banco de Londres y Mexico and the Chihuahua banks opposed them. These banks were politically powerful. The Porfiriato also needed to distribute benefits to the state governors, which might involve weakening Banamex's privileges. In addition, as the 1880s gave way to the 1890s, and economic growth improved the government's fiscal position, the value of future credit from Banamex lessened. The temptation to renege on past agreements grew, and constitutional mechanisms were no guarantee in Mexico. How did Banamex preserve its privileges and structure the banking market to maintain its dominance? Banamex's owners needed to create a political coalition that would allow it to preserve its competitive advantage over its competitors.

Chapter 2

Structuring the Credit Market The new law will no doubt give birth . . . to a sort of banking oligarchy. -Jose I. Limantour

Banamex and the Banco de Londres y Mexico had established themselves as the nation's dominant commercial banks by 1888. What strategies did they pursue in order to structure the banking market and retain their dominance? Both banks gained substantial political advantages during the 1880s, but maintaining the economic benefits conferred by those political advantages during the following quarter-century was neither a simple nor an automatic process. The government needed to satisfy two not entirely consistent goals: maintaining its own access to credit through Banamex and distributing benefits to the state governors. The federal government needed to distribute benefits to the state governors in order to secure its own political position. In other words, it needed to give Mexico's regional warlords a stake in the nation's political stability. Paradoxically, one way to do this was for the states to renounce the right to issue bank charters. In return, the federal government would agree to give each state governor the right to distribute a single federal banking charter to selected cronies. The resulting banks would be prohibited from branching across state lines. With a few exceptions, competition within each state would then be reduced to the state bank and the local branch of Banamex or BLM. Finance Secretary Jose I. Limantour formalized this arrangement in the General Banking Act of 1897. 1 The arrangement defined by the General Banking Act was not stable simply because Congress passed a law. Mexico was not a constitutional gov34

Structuring the Credit Market

35

ernment-and would not be a constitutional government for still another century. Passing a law meant little by itsel£ The arrangement was stable because it met the needs of all the players. The federal government could not unilaterally change the rules, because this would harm the interests of the state governors. A revolt by them could bring down the Porfiriato. No state could unilaterally liberalize its laws, because this would dilute the value of the monopolies granted by the other state governors. They would support the federal government in disciplining a wayward state. The private bankers who obtained charters-and granted a state governor a share of the resulting rents-could rest easy that their property rights were secure. The law also preserved many of the privileges ofBanamex and BLM, offering them limited protection from competition with the state banks. Eliminating their privileges would both harm the government's access to credit and trample on increasingly important political interests in Mexico City. Banamex and BLM disliked the establishment of the state banks, but failed to prevent it. The federal government needed too badly to give the states a stake in the regime. Nonetheless, the privileges of the two national banks conferred them great advantages, which translated into control over a huge t'ercentage of the banking market. In short, the institutionalization of Mexico's concentrated and uncompetitive banking structure was an inherently political process. In this sense, the extraconstitutional nature of the Porfirian regime imposed real costs on Mexico's economic development. A truly federal system under the rule of law would have encouraged regulatory "competition" among the states. States with relatively liberal banking policies would have grown faster and attracted more investment than those that allowed small groups of political insiders to monopolize the provision of capital. Constitutional government, however, was not feasible in a country dominated by regional warlords and with a long and dolorous history of political instability and political predation. Inside arrangements were feasible, and inside arrangements were what Mexico got. It may have been a second-best solution to constitutional democratic government, but the alternative was continuing chaos.

The Chihuahuan Banking Cluster The Chihuahua state government never established a general banking law, or made establishing a bank a matter of administrative procedure. Rather, from Chihuahuan banking's beginning in 1875, obtaining a charter from the state required a special act of the legislature.

36

Structuring the Credit Market

The first banking charter in Chihuahua was issued in 1875. It went to an American, Francis MacManus. Its charter granted the Banco de Santa Eulalia an exemption from state taxes and permitted it to issue banknotes exchangeable at par for copper coins or at an 8 percent discount for silver pesos.2 Management soon passed from Francis MacManus to his sons Tomas and Ignacio and to another American, L. H. Scott. In 1882 the state authorized the bank to issue up to 200,000 pesos of copper-backed notes and an additional 100,000 pesos of silver-backed notes, receivable at par in state offices. The 1882 concession reforms also required the bank to open a 25,000 peso credit line to the state government, at 8 percent interest, and imposed a usury ceiling of 10 percent. Three years later, in 1878, Governor Luis Terrazas issued himself a bank concession, in concert with the merchants Felix Francisco Maceyra and Antonio Asunsulo. The Banco Mexicano de Chihuahua received the right to issue 300,000 pesos in copper-backed pesos and an additional 300,000 pesos of silver-backed notes. 3 A merchant, Inocente Ochoa, founded the Banco Minero Chihuahuense in 1882 in partnership with the Mexican-born son of an American, Enrique Creel, receiving the right to issue Mx$300,000 in copper-backed bills, redeemable in silver at an 8 percent discount. 4 The following year the legislature granted the bank the right to issue a further Mx$300,000 in silverbacked money and required the bank to open a 15,000 peso credit line to the state government at 8 percent interest. Like the Banco de Santa Eulalia, but unlike Governor Terrazas's bank, it was subject to a 10 percent usury ceiling. 5 In 1883 the legislature authorized the creation of the Banco de Chihuahua, authorized to issue up to Mx$300,000 in copper-backed notes, this time exchangeable at market rates rather than a fixed discount, and Mx$500,000 in silver-backed notes. 6 This bank was founded by Henry Muller, a GermanAmerican and business partner ofLuis Terrazas. 7 Terrazas's influence showed in the favorable terms of the new bank's concession, compared to the Banco de Santa Eulalia. Muller raised capital by selling stock in his bank to wealthy landowners and merchants throughout the state, using the proceeds to finance his family businesses. For example, the bank provided credit to a textile operation owned by Muller's brother-in-law, Carlos Moye. 8 The bank grew rapidly, increasing its assets from 152,389 pesos in 1883 to 1,608,398 pesos in 1888, including the purchase of the Banco de Hidalgo del Parral in 1885.9 The state government authorized at least five other charters in 1882 and 1883 (including the Banco de Hidalgo del Parral) in addition to the five

Structuring the Credit Market

37

mentioned above. Chihuahua's banking policy was by no means free of favoritism, as Governor Terrazas's special privileges demonstrate. Nevertheless, Chihuahua was competing with neighboring states-and nearby parts of the United States of America-for investment, and its policy grew increasingly liberal as time went on. Why wasn't this happy experience repeated on the national level?

Federal Banking Policy in 18 86-91 Finance Secretary Manual Dublan tried to undercut Banamex's privileges as much as possible during his tenure. Dublin was putatively an economic liberal. He wanted to "subject all these institutions to a scientific system and consolidate" their privileges into a single law. He also wanted to liberalize banking regulation and make charters relatively easy to obtain. He was limited, however, by the realities of the political situation. The federal government remained dependent on Banamex to fill persistent budget deficits. It could not risk a repeat of1883 or 1885. In addition, both Banamex and the BLM enjoyed growing political influence. In Dublin's own words, he had to avoid "adversely affecting rights acquired" under existing concessions. 10 His attempts to undercut Banamex would ultimately fail. The only banks that would succeed in opening their doors under his "liberal" policies would be those with close connections to powerful local politicians. Dublin's first moves were subtle, consisting of abetting BLM's attempts to undercut Banamex. Before 1888, BLM redeemed its own banknotes only at the branch of issue. On January 3, 1888, Frederick Howes, the Veracruz branch manager ofBLM, announced that his office would now accept BLM notes from any other branch at their face value. 11 Now importers could use BLM notes, knowing they would be good throughout the country at their face value. They could borrow at the BLM, change the notes needed for tax payments into silver coins (or Banamex notes) at the bank's Veracruz branch, and use the rest to finance their transactions. Previously, importers would borrow from Banamex, receiving their loans in Banamex notes, part of which they would use to pay customs duties and the rest to buy merchandise. These notes, however, could only be used in or near Veracruz, since Banamex branches elsewhere would accept them only at a discount. This meant that if a Banamex borrower wanted to transfer their funds or use them to buy products from outside the state they had to either pay the discount, use a money order (which were also discounted), or send specie coins (which was risky and expensive). With the new BLM policy, however, provincial merchants now had an incentive to finance transactions at the local BLM

38

Structuring the Credit Market

branch, and businessmen everywhere with business in Veracruz now had an incentive to prefer BLM banknotes. Banamex, obviously, was not pleased and instructed their Veracruz branch to stop accepting BLM notes. Not only was the bank worried about the proximate advantage the Banco de Londres y Mexico gained from its new policy, but it also worried that this move might presage the loss of their monopoly over tax payments. They hoped that "if the situation can be maintained it will greatly help us counteract the intentions of said institution." 12 Their worries proved justified in 1888, when Dublin pushed a bill through Congress, granting him the authority to issue new federal bank charters. It passed on June 30. Luis Labastida, a lawyer charged by the Finance Ministry (Hacienda) with the design of a new general banking law, described the government's motivations as follows: "The Executive ... rejecting immoral stipulations, and animated by the desire to make real the rights of Man and favor the public interest, gave legal existence to the banks of Chihuahua." The bill undercut Banamex's efforts to force the banks of Chihuahua to abandon their note issues. The Chihuahuan banks may have been small, but they were located in Mexico's fastest growing region. Banamex took their competitive challenge seriously. 13 Banamex had stopped accepting banknotes issued by other banks in May 1887. 14 IfBanamex would not accept banknotes issued by the other banks, then those notes would lose much of their utility, and borrowers from the Chihuahua banks would demand that their loans be made in silver pesos or Banamex notes. 15 After the federal government recognized the Chihuahua banks, however, Banamex's efforts to undercut their note issues fell apart. The federal government needed to retain Governor Terrazas's key support in Mexico's uncertain North. Enrique Creel and Miguel Salas received federal charters for the Banco Minero Chihuahuense and the Banco Mexicano de Chihuahua. 16 Their federal charters guaranteed them all ofBanamex's privileges until1903, except the use of their notes to make federal tax payments and the ability to branch across state lines. The banks could now issue notes for up to three times the amount of their specie reserves or the amount of their paid-in capital, just like Banamex, and they received the same tax exemptions. The charters also repealed the 10 percent ceiling on interest rates and allowed the banks to retire their old issue of copper-backed banknotes and issue new silver-backed paper money. 17 The Banco de Chihuahua signed a federal charter in December 1888, followed by the Banco de Santa Eulalia four months later. Banamex soon discontinued its boycott of their banknotes. Banamex received another blow on August 31, 1888, when Secretary Du-

Structuring the Credit Market

39

blan allowed the Banco Internacional e Hipotecario to issue "certificados del deposito al portador." These CDs had to be denominated in pesos and 100 percent backed by gold or silver coins or bars in the bank's vaults. They were redeemable on demand and paid no interest. 18 Banamex protested that these CDs were essentially banknotes. 19 The 100 percent reserve requirement meant that CD issuance would not be particularly profitable for the Banco Internacional e Hipotecario, but Banamex worried about any assault on its privileges. Dublin rejected Banamex's protests. 20 Dublin also granted concessions for banks of issue in the states of Coahuila, Durango, Nuevo Leon, Sonora, San Luis Potosi, and Yucatan. Miguel Saavedra, a political ally of Dublin's, obtained the concession for a bank of issue in San Luis Potosi, which was also permitted to branch throughout the states of Zacatecas, Nuevo Leon, and Tamaulipas. The Banco de Sonora concession went to the Basque merchant Ricardo Uruchurtu, with a territory covering Sonora, Sinaloa, and Baja California, while the two banks established in Yucatan were allowed to operate in Yucatan, Campeche, and Tabasco. 21 Dublin also granted myriad concessions for "bancos agricolas e industriales," which were denied the right to issue banknotes. Banamex worriedly poured over these concessions, trying to determine how much of a competitive threat they posed. 22 The financial press lauded them, although it also noted that the new banks did not enjoy Banamex's privileges. For example, the Economista Mexicano noted that "the Banco de Sonora has not been conceded the innumerable franchises and exemptions that the privileged Banco Nacional de Mexico enjoys." Nevertheless, the Economista Mexicano concluded, "We applaud this unbreakable zeal of the government for removing obstacles to the nation's progress and encouraging individual initiative." 23 Unfortunately, the press's enthusiasm turned out to be misplaced. None of the bancos agricolas e industriales ever opened their doors. Glowing provincial reports tried to attract interest, but failed. 24 Without the right to issue banknotes, they simply could not compete. Considering the substantial profits from note issue, it is no surprise that the bancos agricolas failed to attract investors. Among the banks of issue Dublin chartered, those which succeeded universally possessed direct connections to powerful state politicians. Consider first the three banks that failed. The Banco de Sonora concession had been granted to a Spanish merchant in 1889 and could not attract investors. The Banco de Coahuila concession went to an American, Francis Relph, and it also failed. The Banco de San Luis Potosi charter went to a personal ally of Dublin, with no local connections, and it failed to open its doors.

40

Structuring the Credit Market

Only one bank, the Banco de Zacatecas, succeeded without known political connections. 25 Contrast this with the four successful new banks. 26 In 1892 the Banco de Nuevo Leon started operations in Monterrey. Among its board members prominently stood Geronimo Trevino, a former governor and former general who still practically ran the state. 27 In Yucatan, both charters went to existing politically powerful families. For example, Dublin granted the Olegario Molina permission to establish the Banco Yucateco. Molina would eventually become governor of the state. Yucatan's other charter, for the Banco Mercantil de Yucatan, went to the powerful Escalante family, a source of many of the state's rulers. 28 The president of the Banco de Durango at its foundation was none other than General Juan Manuel Flores, who was also the governor of the state. When Governor Flores died in 1898, a member of the Banco de Durango's board of directors, Juan Santa Marina, succeeded him as governor. 29

Banking Policy and the Mexican Economy Under Limantour, 1891-96 Dublin died in May 1891. President Diaz appointed Matias Romero to succeed him as finance secretary. Diaz also appointed Jose Ives Limantour as Romero's chief assistant, and Romero placed Limantour in charge of banking matters. 30 Limantour replaced Romero two years later. Limantour would not deal with banking issues until1897. He had bigger fish to fry. First, Mexico's government debt and federal budget were thoroughly disordered. The federal deficit had not been tamed. Second, the alcabalas, or internal tariffs levied by Mexican state governments, seriously limited interstate trade. Creating a banking system to facilitate capital mobility meant little when goods were still not free to circulate within the national territory. In essence, Limantour had to carry out a small revolution in Mexico's economic institutions. He was not about to devote much effort to a new banking law until these other, more basic tasks were substantially complete. Therefore, he placed a moratorium on new banking charters after 1892. Once the federal budget was balanced in 1895, Limantour again turned his attention to banking law.

The General Banking and Credit Institutions Act of 1897 Limantour needed to give the state governors something in exchange for getting them to give up their biggest source of revenue-internal tariffs. Banking charters were an obvious plum. The arrangement devised by Limantour

Structuring the Credit Market

41

in 1896 (the law was not officially passed by Congress until 1897) ratified and extended the status quo that had emerged under Dublan. Under the new arrangement, only the federal government would provide charters. These charters would be limited to one per state, and banks would not be allowed to branch outside their concession territories. Any new entrants into a banking market would be hit by federal taxes on their capital and banknotes. In practice, the governor of the state would select who would receive that state's federal charter, usually in exchange for a seat on the board and substantial quantities of stock. All banks would be required to publish monthly balances, and a highly paid federal auditor would sit in their offices. Banamex's privileges as the government's fiscal agent would be preserved, as would the unique position of its banknotes (good for tax payments) and its ability to branch across state lines. It would also face lower reserve requirements on its note issues. BLM would share the ability to branch nationally with Banamex. The deal was not stable because the federal government passed a law in 1897. Porfirian Mexico was not governed by the rule of law. Rather, the arrangement was stable because once in place none of the actors had an incentive to violate it. The federal government could not unilaterally change the rules, because that would incite the ire of all the state governors, who could collectively bring down the national government. Single states did possess incentives to liberalize banking law within their territories, in order to attract more business, economic activities, and tax revenues. Liberalizing, however, would reduce the value of the quasi monopolies granted the other state governors. They would therefore support the federal government in bringing the liberalizing state back into line. By contrast, banks and entrepreneurs with no connections to the state governors had no security that their property rights would not be unilaterally abrogated by either the federal government or the states. This is why all (save one) of the successful banks founded under Dublan were tied to powerful state governors or politicians. 31 State governors could and did violate federal law with impunity during the Porfiriato. Governor Enrique Creel of Chihuahua (the same Creel who founded the Banco Minero Chihuahuense), for example, decided in 1905 that federal land law did not provide sufficient opportunities to steal land from the descendants of the old Spanish military colonists in the region. He therefore wrote his own land law in 1905, in violation of the Constitution of 1857. When the federal government protested, he obliquely threatened revolt, and the federal authorities backed down. 32 In fact, not only was Creel's land law unconstitutional, but Creel's very governorship was unconstitutional. His father had served as the American consul in Chihuahua City,

42

Structuring the Credit Market

and under the Constitution of 1857 only the Mexican-born children of Mexican-born parents could hold public office. 33 The difference between land law and banking law was simply that unilaterally rewriting land law within a state did not (in this case) negatively impact interests outside the state, but unilaterally rewriting banking law would. Limantour was under pressure from the state governors. In particular, he was under pressure from the governors in the states that already possessed banks to stop Banamex from using all its powers to put the local banks out of business. Banamex, for example, collected and then presented huge amounts of state banknotes for redemption and refused to clear their checks. 34 Limantour also wanted to insure that the Banco de Londres y Mexico held on to its privileges. His family was a major BLM shareholder, and his brother had, in fact, been one of the masterminds behind the Banco de Londres y Mexico's purchase of the Banco de Empleados concession in 1886. 35 Limantour began to discuss the details of arrangements with Banamex well before he received formal authorization from Congress to formulate a new law. 36 On March 31, 1896, Banamex's directors received a memo from the Finance Ministry stating that the secretariat intended to "legalize" the banks of issue already operating in the country and wanted to negotiate "compensation" for Banamex's final surrender over its right to a monopoly on note issue. 37 Limantour could not alienate Banamex completely. First, politicians of considerable power and influence had been appointed to the bank's board of directors. The Mier family (who had governed Veracruz) was one example. The Praxedis family (who governed Puebla) was another. 3 H Second, although Limantour succeeded in balancing the budget in 1895 and would maintain balanced budgets until 1908, the government might need access to foreign borrowing again. In addition, Banamex remained the federal government's primary fiscal agent. Therefore, Limantour made sure to give Banamex representation in designing the new arrangement and had no trouble granting Banamex semilegal favors against certain competitors. In order to reduce Banamex's opposition, for example, the federal government offered to permanently revoke the note-issuing rights of the Montepio. The Montepio's owners were connected with the long-out-of-power Gonzalez presidency, and their property rights could be easily altered for political gain. Banamex, of course, accepted the offer. 39 Banamex was well represented on the committee that wrote the General Banking Act of 1897. Of the committee's eight members, three were associated with Banamex. Hugo Scherer and Miguel Macedo were major shareholders and sat on Banamex's board. Carlos Varona was Banamex's general

Structuring the Credit Market

43

manager. BLM was also well represented. Its general manager, H. S. Waters, sat on the committee and Limantour's family possessed a substantial financial interest in the bank. 40 The General Banking and Credit Institutions Act, which passed Congress on March 19, 1897, institutionalized the arrangement. The national banks (Banamex and BLM) retained the ability to branch nationally, while the law limited all other banks to their charter territories. Banamex and the Banco de Londres y Mexico also retained a duopoly over the Federal District, the center of Mexico's commercial life and the location of the nation's only active securities market. 41 Banamex, in addition, retained its position as the federal government's primary financial agent, although it agreed to reduce its commission on the transfer of federal funds from 2 to 1.75 percent, halve its fees for servicing the domestic debt from 2 to 1 percent, and open a new Mx$4 million credit line to the government. 42 The federal government's continued use of Banamex facilities to collect tax revenues provided Banamex with an easy source of specie reserves for the support of its own operations. Limantour felt the need to provide an economic rationale for the new arrangement. Therefore, he justified the national bank's special privileges by claiming that the national banks would "develop into banks of rediscount, and by that very fact, become true protectors of the local banks." In effect, he claimed that Banamex and BLM would assume central banking responsibilities. 43 He justified the local semimonopolies by stating that Mexicans were unaccustomed to the "habits" needed for the widespread use ofbanks. According to his statements, fly-by-night banking operations might take advantage of unsophisticated depositors and noteholders, and the financial system would prove easily susceptible to banking panics if moneys were entrusted to multiple small banks established by any and all comers. The failure of some of these banks could then destabilize the entire economy. 44 The state banks established under the law differed little in their particulars from the ones already chartered under Dublin. They could issue banknotes up to the lesser of twice their specie reserves minus deposits or three times their paid-in capital, had to putatively limit their loans and discounts to six months, and were prohibited from mortgaging real property. 45 The law permitted branching, although only within a territory fixed in the bank's charter. Banks also had to obtain the finance secretary's permission to issue new stock. The law gave these state banks effective monopolies by imposing a 2 percent annual tax on the paid-in capital of all subsequent banks of issue established in the state. 46 This, combined with an extremely high minimum paidin capital requirement of250,000 pesos, established a practically insuperable

44

Structuring the Credit Market

barrier to entry. The minimum capital requirements were extremely high, almost five times higher than in the United States, an economy thirty times largerY The General Banking Act contained provisions for other types of banks -mortgage banks and bancos refaccionarios-both prohibited from issuing banknotes. Even the law's primary author, however, doubted they would be successful. Limantour wrote, "We must not embrace the illusion that these institutions will promptly multiply." Such institutions would not develop "until there has developed the spirit of enterprise among us; until we appreciate the way these banks function and the benefits to be obtained from them." 48 This was somewhat disingenuous. Limantour knew full well that the economic benefits from the right to issue notes made it impossible for any other type of bank to effectively compete. In a surprising moment of clarity, he wrote, "In following this plan the new law will no doubt give birth, at least in the early years of its operation, to a sort ofbanking oligarchy." 49 All that was required from Banamex in return was cheap credit for the federal government and a vague promise that it would support the state banks in times of need.

The Structure of the Banking Market The General Banking Act of 1897 caused substantial changes in the structure of the Porfirian banking market. The first major change was the proliferation of state banks of issue, encouraged by the protection from competition granted them by the General Banking Act. 50 At the beginning of 1897 ten banks of issue operated in Mexico: the two national banks, three banks in Chihuahua, two in Yucatan, and one each in Durango, Nuevo Leon, and Zacatecas. By the end of the year, after the passage of the General Banking Act, that number had risen to sixteen. By 1903, the number of banks of issue in Mexico reached thirty-one, the highest it was to reach during the PorfiriatoY Practically every state possessed its own bank, and Chihuahua, Nuevo Leon, and Yucatan had more than one 52 (see Figure 2.1). This wave ofbank start-ups steadily drove down Banamex's market share, which fell from 60 percent in 1897 to slightly more than 30 percent in mid1903, before rising again. 53 BLM also lost ground to the smaller banks, with its share of the market eroding from 30 percent to less than 20 percent during the same period. As with Banamex, this decline in its market share halted in 1903, which, not coincidentally, was the last year a new bank of issue entered operation. 54

Structuring the Credit Market

45

70%

- - Banamex -e-· BLM

60% 50% 40% 30%

....•

20% 10% 0%

Figure 2.1.

Banking Market Shares

Calculated from balance sheets published in Economista Mexicano and El Boletin Financiero y Minero.

SOURCE :

The Herfindahl index is a measure of concentration. It is defined as the sum of the squares of the market shares of all the companies in the market (see Figure 2.2). For example, a market with only two equally sized firms has a Herfindahl index value of 0.5 2 + 0.5 2 = 0.5. A market with three equally sized firms has an index value of0.33 2 + 0.33 2 + 0.332 = 0.33, and so on, so that a market with ten equally sized firms has an index value ofO.l and one with fifty such firms a value of0.02. In this way, concentration levels between differently structured markets can be compared. In 1893 Mexico's banking market, with seven banks in operation, was almost as concentrated as a market with only two equal-sized banks. 55 This high level of concentration persisted until 1898. In 1899 and 1900, however, the Herfindahl index fell, as the new banks chartered under the General Banking Act began to sell stock on the Mexico City exchange and attract capital into their operations. Nevertheless, concentration continued to characterize the banking market. With thirty-odd banks in operation, it remained as concentrated as an equally divided market with only five banks. The General Banking Act provided a one-time impetus to reducing concentration in the Mexican banking market, but it could not overcome the effects ofBanamex's privileges or Limantour's refusal to allow the state banks to sell more stock on the Mexico City exchange after 1905.

Structuring the Credit Market

46

1897 1898 1899 1900 1901

Figure 2.2.

1902 1903 1904 1905

1906 1907 1908 1909 1910

Herfindahl Index for Porfirian Banking

Calculated from balance sheets published in Economista Mexicano and El Bolet{n Financiero y Minero.

SOURCE:

Conclusion Porfirian banking law was written to satisfY two different and not entirely consistent aims. One was to secure a steady source of credit for the federal government. The other was to give the state governors a stake in the national political system and the continued rule of Porfirio Diaz. Once freed of the need to depend on Banamex for the day-to-day functioning of the government,Jose I. Limantour designed an arrangement intended to distribute economic benefits to the state governors. Armed with federal charters and federal restrictions on entry, the state governors then proceeded to fuel a wave ofbanking start-ups that diluted, but did not destroy, the dominance of the two national banks over the financial system. Property rights in Porfirian Mexico were inherently politicized. Lacking the institutional checks and balances oflimited government, the only way to protect oneself against predation by the state was to, in effect, pay off powerful political protectors. By binding the interests of the state governors to the interests of the banks, Porfirian Mexico managed to create a credible commitment that the bankers' property rights would not be violated, at least as long as the central government survived. Of course, the federal government could not afford to alienate Banamex, BLM, or the increasingly powerful political interests arrayed around them. The resulting compromise was far from perfect. It did, however, reconcile the government's two conflicting

Structuring the Credit Market

47

goals as much as possible, given the historical constraints faced by the Porfirian government. It must be borne in mind, however, that the historical feasible alternative was not limited government and a competitive financial system-the alternative was continuing political and economic chaos. The Porfirian federal government claimed to be more than merely the chief thug in a political system composed of competing and heavily armed thugs. Rather, a coterie of federal bureaucrats, under the watchful eye of President Diaz, claimed to be objective professionals carrying out public policy according to scientific precepts. That is, they claimed that their actions were taken in accordance with the scientific laws that they believed governed human societies. 56 Limantour, therefore, claimed that the dominance of the national banks was actually salubrious, as they carried out the functions of a "central bank of rediscount" and acted as indispensable stabilizing forces for the national economy and banking system. Did Banamex and BLM actually carry out valuable public functions, or were these claims nothing more than excuses intended to justify the existence of a financial oligopoly?

Chapter 3

Central Banking in Porfirian Mexico A panic, in a word, is a species of neuralgia, and according to the rules of science you must not starve it. -Walter Bagehot

Porfirian legislation deliberately granted the two national banks, Banamex and the Banco de Londres y Mexico, special privileges that allowed them to retain control of half the banking market despite the rapid growth in the number of other banks after 1897. The writers of the new law were well aware of the dominance of the national banks. 1 Limantour justified the national banks' preponderance, however, by claiming that they would ftll an indispensable public role: There can be no doubt that, by the very nature of both kinds of institutions, the general banks, which operate at many points within the Republic with large capital and extensive connections, will develop into banks of rediscount, and, by that very fact, become the true protectors of the local banks, with which they neither should nor can come into conflict, because they complement each other and constitute, in brief, distinct organs of a homogenous and well-balanced system. 2 In short, Limantour claimed that the national banks served a completely different function from the commercial state banks, thereby justifying their advantages under the General Banking Act of1897. Limantour stated that the national banks would eventually become "banks of rediscount," concerned with the health of the banking system as a whole and playing the indispensable role of lender of last resort. In fact, Limantour later claimed that his prediction had come true, and that Banamex, at least, indeed played exactly 48

Central Banking in Poifirian Mexico

49

such a role. In the aftermath of the Panic of 1907, for example, Limantour wrote, "Under the present trying conditions the bank in question [Banamex] has rendered vital service to the commercial and industrial interests of the nation, and to private individuals, thus giving another proof of its great usefulness to the community." 3 Limantour was not alone in this assessment. Charles Conant, in his 1910 report to the U.S. National Monetary Commission, wrote that Banamex exercised "the functions of a central bank." 4 Was Banamex indeed a central bank, or even evolving in that direction? IfBanamex was not a central bank, what other institutions, if any, filled that role? In fact, Banamex filled none of the essential functions of a central bank. The private "Banco Central Mexicano" filled some of the functions that Banamex refused to, but was not a central bank. Unlike American clearinghouses of the late nineteenth and early twentieth centuries, the Banco Central could not assume control over its associated banks' operations in emergencies. The Banco Central also suffered from the contradictions between its commercial role and its role as a clearinghouse for the state banknote issues. Dividend pressures led the Banco Central to involve itself far too deeply in commercial and agricultural ventures: when the state banks needed support, it had no capital to lend, because it was all tied up in relatively illiquid loans. Like the central banks ofEngland and France, which also began as commercial institutions, the Banco Central might have eventually resolved these contradictions. The banks ofEngland and France, however, had decades, even centuries, to resolve them. The Banco Central had scarcely eleven years between its founding and the outbreak of the Mexican Revolution. The government was aware of the problems created by the lack of a lender of last resort and stepped into the breach in 1908, with the creation of the Caja de Prestamos para Obras de Irrigaci6n y Fomento de la Agricultura. The Caja de Prestamos injected almost 50 million pesos, borrowed by the federal government on foreign markets, into the banking system. Had Banamex carried out the functions of central bank, then the Caja de Pn!stamos would not have been necessary. In short, the privileges granted Banamex were the product of a political process begun by the Porfirian federal government's desperate need for funds in the 1880s. Scientific laws and abstract debates did not determine the shape ofPorfirian banking. Politics did.

The Functions of a Central Bank In the late nineteenth century, a central bank operating under a strict specie standard served two basic functions. The first was to establish and maintain

50

Central Banking in Poifirian Mexico

a uniform national currency. The second was to insure a stable banking system and prevent financial panics. Maintaining a uniform national currency, or, in the words of the Porfirian finance ministry, "amply and securely systematizing the fiduciary currency," appears at first glance to be an odd function for a central bank under a specie standard. 5 After all, national currencies already existed during the nineteenth century. The Mexican federal government, for example, specified that a peso was equal to a silver disk of a specific size, and a specific metallic content, with a specific design. That was Mexico's currency, legal tender, and Mexicans accepted it in payment throughout the national territory. What then, in practical terms, did "systematizing the fiduciary currency" mean? Essentially, it meant insuring that all the other instruments that functioned as money maintained a fixed value against the silver peso. The banknotes issued by Mexican banks during this period functioned as money, inasmuch as they were exchanged in payment for goods and services, but they were not legal tender. Rather, they were instruments issued by banks and accepted voluntarily by the public, which the banks promised to redeem for legal tender (silver pesos) on demand, but only at the branch that issued the note. In a sense, they best resembled modern cashier's checks. Since banknotes were only redeemable for silver at the branch of issue, they circulated at a discount outside the immediate area. The discount grew with distance. That meant that a paper peso issued by the Banco de Durango would be worth one silver peso in Durango City, but only 90 centavos in the Federal District. The same problem applied to checks. A check is nothing more than an order to transfer a certain amount of legal tender from one deposit account to another. Since checks are accepted as payment, they are a form of money. Since checks could only be cashed for legal tender at the issuing bank, faraway banks accepted ("cleared") them at discount off their face value. In effect, Porfirian Mexico had several different currencies linked by a fluctuating exchange rate. A Mexican central bank, then, would have the responsibility of insuring that all banknotes and (possibly) checks denominated in pesos be exchangeable for silver pesos at a 1 : 1 rate all across the national territory. A monopoly over note issue would make this job much easier for a central bank. (It would not, however, automatically insure that checks would be cleared at their face value.) Even with a monopoly the central bank must promise to redeem its notes at par at any of its offices, regardless of the branch that issued the note. It is possible, however, to accomplish the task of "systematizing the fiduciary currency" without a monopoly over banknote issue.

Central Banking in Porfirian Mexico

51

The First and Second Banks of the United States (BUS) helped create a national monetary union for a period by insuring that the notes issued by American banks could circulate far from their place of origin without a discount. 6 In fact, the creation of a "national currency" that would be everywhere exchangeable for specie at a fixed rate was one of the primary reasons why the U.S. Congress chartered the Second Bank of the United States in 1816.7 This brings us to the second mission of a central bank: insuring the stability of the banking system. It is always possible that an imprudent or unlucky bank might prove unable to redeem its banknotes or pay checks drawn on it. The public might then mistake an isolated failure for a general problem and rush to withdraw its deposits and redeem its banknotes. Sufficiently large withdrawals and redemptions could cause even healthy banks to close their doors and bring the banking system to a standstill, with an immensely negative effect on the economy. A central bank, by acting as a "lender oflast resort," can help prevent such general crises. In the nineteenth century, a central bank would make its own specie reserves available to the commercial banks in a crisis by exchanging them for the commercial banks' illiquid assets. This is known as "rediscounting." In effect, the commercial banks would sell some of their loans and discounts to the central bank at less than their face value, in exchange for silver pesos that they could then use to reassure or pay off their depositors and noteholders. This opens a central bank to several risks. First, it may be forced to accept loans and discounts of dubious value. Second, since the central bank's primary aim in a crisis has to be getting money out the door, it will likely have to charge an interest rate on these rediscounts that does not adequately compensate it for its risks. Third, in order to insure that it possesses sufficient specie reserves in a crisis, the central bank will have to consistently and systematically pass up potentially profitable loan opportunities during good times. Central banks should contract their own note issues in good times in order to insure that the other banks follow prudent policies. The central bank's greatest tool in this respect is an ability to turn in large amounts of the banknotes of an imprudent bank. To use this tool, however, the central bank must maintain a position as a net creditor and limit its own note issues. Otherwise offending banks could exchange central banknotes for their own and neutralize the central bank's attempts to police them. Limiting one's note issues, however, also means limiting one's profit opportunities.

52

Central Banking in Porfirian Mexico

In the long run the roles of a private commercial bank and a central bank are incompatible. Historically, in Europe, most commercial banks given central banking functions by their governments eventually abandoned their commercial role. 8 (President Andrew Jackson killed the Second Bank of the United States. The Federal Reserve, when established, had no commercial duties.) 9 Did that happen in Mexico? Did Banamex carry out any of these central banking functions? A great deal of evidence suggests that the answer is no. In Mexico, the process that occurred in England and France and elsewhere did not happen, at least during a three-decade period.

Banamex as a Central Bank Banamex was certainly well placed to assume central banking functions. The tax payments and government outlays passing through its hands insured a regular flow of specie through its vaults. This meant that a large chunk of its reserves were made up of federal government funds in transit, an extremely low-risk liability. The risk associated with these resources dropped even further after the government balanced its budget in 1895. Balanced budgets meant steady or growing government reserves, which it kept in Banamex's vaults. In addition, Banamex's ability to branch nationally meant that it could grow larger and could better diversifY its operations than any of its competitors (except perhaps BLM), and that its notes would remain in circulation longer. Banamex's minutes indicate that its owners and managers believed themselves entrusted with ensuring the smooth operation of the financial system, although it is not clear that their actions matched their words. For example, in 1901, Banamex directors stated that during the previous year, acting in the bank's "role as the premier credit institution in the Republic," they combined "prudent and liberal" measures to combat the "violent" growth of business activity, raising interest rates to 10 percent. But Banamex directors also wrote, somewhat contradictorily, that they closed none of their "commercial offices, rather amplifYing their resources" and granted "extensions [on loans] to those who have asked for them under good conditions." 10 In 1902 the bank's managers wrote, "Our industries, and even trade in general, suffered the crisis in the last half of the year. . . . The bad cereal harvest of 1900, whose effects were felt in 1901, aggravated the situation, prolonging the restriction of popular consumption, particularly of cotton textiles, to the point where many factories had to suspend operations. The

Central Banking in Por:firian Mexico

53

situation was delicate for the Bank, because it had to alleviate the crisis." However, Banamex directors went on to say that "without abandoning the prudent rules that always govern its conduct. So it proceeded to keep the interest rate as low as possible, but procuring enough time to call in previously granted credits and assure those that have been renewed." 11 In other words, it kept rates low, but restricted credit to favored customers. This is a sound policy for a commercial bank, but it does little to provide liquidity to a credit-starved market. Another example of the bank's pretensions to the status of central bank can be found in the bank's 1908 report, when the Panic of 1907 in the United States began to hit Mexico in force. It read: Even though [the situation] seemed to indicate that we strengthen our specie reserves and reduce our operations, we have not wanted to do this ... for fear of aggravating the situation and producing a true crisis .... We have continued making [credits] .... Thanks to the efforts we undertook ... we can say today with satisfaction that the Yucatecan crisis has been overcome, that business in that important part of our national territory has returned to normal, and we hope that in the course of this year will continue in its accustomed development. 12 This sounds like a central bank, or at least a bank that has internalized the health of the national banking system. Banamex's contradictory words do not provide enough information to judge the claim that it was a central bank. One must turn to its deeds. First, however, one must first strictly define the role of a central bank under the conditions of a strict specie standard and a plurality of note issue. First, to have a credible claim to the central bank title, Banamex would have had to have encouraged the national circulation of its notes at par. Second, the bank should have attempted, as much as possible, to control the note issues of the state banks. The above implies that Banamex should have held more claims against the state banks than the state banks held against it. Third, Banamex should have acted to supply liquidity to the banking system in times of crisis, implying that its specie reserve ratios should move against the rest of the system.

Encouraging a National Currency For the first decade of its existence, Banamex did nothing to encourage the circulation of its notes at par. For example, in 1891 the bank published the

Central Banking in Poifirian Mexico

54

Table 3.1 Banamex Note Discounts Chihuahua Durango Guadalajara Guanajuato Mazat:Ian Merida Monterrey Oaxaca Puebla San Luis Potosi Veracruz Zacatecas

SOURCE:

0.75% 1.00% 0.75% 0.50% 2.00% 1.00% 0.75% 2.00% 0.25% 0.50% par 0.50%

12/1/91, Aetas de Consejo, vol. 4, AHBNM.

following list of the discounts the head office in Mexico City charged against the note issues of its branches, reproduced in Table 3.1. 13 This is in marked contrast to the policy followed by the Second Bank of the United States (BUS). From 1816 to 1836, the BUS guaranteed the acceptance of the notes of its branches at par throughout the United States. 14 Banamex undertook no such policy for ten years after its foundation. When Banamex finally unified the exchange rate between its branches in 1894, it did not offer an ironclad guarantee. The bank reserved the right to suspend payments when "excessive amounts" ofbanknotes from other branches were presented. 15 This right was exercised. In 1905 the Somellera brothers ofMazatlan complained that the manager there had refused to exchange Banamex notes for silver pesos. A letter from the branch's manager to the head office explained that the Somellera brothers and others had presented excessive quantities of notes for redemption. The head office replied to the Somelleras that the manager was required only to change notes issued at that branch and had acted correctly. 16 Despite Banamex's weak commitment to maintaining a national circulation at par, the percentage of total note circulation accounted for by its note issues rose steadily after 1904, reaching almost 50 percent by 1912 (see Figure 3.1). Three factors combined to cause the state banks to practically cease issuing new notes after 1905. The first was a change in the federal govern-

Central Banking in Por:firian Mexico

7o,ooo 1

-----r===========,--------r o.7 ~

60,000

s; g_ 50,000

c ...~:: "

u

55

Banamex nore 1 ues

-«-· BLM nore 1 ues \

,.

0.6 0.5

40,000

0.4

30,000

0.3

20,000

0.2

10,000

0. 1

'0

§~

Figure 3.1.

Banknote Circulation

Calculated from balance sheets published in Economista Mexicano and El Bolet{n Financiero y Minero.

SOURCE:

ment's policy. The second was the increasing use of checks. The third was the contraction prompted by a change in the silver-gold mint ratio to reflect the ratio in international markets. This change in the federal government's attitude first became evident in 1905, when Congress banned the chartering of new banks of issue. Secretary Limantour jawboned the state banks into restricting their circulation. In 1906 he further pressured the state banks to agree to limit new note issues to changes in their specie reserves.17 When several banks requested permission to issue new stock, Limantour vituperatively denied their petitions, accusing them of seeking to defraud their current shareholders. 1H These actions caused Banamex's (and BLM's) share of circulation to increase. Check clearing became much easier after the establishment of the Banco Central in 1899. By 1906, the volume of checks cleared through the Centro de Liquidaciones each month totaled roughly 20 percent of the total assets in the banking system. The annual turnover of checks reached 5 percent of GDP. 19 Mexican observers noted the rise and correctly predicted that banknotes would become increasingly less important. For example, in 1906 Joaquin Casasus wrote: Should this tendency become more marked and continue for a number of years, may perhaps prove what is already an accomplished fact in other countries, viz, that in proportion as the banking system of a

56

Central Banking in Poifirian Mexico

country becomes more perfected the note issue decreases in importance and comes closer to the level of the cash on hand, because the deposits and checks constitute the greatest force of the metallic circulation movement, a force such as notes payable on sight and to bearer are incapable of giving. 211 Banamex's percentage of total note issues also grew due to the effects of the change in the silver-gold mint ratio from 16: 1 to 32:1 in 1905. 21 Gresham's Law had previously driven gold from circulation within Mexico. That changed once the mint ratio was altered. What the reform's authors did not expect, however, was that the price of silver would rise on world markets, to the point where 32 ounces of silver bought more than an ounce of gold. That caused Gresham's Law to kick in again, but in reverse: now silver left circulation. Since silver coins made up all of Mexico's small denominations and most of its circulating currency, this was highly disruptive, forcing the government to place a tax on exports of silver pesos in December 1906. The five-cent tax on banknotes meant that the banks had no incentive to issue small-denomination notes to compensate for the loss of silver. 22 These disruptions affected the rest of the banking system more than Banamex.23 BLM, for example, hemorrhaged specie, prompting a decline in its note circulation, but Banamex saw little change in its reserves. 24 After the crisis passed, Banamex took advantage of the subsequent boom to increase its note issues, blatantly contravening a November 1905 agreement with the Finance Ministry to limit its note issues to the amount of specie in its vaults. 25 Banamex received no sanctions, although the federal government rigorously enforced similar agreements with the state banks. Banamex, therefore, was able to increase its share of national note circulation after 1906, since the government imposed effective ceilings on the note issues of the rest of the system. Banamex expanded its issues during the boom from 1905 to 1907, at a time when federal policies limited the other banks from following suit, and again when recovery from the effects of the Panic of 1907 began in late 1909. During the crisis, from 1907 to 1909, its note issues were stagnant. This behavior is congruent with what would be expected from a commercial bank unconcerned with the health of the rest of the financial system.

Controlling the State Banks Data published by the Secretaria de Hacienda can be used to calculate a lower-bound estimate of the state banks' holdings ofBanamex notes. Com-

57

Central Banking in Poifirian Mexico

1900

Figure 3.2.

1901

1902 1903

1904

1905 1906

1907

1908

1909

1910

1911

1912

Banknotes as a Percentage of All State Bank Liquid Reserves

Calculated from balance sheets published in Economista Mexicano and El Bolet{n Financiero y Minero.

SOURCE:

paring state banks' holdings ofBanamex notes to Banamex's holdings of the· state banknotes can reveal whether Banamex attempted to maintain the ability to regulate state bank issues by redeeming local banknotes for specie in order to keep the value of the state banknotes close to par 26 (see Figure 3.2). Banknotes never became much more than 25 percent of the state banks' reserves, as Figure 3.2 illustrates. This was because they could not be used to support note issues. How many of these notes were Banamex's cannot be determined, but a lower-bound estimate is that the percentage matched Banamex's share of total note issues. This assumption is improbable. Banamex notes were likely held disproportionately, considering that only they could be used to make federal tax payments. Figure 3.3 is based on the above. From Figure 3.3, there is no indication that Banamex attempted to maintain its ability to regulate the state banknote issues. 27 It allowed its net position to deteriorate against the state banks temporarily in late 1903 and from mid-1905 to mid-1907, and permanently after 1908. Since these are lowerbound estimates of the state banks' holdings, Banamex's real ability to control the local banks was even less than that indicated by these figures.

Lender of Last Resort? Banamex's first interbank operation occurred in 1901 and consisted of arediscount of paper presented by the Banco Oriental. 28 Banamex conducted a

Central Banking in Poifirian Mexico

58

4,500 4,000 3,500

~

"

0.

3,000

;::

~ 3 2,500

._u 0

00

o.

2,000 1,500 1,000 500

1901

1902

1903

1904

1905

1906

1907

1908

1909

1910

1911

1912

- - Banamex holdings ofbanknotes from other banks --· Estimated holdin!,>s ofBanamex notes by other banks (except BLM)

Figure 3.3.

Banknote Holdings

souRcE : Calculated from balance sheets published in Economista Mexicano and El Bolet{n Financiero y Minero.

similarly sized operation with the Banco Oriental the following year. It made no interbank loans in 1903. Mter 1904, Banamex's advances and rediscounts to other banks grew, but never to more than a minuscule percentage of the bank's total operations, as Table 3.2 shows. Most of this business was commercial in nature, often conducted at interest rates higher than those prevailing in the market. Only one operation could be said to have had a noncommercial aim: the June 1904 loan to the Banco Central at 6 percent, two points below the prevailing market rate, which provided the Banco Central with some much-needed liquidity. 29 In addition, Banamex's assumption of responsibility for the note issues of the Banco de Michoacin and Banco de Campeche in 1908 might also be considered to have had the aim of preventing a general panic. These operations, however, were a small part ofBanamex's interbank lending and a negligible part of its overall portfolio. Banamex did not extend much credit to the rest of the banking system directly. It may, however, have indirectly rediscounted paper presented by other banks. Rediscounting would have caused Banamex's other assets to rise, since it would now have the rediscounted paper on its balance sheet, while

Table 3.2 Banamex Advances to Other Banks Loan rate

Market Proportion rate cif porifolio

1,500,000 1,500,000 400,000 400,000 200,000 250,000

6.0% 8.0% 8.0% 8.0% 7.0%

9.0% 8.0% 8.0% 8.0% 8.0% 8.0%

1.2% 1.2% 0.3% 0.3% 0.2% 0.2%

9.0% 9.0% 7.0% 8.5%

Banco de Tamaulipas Banco de Tabasco

900,000 900,000 500,000 400,000 300,000 200,000

8.0% 8.0% 8.0% 8.0% 8.0% 8.0%

0.5% 0.5% 0.3% 0.2% 0.2% 0.1%

Descuento Espaiiol Descuento Espaiiol Banco Americano • U.S. Banking Company Banco de Tamaulipas Monte de Piedad de Morelia

900,000 500,000 400,000 300,000 200,000 150,000

8.0%

8.0% 8.0% 8.0% 8.0% 8.0% 8.0%

0.4% 0.2% 0.2% 0.1% 0.1% 0.1%

BMCI Banco de Nuevo Le6n Internat'l Banking Co. 1'

900,000 400,000 100,000

6.0% 3.0%

9.0% 10.0% 9.0%

0.3% 0.2% 0.0%

1908

Banco de Sonora' Banco de Michoac:ind Banco de Campeche

50,000 370,000 117,000

10.0% 0.0% 0.0%

10.0% 10.0% 10.0%

0.0% 0.3% 0.1%

1909

Descuento Espaiiol Banco Oriental

950,000 950,000

6.0% 6.0%

9.0% 9.0%

0.7% 0.7%

1910

none

Year

Bank

Amount

1904

Banco Central International Banking Company Banco de Jalisco Descuento Espaiiol Descuento Espaiiol Cia. Bancaria Cat6lica

1905

Banco Mere. de Yucatan Banco Yucateco Banco Oriental II

1906

1907

SOURCE:

8.0% 8.5%

Loan dates and size from Aetas de Consejo, AHBNM. Market discount rates

from EM. •AHBNM (1907), Aetas de eonsejo, vol. ?,January 8th. This was actually a deposit in an interest-bearing checking account.

(continued)

60

Central Banking in Poifirian Mexico

Table 3.2 (continued) "This loan was part of a joint operation with the Banco de Londres and Banco Central, whereby the three banks loaned Mx$1.2 million to the Banco Americano for the purchase of a hacienda from one ofBanamex's prominent shareholder's, Juan Llamedo. 'AHBNM (1908b), Aetas de consejo, vol. 7, May 18th. This loan was authorized by the bank's Mazatlan branch. Since Banamex authorized its branches to advance credits up to 50,000 pesos without the approval of the Mexico City HQ, other small advances by branches to the state banks may not be included in these figures. The Mazatlan loan only came to the head office's attention because the branch wished to increase it above the limit. The head office refused. 4Both this and the subsequent credit to the Banco de Campeche were not loans, per se. Rather, Banamex assumed responsibility for these troubled banks' existing note issues, in return for which they would be converted into bancos rifaaionarios and no new banks of issue would be permitted to take their place in those states.

causing the earning assets of the other banks to fall against their specie holdings. One would then expect Banarnex's reserve ratio-the ratio of specie plus banknotes to all other assets-to move against the ratio of the rest of the banking system and to fall during financial panics.

:SJ1ecie l § " " Vi'-' z0 z~ -, ~ ~-,

Monthly Rate oflncrease in the Peso/Dollar Exchange Rate, in Annualized Terms, 1913 -16 Figure 7.2.

soutzCE: Kemmerer 1917:45-46.

verted to an annual rate. Inasmuch as the exchange rate tracked changes in the relative price levels of the United States and Mexico, the numbers give an idea of the inflation's extent. A graph of the exchange rate does not do justice to the chaos caused by the self-destruction of Mexico's currency. A wave of strikes broke out in late 1915. The goal of all of them was that salaries be paid in specie or dollars. Miners in the states of Hidalgo, Mexico, and Michoac:'m walked out over this issue; so did printers in Mexico City, machine shop operatives in Aguascalientes, and smelter personnel in Monterrey. General strikes hit Veracruz in February and May 1916, and railroad traffic in Chihuahua practically ground to a halt as engineers refused to work until they be paid in silver coin. In November 1916 riots broke out in Mexico City when the government tried to pay police and tram workers in paper money. 111 As the hyperinflation galloped along, and people became increasingly reluctant to hold government paper, there were cases of government officials threatening companies and workers. General Manuel Dieguez, in Guaymas, Sonora, called both workers and management at the Cananea mines "agitators" for the company's continued policy of paying its employees in silver and said they would have to be "disciplined." 112

Financing a Revolution

151

The Constitucionalistas also cracked down on those who "shamelessly speculate on the fluctuation of national securities ... persons under the designation of corredores dedicate themselves to speculating on the paper money issued by the Constitucionalista government." Federal authorities took control of the licensing and operations of all exchange operations in the country and required them to deposit 10,000 gold pesos on penalty of fines. 113 Later, in February 1916, the government announced the creation of a US$10 million gold fund to back its paper issues with the following slightly veiled verbal assault on the banks: The Secretaria de Hacienda y Credito Publico should never forget that the Constitutional Government's issue of notes are loans made by the Mexican people in its moment of greatest need, for the winning of its freedoms, and, therefore, represents a sacred debt no government could refuse to recognize .... Given the financial policies of past administrations, that always aimed their efforts at saving the interests of commerce, industry, and banking, the current government should never lose sight of the fact that, in the final analysis, these institutions should not be more than the representatives of the true interest of the Mexican people .... The government ought to aim at improving the economic condition of the people and not only safeguarding and protecting the personal interests of the great capitalists ... which in many cases are in conflict with the true interests of the Mexican people. Mexican finances have, until now, during the past administrations, been entirely dominated, directed, and influenced by certain banking and capitalist groups of great power ... said finances ought to be free and independent of whatever private institution. 114 The banks lost, since debts contracted in gold and silver pesos were now paid back in rapidly depreciating paper money. 115 Those banks that had not yet abandoned checking accounts did so, and all the banks prohibited the specie convertibility of deposits made in fiat money. Most did nothing to withdraw their notes from circulation, however, believing that the laws allowing them to suspend specie convertibility would remain in force. Banamex was the exception: it retired Mx$4,267,400 in notes between July and November 1915, indicating that the bank wished to reduce its potential exposure. 116 The public lost confidence in all paper money, not just government issues. Specie and private banknotes essentially disappeared from circulation, being "quoted and traded like any other merchandise." 117 Banks carefully

152

Financing a Revolution

guarded their specie reserves, refusing to make payments in anything other than government-issued fiat money. The result was total commercial paralysis. As Banamex wrote in its annual report published on June 27, 1916: "In effect, the custom has been generalized among the public of retaining in their power the documents, money orders, checks, etc. against the banks, especially against [Banamex], not presenting them for collection in even the most extreme circumstances, taking them as a means of investment guaranteed against the eventualities of the paper money. In order to contain this speculation, at least partially, [we have ordered] the suspension of money orders from the Central against the branches, and vice versa." Desperate to get the fiat money out of its hands as fast as possible, Banamex began buying up claims against it and paying depositors in government paper money at a premium over the prevailing exchange rate between paper and specie. 118 The Constitucionalista government announced a plan to end the inflation in mid-1916, indicating that it would change all the existing issues of paper money for "infalsificables" at an exchange rate of 10:1. Unfortunately, the government eventually printed 400 million of the infalsificables, far above the 67 million needed to soak up earlier emissions. 119 This, of course, set off another wave of hyperinflation. Capital flight intensified: Banamex, for example, authorized Agustin Legorreta to invest whatever amounts he considered "prudent" in Europe. 12° The hyperinflation ended in much the same way as the Argentine hyperinflation of the 1980s: people simply stopped using paper money. It lost all value as either a store of value or a unit of account. From 1916 onward, all debts had to be denominated in gold. Banamex, for example, made its first credit operations in six months on June 13, 1916, when it lent Jose Peralta 5,000 gold pesos. The bank also signed an agreement with one Marcial del Prado to convert his debt into dollars at the rate of2¢ US to the peso. 121 This was followed by a wave of transformations, changing old credits to dollars, gold pesos, or equity. For example, debts owed by the Compaiiia de Luz y Fuerza and the Ferrocarriles de Pachuca were converted into priority shares paying out a minimum of7.5 percent on their par value. 122 By August Banamex converted old prehyperinflation debts from pesos to dollars at a rate of 12¢ per peso. 123 Banknotes never reentered circulation. From 1916 onward, they traded at large discounts from their face value. Banknotes solely retained value in the hope that the banks would at some point in the future redeem them for specie currency. In effect, they had become a speculative instrument.

Financing a Revolution

153

Direct Predation The seigniorage revenues from the inflation were not enough to satisfy the revenue demands imposed by the military campaigns against the other revolutionary factions. Once confidence collapsed to the point where no amount of threat could convince the public to accept new issues of fiat currency, the Carranza regime turned back to the banks. The first sign of the predation to come arrived on September 14, 1916, when the Montepio (now acting as a branch of the government) presented all the Banamex notes in its possession and demanded they be exchanged for gold. Banamex officials offered the Montepio an interest-free loan in an effort to protect its specie reserves, but the federal government refused. Worried, Banamex contacted federal officials. Venustiano Carranza himself, as primer jefe of the Revolution, reassured the bank in a signed note that this was a one-time-only event and set no precedents. 124 Unfortunately, the government belied its own statements the very next day. On September 15, 1916, the Constitucionalistas effectively assumed control of the banks, appointing juntas de incautaci6n to supervise their daily operations. The legal justification was a decree amending the General Banking Act to increase the banks' legally required ratio of specie reserves to banknotes from 50 to 100 percent. The banks had to either halve their note issues or double their reserves of gold and silver coins in seventy days. Banks that failed to comply would be liquidated by the government and their specie reserves seized. The juntas, ostensibly, would watch over the process of reducing their note issues. In reality, their purpose was to insure that the banks did not hide their specie reserves or spirit them out of the country. Banamex feared that the decree was nothing more than pretext for expropriating their specie reserves. It publicly protested, shut its doors to the public, sent telegraphs to their Paris directors notifying them of developments, and refused to aid the juntas de incautaci6n in any way, "since the Bank does not have to lend its efforts to the execution of the dispositions of the decree of September 15th that fails to recognize the legitimacy and validity of its concession." 125 The government sent Banamex a terse response ordering the bank to reopen its offices to the public. The government then appointed Francisco Pelletier as Banamex's official interventor. Banamex reluctantly reopened its doors "to make payments already in progress and carry out the operations necessary to mobilize the funds to carry out said business," but no new loans or credits were to be authorized, accounts opened, or transactions agreed to. 126 Before they could reopen, how-

154

Financing a Revolution

ever, the finance secretary held a contradictory press conference where he simultaneously declared that "the banks of issue will cease to exist due to their noncompliance with the decree of September 15th" and announced that Banamex and BLM would reopen their doors to the public on Monday. Banamex responded with another protest to the secretary. 127 On September 26, a junta de incautaci6n consisting of Francisco Pelletier, Diego Fernandez, and Francisco Bracho (the latter headed the Finance Ministry's Department of Credit and Commerce) arrived to assume supervision over Banamex's operations. Agustin Legorreta refused to allow them onto the premises, claiming that he had not received any official notification of their appointment. 128 The junta returned later that day with a copy of their official notification stating that Diego Fernandez was a judicial agent capable of bringing up charges against Banamex's directors if they continued to interfere in their work. They warned of"the power of the Junta de incautaci6n," and demanded that their first declaration be posted on the bank's front door as an "example" of their power. Banamex's directors refused to comply and sent another protest. Legorreta threatened to resign if the junta continued to act "arbitrarily." 129 The junta responded with a declaration that it would inventory Banamex's portfolio of "funds and securities" and appealed to the bank's cashier and personnel for help against Legorreta's obstreperousness. Banamex's directors replied "sincerely" that they could not help with the junta's requests, but gave the cashier, Francisco Mijares, permission to cooperate in order to avoid the use of force. 130 The next day, September 28, 1916, the junta ordered the bank to open its vaults, but the cashier and other bank personnel all mysteriously arrived late for work. Legorreta and Subdirector Castello were arrested by military officials because they had "abandoned their posts" without explanation. Castello responded that they had not abandoned anything. "On the contrary, the junta did not arrive on the time that had been officially given for ten in the morning." Soldiers also arrested Francisco Mijares, but he was released minutes after his arrival at the police station. After the roundup ofBanamex officials, troops forcibly occupied the bank's headquarters, and Diego Fernandez posted fliers on the building announcing the arrests under his powers as an agent of the public ministry. The Compaiiia Bancaria de Paris y Mexico offered to help Banamex with any salaries or other payments it had to make while under occupation. 131 The melodrama ended with Legorreta's and Castello's release and the removal of federal troops on September 29, but the government's confrontation with the banks was not yet over. 132 On October 2, the Constituciona-

155

Financing a Revolution

lista government of the State ofVeracruz yanked all its deposits from the Banamex branch in Veracruz City. 133 The following day Banamex shut its branches outside the Federal District. 134 The government responded by pulling all deposits made in the name of the old Porfirian Comisi6n de Cambios and transferring them to the new state-run Comisi6n Monetaria. Banamex's Parisian directors responded by announcing that no overseas money orders drawn on Banamex would be paid. 135 Banamex tried to bolster its precarious financial position by withdrawing Mx$600,000 it had on deposit at the Caja de Prestamos. Unfortunately, the Caja, acting under orders from the Finance Ministry, paid out the deposits in worthless government paper money. 136 By November, the bank was announcing layoffs. 137 These confrontations were softening-up operations, to demonstrate to the banks that resistance to the government's demands was futile. The other shoe finally dropped on November 15, 1916, when the government announced to Banamex and the Londres y Mexico that it wanted a new 10-million-peso loan. 138 The federal authorities wanted to preserve the illusion that this was an ordinary loan request. Therefore, they allowed the bank to process the loan through normal channels and presented the following collateral: 139 Interior consolidated 3% debt bonds Shares in the Ferrocarriles Nacionales Interior amortizable 5% debt bonds Shares in the El Imparcial newspaper

$4,200,000 $4,400 $4,606,800 $550,000 $10,361,200

Banamex's officials initially agreed to the loan. 140 One week later the headquarters in Paris changed its mind, declared that the collateral was not acceptable, and refused the loan under any circumstances. 141 On December 19, Finance Secretary Nieto requested a loan of 3 million pesos, guaranteed by an unspecified combination ofFerrocarriles Nacionales shares, government deposits at the Banco Central, and a mortgage on the Hospital de Pobres on AvenidaJuarez in Mexico City. Nieto wanted an immediate response, but the board was adamant that they had to consult with Paris over a loan of that size. 142 Paris gave a tentative go-ahead, tempered by a suggestion that the bank ship at least 1 million dollars to New York, to protect it from potential expropriation. 143 Once the government promised 30 percent of the revenues from petroleum taxes would go to repayment, the loan went ahead onJanuary 9, 1917. 144 Unfortunately, the government's appetite for funds was whetted rather than satiated. Three days after its approval, federal authorities upped their

156

Financing a Revolution

request from 3 to 5 million pesos, backed by a promise to dedicate 50 percent of the petroleum tax revenues to its repayment. When Legorreta refused, the government threatened to have the junta de incautaci6n simply seize the funds, causing Banamex to promptly change its mind and approve the loan increase. 145 The threat was not idle. The Constitucionalista government had already begun seizing the specie reserves of the other banks of issue after December 14, 1916. The first bank to have its specie reserves stolen was the Banco Peninsular de Yucatan. 146 By the end of 1917, the liquid assets of all the state banks had been confiscated. The regime's revenue needs outstripped the amount of money it could raise from the banks using quasi-legal methods. It now resorted to simple and outright expropriation. In 1917 the government extracted Mx$10 million in forced loans from Banamex. Beginning on February 17, the federal government requested 1 million pesos, backed by an official communique granting the junta the authority to simply expropriate the funds. Banamex's board agreed to the loan, "for there was no way to oppose it." 147 After a week of tense negotiations, the junta was given one of the combinations to the bank's vaults, the board of directors keeping the other. 148 On March 2, the bank received a request for another million pesos. The board of directors appointed a commission consisting ofLegorreta and Castello to try to negotiate with the government. They had no success, and the bank approved the loan, "since the bank cannot effectively discuss this order." 149 Three days later, Banamex received a new request for Mx$500,000. "The Board, in light of the fact that it is still impossible to resist these orders . . . agreed that the new amount will be sent." 150 On March 17, the bank loaned the government an additional 400,000 pesos. 151 By this point, the total amount extorted from Banamex since the beginning of 1917 totaled Mx$7. 9 million. On April 17, the federal government requested a loan of Mx$1.5 million for itself and an additional Mx$500,000 for the Comision Refaccionaria de la Laguna, "to help the farmers of the region." 152 It is not known how much, if any, of the Comision Refaccionaria's resources actually got to the farmers of the Laguna. On April28, the government requested an additional800,000 pesos, and Legorreta traveled to New York to meet with Jose Simon, Banamex's president-in-exile, to discuss the situation. 153 By this time, the government's debt totaled Mx$10.7 million. On May 15 the government demanded another loan ofMx$500,000, ostensibly for the Comision Refaccionaria, offering Banamex a share of the profits from the Comision's lending activities rather than interest. Subdirec-

Financing a Revolution

157

tor Castello, negotiating in the bank's name, asked for 6 percent. 154 The government rebuffed his position. 155 In May the government liquidated the Banco de Coahuila and the Banco Mercantil de Veracruz. After confiscating the two banks' specie reserves the government withdrew the 140,000 pesos that the Banco de Coahuila had deposited at Banamex's Torreon branch, as well as an additional million pesos that had been in the Banco Mercantil de Veracruz's checking account. The two withdrawals reduced Banamex's specie reserves by almost hal£ 156 When Legorreta returned from New York, the government demanded a loan of Mx$1.5 million in gold or silver pesos. Banamex refused, considering that its Mexican hard currency reserves had dropped to only Mx$1,548,884. 157 The government changed its demand to a loan of Mx$200,000 in Mexican gold pesos for the Comision Monetaria and a loan ofUS$760,000 for the government. Banamex's board "agreed ... considering it was not possible to discuss the order." 158 At this point, valuing the US$760,000 loan at Mx$3,534,000, Banamex had made a total of 14.9 million pesos of forced loans to the Constitucionalista government. 159 Its dollar reserves were gone. On June 23 the government confiscated Banamex's remaining gold reserves with the stroke ofa pen. Banamex's final loss totaled Mx$16.5 million. This left the bank with silver reserves valued at approximately a half million pesos. Unfortunately, by the end of 1917 this too disappeared, as the government liquidated more of the state banks under the provisions of the decree of September 15, 1916, and withdrew their Banamex accounts. By October Banamex was completely denuded ofliquidity. 160 The entire financial system established during the Porfiriato had been shut down by the end of 1917. The banks existed in a state of suspended animation, still existing as legal units but lacking the ability or authority to conduct any business. The Banco Central, one of the linchpins of the Porfirian system, had gone bankrupt and disappeared. The check-clearing operations in Mexico City had ceased. The banks' day-to-day operations were run by government-appointed officials. By the time the government passed the Payments Act of 1918, which suspended all debt payments, the financial system had already shut down. Not including the seigniorage gains from the Huerta and Constitucionalista inflations or the smaller seizures and sackings by other revolutionary factions, the regimes of the period expropriated at least 85 million silver pesos from the banks, broken down by bank in Table 7 .1. In terms of Mexico's GDP, this might not have been that high. In terms oflost confidence, it was huge.

Financing a Revolution

158

Table 7.1 Forced Loans from Revolutionary Governments Constitucionalista

Huerta Banamex* Banamex** Londres y Mexico Coahuila Durango Estado de Mexico Guanajuato Guerrero Hidalgo Jalisco Minero de Chihuahua Morelos Oriental Peninsular Queretaro San Luis Potosi Tabasco Tamaulipas Zacatecas TOTAL

14,934,000

11,995,400 19,471,377

17,526,376 100,000

11,705,752 1,852,752 1,550,000 1,470,194 1,630,378 215,989 985,157 2,100,531

1,023,949 690,920 1,212 492,358 12,056 590,000

300,000 3,949,000 4,249,996 555,147 655,105 400,070 2,080,000 190,046 33,980,320

6,477,997 1,940,209 851,000 396,506 925,520 549,274 19,471,377

31,577,377

souRcE: Bolet{n Financiero y Minero, 11110/20. *These estimates ate derived by adding up the various loans mentioned in Banamex's Aetas de Consejo. They do not include the final confiscation ofBanamex's specie reserves on June 23rd, 1917. **This estimate comes from the same source as the rest of the chart. The Boletin Financiero y Minero based its chart on the official declarations of the incoming Obregon government in 1920. It did not distinguish between the various governments for Banamex's debt.

Conclusion The Mexican Revolution turned Mexico's economic life upside down. This was not done through the destruction of physical capital and equipment. In fact, Mexico's manufacturing plant, irrigation systems, mining installations, and railway network came through the fighting relatively intact. Rather, it came about through the widespread violation of property rights.

Financing a Revolution

159

The greatest challenge facing Venustiano Carranza's government, and his successor Alvaro Obregon, was to convince bankers and investors to begin betting on Mexico's future again, rather than hiding their wealth overseas or under mattresses. If the private sector was to be brought back to life, the government would have to make some sort of credible commitment that the property rights of investors would be protected.

Chapter 8

Reconstructing Confidence There is an inevitable tendency to think in terms of obtaining sufficient capital from the state. -Alberto Pani

The violence of the Mexican Revolution shook the confidence of the nation's business class to the core. During the Revolution, unstable governments succumbed to strong incentives to steal bank reserves, force financial institutions to make them loans, engage in the unrestrained printing of currency (thereby setting off an inflation that will essentially be a tax on holding cash), and change the rules that regulate banking and the securities markets so as to maximize the government's access to funds. The Mexican financial system was negatively affected by political instability, but only in the short run. As Chapter 7 discussed, the civil wars of191317 brought about a near total collapse of the banking system. The governments that came to power after 1917, regardless of their stated ideologies, all recognized that they needed to come to an accommodation with the bankers. First, they understood that they needed a source of credit in order to restore political order. Second, they understood that without a functioning financial system, there could be no economic growth, and without economic growth, there would be no tax revenues. In short, they realized that restoring the financial system was crucial to their own political survival. The governments of the 1920s therefore tried to recreate the arrangements that underpinned banking during the Porfiriato. Just as Diaz had done, they selectively enforced property rights and allowed the bankers themselves to write the laws regarding entry into banking. To make its commitments

160

Reconstructing Confidence

161

stick, the government had to create a hostage: a government-owned commercial bank, the Banco de Mexico, or Banxico. Banxico served to link the interests of important politicians to the interests of the bankers and to insure that the government's promises would be kept. As a result, the domestic banking system grew rapidly during the 1920s. One implication of the selective enforcement of property rights, however, is that there must have been some group oflosers. In this case, the losers were Mexico's foreign creditors. In order to create a credible commitment to the country's domestic financiers, the government abrogated the foreign debt from the Diaz and Huerta regimes.

Failed Attempts to Reestablish Confidence By the middle of 1917, the government had squeezed the banks for all they were worth. Unfortunately, the government's need for revenue remained. Reconciliation with the private sector was necessary if the Constitucionalistas were to reestablish the credibility needed to reopen Mexico's access to credit. Unfortunately, a combination of corruption and continuing penury led the Constitucionalistas to continue arbitrary and confrontational policies toward the banks. President Carranza authorized Alberto J. Pani to organize a national conference of manufacturers in 1918. 1 The industrialists' vituperative opposition to the new revolutionary constitution surprised the government. Opposition centered around Article 27 (which reserved all subsoil rights to the state and banned the corporate ownership ofland) and Article 123 (which established progressive labor norms). 2 Changing the constitution would alienate the government's radical labor and agrarian supporters; therefore it was not an option. The government could, however, ignore the constitution's radical provisions and return occupied or confiscated properties to their owners. 3 To curry additional favor, when possible, the government also granted favors to important capitalists. The State of Mexico, for example, assumed the debts of Arturo Braniff, a prominent industrialist and Londres y Mexico board member. 4 Similarly, Nuevo Leon's revolutionary government agreed to take over Mx$9,000 in annual payments on the Fundidora Monterrey steel plant's debts. 5 In September 1918 Finance Secretary Rafael Nieto asked Banamex's help in guaranteeing a wheat shipment to Veracruz valued at US$300,000. He asked Banamex to guarantee the payment of a money order to the Equitable Trust Company of New York for 4,000 tons of wheat, for which the bank

162

Reconstructing Confidence

would receive a 1 percent commission. In what appeared to be a first step toward reconciliation, the board approved the operation. 6 Following the wheat operation, the Ministry opened talks with Banamex over the removal of the juntas de incautaci6n. Banamex was eager to free itself of the government's control over its operations. The proposal the two parties worked out in two weeks of negotiations contained the following provisos: 1.

2.

3.

7. 9. 10. 11. 12. 13. 14. 15.

Banamex would agree to cease all future banknote issues unless the constitution was changed in the future, but would retain the rest of its original concession. Banamex would agree to receive its notes at par in payment of any debts contracted before April 1, 1913, and all credits could be paid in gold, silver, or foreign money, subject to the exchange rates published in the decree ofDecember 24, 1917. The bank would withdraw all its notes from circulation during an eight-year period. (Provisos 4 to 6 contained the details for the repayment of federal debts.) The government would pay 10 percent interest on its debt. Any future foreign debts contracted by the government would be used to pay off the interior debt. All note reimbursement would be supervised by an interventor appointed by the government. All other official intervention in Banamex operations will cease. The government will allow the bank to reestablish its branches and business relationships. Any special privileges given currently existing banks will also be granted Banamex. The government will pay Banamex compensation for the right to pay part of its debt in fiat currency. The government will pay the 6 percent ten-year debt of 1913, or equivalent compensation. 7

The bank appeared satisfied by the substance of the contract. 8 Unfortunately, the government failed to ratify the agreement. 9 Additionally, the wheat guarantee turned into a fiasco. The wheat was turned over to the government directly, rather than to the Banamex office in Veracruz, resulting in the nonpayment of the money order and series of angry letters from Equitable wondering what was going on. 10 The government paid Mx$545,454 for the wheat, at an exchange rate of 55¢ per peso, but Banamex had to shell

Reconstructing Confidence

163

out Mx$576,923 to cover the US$300,000 money order at an exchange rate of 52¢ per peso, the difference eating up Banamex's US$3,000 commission and souring the bank on its ability to do business with the government. 11 Other banks' relations with the government were equally contentious. For example, before the Revolution the Banco Oriental de Mexico held a majority stake in Descuento Espaiiol. 12 In 1916, the government took control of the Banco Oriental. The Constitucionalista government then put the Cornisi6n Monetaria, as the legal successor to the old Cornisi6n de Cambios y Moneda, in charge of the bank and appointed a whole new board. Descuento Espaiiol's management called off its 1917 shareholder's meeting, in order to avoid conflicts with the Oriental's new management. Unfortunately, the Descuento Espaiiol could only avoid conflict for so long. On December 16, 1918, the Cornisi6n Monetaria called an assembly ofDescuento Espaiiol's shareholders. The government wanted to appoint a new board in order to seize the securities and real estate entrusted to Descuento Espaiiol by the Banco Oriental. The Descuento Espaiiol's management refused their request for an assembly, angrily accusing the Cornisi6n of complicity with those who had seized the specie reserves of the banks of issue. Nevertheless, the assembly took place, installed a new board of directors, and turned over the goods held in trust. Nicanor Gurria Urgell, the Descuento's former director, protested in no uncertain terms. The Cornisi6n Monetaria, he wrote, turned itself into a judge and, without any legal foundation or the correct formalities, convened for itself and before itself the meeting it wanted. It surely forgot that the Commerce Code and the Descuento's statutes reserve that right only to the Board and Board members .... This simulacrum of a Board ... began by declaring the meeting good and legal. With what authority? With what powers? Since when has the law been changed to permit one to dispose of the house of another by simply affirming that one is indeed the owner? He went on to accuse the Cornisi6n's representatives of preventing smaller shareholders from speaking freely about the "illegalities," asking them to leave when they tried. 13 The Cornisi6n Monetaria's president, C. Basave del Castillo, responded to the accusations first by stating that the Descuento Espaiiol's refusal to call a meeting in 1917 meant that they recognized the Cornisi6n's legitimacy. 14 A few days later it published a statement stating, "Among the goods that made up the assets of the Banco Oriental de Mexico and other banks were a considerable number ofDescuento Espaiiol shares, and the Cornisi6n Mone-

Reconstructing Confidence

164

250 240 230 220 210 200 190 180 170 160 150 140 0 130 rl 120 0 0 ;::: 110 100 90 80 70 60 50 40 30 20 10 0

...

Figure 8.1.

\ \

\ \

\

\

\ \

\ \

\ \ \ \ \

\ \ \ \ \

\

\ \

\ \ \

\ \

\ \ \ \

\ \

\ \ \ \

\

\

\

\ \

\ \

\

Real Banking Stock Index

taria, as the representative of these institutions, has the unavoidable obligation to monitor the conservation of their goods." 15 That was the end of the debate: the federal government gained control of the Descuento Espafiol's assets. Stock prices show the lack of confidence investors had in the banks under postrevolutionary circumstances. Figure 8.1 is a weighted index of the end-of-month price of all bank stock on the market. Buy prices were used except when transactions were made. 16 It indicates the scale of the losses suffered during the Revolution, and the serious investor doubts that they might

Reconstructing Confidence

165

soon be overcome. On average, banking stocks traded for less than 10 percent of what they had before the Revolution. In 1918 the only bank to pay any dividends was the Banco de Zacatecas: the yield was 63 percent.'7 In 1922 the Banco de la Laguna paid a dividend yielding 17 percent, while Banco de Nuevo Leon (BNL) stock yielded 10 percent. 18 In 1923, BNL yielded 13 percent; in 1924 its yield dropped to 9 percent. Credito Espaiiol (distinct from Descuento Espaiiol), a bank founded after the Revolution, yielded 24 percent in 1922, 14 percent in 1924, and 13 percent in 1925. 19 The value ofbank stock after 1918 drifted up and down, but despite substantial increases in the value of certain stocks (BNL shares, for example, rose from Mx$15 to Mx$57.50 right before the announcement of its 1922 dividend), banking shares basically stagnated. Even a 1921 announcement that the government would remove the juntas de incautaci6n from selected banks failed to induce a rise in the price of banking stock. Low investor confidence in the banks reflected low confidence in the economy and political environment. Banamex, for example, almost completely divested itself of all its long-term investments in the wake of the hyperinflation and Constitucionalista forced loans. In 1917, it sold large shares of its securities holdings at discounts ranging from 5 to 89 percent and liquidated long-term loans to the Fundidora Monterrey steel manufacturer in Nuevo Leon, a major lumber producer in Durango, and Mexico City's main streetcar company. 20 In 1918, it sold off its equity stakes in Fundidora Monterrey and the San Antonio Abad textile conglomerate. 21 In 1919 it divested itself of its ownership of the Cigarrera Mexicana cigarette company and sold off the El Cuyo y Anexas sisal plantation in Yucatan. 22 Companies affiliated with Banamex begged for credit, but the bank consistently refused. In some cases, the bank approved an operation, but it could not find the cash needed to carry it out. 23 In others, like its refusal to finance a new hydroelectric power station in 1917, it decided the risks were too high. 24 Agustin Legorreta himself pleaded with the board of directors to approve a loan to the Cemento Portland company to fix broken and worn-out machinery, but the board said no. 25 Other institutions also pursued extremely conservative policies. The Compaiiia Bancaria de Paris y Mexico, a private bank started up in the final years of the Porfiriato, kept 21 percent of its assets in the form of gold pesos in its vaults in 1919. That figure is almost certainly an underestimate, since the Compaiiia admitted in the financial press that its assets included "old accounts in the process ofliquidation, and therefore does not represent the true state of the company." 26 The Banque Fran~aise du Mexique, which got

166

Reconstructing Confidence

its start after the Revolution trading banknotes on the Bolsa, kept 17 percent of its assets in gold and silver coins or U.S. dollars in 1920 and maintained a further 15 percent as deposits in overseas banks. 27 Commercial loan rates oscillated between 24 and 36 percent annually, when credit could be had at all. 28

Incredible Commitments Under Obregon On April22, 1920, General Alvaro Obregon called for President Venustiano Carranza's ouster and replacement. By the end of May, Carranza had "committed suicide" by riddling his own body with machine-gun bullets in the hills of the state of Puebla. After a brief power struggle with General Luis Gonzalez, General Obregon became president on September 5. Many bankers reacted with barely disguised jubilation at the fall of a government they had long considered their nemesis. Agustin Legorreta of Banamex called a special board meeting to discuss the new developments, which have "totally changed the situation in the country." 29 Legorreta and the new finance secretary, Salvador Alvarado, opened discussions about regularizing the financial situation. Legorreta suggested that the new government's first order of business ought to be to abolish the juntas de incautaci6n. 30 The new regime agreed. On November 5, 1920, Secretary Alvarado announced his intention to repeal the Payments Act of 1918. 31 The following day he publicized a proposed law to regularize the old banks of issue, liquidating those banks with insufficient assets to cover their liabilities, but permitting solvent banks to reopen and requiring them to redeem their outstanding note issues within five years. On November 10, he announced that the government would fully recognize its debt to the banks. 32 This was a commitment, but not a credible one. The proposed law was not well-received. The Bolet{n Financiero y Minero editorialized: The truth is that the authors of the law appear to have become attached to the idea that credit is something forged in the clauses of an ordinance, showing an eagerness to solve the problem through the continuing intervention of the Finance Ministry. It is indisputable that everyone is hoping that the Republic's immediate future will bring a better administrative morality than we have enjoyed in past epochs, but it has occurred to no one that credit and confidence might not be restored barely the day after the successive blows that destroyed them. 33 The paper's editorial writers more specifically criticized clauses in the law that regulated the banks' ability to choose how to distribute their profits and a

Reconstructing Confidence

167

prohibition on rediscounting by any bank other than an as-yet-nonexistent federally owned single bank of issue. They also worried that the law did nothing to ease the problems of the old banks of issue, "but rather makes the rehabilitation of the majority highly improbable." 34 On November 15 the Bolet{n revealed that a series of ongoing bankers' meetings, including every bank except Banamex and the Banco de Morelos, had been unable to come to an agreement over how to pay off the holders ofbanknotes or how to liquidate liabilities made after "the beginning of the banking disorder in 1910 or 1912." All agreed that the government would have to return the assets seized under Carranza, and most agreed that a staterun single bank of issue was thoroughly infeasible, but other than that the bankers held "diametrically opposed" opinions. 35 After consulting with representatives of the Banco de Sonora, Banco Occidental, Banco de Londres y Mexico, and the Banco Minero de Chihuahua, President-elect Obregon proposed the creation of six regional and one national bank of issue. Obregon's point was political, not economic. The motivation was that a plan for seven concessions for banks of issue gave Obregon more favors to distribute than a single bank of issue. 36 It was not clear, however, why the plan would be credible. Six creditors would have trouble coordinating the punishment of the federal government should it renege on its debts. The bankers faced another dilemma regarding their outstanding banknote issues. By the end ofNovember, they had agreed to pay all their creditors who had taken out accounts before 1913 in gold pesos, but they could not decide whether to accept their own banknotes at par or at their discounted market values. 37

The Panic of 1920 The combination of a cotton crop failure in northern Mexico (caused by extensive flooding) and the collapse of international copper prices provoked a banking panic in December 1920. 3R The trigger was a rumor that the Compaiiia Bancaria de Paris y Mexico had made important advances to cotton growers. The Paris y Mexico was now one ofMexico's most important banking institutions, considering that most of the old banks of issue remained either closed or subject to close government supervision. 39 The bank closed its doors on December 30 with the excuse that it needed to make its end-ofyear inventory. On January 3, 1921, authorities sealed up the Paris y Mexico's building, announcing that it was now in bankruptcy proceedings. 40 Bank runs spread.

168

Reconstructing Confidence

A rumor proliferated that the manager of the Mercantile Banking Corporation, a prominent American-owned bank, had sent almost half-a-million pesos to the United States in an embezzlement attempt. Withdrawals forced the Mercantile Banking Corporation to shut on January 7. 41 Daily clearances on the Centro de Liquidaciones (which had only just restarted operations) fell by half, from Mx$1.6m to Mx$0.7m, and the faint revival in the credit market triggered by Obregon's accession to power collapsed. Judicial interventors soon found both the Paris y Mexico and the Mercantile Banking Corporation to be in fine shape.42 In Agustin Legorreta's words, "The financial crisis only existed in the minds of those who started the panic." 43 The Boletfn Financiero wrote on January 10, "There were no solid data, and the habitual silence of our banking institutions aggravated the mystery." The crisis, however, showed the weakness of public confidence and brought a sense of urgency to the government's banking reform project. Obregon moved quickly to free the old banks of issue, at least the solvent ones, from the burden of the juntas de incautaci6n. On January 31, he announced that Banamex, the Banco Occidental, Banco Mercantil de Veracruz, Banco Mercantil de Monterrey, and the banks of Nuevo Leon, Tabasco, Mexico State, Guerrero, Sonora, Hidalgo, and Zacatecas would be allowed to reopen, while the Banco Oriental and Banco de Queretaro would be shut down for good. 44 President Obregon also authorized the Comision Monetaria to rediscount the Mercantile Banking Company's paper. 45

Institutional Obstacles Unfortunately, the government's plans came to little. The private banks-ofissue proposal, losing favor among the bankers, soon died quietly in Congress. 46 Bureaucratic confusion allowed only Banamex, BLM, the Banco Mercantil de Monterrey, the Banco Occidental, and the banks of Sonora and Nuevo Leon to reopenY Seesawing over whether or not the federal government would recognize the Huerta regime's bank debt rollicked the markets. 48 Meanwhile the rest of the banking system ground to a halt for three reasons. The first was uncertainty over whether Article 27 of the Constitution of 1917 allowed banks to own farmland. Article 27 gave the state ownership rights over all subsoil or water resources, banned corporate landownership, and mandated the expropriation of large estates and their subdivision into communal landholdings. 49 The Banco de la Laguna, which had operated through the worst of the revolutionary violence and survived the 1920 loss

Reconstructing Confidence

169

of the cotton crop in its home region, debated shutting down over this issue. 50 A government official, Miguel de la Peiia, blamed the almost complete disappearance of agricultural credit on Article 27 _51 The second problem was political instability. Fears proliferated about the stability of President Obregon's regime and the outbreak of a new period of civil disorder. When the government finally issued eight-year bonds with a 6 percent coupon in exchange for its domestic debt, rumors flew that several banks tried to unload them on various oil producers at 40 percent of their nominal value. 52 The Bolet{n Financiero y Minero pinned the lack of credit on instability and the looming threat of violence. In an atmosphere in which property is continually assaulted and threatened, under social and political conditions under which no one can offer an absolute guarantee which is theirs in order to obtain credit, the functions of this particular factor ... are impossible to carry out. So we are living a precarious and anguished life, practically vegetative .... We do not believe this can be easily changed.... It is not possible to reestablish credit in Mexico while there persists a lack of confidence and insecurity about property, which is the foundation of credit. 53 Monetary uncertainty exacerbated the situation. Gold coins had been Mexico's primary circulating medium since the end of the hyperinflation, but in late 1921 an unexpected drop in the price of silver caused gold to leave circulation. The mint began to coin more silver pesos, but never opened an office in which gold could be exchanged for silver at par. As a result, the financial pages began quoting silver's discount against gold pesos, and Mexico found itself, in effect, with two currencies whose day-to-day value fluctuated randomly. 54 These problems combined to transform the banks into vehicles for spiriting capital away overseas. In early 1922 the Bolet{n Financiero y Minero wrote, "Without fear of correction, we can state that the majority of Mexican deposits-cash wealth-are found in the United States. American banks have made and continue to make formidable efforts to obtain deposits, and some of them maintain permanent agencies here with the sole object of soliciting deposits." 55 Joseph Sterrett and Joseph Davis, two American economists commissioned in 1927 by the International Bankers Committee to prepare a report on Mexico's economic condition, hinted at the persistence of this problem in their report when they pinned the low level of deposits in Mexican banks on "a tendency, on the part of many concerns, to keep deposits abroad rather than in Mexico because of the fluctuations in currency and exchange rates and fear of untoward developments of various kinds." 56

170

Reconstructing Confidence

What little banking activity there was before 1925 was concentrated in the petroleum zone around Tampico, often by American-owned banks employing American managers. 57 Tampico, however, was so disarticulated from the rest of the Mexican economy that its Chamber of Commerce lobbied the Finance Ministry to allow dollars to circulate freely within the region. sH

More Incredible Commitments-the Caja de Prestamos The federal government attempted to use the parastatal Caja de Prestamos para Obras de Irrigaci6n to reestablish credibility. This bank had been created in 1908 in order to inject liquidity into the banking system after the Panic of 1907. By 1922 it was moribund, but still existed as a legal entity. The government's idea was to issue bonds in order to purchase Caja stock from the banks and then use revenues from Caja operations to begin repaying those domestic bonds. 5 9 The government's consulting lawyer, Sotelo Regil, stated that as a parastatal institution the Caja de Prestamos would not be subject to the constitutional prohibitions on landownership, and would charge only 12 percent on its loans, less than half the going commercial rate. 60 In this way, the government would promote economic activity and at the same time make a commitment to respect property rights by giving itself a stake in the maintenance of stable private property rights over land. 61 The government named General Amado Aguirre manager of the institution. 62 The Caja de Prestamos experiment was based on the experience of the Comisi6n Refaccionaria de la Laguna, a state-owned body empowered in 1919 to take deposits and make loans to landowners in northern Mexico. The Comisi6n Refaccionaria gave the northern revolutionaries who owned its stock a direct stake in the maintenance of private property in the Laguna region. If property rights were to be threatened, so would the Comisi6n's profits. The knowledge that powerful politicians in the state government had a stake in the maintenance of private property gave the managers of the Banco de la Laguna sufficient confidence to restart their own lending operations. 63 The federal government hoped to repeat this salubrious experiment nationwide. Unfortunately, the commitment was not credible. Politically powerful interests did not buy Caja bonds. Rather, the bonds wound up in the hands of powerless domestic bankers or New York investors. The federal government therefore never paid interest on the bonds. By June 1921, outstanding interest had risen to Mx$21.2 million, and the holders of its New York bonds began judicial proceedings. 64 In point offact, not only did the Caja notrepay its creditors, but also it did not try to maximize income from its assets.

Reconstructing Confidence

171

When it did repossess a mortgaged property, it did not sell it. Instead, it transferred that property to the government so that it could be redistributed to the politically powerful. A private commission recommended that the Caja de Prestamos's management should slow these transfers, but General Aguirre, the director, publicly announced that he did not agree with the recommendations. 65 That meant that the Caja de Prestamos could not cut its losses from bad credits through the resale of mortgaged properties. 66 The Caja collected few outstanding loans from the pre-1913 period. 67 Most of its post-1921 loans were given to politically favored clients for far more than the value of the collateral, wound up in litigation, and were never repaid. 68 By October 1922 the organization had been placed in liquidation, and the federal attorney general began criminal proceedings against its management. 69 In short, as an attempt to reestablish faith in the government's commitment to respect private property, the Caja was an almost total failure. By the end of 1922, the government's credit problem regained urgency for three reasons. The first was the federal government's worsening fiscal crunch. In October 1922, the treasury admitted that it could not make the debt payments due in January 1923. 70 It was so short of cash that it required support from private moneylenders to carry out the day-to-day operations of the national railway company. In 1923 Obregon imposed new oil taxes, collected at the point of production and paid in cash, but the federal government lacked the administrative capacity to fully collect even these. 71 Government revenues increased, but far less than expected. 72 Political instability exacerbated the government's need for cash. Obregon's recognition agreement with the United States on September 3, 1923, initially engendered optimism among Mexico's financiers. 73 Unfortunately, many factions strongly disapproved of the agreement, believing it made far too many concessions to American interests. Obregon resorted to rather heavyhanded lobbying tactics, including bribery, terrorism, and the assassination of an outspoken opponent of the accords, Senator Francisco Field Jurado. Obregon's political strategy provoked a large-scale rebellion on December 5, 1923, led by Obregon's former confidant and finance secretary, Adolfo de la Huerta. Almost 70 percent of the regular army sided with the rebels. The resulting civil war would last four months before De la Huerta's forces were finally defeated. The revolt destroyed the government's plans to restart payments on its debts. On the same day that the revolt began, Banamex transferred US$13.5 million to the International Bankers Committee to restart payments on the foreign debt.7 4 The government then borrowed 10 million pesos from the Huasteca Petroleum Company against future oil tax liabilities, with the in-

172

Reconstructing Confidence

tention of using the funds to make additional payments on the foreign debt. 75 Instead, the money went to finance military campaigns against the rebels. 76 That meant that the government had to borrow from Banamex to pay foreign creditors. 77 Loan payments remained subordinated to current expenditures. 78 Memories of revolutionary confiscations fresh in their minds, many bankers contemplated sending their gold reserves overseas. 79 The revolt's defeat did not increase the government's ability to pay its foreign debt. It officially suspended foreign debt payments on July 2, 1924, and prohibited Banamex from loaning out any funds deposited in payment of federal taxes by the Ferrocarriles Nacionales company or the oil producers. 80 Despite unilaterally revoking its commitments with foreign creditors and Banamex, there was no way to enforce a credit embargo. The government continued borrowing money short-term from Banamex and the Huasteca Petroleum Company-albeit at exorbitant rates. 81 The third reason for the government's urgency was pressure from the private sector. Monetary conditions remained extremely tight. In 1922 four banks failed, and credit dried up almost completely. The Ministry of Finance hinted in private correspondence that the regime might not last much longer if the economic situation did not improve. 82 Concamin, the national manufacturers' association, called for the establishment of banks of issue to ease monetary conditions. The Chamber of Commerce called for intervention to ease the circumstances of otherwise solvent borrowers who failed to pay because they lacked liquidity. The secretary of industry, commerce, and labor published a report on Puebla's economy in 1923, one of Mexico's most important textile centers, blaming the recession on extremely tight credit. 83 Unfortunately, the De la Huerta revolt meant that the government was powerless to change the situation in 1924. Obregon's handover of power to his handpicked successor, Plutarco Calles, only exacerbated the uncertainty. Plans to create a single bank of issue were dead for the duration. 84

Plans for a Single Bank Mexico had no lack of plans for a private mechanism to coordinate loans to the government. Unfortunately, Mexico's foreign creditors could not agree among themselves. Proposals by French bankers and economists to return Banamex to the role it had enjoyed during the Porfiriato floundered on American opposition. 85 By the end of 1922, a deal with Banamex was off the table. 86 J. P. Morgan proposed the establishment of a "Banco MexicoAmericano," on the basis of a US$50 million loan to the Mexican gov-

Reconstructing Confidence

173

ernment. The Banco Mexico-Americana would keep substantial reserves in U.S. banks. Therefore, if the Mexican government reneged on its commitments, it would face both a credit boycott and the loss of these reserves. 87 These plans sank with the outbreak of the De la Huerta revolt. Speyer & Company, another American creditor to the Mexican government, had different ideas. In 1922, it engineered the appointment of Eduardo Iturbide as manager of the Banco Mexicano de Comercio e Industria, and instructed him to purchase a majority interest in the bank from its Mexican shareholders on Speyer's behalf Speyer hoped the BMCI would form the basis for a privately financed central bank to the new government. This plan too came to naught. 88 Once the De la Huerta revolt forced the government to suspend all payments on the foreign debt, any plans involving foreign capital were worth no more than the paper on which they had been written. 89

Establishing a Credible Commitment The Calles government, therefore, was left in a quandary. Domestic manufacturers were unhappy because there was no payments system and they could not get credit. Foreign bondholders were unhappy because the government was not making payments on its prerevolutionary and revolutionary-era debts. Domestic bankers were unhappy because the government had abrogated their property rights and had made no commitment to them that it would not do so again. Last but not least, the government needed cash, and needed it badly, in order to restore political order. The problem was that the government could not simultaneously satisfy all of these groups and still provide itself with adequate resources to defend itself against potential rivals. If it had possessed the capability to increase tax revenues rapidly, then it could have satisfied all parties. Tax increases were not, however, an option. The most likely source of such an increase would have been the oil sector. Obregon had tried, in fact, to raise oil taxes. The oil companies rebuked him, however, by shutting down production until he compromised on a lower tax rate. Moreover, even if the government obtained a higher tax rate, total oil tax revenues were falling because Mexico's easily tapped deposits were running out for purely geological reasons. 90 The only way out, therefore, was to make a commitment to some of the interested parties that their property rights would be protected, while leaving other interested parties without property rights protection. That is, the government was going to have to selectively enforce property rights.

174

Reconstructing Confidence

Calles's political calculus was straightforward. He essentially decided to focus on restoring domestic financial credibility while abandoning Mexico's foreign creditors. Given the problem Calles faced, a de facto default on the foreign debt was a logical move. Picture a sovereign borrower, like Mexico, who promises to repay a loan over a period of years at a certain interest rate. Each year, when the interest and principal payments come due, the sovereign must decide whether to repay or renege. The lenders must then decide whether to impose a penalty. The key insight is that lenders will not lend more than the present value of the penalty they can credibly impose. In other words, there is a maximum amount that they will lend, regardless of the sovereign borrower's financial prospects or the interest rate offered on the loan. Therefore, if sovereign governments want to increase their debt limit, they must increase the ability of lenders to impose penalties on them. 91 In 1923, Mexico was subject to few penalties by foreign lenders short of sending in the U.S. Marines to dislodge the government, something the United States was in no way prepared to do. The bottom line was that even if Mexico repaid its outstanding loans, it was extremely unlikely that foreigners would extend more credit. The government could, however, substantially and credibly raise the penalties for reneging on its domestic debt. It did so by delegating policymaking to the bankers, and creating a forum through which the bankers could coordinate their actions. It also created a single bank of issue that acted as the Treasury's financial agent and provided a mechanism by which politically crucial individuals received a stream of rents from the policies designed by the bankers. The government's first move was to allow the bankers to write a new and favorable banking law. With the aim of"harmonizing public and private interests" and ending "the disjunction between banking legislation and actual circumstances," Finance Minister Alberto J. Pani called a national bankers' convention in late 1923. 92 Pani was by this time an old hand at organizing "conventions" such as this one in order to form coalitions with wealthy asset holders. In fact, Pani had earlier run a similar convention of industrialists in the fall of 1918, while Carranza was still in power. 93 One ofPani's crucial moves was to signal the bankers that they were being invited to steer legislation back to the status quo ante. He accomplished this by inviting two of Porfirian Mexico's most visible, and perhaps notorious, bankers to "advise" the convention. The first, Pablo Macedo, was a highly placed Banamex board member who had also been the president of Congress under Porfirio Diaz. The second, Enrique Creel, was one ofPor-

Reconstructing Confidence

175

firian Mexico's most brutal (and greedy) governors. Creel and his father-inlaw, Luis Terrazas, had run the state of Chihuahua almost as a private business enterprise. His holdings included, among a wide variety of other things, Chihuahua's most important bank, and he had been a founder of the nowdefunct Banco Central. 94 Unsurprisingly, the resulting legislation was tailor-made to protect the interests of the existing banks. The new law, enacted in 1925, established extremely high barriers to entry. Banks operating outside the Federal District required a minimum capital investment of 500,000 pesos, wl1ile those that wished to operate a branch or office in the capital had to invest at least 1 million pesos before they could open their doors. In addition, no banks could operate without the specific approval of the finance secretary and the president. 95 This last clause allowed the finance secretary to allow the new small institutions that had sprung up during and after the Revolution to remain in operation despite the fact that most of them did not meet the minimum capital requirement. 96 The resulting system was therefore more competitive than the Porfirian banking system in the short run, but protected from new competition in the long run. In other words, all the existing bankers joined Calles's coalition, and they used their position to prevent the possibility of future competition. The banking convention also created a forum that allowed the bankers to coordinate their actions vis-a-vis the government: the Cornisi6n Nacional Bancaria (CNB). Prominent bankers were named to the commission, which became an informal "legislature" responsible for all banking and credit policy. 97 Of course, the CNB's decisions could be overturned or ignored by President Calles. What the CNB did, however, was give the bankers an officially sanctioned way to coordinate their actions and pressure the government. The Calles government understood that its good word alone would not be a credible commitment to the bankers. The actions of every Mexican government since 1913 provided abundant evidence that the country's rulers would behave opportunistically toward the bankers if given the chance. The government therefore created a hostage-the Banco de Mexico (Banxico). Banxico was given a monopoly over government lending and a monopoly over paper money. The monopoly over paper money was not important, since the public resolutely refused to accept Banxico's banknotes. Rather, Banxico's importance lay in how it aligned the interests of the federal government with the interests of the bankers. It did so in two ways. First, the federal government capitalized Banxico with 50 million pesos of tax reve-

176

Reconstructing Confidence

nues. A good portion of these reserves was then lent to the banks. The banks proceeded to stash much of this gold overseas. If the government changed the rules, it would lose those reserves. Second, Banxico conducted an active and quite profitable business as a commercial bank. It then shared the rents from that business with a wide variety of important and powerful federal politicians, including Alvaro Obregon and Plutarco Calles. Any change in the rules that allowed greater banking competition would threaten those rents. In short, the strategy behind Banxico was not qualitatively different from the strategy behind the Caja de Prestamos, but it was applied more effectively. Mter suppressing the De la Huerta revolt, the government decided not to use the funds released by cessation of hostilities to restart payments on the foreign debt. Instead, on December 24, 1924, President Calles ordered that the receipts from the 10 percent railroad tax and the various petroleum taxes would henceforth go directly into federal coffers, rather than through Banamex's. He then reorganized the agency that collected the revenues, the Comision Monetaria, into a state-owned corporation under the management of Alberto Mascarenas. This gave the government both a source of finance (the tax receipts) and an organization that could serve as the basis for a bank. Banamex's president, Agustin Legorreta, pointed out that Calles's actions violated Mexico's agreements with the International Bankers Committee. He also argued that the money freed up by the suppression of the De la Huerta revolt should go to paying the international debt. These arguments went unheeded. 98 Calles's next move, on January 1, 1925, was to incorporate the Comision as a bank. 99 Three weeks later, the government promulgated a new banking act, based on the decisions of the Convencion Nacional Bancaria. 100 Finance Secretary Pani then appointed a committee to manage the new bank, consisting of Manuel Gomez Morin, Fernando de la Fuente, and Elias S. A. de Lima. Gomez Morin had studied the U.S. Federal Reserve system and central banking practice at Columbia University in New York in 1921. De Lima had managed the Banco Mexicano de Comercio e Industria during the Porfiriato, and General Obregon considered him one of the "most able financiers" in the country. De la Fuente was a Sonoran lawyer working in the Finance Revolution and enjoyed an extremely close personal relationship with President Calles and General Obreg6n. 101 The committee managed the Comision Monetaria until it accumulated sufficient resources to open its doors as the Banco de Mexico (Banxico) on September 1, 1925. The government immediately ordered Banamex to transfer over all funds and securities held as deposits for contracts with the

Reconstructing Confidence

177

federal government. 102 Four months later, the government transferred the revenues from the oil and railroad taxes to Banxico. 103 Banxico was chartered as a commercial bank, even though some of the stock was held by the federal government. Private banks purchased some Banxico stock in order to take advantage ofBanxico's rediscount facilities. That is, they could sell Banxico commercial paper and other receivables that they had themselves acquired. These banks could then stash the funds so acquired overseas, leaving Banxico with assets that would lose their value should the government begin to alter the structure of property rights. These banks included the Banco de Londres y Mexico and the Banco de Sonora. The first bank was lured into supporting the new central bank with a transfer of 5 million pesos worth of negotiable securities held by the Cornisi6n Monetaria as payment for the government's outstanding debt. 104 The Banco de Sonora associated with Banxico because of its longstanding political connections to the Sonoran clique now in power. Banamex ultimately also associated itself with Banxico. In late 1926 it affiliated with Banxico on the condition that the 10 percent deposit the law required it to place in Banxico be compensated by an equal deposit from Banxico to its coffers. 105 Other banks eventually followed. Banxico was, in short, a punishment coordination mechanism. It served this function in three ways. First, it coordinated private lending to the federal government. This provided a coordination mechanism by which domestic lenders could cut off future credit should the government renege on any new debts. Second, by making Banxico's rediscount facilities available to the private banks, the government reinforced its promise to protect bankers' property rights in the future. In essence, the government put up a hostage to guarantee its good behavior. When the banks rediscounted their receivables at Banxico, they received gold in exchange, which they could and did spirit out of the country (see Table 8.1). This gold would be lost if the government moved to seize the banks' assets. 106 In short, the government would have to weigh the gains from predation against the almost certain loss of most of Banxico's capital. Third, Banxico's profits served to guarantee that the government would not unilaterally liberalize banking law. Banxico acted as a commercial bank. Its profits from commercial lending would be dissipated if the government allowed greater competition in banking. Since Banxico passed on much of these profits to major politicians in the form of cheap (or free) capital, anything that reduced these profits would directly damage the interests of powerful political actors.

Reconstructing Confidence

178

Table 8.1 Assets Denominated in Foreign Currency, 1925-31, (1000s of pesos)

1925 1926 1927 1928 1929 1930 1931

Deposits in Foreign Banks

%of Assets

24,438 28,793 37,319 50,122 48,988 48,892 19,063

5.9% 5.6% 7.2% 8.7% 8.2% 7.7% 3.1%

SOURCE: Estadlstica Nacional (1925-29) and Comision Nacional Bancaria (1930-31).

The Banco Nacional de Credito Agrfcola Banxico was not the only state-owned financial institution created in the 1920s. The government accompanied Banxico with the Banco Nacional de Credito Agricola (BNCA), whose doors opened in 1926. The BNCA's institutional history in its early years was remarkably similar to its larger cousin. Its founders intended it to move capital into the credit-starved agricultural sector. They knew full well that attracting private investment would be wellnigh impossible under the circumstances of the 1920s, namely, continuing political instability, sporadic civil war, and a constitution that threatened private property rights in land. Agustin Legorreta, Banamex's general director, told the government in 1926 that the nation's capitalists would not invest without ample guarantees protecting their investments. 107 Manuel Gomez Morin, a prominent lawyer, designed the BNCA. On paper, the plan was that private banks and individuals would provide most of the BNCA's equity capital. Then, hoped Gomez Morin, the BNCA would issue mortgage bonds and debentures guaranteed by dedicated tax revenues and use the credit to fund rural credit cooperatives. 108 Higinio Alvarez, a congressional deputy, wrote that the bank would promote "the construction and development of irrigation works, hydroelectric projects, reservoirs, the installation of agricultural and industrial machinery, road construction, the development oflivestock and livestock products, the acquisition of vehicles and other means of transporting agricultural products, the increase of cul-

Reconstructing Confidence

179

tivation." 109 The Agricultural Credit Act of 1925 authorized a capital of 50 million pesos, of which 20 million in A-series stock would come from the federal government and 2 million in B-series from the states, supplemented by 28 million pesos of private capital. In reality, investment in the BNCA was much lower than the act authorized. The BNCA attracted only 2.3 million pesos in private investment. Fully 89 percent of that came from Banamex. State governments were even less enthusiastic. Only three states, Guanajuato, Tamaulipas, and Yucatan, bought any B-series stock at all, and their apportions altogether totaled only Mx$55,000. 110 Even the federal government's stake was less than it seemed. The government paid cash for only 53 percent of its investment. An additional 3 percent consisted of the offices of the old Banco Central, confiscated during the Revolution. The rest consisted of the assets of the Caja de Prestamos, which were mostly made up of unrecoverable debts and repossessed rural estates. 111 The point of the BNCA, however, was not to mobilize federal capital for agriculture. Rather, it was to give important and powerful politicians a stake in the maintenance of the existing structure of property rights. Banxico earned high profits off its commercial operations. Those profits subsidized cheap credits to the BNCA. The BNCA then loaned those credits out to politically connected individuals, "de favor," meaning without collateral. Access to cheap credit, therefore, depended on the existing banking laws and the existing structure of property rights. In 1926, for example, the bank authorized 355 credits, averaging Mx$48,270 each; in 1927 it made 809 loans, averaging Mx$11,246. Only 19 percent of these credits went to the rural peasant cooperatives. Alvaro Obregon himself was the single largest recipient of funds granted "de favor," which he used to become the largest producer of chickpeas in northern Sonora. 112 The fact that inside politicians received the lion's share of the credit produced a great deal of controversy, and the Mexico City newspapers continually accused the bank's management of favoring well-connected politicians. Nevertheless, the BNCA remained a going operation and succeeded in its aim of reinforcing the government's commitment to private property rights in land. 113

Banxico in Operation Banxico entered operation at a desperate time. In 1926, less than a year after the bank's founding, President Calles repudiated all federal debts contracted under General Huerta. 114 Although long-expected, the repudiation provoked a short-lived American boycott of all financial dealings with Mex-

180

Reconstructing Confidence

ico, which prompted bank runs in the middle of 1926. 115 Rumors spread that the government intended to shut down and seize the reserves of three Chihuahuan banking houses for illegally exporting gold to the United States. The rumors caused more runs and forced the Finance Revolution to declare a bank holiday in Mexico City in April. 116 Toward the end of the year, President Calles's attacks on the Catholic Church caused church supporters (later called "Cristeros") to take up arms in a civil war that would not end until 1929. In January 1927, when the magnitude of the Cristero Rebellion became apparent, additional rumors flew that President Calles had nationalized Banamex. 117 In 1928, a Cristero supporter assassinated Alvaro Obregon just as he was preparing to reassume the presidency from Calles, forcing Calles to declare himself ''jefe maximo de Ia revoluci6n" and rule through puppet presidents. Maintaining confidence in this atmosphere would not be easy. Banxico, therefore, moved fast. By the end of 1927, Banxico had loaned more than 31.3 million pesos to private borrowers, and 8.2 million pesos to its associated commercial banks. 118 Much of this private credit went to important politicians and their business interests. President Calles's sugar refinery, the Compaiiia Azucarera El Mante, received massive infusions of credit, causing a short-lived political scandal. 119 Other politicians and relatives connected with Calles received credit from the new bank. One of Calles's son-in-laws, Fernando Torreblanca, received large loans, as did Torreblanca's two brothers, Edmundo and Rodolfo. A sugar producer owned by another one of the president's sons-in-law, Jorge Almada, the Compaiiia Azucarera Almada, took out several long-term loans. Additional credits went to the politicians Moises and Aaron Saenz, Secretary oflndustry, Commerce, and Labor Luis M. Morones, Finance Secretary Alberto ]. Pani, former Finance Secretary Luis Cabrera, and, of course, former President Alvaro Obregon himself 120 As the bank's president, Manuel Gomez Morin, wrote in the 1928 annual report, "Although it appears, to those in a great hurry, that the Banco de Mexico has been slow and limited, the above data and considerations will show ... the bank has worked to full capacity as an ordinary credit institution." 121 Banxico succeeded in channeling benefits to powerful politiciansparticularly the president-therefore assuring the private bankers that the government would find it politically costly to unilaterally change the rules. Banxico also generated substantial returns for the government directly. Its returns-on-assets were low in its first three years of operation, but accelerated thereafter. 122 Mter 1928, it was the most profitable bank in Mexico once adjusted for risk. Liberalizing entry into banking would jeopardize these profits and damage the interests of the politicians who reaped the benefits.

Reconstructing Confidence

181

The only place where Banxico completely failed was at getting the public to accept its banknotes. Between September 1, 1925, and the end of the calendar year, Banxico issued 3.2 million pesos of paper money, much less than the 113.8 million pesos the bank's charter entitled it to issue. 123 During the next two years, Banxico's note circulation fell to 1.8 million pesos. 124 This is not surprising, considering the memory of the revolutionary hyperinflation. It must be remembered, however, that putting paper money into circulation was not the bank's primary aim. It's primary aim was to allow the government to credibly commit that it would not confiscate the bankers' assets.

The Success

if Financial Reconstruction

How successful was the government's strategy? How quickly and how efficiently did the banks recover from the devastation of the Revolution? Was the commitment to respect the banks' privileges as credible as it had been during the Porfiriato? Were the banks more or less efficient? SIZE OF THE BANKING SYSTEM

Perhaps the most straightforward way to determine the rate at which the banking system recovered from the depredations of Huerta and Carranza is to look at the real (inflation-adjusted) value of assets in the Mexican banking system. The analysis of total assets is complicated by the fact that once the government started to prey, it stopped gathering data in a systematic fashion. Thus, there exists a sizable gap in the series, stretching from the end of fiscal year 1911-12 to 1925, at which time the Comisi6n Nacional Bancaria once again started to gather data on total assets. Even with this gap, however, it is clear that there must have been a substantial drop in the size of the banking system. Total real assets in June 1912 were 450 million pesos (see Figure 8.2). As late as December 1925, total real assets stood at roughly 34 percent of their 1912 level. The data also indicate that, at least as of 1922 the banking system was growing rapidly. By 1930, the value of total assets had reached 61 percent of its 1912level, which is to say that the banking system grew by roughly 90 percent in the five years 1925-30. The data sets we have put together on total bank assets indicate that the rapid growth of the Porfirian banking system was largely over by 1909. Neither measure of banking sector size indicates much in the way of growth during the next two years. The reason for this is fairly straightforward: the regulatory structure laid down during the Porfiriato essentially prevented

$800,000 - - Tot~l assets, current - - To1.1l assets, 1900 pesos Cash and loans, current • • • Cash and loans, 1900 pesos

$700,000

'

''

$600,000

''

''

~ $500,000

''

''

" ......

c.. 0

-g

$400,000

''

i;1 ::l 0

;:; $300,000

''

''

''

''

$200,000

''

''

''

''

''

$100,000

''

'

''

''

*...

''

''

''

*••*···

•**

.. --···

'

$0 ,_.,_,,_.,-,,-.,-,,-.,ro-.~~-..-ro-ro-TO-o.-.,ro-.~~-, ~

~~~00~0-N~ ~

~~~OO~

O~

~~~~~~oooooooooo --

N~~~~~OO~O

-- - - -

-

-

N~V~~

-- N N NNNNN

~ OO~O -

N N N

N

~~~

00~~~~~ ~ ~~~~~~~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~

Figure 8.2. NOTE:

Total Size of Banking System

Real values calculated using the G6mez- Galvarriato-Musacchio price index.

Reconstructing Confidence

183

the system from growing. In order to issue new equity, state banks had to obtain the permission of the secretary of the Treasury-and he usually said no after 1905. In short, there is strong reason to believe that had there not been a revolution, the Porfirian banking system would have grown at a modest pace. It would have been larger than the observed banking system circa 1928, but the weight of Mexico's political institutions would have held it back. SURVIVING THE REVOLUTION

There were, in fact, a number ofbanks that survived even the depredations of Huerta and Carranza. Indeed, what is surprising is that circa 1921, after five years of forced loans and legalized theft, there was any banking system left at all. In point of fact, ten of Mexico's forty-two banks survived to 1921. Was there any systematic pattern that explains why some banks failed and some banks survived predation? Probit and logit analysis can help answer that question. Two hypotheses can be tested. First, that political connections with the faction that came to power in 1920 determined bank survival. Second, that a structure of assets that was illiquid (that is, that could not be easily seized and sold) determined bank survival. Region is used as an instrumental dummy variable for political integration. The rationale is that Alvaro Obregon and most of the politicians around him came from the northern border states and possessed-or came to possess after the Revolution-significant business interests in that region. In the case of one northern bank, the Banco de Sonora, this is true. We use the actual structure ofbank assets (coding bank balance sheets by liquidity of assets) in order to proxy the second variable. We defined liquid assets as securities holdings (primarily government bonds), secured credit lines, cash, and mortgage loans. Illiquid assets were discounts and collateralized loans. We included discounts among illiquid assets because they were often used to finance long-term loans. That is, the discounts tended to be rolled over at the end of their six-month term. Collaterized loans were generally made to individuals with some personal connection to the bank. The fictitious "collateral" was made up of shares in the business being financed. Both were ways to avoid the Porfiriato's legal restrictions on long-term loans. Porfirian bankers considered them riskier than other earning assets. 125 We include other conditioning variables in the regressions, including relative size (measured by share of total banking assets), age, and the riskiness ofliabilities. 126 The results are presented in Table 8.2. Only two variables had a significant effect on bank survival. The first was region. Chartered banks located in the northern border states were twice as likely to survive the Revolution.

Reconstructing Confidence

184

Table 8.2 Survival Regressions Li!git Estimates

Probit Estimates

Constant t-stat Equity ratio t-stat Ln (Risk: safe asset ratio) t-stat Ln (age) t-stat Ln (total assets) t-stat Center dummy t-stat North dummy t-stat

-2.87 (-0.48) 1.54 (1.14) 3.63* (1.81) 0.76 (0.91) 0.35 (0.58) 2.78 (1.37) 2.88* (1.94)

-1.90 (-0.54) 0.89 (1.13) 2.17* (1.91) 0.45 (0.92) 0.24 (0.67) 1.52 (1.42) 1.59* (2.06)

chi2(6) Prob > chi2 Log Likelihood Pseudo R2

15.99 0.0138 -12.59 0.3882

16.21 0.0127 -12.49 0.3937

Obs 32

*Significant at the 95 percent level

The banks in Nuevo Leon, Tamaulipas, and Sonora all reopened, making up 40 percent of the surviving banks. The liquidity of banks' asset portfolios also had a large effect. That is, banks that had a higher ratio of illiquid assets had a higher probability of surviving the Revolution than banks with more liquid portfolios. During the Revolution, firms defaulted on their secured credit lines, securities lost their value, mortgages proved impossible to repossess, and most other liquid assets were either seized by the government or became valueless as creditors defaulted. Illiquid assets, however, such as long-term loans to textile companies for plant and machinery, were much more likely to retain value or avoid confiscation. 127 Therefore, banks with more illiquid assets in their portfolios were less likely to have been completely liquidated by 1921 and more likely to reopen.

Reconstructing Confidence

185

NEW ENTRANTS

A large number of new banking institutions entered the market between 1920 and 1925. The number of banks in the system exceeded Porfirian levels by almost 40 percent by 1928, despite a wave ofbank failures in 1921 and 1922. 128 Many of the new entrants were Porfirian bankers starting under another name. For example, Jose Castello, a former Banamex board member, took over many of Banamex's former clients in Mazatlan, Sinaloa. Another former Banamex member, Bernardo Zorrilla, did likewise in Ciudad Victoria, Tamaulipas, near Mexico's primary oil zones. 129 The growth in the number of banks was not solely due to new domestic start-ups. Several foreign banks entered the market. Porfirian law had banned branches of foreign banks. The 1925 law removed these restrictions. Foreign banks entered, including the Canadian Bank of Commerce, the Bank of Montreal, the German Bank of South America, and the AngloSouth American Bank. 130 They were joined by the National City Bank in 1929 and the Chase Bank in 1930. 131 INVESTOR EXPECTATIONS

If the banking system was growing, one would expect asset prices to reflect rising investor expectations about future profits. In order to see whether changes in the prices of bank securities support the hypothesis that investor expectations were improving, a weighted index was constructed of the stock prices of all banks and financial companies traded in Mexico City. Figure 8.2 presents the results. Buy prices were used except when transactions were made. 132 In the early 1920s, banking stocks traded for less than 10 percent of what they had before the Revolution, regardless of the price index used. The value of the index drifts up and down, but despite substantial increases in the value of certain stocks (BNL shares, for example, rose from 15 pesos to 57.50 pesos right before the announcement of its 1922 dividend), banking shares stagnated. Even the 1921 announcement that the government would remove the juntas de incautaci6n failed to cause banking stock to rise. This suggests a paradox: investors were plowing resources into the Mexican banking system, but they had low expectations about future profits. This paradox can be resolved by the fact that most banks were no longer publicly traded. Only fifteen of Mexico's fifty-odd banks chose to list. Of those fifteen, seven went out of business by 1934. Two of the remaining eight, Banamex and BLM, remained going operations, but lost their special privileges. The post-1919 stock index, therefore, is biased toward the least profitable banks.

Reconstructing Confidence

186

0.400 . . . . . - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ,

0.375 0.350 0.325 0.300 0.275 0.250 0.225 0.200 0.175 0.150 0.125 0.100 0.075 0.050 0.025 0.000

+-----,-----,,....---~----.-----.---,.----.------i

1895

Figure 8.3.

1900

1905

1910

1915

1920

1925

1930

1935

Banking Market Herfindahl Index

Market Structure and Competition What effect did the reconstruction have on banking competition? Figure 8.3 presents estimates of the Herfindahl concentration index. The data indicate that Mexican banking was highly concentrated both before and after the Revolution. During the Porfiriato, the index never fell below 0.13. In 1928 it stood at 0.11. 133 By 1932, however, the Herfindahl had drifted back up to 0.16, a level of concentration roughly equal to the late Porfiriato. 134 The new banking hierarchy differed from the Porfiriato in detail. Banamex, for example, lost its dominance. In 1910, Banamex controlled 39 percent of all banking assets, 23 points more than the second largest bank, the Banco de Londres y Mexico. By 1928, Banamex's share was only 19 percent.

Reconstructing Confidence

187

Banxico was now the largest bank, with 25 percent. Banamex was second, and the Bank of Montreal third, with 10 percent. 135 The Banco de Londres y Mexico fell to fourth place with 5 percent of all banking assets. The Revolution also altered the geographic distribution of banks. Porfirian banks were deliberately widely distributed across the country. Northern banks disproportionately survived the Revolution, however, and new entrants concentrated in Mexico City or the Tamaulipas oil zone. 136 The far south-Chiapas, Oaxaca, Guerrero, Tabasco, and Morelos-lost all their banks. Banking became far more geographically restricted. RISKS AND RETURNS

Were banking profits higher after the Revolution than during the Porfiriato? The answer is no. This is consistent with the stock market data. Did Banamex lose its privileged position? The answer is yes. Once adjusted for risk, Banamex's profits no longer appeared to be particularly high. Did Banxico earn supernormal profits, compared to the rest of the banking system? The answer to this question is also yes. Table 8.3 shows the real returns on banking equity in 1901-10 and 192831. We use 1928-31 as the basis for our postrevolutionary comparison because no systematic data for individual banks is available for 1920-27. Returns were calculated as dividends plus changes in net worth, calculated in 1900 pesos using the G6mez-Galvarriato Mexico City price index through 1929 and the INEGI cost-of-living-index thereafter. Average returns-onequity were less than half the level of the late Porfiriato. The banking business was also riskier in the 1920s than during the Porfiriato. The standard deviation of returns-on-equity rose ten percentage points, despite the drop in average returns. In other words, risk-adjusted returns fell by more than the absolute drop in profit rates. The loss of Banamex's privileges affected its ability to make profits. Figure 8.4 graphs the banks' average return on equity against the standard deviations of those returns. In other words, they show how profitable banks were relative to their risk. In 1901-10, Banamex clearly stands out from the pack. By 1928-31, this was no longer the case. Real (inflation-adjusted) returns on bank equity were lower and risks higher after the Revolution than before. Inasmuch as high Porfirian profits represented rents accruing to the bankers, banking became more competitive after 1925. There were positive benefits to the government's decision to integrate with a broad spread ofbankers in 1924. This is not to say that the bankers reaped no benefits from restrictions on competition. Rather, for most banks, the benefits were less than they had been during the Porfiriato.

Table 8.3 Banking Profits Average real returns on assets

1901 Weighted average Unweighted average Standard deviation

2% -2% 16%

1902

1903

1904

1905

1906

1907

1908

1909

1910

5%

10%

-1%

13% 15%

1% 8%

18% 13%

7% 5% 3%

3%

6% 11%

6% 5% 6%

0% -3%

0% -1%

13%

3%

15%

4% 6%

AVG. AVG. 1907-10 5% 4% 10%

2% 1% 7%

Average real returns on equity

Weighted average Unweighted average Standard deviation

1901

1902

1903

10% 1%

27% 26%

34% 42%

33%

61%

41%

1904

1905

1906

1907

1908

1909

1910

1%

66% 49%

31% 18% 16%

13% 17% 36%

22% 76% 154%

43% 54% 191%

1% -3%

5% 23%

56%

10%

AVG. AVG. 1907-10 25% 29% 62%

20% 36% 98%

Average real returns on assets

Weighted average Unweighted average Standard deviation

1928

1929

7%

6%

6%

6%

29%

35%

1930

1931

AVG.

2%

5%

5%

5%

2%

5%

24%

42%

32%

Average real returns on equity

Weighted average Unweighted average Standard deviation

1928

1929

1930

1931

AVG.

13%

-1%

9%

13%

9%

21%

0%

9%

31%

15%

112%

121%

56%

143%

108%

Reconstructing Confidence

190

80%

·s0"

c

60%

"0

"'E

50%

•

,

70%

•

• •

• •

B 40%

•

~

"~