Corporate financing and governance in japan the road to the future

Theauthors argue that many of the current difficulties of the Japanese finan-cial system result from the mismanaged process of financial deregulationthat allowed borrowers to shift away

Governance in Japan

The MIT Press

Cambridge, MassachusettsLondon, England

Governance in Japan

The Road to the Future

Takeo Hoshi and

Anil K Kashyap

All rights reserved No part of this book may be reproduced in any form by any electronic

or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher.

This book was set in Palatino by The MIT Press.

Printed and bound in the United States of America.

Library of Congress Cataloging-in-Publication Data

Hoshi, Takeo.

Corporate financing and governance in Japan : the road to the future / Takeo Hoshi and Anil Kashyap.

p cm.

Includes bibliographical references and index.

ISBN 0-262-08301-9 (hc : alk paper)

1 Finance—Japan—History 2 Banks and banking—Japan—History 3 Corporations— Japan—Finance—History 4 Corporate governance—Japan—History I Title.

HG187.J3 C67 2001

Foreword by Stanley Fischer vii

6 Benefits and Costs of Keiretsu Financing 185

7 Transformation through Deregulation 219

8 The 1990s: Crisis and Big Bang 267

9 The Future 305

Bibliography 329

Author Index 347

Subject Index 351

Stanley Fischer1

It has been clear since the early 1990s that the future of the Japanesefinancial system is critical to the future of the Japanese economy, othereconomies in Asia, and indeed the global economy Much has beenachieved in reforming and strengthening the banking system, but muchremains to be done Nor is it obvious what type of financial system andsystem of corporate finance will emerge when the Japanese Big Bangderegulation program is completed

Hence this book by Takeo Hoshi and Anil Kashyap—whom I had thepleasure to meet, teach, and learn from, when they were students atMIT—is especially timely It is also exceptionally interesting Its theme isthat the past is prologue—and that the future is more likely to be similar

to the pre–World War II Japanese financial system, and to the currentUnited States and London-based financial systems, than to thepost–World War II Japanese financial system that many think of as anessential part of the Japanese economy

To establish this, Hoshi and Kashyap take us through the history of theJapanese financial system since the Meiji era They divide the historyinto four periods The first runs from the Meiji restoration through somepoint in the 1930s, when Japan began preparations for war They arguethat the Japanese financial system during that period was more like themodern United States system than the post–World War II system inJapan Firms, including medium-size enterprises, received most of theirfunding through the capital markets, both bond and stock markets Thebanks were important but not dominating

The second period ended after the postwar restructuring of theeconomy that resulted from the Allied insistence that the Japanese gov-ernment not repay war debts and the attempt to restructure the zaibatsu

1 First Deputy Managing Director, International Monetary Fund Views expressed are those of the author, and not necessarily of the International Monetary Fund.

In the 1930s and through the war, the Japanese government took control

of the allocation of credit, restricting the rights of shareholders, andusing the banks to implement its preferences And then at the end of thewar, the banks took the lead in establishing business plans as companieswere restructured, and in helping implement them It was during thisperiod that the banks became the dominant financial institutions.The third period ran from the early 1950s to the early 1970s This wasthe period of the Japanese miracle, at the end of which the Japaneseeconomy was the second largest in the world, despite its relativelyunderdeveloped financial system Savings options for individuals werelimited, so most savings went to banks, which intermediated them most-

ly to large industrial companies The government ran balanced budgets,and did not need to develop the bond market

The fourth period, a period of slow deregulation, is ending now:Hoshi and Kashyap date its end at March 2001, when the Big Bangderegulation process was completed During this period market financ-ing options were permitted for firms, but savings options for individualswere still relatively restricted Corporations began to borrow in thebond markets, but the banks and other corporations bought most of thebonds Operating restrictions on banks were lifted only gradually, sothat even though they were losing their blue-chip customers, the bankscould not offer new products and services to try to retain them

Hoshi and Kashyap argue this is why the banks went into new lines ofbusiness in the 1980s, including lending to smaller firms and the realestate sector, and hence why the banks got into such trouble in the 1990s.The banks also moved abroad, as capital controls were lifted Theauthors argue that many of the current difficulties of the Japanese finan-cial system result from the mismanaged process of financial deregulationthat allowed borrowers to shift away from bank financing, but pushedsavers to continue using banks, while denying banks the right to diversi-

fy their lending activities

The analysis is buttressed with a wealth of data and case studies,which make this not only a persuasive story, but also an interesting one

It ends in Chapter 9 on the brink of the fifth period Savers’ options will

be fully liberalized, but at the same time the banks will be able toexpand their business lines The authors recognize that there are manyways the deregulated system can develop, but make the case that themost likely outcome is a system closer to that of the first period, and theUnited States and United Kingdom systems, than the system of the third

period, which many people think of as the Japanese system.

However, they do not see a smooth path ahead for the banking system.The banks, they say, “are destined to shrink massively”—the Japanesebanks cannot remain the largest in the world while being among theleast profitable

It is of course far from certain that Hoshi and Kashyap’s scenario willeventuate But even if it does not, the reader of this book will have beengiven the essential elements of history, and the data and trends in the sys-tem, that will make it possible to follow and understand the unfolding ofthis critical story in the coming years

This book could not have been written without the help of numerouspeople.

We worked on this manuscript all over the world, and are particularlygrateful to the institutions that hosted and supported us These includeOsaka University, the International House of Japan, the Bank of JapanInstitute for Monetary and Economic Studies, the Center forInternational Research on the Japanese Economy (University of Tokyo),the Tokyo Center for Economic Research, the University of California atSan Diego, the Federal Reserve Bank of Chicago, the George J Stiglercenter for the Study of the Economy and State (University of Chicago),the Center on Japanese Economy and Business (Columbia University), theNational Science Foundation, the National Bureau of Economic Research,the London School of Economics Financial Markets Group, and theBanca D’Italia

Over the years, ongoing conversations with many of our colleaguesand co-authors were very valuable We specifically thank DouglasDiamond, Peter Gourevitch, Yasushi Hamao, Randy Kroszner, JohnMcMillan, Merton Miller, Minoru Sawai, Ulrike Schaede, Jose Scheinkman,and Jeremy Stein

We also benefited tremendously from detailed feedback on the script from Tetsuji Okazaki, Juro Teranishi, Hugh Patrick, RaghuramRajan, and Luigi Zingales Our students at Chicago, Osaka, and UCSDwere exposed to drafts of this book and provided valuable comments.Our first papers on Japanese corporate finance were co-authored withDavid Scharfstein This collaboration shaped our thinking on the postwarsystem, and we thank David for all he taught us

manu-Accounting data for Japanese firms that we used for our research withDavid Scharfstein were provided by Toshiro– Fujiwara, Yoshiki Ikeda,and Sumio Saruyama, to whom we express our thanks

Satoshi Koibuchi and Ki Young Park provided research assistance,and Cynthia Bailey handled administrative aspects, in the final stage ofputting together the book.

Larry Meissner deserves an immense amount of credit for his tance in producing the final manuscript Larry tirelessly read multipledrafts of the book His critical analysis pushed us to sharpen our argu-ments, and this book is much better because of his help

assis-Terry Vaughn found our proposal for the manuscript interesting,encouraged us, and waited patiently for us to deliver After Terry leftthe MIT Press, Elizabeth Murry directed the final stages of the publishingprocess

We have been working together for 15 years, and have spent parts of 7

of them on this project Without the tremendous support of our families,none of this would have been possible We thank Katie, Tamaki, Banri,Julie, Laurie, and Momo for their understanding and patience The man-uscript was completed on Halloween, and it was with great pleasure wewent trick-or-treating with our children We hope this book will be, if not

a “treat,” then at least intellectually filling

Figure 2.1 Bank Loans by Type of Collateral

Figure 7.1 Annual Growth Rate of GDP, 1956–98

Figure 7.2 Changing Patterns of Financial Surpluses

Figure 7.3 Land and Stock Prices, 1970–99

Figure 8.1 The Japan Premium

Figure 8.2 Loans to Small Enterprises as a Percentage of All Loans,

1970–97

Figure 8.3 Loans to the Real Estate Industry as a Percentage of All

Loans, 1970–97

Table 2.1 Allocation of Private Sector Financial Assets: 1900–31Table 2.2 Funding Pattern, Flow Estimates: 1914–35

Table 2.3 Funding Patterns, Level of Claims Estimates: 1902–40Table 2.4 Size of the Tokyo Stock Exchange: 1920–40

Table 2.5 Relative Stock Market Capitalization at Year-end 1998

Compared to Prewar Japan

Table 2.6 Bank Underwriting of Corporate Bond Issues, 1924–35 Table 2.7 Corporate Bond and Financial Debenture Issues, 1902–31Table 2.8 Profitability and Composition of Assets of Ordinary Banks,

1913–36

Table 3.1 Wartime Financial Controls Related to Finance

Table 3.2 Relative Size of the Five Largest Banks

Table 3.3 Repudiated War-Related Government Debts

Table 3.4 Munitions Company Assignments to Financial Institutions

and Postwar Main Bank Relationships

Table 3.5 Allocation of Private Sector Financial Assets: 1931–57Table 3.6 Sources of External Funds for Industries: 1931–57

Table 4.1 Interest Rates, 30 April 1973

Table 4.2 Allocation of Household Financial Assets: 1955–75

Table 4.3 Allocation of Private Sector Financial Assets, 1955–70Table 4.4 Sources of External Funds, Flow Data: 1954–75

Table 4.5 Sources of External Funds, Level of Claims Data: 1954–75Table 4.6 Japanese Bond Market Issuers in 1964

Table 4.7 Financial Firms Belongs Presidents’ Councils at the End of

the High Growth Era

Table 4.8 Bank Profitability and Assets Compositions: 1955–75Table 4.9 Patterns for Japanese Manufacturing Firms, 1975

Table 5.1 Export Shares of Major Japanese Auto Companies in 1984Table 5.2 Corporate Restructuring: A Comparison of 1977 and 1992

Table 6.1 Delisting and Bankruptcy of Firms Listed on the New York

and Tokyo Stock Exchanges

Table 7.1 Expenditures and Deficits in the Government General

Account as a Percentage of GDP: 1965–83

Table 7.2 Number of Companies Qualified to Issue Unsecured BondsTable 7.3 Security Financing by Listed Firms, 1972–98

Table 7.4 Bond Financing by Listed Firms, 1972–98

Table 7.5 Stock Financing by Listed Firms, 1972–98

Table 7.6 Allocation of Household Financial Assets: 1964–96

Table 7.7 Sources of External Funds, Flow Data: 1971–95

Table 7.8 Funding Patterns, Level of Claims Data: 1975–95

Table 7.9 Bank Debt as a Percentage of Total Assets for Publicly

Traded Japanese Firms, By Sector, 1970–97

Table 7.10 Structure of Bank Assets and Deposits, 1975

Table 7.11 Structure of Bank Assets and Deposits, 1996

Table 7.12 Distribution of Loans by Borrower and Duration, 1975 Table 7.13 Distribution of Loans by Borrower and Duration, 1996 Table 7.14 Bank Subsidiary Underwriting of Corporate Straight Bonds:

94Q1–97Q4

Table 7.15 Changing Patterns for Japanese Manufacturing Firms, 1975

and 1997

Table 7.16 Firms with Zero Bank Borrowing, by Sector

Table 7.17 Cross-Shareholding by Presidents’ Council Members

Table 8.1 Level of Bad Loans at All Banks by the Various DefinitionsTable 8.2 Bad Loans as a Percentage of GDP

Table 8.3 Profitability and Non-Interest Income of Major US Banks,

1976–96

Table 8.4 Interest Income, Fee Income, Return on Assets, and Return

on Equity for Japanese City Banks

Table 8.5 Key Elements of the Big Bang

Table 8.6 Scope of Businesses Open to Banks After the Big Bang

Table 9.1 Comparative Allocation of Household Financial Assets,

Box 2.1 The Financial System Prior to the Military Build-up in the

1930s

Box 2.2 The Japanese Abroad and Foreigners in Japan

Box 2.3 The Samurai as Investors

Box 3.1 The Financial System During the War Years and OccupationBox 3.2 The Japanese Government in 1931–45

Box 3.3 Rising Militarism

Box 3.4 Governing Japan Under the Occupation

Box 3.5 The Scale of the Major Zaibatsu at War’s end

Box 3.6 Distribution of Stockholdings

Box 3.7 Major Changes in Financial Structure During the Occupation

Box 4.1 The Financial System During the Rapid Economic Growth

Period

Box 4.2 Corporate Finance and Main Banks

Box 4.3 World Bank Borrowing By Japan

Box 4.4 Securities Investment Trusts (Sho – ken to – shi shintaku)

Box 4.5 Trust Banks (shintaku ginko –)

Box 4.6 Long-term Credit Banks (cho – ki shin’yo – ginko –)

Box 4.7 The Convoy System at Work in the High-Growth PeriodBox 4.8 The Convoy System at Work in the 1980s: Heiwa So–goBox 4.9 General Trading Companies (so – go – sho – sha)

Box 4.10 Cross-Shareholding

Box 6.1 The Modigliani-Miller Proposition

Box 6.2 Benefits of Keiretsu Finance

Box 6.3 Keiretsu and R&D Investment

Box 6.4 Costs of Keiretsu Financing

Box 6.5 Market Failure

Box 7.1 The Financial System After Deregulation (Early 1990s)Box 7.2 Blowing and Popping Bubbles

Box 7.3 Examples of New Products

Box 8.1 The Ju – sen

Box 8.2 The First Wave of Banking Troubles

Box 8.3 LTCB and SBC

Box 9.1 The Financial System in the 21st Century

Box 9.2 The Financial System’s Changing Patterns

As the 1980s ended there was talk of Tokyo becoming the financial tal of the world, and it was fashionable to ask whether other countriesshould be adopting Japan’s industrial policy and financial system Bythe late 1990s many analysts were arguing that Japan’s problems couldnot be solved using the policies that had become commonplace duringthe preceding four decades of mostly boom To some observers, thismeans Japan should drop everything that characterized its “traditional”economic system All the “Japanese” aspects of the system were to bepurged, so that the economy could become more open, free, and inter-national In short, it is common to assert that Japan’s economy mustundergo a “globalization,” which often is meant as the equivalent of

1.1 Focus

This book focuses on the Japanese financial system, because it was atthe center of the difficulties in the 1990s and is poised for major change inthe 21st century Special attention is given to the close nature of thebank-firm relations that characterized Japanese finance in the period ofrapid economic growth (roughly 1955–73) We call Japanese corporatefinance based on such a close bank-firm relationship “bank-centeredfinancing.” It has been most clearly observed in the major corporategroups in Japan, whose nature and antecedents are discussed inAppendix 1.1 We examine how bank-centered financing developed

over time, what roles it played in the rapid economic growth, when itstarted to change, and how it will be transformed in the future.

Our analysis of the evolution of the financial system is framed interms of the answers to four questions about each of the four regimes thesystem has evolved through since the late 19th century and the regimethat is emerging in the 21st century The questions are: How do house-holds hold their savings? How is business financing provided? Whatrange of services is being provided by banks? What is the nature andextent of bank involvement in corporate governance? This provides aconsistent analytic framework for what has happened since the (1868)Meiji Restoration These are also the questions that are the core of thedebate about reforms for the 21st century

A guiding principle in our analysis is that whenever possible we willbring data to bear to support our argument While most of the conclu-sions in this book are the outgrowth of rigorous academic research that

we and others have published previously, we believe that the readershould not have to rely on “appeals to authority.” Too often discussions

of Japan are clouded by unsubstantiated claims Our view is that if wecannot present some simple numbers to support our analysis, we havenot done our job

Understanding past and coming changes depends on an appreciation

of the rationale for the antecedents of the incumbent system Hence thestress on historical perspective and historical background World War IImarked a major discontinuity; for practical reasons and (at the time,widely held) theoretical beliefs about the process of economic develop-ment, the Occupation and Japanese authorities continued a number ofwartime practices (For example, mobilization for development was seen

as not all that different from mobilizing for war.) For a long time the tem worked, and the carried-over ideas received much of the credit Thissuccess colored the nature and pace of changes

sys-1.2 Lessons

What are the lessons from our look at the 130-plus years of evolution ofthe Japanese financial system? We see two recurring themes

First, the behavior of financial institutions, savers, and borrowers can

by and large be explained using standard economic analysis One doesnot need a new set of tools to describe and appreciate the superficiallyunique financial system that evolved in Japan Rather, one has just torecognize the shocks and the prevailing regulatory environment that put

the Japanese financial system onto a specific evolutionary path Fromthere, the patterns become familiar and easy to understand.

Second, the major shifts in the system involved regulatory changes,many of which were reactions to important economy-wide disturbances.These were, in turn, often related to large external shocks Thus, the dis-ruption of the China and Pacific wars and subsequent Occupationdirectly affected the nature of Japanese finance The 1970s oil shock trig-gered the deregulation that gradually transformed the system over thenext three decades, although the maturing of the Japanese economymeant that this transformation was bound to happen even without theoil shock Finally, the Big Bang of the mid-1990s was formulated as thefinal step in the deregulation, moving at least the nominal structure ofJapan’s financial system to the forefront of the global system that hasevolved centered on New York and London

1.3 Overview

Between the beginning of Japan’s modernization in the late 19th centuryand the end of the 20th century, Japan’s financial system has gonethrough four distinct periods The following provides a brief explanation

of how the four eras flow into each other The largely chronologicalstructure of the book means it can be usefully read as a history ofJapanese finance, but our larger intent is an analysis of the system’s evo-lution and response to changing circumstances Having understood thepast four regimes, the fifth one, into which Japan is now moving,becomes much easier to understand.1

The first financial regime starts in the 19th century and continues untilthe beginning of Japanese hostilities with China in the late 1930s InChapter 2 we show that in comparison with the postwar period, this erawas characterized by the relatively low importance of banks in thefinancing of corporations For instance, bank financing of large firmswas often less important than bond financing Even among the largeindustrial alliances known as zaibatsu, bank financing was not a veryimportant funding source during this period

In contrast, securities markets were quite active New shares wereroutinely issued by the leading corporations and shares traded actively

on stock exchanges and over-the-counter The trading was done by a

1 Another author with a similar periodization is Patrick (1972) He divides the first regime into a formative period (to the time of the Russo-Japanese War) and the period during which the prewar system matured.

diverse group; the banks as a rule did not own much equity and thenotion of “shares held in friendly hands” was rarely mentioned Bondmarkets were also deep and vibrant It was not unusual to see yearswhere more net corporate funding was done in bond markets thanthrough bank borrowing.

Understanding this era is important because it dispels several mythsabout the Japanese financial system Most importantly, prewar historydemonstrates that Japanese are not inherently averse to relying on capitalmarkets for financing Quite the contrary Thus, in Chapter 3 we explainhow regulations passed as part of 1930s militarization worked to over-turn the prewar financing system The central feature of these regula-tions is that they progressively transferred more and more power to thebanks in general and to zaibatsu banks in particular

The late 1930s and early 1940s was the time when bank financingbecame the dominant funding source for most of the industrial firmsinvolved in the war effort During this period the depth of the tiesbetween specific firms and banks increased noticeably The shift wascompleted with the designation of specific banks as being responsible forthe financing of specific militarily important firms This meant that thelending coalitions that had previously characterized banking relationswere replaced by one-on-one lending In many ways this shift markedthe start of the tight ties between firms and banks that are the hallmark ofbank-centered financing

That the exigencies of war upset peacetime practices is not itselfremarkable But, during the late 1940s and early 1950s, a number ofevents reinforced the patterns established during the war and turnedthem into the practices of postwar bank-centered financing Probablymost significant among these actions was the method used to settle theinsolvency that plagued the immediate postwar economy Japanese firmsended the war with a large amount of debt owed to banks and to eachother Most of this was incurred to facilitate production of items essential

to the war effort, and both the trade credit and bank lending had beenbacked by government guarantees After the war the Occupation insistedthat the Japanese government default on its guarantees

The banks played an active role in deciding how the firms would digout from their debts Their role in the reorganization planning helped tosolidify their status as the dominant financial players in the economy.Furthermore, we show how this process transformed the wartime lend-ing relationships into funding arrangements that worked in a growing,

industrial economy This transformation left Japan with a completelydifferent financial system in 1955 than it had had in 1935.

As clean up from the war era ended, the economy was still in a weakcondition Domestic funds were insufficient to cover investment needs.The government had decided that foreign investment would be con-trolled, so capital was rationed An important way that the governmentinfluenced the direction of credit was to force savings to flow through thebanking system The combination of a banking system that was success-fully attracting money from the public and a starved and rebuildingcorporate sector put the banks at the center of the financial system asgrowth took off

As Chapter 4 shows, during the remainder of the 1950s, throughoutthe 1960s, and into the early 1970s a key part of the financial system waswhat became known as the main bank system Just what is meant by a

“main bank” evolved through time (and varies with users of the term),but at its most basic it means that a firm looked to a single bank to takethe lead in organizing its financing and providing the majority of itsother banking needs.2

A main bank usually limited its direct lending exposure to a client byacting as a leader within a group of institutions that provided funding.Nevertheless, it was widely understood that the main bank would pro-vide an anchor Thus, if firms ran into financial difficulty the main bankwas expected to step up and organize a workout Chapter 5 presentscase studies of workouts

The banks continued some of the patterns that had become commonduring the postwar corporate reorganizations These included rotatingsenior bank officials through the managements of firms As a rule, out-siders were uncommon on boards of directors; when there were outsidedirectors, banks were the most likely supplier Similarly, during thehigh-growth era and into the 1990s it was common for the banks to haveequity links with their customers For instance, in Chapter 4, we showthat even as late as 1975, a bank was the largest shareholder of about one-sixth of the firms listed on the first section of the Tokyo Stock Exchange

A hallmark of the high-growth era was the tight regulation of the cial system Bond issuance was controlled, albeit through an “informal”

finan-2 So stated, it is obvious that the term could be applied equally well in other financial tures, especially for small and medium firms So, too, can “relationship” banking in its most general sense There are special aspects of the Japanese versions of these, but their

cul-“uniqueness” is not as extreme as sometimes painted, as the evidence in this book will show.

system, so that bond financing was not possible for many firms Newshare issuance also was impeded by similar formal and informal regula-tions Thus, up through the late 1970s, the banks were effectively theonly game in town for firms in need of external capital and going off-shore to raise money in foreign markets was prohibited On the supplyside, savers had few choices other than bank deposits, and the interestrates on deposits were kept low by regulations.

In Chapter 6, we argue that the proper way to view bank-centeredfinancing is as a system with clear costs and benefits We describe thesecosts and benefits and conclude that during the heyday of the system, thebenefits outweighed the costs

The oil shock marked the start of the fourth era for the Japanese cial system The shock was significant because it threw the governmentbudget sharply into deficit The government, in the aftermath of WorldWar II, had made it a policy not to rely on deficit financing and thus hadnot previously had to sell large quantities of debt However, in the wake

finan-of the oil shock it became clear that the government was going to need torun significant deficits At that point, the lack of a well-developed bondmarket became a problem

In Chapter 7 we explain how this opening up of the government bondmarket started the move to deregulation that transformed the system In

a series of changes, the restrictions regarding corporate debt issuancealso (slowly) were relaxed So, too, were regulations regarding foreignexchange Thus, the 1980 Reform of the Foreign Exchange Act freedinternational capital flows, allowing progressively more Japanese firms

to seek funds abroad More slowly, regulations regarding stock issuancewere loosened

Deregulation of the bond issue market had a significant impact onbank-centered financing Large and internationally known firms sub-stantially reduced their dependence on bank financing by issuing bonds.Banks started to lend to small and medium companies, which did nothave established relations with the big banks Thus the banks werereplacing customers they understood with unfamiliar ones In this sense,the system, based on long-term bank-firm relationships, started to crum-ble For many of the firms, it appears that the benefits of the system nolonger outweighed the costs

The Financial System Reform of 1993 allowed financial institutions

to branch into new niches by using subsidiaries For example, bankswere able to establish subsidiaries to enter the securities business, whilethe securities houses were able to establish subsidiaries to enter trust

banking Thus, the segmentation that had characterized the system alsobegan to disappear.

In Chapter 8 we look at the fallout from all of these changes Many ofthe banks’ new customers in the late 1980s were in real estate, in con-struction, or were non-bank financial institutions, which often in turnlent to real estate developers With rapidly increasing land prices andstock prices, those industries looked very promising ex ante As assetprices collapsed in the 1990s, many of the loans went bad Bad debt inthe banking sector became a major problem

As the crisis unfolded, deregulation continued We outline the package

of reforms, which came to be called the “Japanese Big Bang.” Whencompleted in the spring of 2001, Japanese financial markets were fullyliberalized, with open competition permitted among all types of financialservice firms

What will remain of the old system in the post Big-Bang Japaneseeconomy? Chapter 9 speculates on the shape of the emerging fifthregime The 21st-century system will not be bank-centered Large andinternationally known companies now can raise funds anywhere in theworld without relying on banks The benefits of the system may still beattractive to small and medium firms, but whether banks can developnew relationships with these firms is an open question Further, thebanks’ bad debt problems during the 1990s impeded this kind of devel-opment At some point the banks will have recovered from their loanlosses, and then they will face a decision on the extent to which theywant to pursue this business Banks are sure to be tempted to enter non-banking financial businesses, such as bond underwriting and securitydealing Thus, one must consider whether the banks will even seek torevive the old system We think not

The diminished capacity of the banks to be the center reflects not somuch the banks’ failures (though they are an element) as the availability

of alternative sources of funding and the changing nature of who needsfunding Unshackled, security markets have reasserted themselves Thestock market collapse of the early 1990s has delayed the emergence of ashare-owning culture in Japan, but most of the pieces for its emergenceare being put in place

Reflecting on the lessons from studying the development of the fiveregimes, the overarching conclusion is that standard economic analysiscan easily explain most of what has happened One does not need a newtoolkit to understand the Japanese experience Rather once one accountsfor the regulatory environment and macroeconomic developments, most

of the transitions and arrangements of finance and governance, includingthose for the 21st century, are comprehensible and even predictable.

APPENDIX 1.1: Zaibatsu, Keiretsu, and Kigyo– Shu – dan

Business groups have played an important role in the Japanese economysince its modernization began This appendix explains their nature andthe terminology used in referring to them, both of which have changedover time Although similar in many ways to contemporary westerngroupings, Japanese groups in both the prewar and postwar period alsohave many unique characteristics

Zaibatsu

Prior to the end of World War II, the dominant groups were

conglomer-ates called zaibatsu (commonly translated as financial clique) The word

came into widespread use in the early 1930s and was used universallyinto the 1960s In prewar Japan, it was applied almost exclusively to

family-owned groups, which were thus a subset of kontserun (adapted

from the German term “conzern” (combines).1

Combines generally were organized around holding companies

(mochikabu-kaisha) but they also involved other relationships that gave the whole a unity of action The family held all the equity of the honsha

(literally “main company,” but more appropriately translated as ing company) The honsha then held shares in a set of industrial, finan-cial, and trading subsidiaries and exerted significant control over all ofthe operating businesses Further, the employees of the zaibatsu typicallyfelt a strong personal devotion to the House (family) and to some extent

hold-to the other employees within the combine.2

Mitsui puts the group’s origins in the 17th century, and Sumitomotraces its roots to 1590 when the family copper-crafting shop was estab-

1 This usage was reflected in Occupation policy when the zaibatsu were dissolved Hadley (1970, p 20 note 3) reports being told one firm was not included because, its founder having died in 1944, it was no longer a family-dominated combine.

2 The holding companies usually were partnerships until the 1930s Shares were sold to outsiders in the late 1930s The Occupation counted just 31 families as comprising the four leading zaibatsu (US Dept of State 1945, p 208).

The Mitsui had employed outsiders as senior managers since the Meiji period, and as the Meiji-era generation of the other three majors passed from the scene, they generally were replaced by non-family managers These mangers earned tremendous salaries and bonuses,

so zaibatsu wealth was not restricted just to family members.

lished Mitsubishi, was founded by Iwasaki in 1871 as a shipping pany All three became quite diverse and large Indeed, unlike prewarcombines in the West, which typically sought to dominate a specificindustry, these three zaibatsu were conglomerates that sought “oligopo-listic positions running the gamut of the modern sectors of the economy”(Hadley 1970, p 23) The last major group, the finance-based Yasuda,grew out of a late Tokugawa-era money lending firm Somewhat smallergroups, such as O–kura, Furukawa, Asano, Fujita, and Kawasaki, alsoformed by the end of the 1870s and were called zaibatsu.

com-There are two other sets of groups One formed a bit later and grew

very rapidly around World War I These were collectively called Taisho –

zaibatsu (after the emperor’s reign name) and included Suzuki, Kuhara,

Matsukata, Shibusawa, Iwai, Nomura, and Murai Some of them (mostnotably Suzuki) failed during the financial crisis in 1927 The last set, the

Shinko – (New) Zaibatsu, focused on chemicals and other heavy industries,

and expanded during the 1930s Nissan, Nicchitsu, Mori, Nisso–, andRiken are included in this group

Not all of the groups that the Japanese called zaibatsu were identified

as zaibatsu by the Occupation forces In fact, only 10 (the big four plusNissan, Asano, Furukawa, O–kura, Nakajima, and Nomura) were desig-nated as zaibatsu companies and dissolved The others avoided the dis-solution, although many large companies in those groups were broken

up under holding company dissolution.3

It is difficult to judge just how large and influential zaibatsu, even thebig three, were Indeed, despite significant effort, the Occupation’szaibatsu-busters could not reach consistent conclusions on the matter It

is generally recognized they exercised power over other companiesthrough financing and trading, so it is not particularly useful simply toadd up capital or revenue; doing so would also be problematic becausethe multiple-counting arising from inter-firm transactions would not benetted out, as would be the case if the group had a single consolidatedset of accounts Mitsui was by far the largest, representing upwards ofone-tenth of the “incorporated and partnership business of the nation”(Hadley 1970, p 32) Measured in terms of “capital controlled,”

3 Mitsui commonly was described as the oldest zaibatsu as recently as 1973 when Roberts published an English-language history of Mitsui Surprisingly this—along with a transla- tion of Mishima’s 1989 Mitsubishi zaibatsu-shi—were the only book-length English histo- ries of the major zaibatsu that we could locate There are some pamphlets produced by the groups: Sumitomo Corp (1979) and Mitsui (1968).

Mitsubishi was considered about one-third as large as Mitsui, andSumitomo one-ninth (Allen 1940, pp 628–29).

Moreover, as explained by Morikawa (1992), in his review of all theleading groups, the zaibatsu during the 1920s and early 1930s were by nomeans too big to fail For example, by the mid-1920s Suzuki had built, inless time, a larger trading company than Mitsui and was active in steeland textiles In the 1927 panic, the firm collapsed While this is the bestknown case, many of these family combines fell by the way-side longbefore the Allies arrived in Japan.4

Breaking up the zaibatsu and other major holding companies was agoal of the Occupation (discussed in Chapter 4) This reflected the factthat they had contributed significantly to Japan’s war effort and colonialactivities, especially in Manchuria In 1948 a law was passed prohibitingthe use of the names, trademarks, and logos of zaibatsu Occupationauthorities also sought to have the senior managers of the zaibatsu dis-missed The Japanese government resisted this effort, but ultimatelyabout 1500 executives were removed from their jobs Holding companieswere banned Finally, the shares of the main firms were sold off Thistransfer was remarkable in that it involved over one-third of the paid-incapital of all corporations in Japan in 1945 Together these measuresdestroyed the zaibatsu The families were not left destitute, but in thepostwar period only the head of Sumitomo was among the country’stop taxpayers.5

Keiretsu and Former Zaibatsu Companies in the Postwar Era

Once the Occupation ended, the companies of the four major zaibatsubegan to coalesce into groups that in English are now commonly identi-

fied by the Japanese word keiretsu (Into the 1970s the regrouped firms of

4 For further information in English, Morikawa (1992) is the most comprehensive source Hadley (1970) focuses on their dissolution, but in Chapters 2 and 3 she quantitatively doc- uments their sheer size and diversity during the war years Ownership (Chapter 4) and per- sonnel ties (Chapter 5) are also treated extensively A contemporary account of their emergence is Allen (1940), who takes a somewhat more benign view of them than does Hadley Also see Allen (1962, pp 132–35) who provides a succinct summary of their pre-

1932 development and Lockwood (1954, especially pp 214–35) However, aside from Morikawa all of these limit themselves to the firms that had become successful by the 1930s, or that became successful during the 1930s Takeda (1995) and Yasuoka (1998) are among Japanese scholars taking a broader look at the emergence of zaibatsu.

5 The Japanese publish a list each year, and it has been considered something of an honor

to be on it Sumitomo’s “luck” is attributable in part to the fact forest land was not included

in postwar land reform.

the three prewar majors often were called zaibatsu both in English andJapanese, and sometimes still are for a pejorative implication.)

The term keiretsu actually covers two quite different types of groups.

One is vertical (a supply-chain with one dominant company), another ishorizontal (a group of peers) Horizontal groupings are more precisely

termed kigyo – shu – dan (enterprise groups) This has not been widely

adopted outside of Japan, but we think the distinction is important andthus use “enterprise group” (or simply “group”).6

There are fundamental differences between zaibatsu and enterprisegroups In particular, the firms in a group have significantly more inde-

pendence than under the honsha The reason for this is ownership and

control: there is no single family controlling each group and no holdingcompany with the power to direct the other firms Rather, each firm isindividually traded on public stock exchanges

The prohibition of inter-corporate stock ownership was withdrawn in

1949, and firms bought each others’ stock as part of the recapitalizationprocess They subsequently added to their holdings to preclude hostiletakeovers and increase equity The result has been significant cross-shareholding, but this is very different than a parent company holdingshares in a subsidiary

The law prohibiting the use of the names, logos, and trademarks of thezaibatsu was repealed in 1952, and many companies quickly restoredtheir old names In some cases new firms even adopted the zaibatsuname For instance, Japan Construction became Sumitomo Constructioneven though it had been established after the war Attempts were made

to repeal the ban on holding companies, but these failed Holding panies continued to be banned until 1998

com-In the early 1950s, Sumitomo’s Hakusui-kai (White Water Club, 1951)and Mitsubishi’s Kin’yo-kai (Friday Club, 1954) formed and began meet-

ing once a month These gatherings (shacho – -kai or Presidents’ Councils)

brought together the heads of key former zaibatsu firms, along withother large associated firms, for informal discussions about matters ofmutual interest Mitsui’s Ni-moku-kai (Second Thursday Club) started in

1961, Fuji Bank’s Fuyo–-kai in 1966

6 A useful summary of contrasts among various types of groupings is Gerlach (1992) and Clark (1979, pp 73–95) Matsushita, Hitachi, Nissan, and Toyota are examples of large ver- tical groups For more details on vertical groupings, see Aoki (1988, particularly pp 208–23), Asanuma (1989), and Asanuma and Kikutani (1992).

Before being adopted to describe business groups, keiretsu meant order or succession (in

a lineage).

When the groups were forming in the early 1950s, the trading nies of Mitsui and Mitsubishi were being reassembled (Sumitomo andYasuda did not have major trading companies in the prewar period.)The traders were considered by many to be the heirs to the status of the

compa-honsha because of their intimate interaction with all the other group

members However it is not clear how much power they actually exerted

on group members (See Box 4.9 for further discussion of the tradingcompanies.)

What was clear is how dependent for funds the groups were on theirbanks The banks had been largely untouched by reforms during theOccupation and had helped reorganize many of the firms in the after-math of the war (see Chapter 3) Moreover, presidents’ councils hadformed among firms that had no prewar affiliation but were clearlyassociated with a major city bank These included Sanwa’s San-sui-kai(Third Wednesday Club), which started in 1967 Similarly, Dai-IchiKangyo Bank’s Presidents’ Council, Sankin-kai, was established in 1978

by merging the Presidents’ Councils for the former Furukawa zaibatsu, Kawasaki zaibatsu, (these two formed the core of the Dai-Ichi Bank

group) and the Presidents’ Council for the Kangyo Bank group

Thus it has become common to speak of the bank as the leader withineach group But, even though the banks often are the largest shareholder,they are a long way from exercising much direct control over the firms’operations In any case, by the mid-1960s there were six sets of manufac-turing firms affiliated with each other and a set of financial institutions in

what are called enterprise groups (kigyo – shu – dan) or horizontal keiretsu.

The three largest are related to the largest prewar zaibatsu Banks arespecifically identified as the core of the other three The largest of these,around Fuji Bank, is the successor to the Yasuda zaibatsu The other twoare postwar creations, although the banks themselves are prewar Thesesix groupings and their relations with financial institutions are the focus

of much of our analysis of the postwar financial system

Are Keiretsu Anti-Competitive?

The issue of whether keiretsu ties distort the competitive landscape isworth examining briefly as part of establishing a general understanding

of the groups There are studies for a number of different industries thatexamine this question Assuming the alliances generate extra profits, onecan argue that these profits can be used to allow the members to under-take other activities, even those that are potentially unprofitable For

example, using a cushion of profits from home markets, member firmscould take losses in foreign markets in order to gain market share Otherstudies look at whether the presence of keiretsu firms leads to collusionthat excludes foreign imports and raises prices (Lawrence (1991) is thebest-known example.) Another variation on the argument holds thatcross-shareholdings insulate the keiretsu from hostile takeovers, therebyimpeding foreign direct investment (FDI).

For several important reasons we find almost all of this evidenceeither unconvincing or ambiguous A major problem is the premise ofrents: in fact, keiretsu firms are not particularly profitable Weinstein andYafeh (1995) show that industries with a high keiretsu presence tend tohave lower price-costs margins and to be more competitive This canexplain why FDI might be low, and undercuts the arguments overprices (Weinstein (1997) points out a number of measurement issuesthat cloud the debate regarding FDI.)

Using firm-level data and controlling for many firm characteristics,Ueda and Sasaki (1998) show that the import behavior of keiretsu firmsdoes not look different than that of other firms

A more general concern is that the mechanism by which the keiretsucoordination is supposed to operate is often unclear As Drysdale (1995)explains, it is very hard to see exactly how firms across the differentgroups would cooperate to exclude firms or to keep prices up

Of the 1,313 manufacturing firms listed on the first section of the TokyoStock Exchange in March 1997, only 109 belonged to one of the councils.These firms had average sales of about eight times other listed manufac-turing firms

Unfortunately, even this narrow definition has some problemsbecause it is generally agreed that at least some current council members

no longer have special ties to the group Also, a few (for example,Hitachi, Nippon Express, Kobe Steel, and Nissho Iwai) have belonged to

more than one council Thus, not all council members have the kind offinancial connections that characterize postwar bank financing in Japan.Fortunately, the presidents’ council definition is sufficient to show themain characteristics we seek to highlight Moreover, the basic charac-teristics on which we focus are also apparent when other common defin-itions, such as the largest lender being one of the six major group banks,are used.

Instead, it appears that banks focused more on serving the market needs of firms and their shareholders Thus, the banks worked toestablish a dominant position in the underwriting of corporate bonds.They did not underwrite shares, but they did make substantial loanscollateralized by corporate stock The shares generally were new issues

securities-or shares previously issued that were being paid fsecurities-or in installments

1 Adams (1964) is the best book-length study in English on the prewar financial system as

a whole However, the book is mostly descriptive and lacks an overarching theme; it is also difficult to find Among mainstream sources Goldsmith (1983) is the best known His study of Japan’s financial development through 1977 is half-devoted to the prewar period and offers extensive data As he suggests in his preface (p xiv), further research has refined and changed some of his data and conclusions Ours is one such study, but there is still much one can draw from his analysis and data collection.

Minami (1994) is an excellent analytic overview of the pre-World War II economy Nakamura (1983) is the “classic” single volume on the economy in the period from the Meiji Restoration through 1941 Less technical economic history is provided by Crawcour (1989a) for the 19th century and (1989b) for 1885–1920, and by Nakamura (1988) for the 1920–45 period Bank of Japan (1966) and Ohkawa and Shinohara (1979) are basic statistical reference The Bank of Japan tables are in English as well as Japanese, and the Footnotes and Data Sources and the Explanatory Notes have each been published in English as supplements The Explanatory Notes provide concise descriptions of the evolution of the various finan- cial institutions and includes a chronology of important financial developments (pp 190–93 for the period covered in this chapter).

In short, the early Japanese financial system was more similar to the

US financial system of the postwar era than to Japan’s own postwar tem in the sense that financing through capital markets played a moreimportant role than bank lending

sys-This history, supported by data presented later in the chapter, shoulddispel the perception that market-driven financing is inherently impossi-ble in Japan Indeed, the evidence shows that, given a suitable regulatoryenvironment, Japanese corporations would be willing to rely on stockand bond market financing and that there is no inherent Japanese prefer-ence for a bank-dominated financial system

The next two sections look at early institution building and trace the keydevelopments in Japanese finance into the 1910s and then through the mid-1930s With this historical context, the analytic structure presented in Chapter

1 is applied in sections looking at the distribution of household savings,quantifying the financing of businesses, assessing the ranges of servicesbeing offered by banks, and examining the corporate governance system

2.1 Key Developments to World War I

In 1868 the 260-year Tokugawa era ended and the Meiji period began—together with Japan’s “modernization.” In the words of a leading histo-rian of the period, with the Meiji Restoration the Japanese “executed a

Box 2.1

The Financial System Prior to the Military Build-up in the 1930s

1 Allocation of household financial assets

• Shift from currency to deposits, until mid-1910s

• Securities (stocks and bonds) half or more of gross holdings

• Securities widely used as collateral, especially after 1915

2 Provision of funds

• Equity financing dominant

• Bond and bank financing about equally important

3 Range of services offered by banks

• Underwriting of corporate securities dominated by banks

• Lending to individuals more common than lending to corporations

4 Corporate governance

• Shareholders lead in corporate governance

• Corporate boards of directors typically do not have bank representatives

• Workouts for distressed corporations typically not lead by banks

revolutionary change in political structure and the distribution of powerwithout carrying out a revolution” (Hall 1971, p 272), although it wasnot entirely bloodless.

At the time, the Tokugawa economy had a reasonably well-developedmonetary system (see Yamamoto 1989) The Tokugawa governmenthad monopolized the right to mint gold, silver, and other coins, and hadestablished fixed exchange rates between them Those coins were sup-plemented by paper money printed by local lords and bills issued

by money changers The Meiji government exchanged most of this papermoney with its own paper money, and issued more to finance thegovernment expenditures of the early Meiji period before the land taxsystem was established

Japan also already had a relatively sophisticated financial system inplace As a classic article notes, this meant “The Meiji economy inheritedconsiderable financial expertise, though with a rather traditional, com-mercial orientation, from the Tokugawa period” (Patrick 1967, p 245).Among the institutions in the legacy were well-organized rice exchanges

that included futures trading, and rotating credit associations (mujin).

There were houses (firms) that specialized in exchange and banking, aswell as those that combined exchange business with one or more of mer-chant operations, warehousing, financing of inventories, handling gov-ernment funds, and silversmithing Many of these exchange firms paidinterest to attract deposits.2

Despite this base, Japan was essentially an agrarian country with ahierarchical, hereditary ruling class Its population of around 37 millionmade it slightly smaller than the United States at the time

Japan’s new political leaders, drawn largely from the previous rulingclass, recognized a need to change the economic and social system toachieve development—especially of military strength—so that the countrywould not be colonized, even economically (as was happening inChina) They realized that the West’s powers related to its institutionsand regulations, and they were willing to draw on the West’s experience

in achieving the “rich nation, strong army” (fukoku kyo – hei) that was the

aspiration (and a slogan) of the time

Democracy was not a priority: the country was an absolute monarchyuntil a constitution was granted in 1889 and suffrage for the Diet (parlia-

2 Hanley (1997) and Hanley and Yamamura (1977) examine economic conditions and development in the Tokugawa period, which is also called the Edo period Goldsmith (1983, pp 2–8) summarizes the Tokugawa heritage Crawcour (1989a) covers most of the 19th century.

Patrick (1967) analyzes finance in the 1868–1914 period.

ment) remained quite limited until being extended to all males over 25 inMay 1925.

The economy lacked certain institutions—notably (as regards finance)the commercial banks, joint-stock companies, and stock exchanges—found in more advanced countries, so the Japanese set out to createthem This was part of the process to “develop industry and promote

enterprise” (shokusan ko – gyo –), as the 1870s slogan had it

To this end, the new government sent many missions and students tothe United States and Europe in search of the best practices for eachaspect of the economic, political, and social system It also hired foreignexperts (See Box 2.2.)

Japan had no interest in becoming “Western,” rather, it was seeking toremain Japanese—the slogan was “Japanese spirit, Western technology”

(wakon yo – sai) This was made easier by the diversity of Western models,

which allowed the Japanese to pick, blend, adapt, and even change course

2.1.1 Banks

The first significant development with respect to modern banking wasthe promulgation of the National Bank Act on 15 November 1872 Theform the Act should take had been seriously debated, with some arguing

in favor of an English model where only a central bank issued papermoney, which would be convertible into specie (gold or silver) Instead,the law was patterned after the US system, which at the time did notinclude a central bank but instead authorized national banks that couldissue bank notes The aim was to supply funds to facilitate commerceand to redeem the government paper money One aspect of the Act was

selection of ginko –to be the Japanese for “bank,” distinguishing themfrom the exchange firms set up in 1869 (and effectively out of business by1871) and other bank-like operations.3

On 20 July 1873 the Dai-Ichi Kokuritsu Ginko– (First National Bank)started business in Tokyo, with the Mitsui and Ono as principal share-holders These well-established rival financial houses had been serving asgovernment fiscal agents since the Tokugawa and were threatened withthe loss of the new government’s business if they did not participate

3 Many scholars, including us in past work, have translated the Japanese term Ho–as Law Following the forceful argument by Ramseyer and Nakazato (1999, p xix), we adopt the convention of translating Ho–as Act and jo–as section (rather than as article).

For a summary account of early banking see Adams (1964, ch 1) Cargill, Hutchison, and Ito (1997, ch 2) covers money and banking developments up to 1950 Tamaki (1995)

describes the century through 1959, and includes paragraph-length biographies of 30 bankers from the period (pp 218–22)

Teranishi and Patrick (1978) provides an analysis of the development of commercial ing up to World War II On the establishment of national banks, see Patrick (1967, pp 250, 255–60), Fuji Bank (1967, pp 15–19), Goldsmith (1983, pp 24–27), Tamaki (1995, pp 28–39).

bank-On the 1869–71 exchange firms (kawase kaisha), see Fuji Bank (1967, pp 15–16), Tamaki

(1995, pp 25–27).

Box 2.2

The Japanese Abroad and Foreigners in Japan

Official foreign missions began in 1860 (in the Japanese calendar, Man’en 1) when the Tokugawa government sent a delegation to the United States for the nominal purpose of signing a commercial treaty The Iwakura Mission of 1872–73, which included over 40 members, was the principal official trip during the Meiji period.

Many of the early Meiji leaders were on one or more of these, or wise went abroad Two examples Masayoshi Matsukata, the principal figure in Japanese finance in the last two decades of the 19th century, spent nine months in Paris in 1878, part of the delegation to the Paris Exposition (at which Japan exhibited) Hirobumi Ito, an advocate of the

other-US approach to national banks and early proponent of the gold standard, led a delegation to the United States in 1871 to study the monetary system there, and was in the Iwakura mission.

It is estimated that some 3,000 foreign government advisors were brought to Japan before 1890 “They were terminated as quickly as the Japanese felt they could manage by themselves” (Hall 1971, p 287).

Japan was seeking useful knowledge across the board—military and industrial technologies, the framework of government and financial insti- tutions, and commercial procedures Having been largely closed to foreign ideas for the two centuries before the 1850s, there was simply a lot to be learned, or at least looked at.

As to Western banks in Japan, a British bank opened in Yokohama in March 1863, and another followed in April By the end of 1867 six British and one French bank had opened offices, but three of the British ones closed in the wake of the 1866 financial crisis in London Although British banking practices were taken as models by the Japanese, foreign banks were peripheral participants in the emerging financial system.

Further reading

On Japanese overseas, see Beasley (1995) Pedlar (1990) discusses ers in Japan Burks (1985) covers both groups On the 1860 (Man’en) mis- sion, see Bush (1968) and Miyoshi (1994); on the Iwakura mission, see Soviak (1971).

foreign-Foreign banks in Japan are discussed by Tamaki (1995, pp 17–18).