Directed Credit? The Loan Market
in High-Growth Japan
YOSHIRO MIWA
University of Tokyo
Tokyo, Japan
[email protected]
J. MARK RAMSEYER
Harvard University
Cambridge, MA
[email protected]
Observers routinely claim that the Japanese government of the high-growth
1960s and 1970s rationed and ultimately directed credit. It barred domestic
competitors to banks, insulated the domestic capital market from international
competitive pressure, and capped loan interest rates. In the resulting credit
shortage, it promoted industrial policy by rationing credit.
As much as the government purported to ration and to direct credit, it
apparently accomplished nothing of the sort. It did not block domestic rivals to
banks successfully, did not insulate the market from international forces, and did
not set maximum interest rates that bound. Using evidence on loans to all 1,000odd firms listed on Section 1 of the Tokyo Stock Exchange from 1968 to 1982, we
find that observed interest rates reflected borrower risk and mortgageable assets
and that banks did not use low-interest deposits to circumvent any interest
caps. Instead, the loan market seems to have cleared at the nominal rates.
We follow our empirical inquiry with a case study of the industry to
which the government tried hardest to direct credit: ocean shipping. We find
no evidence of credit rationing. Despite the government programs to allocate
capital, nonconformist firms funded their projects readily outside authorized
avenues. Indeed, they funded them so readily that the nonconformists grew
with spectacular speed and earned their investors enormous returns.
The authors gratefully acknowledge the helpful comments and suggestions of John de
Figueiredo, Michihiro Kandori, Takao Kobayashi, Tatsuya Kubokawa, Naoto Kunitomo,
Akihiko Matsui, Toshihiro Matsumura, Tom Miles, Takashi Obinata, Yasuhiro Ohmori,
Eric Rasmusen, Roberta Romano, Mark West, and the participants at workshops at
Harvard University, University of Tokyo, and Vanderbilt University. Generous financial
assistance was given by the Center for the International Research on the Japanese Economy
and the Business Law Center at the University of Tokyo, the John M. Olin Program in Law,
Economics & Business at the Harvard Law School, and the Sloan Foundation.
c 2004 Blackwell Publishing, 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington Road,
Oxford OX4 2DQ, UK.
Journal of Economics & Management Strategy, Volume 13, Number 1, Spring 2004, 171–205
172
Journal of Economics & Management Strategy
1. Introduction
“There is no question,” writes Paul Krugman (1994, p. 142), “that before
the early 1970s the Japanese system was heavily directed from the
top, with the MITI [Ministry of International Trade & Industry] and
the Ministry of Finance [MOF] influencing the allocation of credit and
foreign exchange in an effort to push the economy where they liked.”
There is no question, indeed. By standard economic theory,
Krugman’s Japanese government (and in truth, it is not his Japanese
government, for he accurately reflects the secondary literature) would
have accomplished a spectacular feat. Try as they might, governments
throughout the world have shown themselves congenitally unable to
ration gasoline, electricity, medicines—even rice.
Yet by Krugman’s account, the Japanese government rationed
money with ease. By his account, it did not just adopt programs ostensibly designed to ration credit. It actually accomplished what it claimed.
Durable, invisible, fungible by definition, traded in the most fluid of
markets, available from anyone with assets of value, and arbritrable
on a moment’s notice—because money is all this and more, it should
be the hardest asset of all to ration. Notwithstanding, according to
Krugman and to most observers, the Japanese government suppressed
competitors to bank loans, insulated the economy from foreign funds,
kept interest rates low, and manipulated the resulting credit shortage to
route capital as it pleased.
In fact, these accounts mislead. At least according to all the evidence we could assemble, the Japanese government managed nothing
of the sort. Despite the massive secondary literature reciting the government claims as if they were true, the Japanese government did not
block firms from raising funds through stocks, through trade credit, or
(for many firms) even through bonds. It did not insulate the domestic
market from foreign investment. And it did not keep nominal interest
rates at submarket levels. All told, it did not create a capital shortage,
did not stop arbitrage, did not ration credit, and had virtually no say in
who invested how much in what. The Japan of the high-growth 1960s
and 1970s was not a world of directed credit. It was a world where
firms raised their funds through decentralized, competitive capital
markets.1
1. Much of the rest of our collective vision of the Japanese economy similarly is mythic.
In Miwa and Ramseyer (2000, 2002b, 2002d), we show that banks did not dominate prewar
Japanese capital markets. In Miwa and Ramseyer (2002a), we explain how “main banks”
in post-war Japan neither dominated nor monitored their borrowers beyond what rational
lenders would do in any advanced capitalist economy. In Miwa and Ramseyer (2002c),
we demonstrate that the post-war bank-centered keiretsu never played a substantive role
in the Japanese economy.
The Loan Market in High-Growth Japan
173
We begin by surveying the literature on which Krugman and others
rely (section 2). We then explain the contours of the actual—not mythic—
regulatory structure (section 3). Using data on all firms listed on Section
1 of the Tokyo Stock Exchange (TSE) (the largest firms), we estimate
the supply function for bank loans. We explore the determinants of
both interest rates and deposit balances (section 4). Finally, we use the
ocean shipping industry to illustrate how the regulations did not bind
(section 5).
2. Tales of Controlled Finance
Krugman (1994) accurately captures the secondary literature on Japan.
Granted, most economists have jettisoned the notion that omniscient
and omnipotent bureaucrats masterminded the high growth in 1960s
and 1970s Japan. Yet even they retain the idea that bureaucrats controlled the allocation of credit. By disabling the securities market, they
funneled the demand for capital to banks. By insulating the market from
foreign competition, they protected the domestic capital market from
international competitive forces. And by capping loan interest rates,
they created an environment where they could ration funds. Within the
rationed environment, bureaucrats determined—not just purported to
determine but did determine—which firms borrowed and which went
without.
2.1 The Structure of the Controls
2.1.1 The Tale. In his National Bureau of Economic Research
(NBER) study, Meerschwam (1991) puts the standard account straightforwardly. “[B]y allowing a complex form of rationed capital allocation
within the context of steering and guidance by authorities,” the government “provided the beneficiaries of the preferential funds [the ability] to
embark on growth strategies without having to rely, to a large extent, on
impersonal capital markets” (p. 206). Lincoln et al. (1998) echo the tale.
“Japanese corporations have been extraordinarily dependent on bank
debt as a vehicle for financing investment,” they write. As a result, the
“Bank of Japan and the Ministry of Finance were able to leverage this
dependence into a high degree of control over the financial sector. . . .”
(p. 324).
According to Cargill (2000), “the domestic financial sector, the
Ministry of Finance, the Bank of Japan, and politicians” all “assumed a
specific set of objectives” (p. 39). Crucially, those objectives included
“reindustrialization” and “export-led economic growth.” They then
implemented their “industrial policy” through “a rigidly regulated and
administratively controlled financial system.”
174
Journal of Economics & Management Strategy
2.1.2 Domestic Limits. To accomplish all this, explain most observers, the government manipulated a portfolio that included three key
policies: securities market restrictions, foreign exchange controls, and
interest-rate caps. First, it effectively banned most domestic alternatives
to bank loans. According to Weinstein and Yafeh (1998), “capital markets
in Japan were highly regulated and immature” (p. 636). “Securities markets were not well developed, and issuing conditions. . . were onerous,”
explains Calder (1993, p. 137). Weinstein and Yafeh (1998) conclude that
“[f]irms could raise only limited amounts of capital through commercial
flotation of debt or equity” (p. 636).
Corporations simply “did not have alternative sources of funding
until the mid 1970s,” writes Ito (1992, p. 119). “[T]he domestic securities
market was underdeveloped, and loans from abroad were not allowed.”
Consequently, “Japan’s financial system was one of the most regulated
and administratively controlled in the world” (Ito, 2000, pp. 95–96).
Banks “were the only game in town,” declare Hoshi and Kashyap (2001,
p. 310). 2
2.1.3 Foreign Exchange. Second, using its control over foreign
exchange, the government insulated the Japanese financial market
from international competition. Throughout the 1960s, explain Itoh and
Kiyono (1988), the Japanese government exercised “strict control” over
foreign direct investment (pp. 166–67).
The regulations did seem labyrinthian in the extreme (Smith, 1984).
As Cargill (2000) writes, “[F]oreign financial institutions were prohibited
or restricted to limited participation in the financial system” (p. 40).
According to Henderson (1986), the exchange controls helped “to shield
from international market forces Japan’s high savings managed by a
controlled banking system, fixed interest rates, and preferential credit
allocations (discriminating against small producers and consumers)
essential to the large export firms” (p. 132).
In sum, concludes Calder (1993), the exchange controls “gave
Japan the crucial freedom to determine its interest rates in isolation
from the rest of the world” (p. 35). The claim, in other words, is not just
that the controls reduced the flow of foreign funds. It is that the controls
cut the flow to levels so low that the government could direct domestic
investment without fear that arbitraged foreign funds would undo its
capital-allocation policies.
2. Similarly, Cargill and Yoshino (2000) write that “corporations had no alternatives
in the form of domestic money and capital markets or external markets” (p. 214). Pempel
(1978) claims that “the virtual absence of a private capital market” made the Bank of Japan
“the single tap through which virtually the entire Japanese monitary and credit supply
must flow” (p. 152).
The Loan Market in High-Growth Japan
175
2.1.4 Interest Rates. Last, the government suppressed interest
rates at artificially low levels. We return to these rates immediately, but
Patrick (1972) reflects the consensus when he asserts that the “interest
rate structure [was] extremely inflexible” and that the rates were “set
below that which would have resulted solely if market forces had been
relied upon to determine them” (p. 114). “It was,” he more recently
explained, “a situation of credit rationing” (Patrick, 2001, p. 3).
As Hamada and Horiuchi (1987) put it, “At least on the surface,
most interest rates in Japan have been rigidly regulated, and price
mechanisms do not appear to have been effective in financial markets”
(p. 236).
2.2 Compensating Deposits
Although most scholars writing about Japan apparently accept the
aforementioned account, a few argue that banks and firms partially
circumvented the loan interest ceiling. Some banks sometimes, they
explain, required borrowers to take more than they needed and to
deposit the “compensating balance” in a low-interest-bearing account
at the bank. Through the ploy, they raised the effective interest rate on
the loan.
If so, the credit allocation debate turns on the effect of the balances.
Did the banks indeed use them to raise artificially low rates closer to
market levels? Or did they merely use the balances—like banks in other
countries—to lower risk by monitoring cash flow patterns at the firm?
2.2.1 The Debate. On the one hand, Hamada and Horiuchi (1987)
claim that such balances “raised the effective interest rates on bank loans
to a level much higher than the regulated nominal rates” (pp. 236–
37). In his recent text, Flath (2000) similarly concludes that although
“the interest rates on bank loans [were] nominally suppressed,” they
“were effectively raised toward market-clearing levels by the device of
compensating balances” (p. 274).
Yet most scholars assert that at least residual capital rationing
remained. In one of the first studies in English on point, Patrick (1972)
reasons that “[c]ompensatory deposit ratios probably do not increase
effective interest costs sufficiently to restrict demand to the level of
supply. Since the price mechanism does not clear most financial markets,
the system relies extensively on credit rationing. . . .” (p. 116).3 More
3. In their early work, Ackley and Ishi (1976) thought the point “obvious”: “Obviously,
the use of rationing instead of price during periods of restriction—which has been in
effect during about 40% of the time over the past 20 years—means that the average rate
of interest could be kept considerably lower than it would otherwise have been” (p. 205).
176
Journal of Economics & Management Strategy
recently, Ueda (1994:94) reaches similar results: Although “the practice
of requiring compensating deposits on loans made the effective interest
rate more variable than the official rate, . . . the effective interest rate
[remained] below market clearing levels in the business loan market.”4
2.2.2 Rationing. Whether the government could allocate credit
depends crucially on whether the credit market cleared. If interest
rates (either formally or effectively through compensating balances)
remained at market levels, anything the government did to direct funds
simply would have produced offsetting shifts elsewhere. Because the
marginal cost of funds generally would have stayed at market levels,
even loan subsidies seldom would have affected investment patterns.
Many observers who argue that the interest rates stayed below
market-clearing levels also claim that only firms favored by the government could raise funds. According to Pempel (1978), for example, “credit
for firms [was available] only in accordance with broader policies of the
Bank of Japan and the government” (p. 153). Likewise, asserts Cargill
(2000), “[g]overnment credit allocation policies. . . played a major role
in allocating funds through intermediation markets” (p. 40).5
In this world, the large firms were the favored firms. As Meerschwam (1991) puts it, “The authorities [used their] influence to steer
the system through a rationed capital market that favored established
corporations” (p. 199). By Milgrom and Roberts’ (1994) account, the
“government restrictions and regulations on financial institutions and
financial markets [gave] large, established Japanese firms a lower cost
of capital. . . .” (p. 25).6
Proceeding along the same line, Wallich and Wallich (1976) claimed that “[d]iscrimination
among lenders and among borrowers, rationing, and subsidies have been the rule”
(p. 251).
4. After surveying the literature, Cargill and Yoshino (2000) likewise determine that the
government’s “objectives were achieved by rigid regulation and administrative control;
market forces played only a small role” (pp. 209–10).
5. See also Meerschwam (1991): Firms who wished to borrow “had to force a consistency between their own goals and those of the authorities” (p. 205); and Ito (1992):
“During the 1950s and the 1960s, the Japanese financial markets were heavily regulated
and isolated from the world financial market. . . . The monetary authorities explicitly or
implicitly fixed most interest rates at low levels. Lending from banks was often rationed.
Consumers and small businesses often had difficulties obtaining mortgages and business
loans” (p. 114).
6. The alleged role that the so-called keiretsu corporate groups supposedly played is
more mixed. Some scholars argue that members could raise funds preferentially. Others
claim that the keiretsu preference came at a high price, as the powerful banks in the groups
used their control to extract rents. In fact, the entire debate is misplaced, for the keiretsu
groups played no role at all. Using variables indicating membership in the various groups
in our regressions, we find that they yield almost entirely insignificant results (Miwa and
Ramseyer, 2001b). On why keiretsu membership proxies for nothing of substance, see Miwa
and Ramseyer (2002c).
The Loan Market in High-Growth Japan
177
3. Regulation in the Loan Market
We do not quarrel with claims that the Japanese government purported
to direct credit and industrial investment. After all, it did adopt most
(though not all) of the restrictions scholars recite. It did claim that
these restrictions let it direct credit and investment. Too often, however,
observers of the Japanese economy do not merely recite its claims.
Instead, they assert that it accomplished what it claimed.
It is this assertion of programmatic effectiveness that we contest. By
basic economic logic, the potential for arbitrage should have doomed
any chance that the government could use these measures to direct
investment. Despite the restrictions on domestic capital markets, firms
raised large amounts of both equity and nonbond debt (section 3.2.);
if they raised those funds, the Japanese government could not have
leveraged any control over banks into controls over industrial investment. Despite the controls on foreign investment, by the 1970s foreigners
were investing massive amounts in Japanese firms (section 3.3.); if they
invested so heavily, Japanese capital markets would have remained
subject to international competitive forces. Despite the mandated caps
on loan interest rates, banks charged rates that varied by risk and did
not bunch at the regulatory maximum (sections 3.4. and 4); if they had
lent at market rates, the government could not have used the caps to
ration credit to favored projects.
We start by describing the actual scope of the programs in place.
We explain why, by standard economic theory, the programs would
not have affected investment patterns substantially. Finally, we survey
a wide variety of evidence and find it consistent with the absence of
binding constraints but fundamentally inconsistent with the notion that
the government directed investment.
3.1 The Domestic Capital Market
3.1.1 Equity Issues. Crucially, the Japanese government never
seriously regulated domestic equity issues. Indeed, despite the pervasive references in the English-language literature to highly regulated
Japanese securities markets (e.g., Weinstein and Yafeh, 1998, p. 636), the
government never even tried to restrict equity issues. Subject to routine
corporate (e.g., par value) and securities (e.g., registration) rules, it let
firms sell stock as they pleased.7 And sell stock the firms did. In 1964,
7. Granted, the stock they issued often came (by voluntary corporate practice) with
preemptive rights to future issues. Yet preemptive issues at submarket prices do not
increase the cost of an issue. Although they increase the number of shares necessary to
raise a given amount, they do not increase the cost of raising it. Instead, they simply let
the firm raise equity at a price all investors consider fair when investors disagree about
the market value of existing stock.
178
Journal of Economics & Management Strategy
TSE-listed firms raised 531 billion yen through 533 issues; in 1970, 681
billion yen through 537 issues; and in 1975, 1,001 billion yen through
285 issues.8
Although cross-country comparisons typically show much lower
leverage among American firms than among Japanese firms, the typical
comparisons mislead. According to the recent consensus, the lower US
ratios merely reflect “differences in accounting” [Myers, 2001, p. 83;
see also Rajan and Zingales (1995)]. Although US firms have a book
debt/asset ratio of 37% (1991 data) where Japanese firms have 53%,
adjusted for basic accounting differences the ratios fall to 33% (US) and
37% (Japan). Although US firms have a market debt/asset ratio of 28%
where Japanese firms have 29%, adjusted for accounting differences the
ratios fall to 23% (US) and 17% (Japan) (Myers, 2001).
This similarity between US and Japanese capital structures is longstanding. In effect, the modern consensus simply brings to Englishlanguage readers what Kuroda and Oritani (1979) showed Japanese
scholars two decades ago. Based on their study of mid-1970s firms,
Kuroda and Oritani (1979) estimated equity/asset ratios of 33.0% for
US firms and 47.4% for Japanese firms. They calculated intermediated
financing ratios of 50.4% for large US firms but only 46.7% for Japanese
firms.
3.1.2 Bond Issues. Neither did the government restrict the bond
market. To be sure, the major banks did try collectively to restrict the
firms that could issue bonds. Notwithstanding, many large companies
still used bonds to raise enormous amounts. In 1965, TSE-listed firms
raised 324 billion yen through 467 bond issues; and in 1970, 509 billion
yen through 306 issues. By 1975, they raised 1,406 billion yen through
306 straight bond issues; 408 billion through 57 convertible issues; and
372 billion through 52 foreign issues (Tokyo, 1985).
Although these bank-driven restrictions did reduce bond issues,
firms could—and did—circumvent them readily. They just borrowed
directly from the institutional investors who bought bonds elsewhere.
In the United States, investors like insurance companies traditionally
bought most of the bonds firms issued. In Japan, such firms lent funds
to industrial firms directly (Kuroda and Oritani, 1979).
8. Tokyo (1985, p. 110, tab. 37). The year 1965 was a low one—only 122 issues totaling
117 billion yen. In the 1950s and 1960s, the Bretton-Woods system pegged the yen at
360 yen per dollar. Obviously, one cannot gauge the relative importance of debt and
equity in corporate finance by comparing loan amounts with equity amounts. An equity
investment is a relatively permanent investment. By contrast, because loans are for limited
periods, loan amounts include a high fraction of renewals of loans made in previous years.
The Loan Market in High-Growth Japan
179
3.1.3 Implications. Although the banking industry restricted
bond issues, neither it nor the government restricted stock issues, and
neither it nor the government prevented nonbank financial intermediaries from lending directly. If a firm was indifferent between debt
and equity, it could issue stocks instead of bonds. If it did prefer debt
to equity, it could borrow from insurance companies directly. Given
the absence of regulatory hurdles to equity issues, the absence of
regulatory hurdles to nonsecuritized debt, and the pervasive possibility
of arbitrage, no government could have used the Japanese restrictions
on bonds to direct investment.
3.2 The Overseas Capital Market
Throughout the 1950s and 1960s, the government did regulate foreign
investment in Japan. It banned it in principle but subjected the ban to
various exceptions. The question is whether the residual ban prevented
foreign firms from investing in Japan and arbitraging away the effect of
any domestic policy measures.
From 1952 to 1960, foreigners invested relatively little. All told,
they invested only $1.01 billion—and only 16% of that as equity. Of the
debt, the World Bank loaned 43% (the largest amounts to electrical utility
firms), and the Washington-based EXIM Bank loaned 21% (Tsusho,
1990). For most foreign investors, firms in 1950s Japan just did not
present very attractive prospects.
At the beginning of the 1960s, foreigners began to see serious economic potential in Japan and invested amounts that quickly swamped
earlier levels. In 1961 alone, foreigners invested $581 million; in 1963,
$904 million; in 1965, $549 million; and in 1967, $880 million (Nihon
ginko, 1974). By the 1970s, they had hiked their investment levels
higher still and increasingly took equity positions: $3.5 billion in 1969
(71% as equity); $4.3 billion in 1971 (63% equity); and even $2.9 billion
in recessionary 1973 (70% equity) (Tsusho, 1990). By the 1970s, any
notion that the Japanese government had insulated the financial market
from international competition stood subject to an annual $3–5 billion
exception.
These amounts are not necessarily the amounts that would have
prevailed in an unregulated market. They are, however, amounts
that are incompatible with a capital market insulated from international market forces. Given the other evidence (presented here) on the
porous nature of the domestic capital controls, they also are incompatible with any claim that the government could have directed credit
effectively.
180
Journal of Economics & Management Strategy
3.3 The Interest Rate Ceiling
3.3.1 The Cap. During the high-growth period, Japanese banks
faced a cap on the interest they could charge their commercial borrowers:
subject to a variety of qualifications, a maximum of about 8.4% to 9.2%
on loans of more than 1 million yen for less than one year. In addition,
they faced a potentially more binding limit imposed by the banking
industry association (again on loans of more than 1 million yen for less
than one year). The cap applied to all banks and ranged from about 5.5%
to 8.4% during 1960–68 (Table I). The actual applicable cap depended on
Table I.
Distribution of Loan Amounts by Interest Rate
Charged: 1960, 1965, 1970a
Interest
Rate (%)
From
0.00
5.48
5.84
6.21
6.57
6.94
7.30
7.67
8.03
8.40
8.76
9.13
9.49
9.86
10.22
10.59
10.95
11.32
11.68
12.05
12.41
To
5.48
5.84
6.21
6.57
6.94
7.30
7.67
8.03
8.40
8.76
9.13
9.49
9.86
10.22
10.59
10.95
11.32
11.68
12.05
12.41
and over
Regional
Banks
1960
0.2
0.1
0.3
3.0
6.0
3.6
9.0
7.3
8.9
13.8
13.1
10.6
10.2
5.9
5.5
0.9
1.1
0.1
0.2
0.1
0.1
1965
0.5
0.4
2.1
6.9
5.4
3.8
5.5
8.1
13.7
17.2
14.8
11.0
7.0
2.2
1.3
0.1
0.1
0.0
0.0
0.0
0.0
Money Center
Banks
1960
0.6
0.4
0.5
3.0
2.8
1.4
30.7
12.4
12.4
12.8
9.8
6.3
4.7
1.6
0.5
0.1
0.0
0.0
0.0
0.0
0.0
1965
1.5
0.5
2.4
9.0
6.4
9.7
11.7
10.3
14.5
13.6
10.0
6.4
2.9
0.7
0.4
0.0
0.0
0.0
0.0
0.0
0.0
Interest
Rate (%)
From
To
Reg. Bk.
1970
M.C. Bk
1970
0.00
4.50
5.00
5.50
5.75
6.00
6.25
6.50
6.75
7.00
7.25
7.50
7.75
8.00
8.25
8.50
8.75
9.00
9.25
9.50
9.75
10.00
10.50
4.50
5.00
5.50
5.75
6.00
6.25
6.50
6.75
7.00
7.25
7.50
7.75
8.00
8.25
8.50
8.75
9.00
9.25
9.50
9.75
10.00
10.50
and over
0.2
0.1
0.1
0.2
1.0
4.8
0.8
6.1
4.6
5.8
7.7
7.9
8.3
8.8
14.3
10.3
7.0
6.1
2.5
1.6
0.3
1.5
0.0
0.6
1.3
0.4
0.2
2.9
0.8
2.5
14.1
10.0
8.7
9.2
7.7
6.3
9.6
11.0
6.1
4.2
2.1
0.9
0.4
0.3
0.7
0.0
Notes: The table gives the percentage of loans at a given interest rate range, as of March of a given year. Because the
source recatalogs interest rates for 1970, we align the columns by the 8% level.
a Zenkoku chiho ginko kyokai (1960, 1965, 1970).
181
The Loan Market in High-Growth Japan
Table II.
Distribution of Loan Amounts Relative to
Trade-Association-Imposed Interest Rate Cap,
1960–1986a
Regional Banks
Money Center Banks
Year
Max Ab , %
Max Bc , %
%<A
%<B
%<A
%<B
1960
1961
1962
1963
1964
1965
1966
1967
1968
7.30
6.57
7.30
6.21
6.57
6.21
5.48
5.48
6.21
8.40
7.67
8.40
7.67
8.03
8.03
7.30
7.30
8.03
13.20
4.30
13.10
2.90
12.00
3.00
0.40
0.50
3.50
38.40
28.70
38.20
25.90
42.20
32.70
25.70
31.70
52.00
8.70
5.00
6.80
2.50
20.00
4.40
1.70
1.80
7.30
64.20
51.90
58.80
45.60
67.20
51.50
46.30
56.30
71.00
Note: Table gives percent of amounts outstanding from regional and money-center banks in March 1960–68, at rates
below Max A and at rates below Max B.
a Zenkoku chiho ginko kyokai (1960–1968); Nakabayashi (1968).
b Maximum interest rate chargeable on lowest risk loans, by trade association, in %.
c Maximum interest rate chargeable on highest risk loans, by trade association, in %.
a variety of factors.9 It was a bizarre cap indeed. One can imagine banks
trying collectively to impose an interest-rate floor. One can imagine firms
trying collectively to obtain an interest-rate cap. Why banks would want
a cap is a tougher question.
Perhaps, however, it is also a needless question, for perhaps they
imposed no such thing. If the cap bound lenders, observed nominal
interest rates should have bunched at the mandated level. They did not.
Instead, as Tables I and II show, they varied broadly. In 1965, the moneycenter banks (known as the “city banks,” they lent disproportionately
to larger firms) charged modal rates of 8.0–8.4% within a range of 6.2–
9.5%. With their smaller clients, the regional banks charged modal rates
of 8.4–8.8%, given a range of 6.2–9.9% (Table I). In no case did lenders tie
lower-interest loans to whether a firm followed “government policy”—
whatever that might mean.
9. Primarily, the rate depended on the type of loan—whether it took the form of a
discountable note and so forth. For account overdraws, the rate was slightly higher, and
separate schedules applied to loans connected with international trade. The rates applied
to some nonbanks and varied by lending institution type. Slightly higher caps, for example,
applied to trust banks and insurance companies.
The association, known as the Zenkoku ginko kyokai, had a membership that covered all
principal banks, including the long-term credit and trust banks. After 1975, the association
removed itself from the rate-determination process. See Nihon Ginko (1977, pp. 168–90);
Nakabayashi (1968, p. 43); and Kitahara (1970, pp. 70–71).
182
Journal of Economics & Management Strategy
From 1960 to 1968, the money-center banks generally lent less than
10% of their loans at rates below the lower end of the trade-associationimposed cap (Table II).10 They lent one-third to one-half of their loans at
rates above the higher end. With their smaller customers, the regional
banks charged even more. From 1960–68, they lent fully one-half to
three-quarters of their loans at rates above the highest cap.
Observed interest rates exceeded the mandatory caps because the
caps applied only to large, short-term loans. Suppose a bank wanted
to lend at rates beyond the maximum. First, for a small client it could
simply cut the amount of the loan. It could either split it into several
smaller loans, or lend some of the money and tell the firm to go elsewhere
for the rest. Large firms did borrow only 10–20% of their debt from
their lead lender anyway (Miwa and Ramseyer, 2001b). Second, it could
extend the loan term beyond the one-year limit. The one-year loan
term was arbitrary, after all. Most banks regularly rolled over shortterm loans. Given this porous character to the regulations, money-center
banks made only 60–80% of their loans on terms that formally subjected
them to the caps. Regional banks made only 50–60% on such terms (see
Table III).
3.3.2 Compensating Balances. If banks could freely avoid
the cap by adjusting the size or term of a loan, they should not have
demanded deposits to adjust the effective interest charge. For the most
part, they did not. Because the larger firms were so safe, banks seldom
demanded deposits from them. With the small firms, they largely
stopped demanding the balances by the mid-1960s.11
In the mid-1960s, MOF surveyed the use of compulsory deposits
at various banks. Starting at the same time but continuing haphazardly
through the 1970s, the Fair Trade Commission (FTC) surveyed the use
of the balances among smaller firms. We reproduce the two surveys in
Table IV. According to Panel A (MOF), the banks used the practice only
with the smaller firms and abandoned it after the mid-1960s. According
to Panel B (FTC), banks similarly cut the use of the balances to the smaller
firms but never abandoned them entirely. By 1979, 20% of the FTC’s 3,600
respondent small firms maintained formally compulsory deposits, and
35% maintained informally compulsory deposits. The former averaged
1.4% of a firm’s outstanding debt, and the latter 8.2% (Kosei, 1979).
10. The exception on Table II is 1964. That year, the association hiked the maximum
rate on March 23, yet to construct the table we examine data from March 31 of each
year. Under the rates in effect prior to March 23, 1964, only 1.6% of the 1964 loans from
the money-center banks would have fallen below the maximum rate applicable to the
lowest-risk loans.
11. The appropriate measure of the compensating balances in these surveys is not the
kosoku sei yokin but rather the jishuku taisho yokin.
183
The Loan Market in High-Growth Japan
Table III.
Regulated and Unregulated Loansa
Tot Lnsb
Reg’dc
Unreg’dd
(LT)e
(SmA)f
4,053
4,875
5,660
7,113
8,513
9,804
10,979
12,398
13,857
15,995
19,118
22,472
27,737
34,703
39,504
43,481
70.0
70.1
71.6
71.7
75.2
76.1
75.6
77.3
77.4
76.7
75.4
73.7
69.0
64.2
63.9
63.4
30.0
29.9
28.4
28.3
24.8
23.9
24.4
22.7
22.6
23.3
24.6
26.3
31.0
35.8
36.1
36.6
10.6
11.4
11.9
13.4
13.4
13.2
14.4
13.3
13.9
15.6
17.9
20.4
26.2
32.1
33.3
34.1
17.5
16.4
14.5
12.7
11.4
10.7
10.0
9.4
8.7
7.7
6.7
5.9
4.8
3.7
2.8
2.5
B. Regional Banks
1960
1,897
1961
2,346
1962
2,816
1963
3,476
1964
4,269
1965
4,820
1966
5,561
1967
6,597
1968
7,734
1969
8,988
1970
10,015
1971
11,944
1972
14,526
1973
18,348
1974
21,482
1975
24,044
52.8
54.9
57.5
58.4
62.1
63.2
63.7
63.7
64.5
64.9
64.1
63.6
60.9
59.3
60.8
60.6
47.2
45.1
42.7
41.6
37.9
36.8
36.3
36.3
35.5
35.1
35.9
36.4
39.1
40.7
39.2
39.4
10.0
10.2
10.5
10.6
11.8
12.9
14.5
16.4
17.5
19.3
21.6
23.5
28.5
32.4
32.8
33.8
34.4
31.8
28.9
27.6
26.1
23.9
21.8
19.9
18.0
15.8
14.3
12.9
10.6
8.3
6.4
5.6
A. City Banks
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
Note: For 1960–63, some 1–4% of the loan amounts were unregulated because they were both small and long term.
a Zenkoku chiho ginko kyokai (1960–1975).
b Total loans, in billion yen.
c Gives the percentage of total loan amounts subject to the interest rate cap.
d Gives the percentage of total loan amounts not subject to the interest rate cap.
e Gives the percentage of such amounts not regulated because the stated term was one year or longer.
f Gives the percentage unregulated because the face amount of the loan was 1 million yen or less.
In order to reduce risk by letting the lender monitor cash flow,
firms still might have routed their transactions through their lending
bank. If so, then cap or no cap the higher risk firms voluntarily would
have kept deposits at their lead lenders. And if firms had kept deposits
184
Journal of Economics & Management Strategy
Table IV.
Compulsory Deposits/Loans (in %)a,b
A.c Ministry of Finance
5/64
11/64
5/65
11/65
City Banks
Large Firms
Small and Medium Firms
Regional Banks
Large Firms
Small and Medium Firms
Mutual Banks
Cooperatives
3.6
1.7
9.4
5.1
3.8
6.1
13.2
11.7
1.0
0.4
2.7
1.3
0.6
1.8
8.3
6.8
0
0
0
0
0
0
5.1
4.1
0
0
0
–
0
–
2.6
2.5
B.d Fair Trade Commission
(Small and Medium Firms only)
Formally
Compulsory
Informally
Compulsory
14.5
9.8
8.5
4.1
2.7
1.5
25.1
25.3
17.6
16.8
16.8
10.3
1966
1968
1971
1973
1976
1978
a Okura sho (1965, 1966); Kosei torihiki iinkai (1979).
b The fraction of compulsory deposits to total loans, in percent.
c The results of four Ministry of Finance surveys.
d The results of six FTC surveys of small- and medium-sized firms.
Table V.
Ratio of Deposits to Bank Loans for Large Firms∗
All Industries
Manufacturing Firms
1965
1970
1975
1980
1985
1990
1995
29.9
32.2
30.9
33.0
29.7
31.9
30.5
34.6
31.8
47.0
42.2
88.3
28.2
59.9
Source: Okura sho (1965, 1966).
∗ The table gives the average ratio of deposits to bank loans for firms capitalized at 10 billion yen or more.
at their lending bank voluntarily, then the level of bank deposits would
not have fallen with the 1980s deregulation. If the earlier deposits had
been compulsory, however, then presumably they would have fallen
during the 1980s. But they did not fall. Instead, if anything, average
deposits/bank-loan ratios (cash and deposits, divided by short- and
long-term bank borrowings, in %) for large firms (capitalized at 10 billion
yen or more) increased. See Table V for details.
All this is exactly what we predict. Deposit/bank loan ratios did
not fall during the deregulated 1980s, because banks and firms had
The Loan Market in High-Growth Japan
185
not been using compensating balances to avoid interest rate caps; they
had not been doing this because, contrary to virtually all academic
commentary, the caps never bound.
4. Bank Loans in the Japanese Capital Market
4.1 The Determinants of Interest Rates
By standard micro-economic theory, such a haphazardly binding institutional framework should not have generated either credit shortages or
rationing. To explore what factors did determine the allocation of credit,
we assemble a data set of over 1,000 large Japanese firms in the 1970s
and estimate the supply function for bank loans. To date, most empirical
studies of the Japanese credit market have used Bank of Japan data
to estimate the determinants of average interest rates. Obviously, that
tells us nothing about the distribution of credit among firms. By using
firm-level data instead, we ask whether the observed credit allocation
patterns differ from what one would expect in competitive capital
markets.
4.1.1 The Model. To estimate the loan supply function, we begin
with a pair of structural equations in which the observed interest rate is
a function of both the demand for and the supply of loans:12
id = f (bank debt, profit volatility, Q, sales, industry)
is = g(bank debt, profit volatility, mortgageable assets, leverage,
financial shareholdings)
We posit, in other words, that potentially, the amount of debt a
firm demands at a given interest rate could depend on (1) the volatility
of its performance; (2) its business opportunities (Tobin’s Q); (3) its sales
(reflecting its transactional demand for cash); and (4) the industry in
which it competes. We further posit that potentially, the amount a bank
supplies at a given interest rate could depend on (1) the volatility of the
firm’s performance; (2) the security interests the firm can offer; and (3)
its equity cushion (leverage). Arguably, the amount a bank supplies also
could depend on the shares it holds in the firm.13
Suppose the market for bank loans is competitive and is subject neither to rationing nor to disguised interest charges through
12. Ordinary least squares (OLS) (used for this purpose by, e.g., Caves and Uekusa,
1976) does not disentangle the effects of demand and supply shifts.
13. In a recent review of corporate financing literature in the 40 years after MillerModigliani (MM), Myers (2001) reports empirically that debt/equity ratios tend to be
lower when a firm is smaller, is riskier, has more intangible assets, and is more profitable.
He also finds considerable industry-specific variation.
186
Journal of Economics & Management Strategy
compensating balances. In such a market, the interest rate at which
a bank supplies funds primarily will reflect the volatility of a firm’s
performance. It also may reflect the steps the bank can take to mitigate
that risk, such as obtaining a security interest or providing an equity
cushion.
We estimate is through two-stage least squares (We include selected summary statistics in Table VI; for reference, we include comparable OLS results in Table VII, Panel B.3.). We include surrogates for
Tobin’s Q in order to capture the business opportunities a firm faces—
important in determining a firm’s demand for funds. Because we lack
the data necessary to calculate Q itself, we use two proxies: market capitalization/equity (denoted Surrogate Q) and operating income/total
assets (denoted Profitability). To ask whether (as sometimes asserted)
bank dominance affects a bank’s willingness to lend, we include the
percentage of a firm’s shares held by financial institutions. We use
a vector of industry dummies to reflect industry-specific variation in
demand patterns. 14
4.1.2 Variables. We use the means and variance of the financial
variables but calculate them separately for 1968–75 and 1976–82 (For
reference, we include regressions using the means and variance of the
financial variables over the entire period 1968–82 in Table VII, Panel
B.2.). We do so because the “oil shock” of the mid-1970s divides the
period into distinct economic environments. During 1968 to 1975, real
gross national expenditures (GNE) in Japan rose 54%. By contrast, from
1975 to 1982 it rose only 22%.
We define the following variables:
r Interest Rate: The ratio of a firm’s interest expenses (#103 of the Nikkei
NEEDS database) to the sum of its bank debt (#s 46, 47, 63), longterm notes payable (#64), bonds (#s 48, 62), and discounted notes
receivables (#142).
r Bank Debt: The sum of a firm’s short- (#46) and long-term (#s 47, 63)
borrowings.
r Volatility: The variance of the ratio of a firm’s operating income (#95)
to total assets (#89).
r Mortgageable assets: The ratio of a firm’s tangible assets (#21) to total
assets (#89).
r Leverage: The ratio of a firm’s total liabilities (#77) to total assets (#89).
r Surrogate Q: The ratio of a firm’s stock market capitalization (at the
end of a firm’s April–March fiscal year) over its equity (#78).
14. On whether loan rates also would depend on “keiretsu affiliation,” see footnote 6,
supra, and Miwa and Ramseyer (2002c).
187
The Loan Market in High-Growth Japan
Table VI.
Selected Summary Statisticsa
1968–75
Interest Rate
Bank Debt
(billion yen)
Volatility
Mortgageable
Assets
Leverage
Financial
Shares
Surrogate Q
Profitability
Sales
(billion yen)
n
Min
Mean
1003
1022
.022
0
24
1022
1022
0
0
1976–82
Max
.086
871
.001
.277
1022
965
.147
.001
267
1022
1022
.06
−.331
.068
Industry Dummies
Construction
Trade
Service and Finance
Utilities and transportation
Light Industry
Chemicals
Metals
Machinery
.762
.278
n
Min
Mean
.468 1094
1138
0
0
44
.029
.857
0
1138
1138
1.154 1138
.754 1068
1.7
8.2
.081
.554
78
5204
760
1138
1138
Max
.091
.463
1940
.002
.001
.232
.045
.832
.124
.004
.746
.293
1.518
.755
.3
−.097
.920
2.1
8.7
.069
.524
165
11100
n
mean
n
mean
1022
1022
1022
1022
1022
1022
1022
1022
.100
.100
.038
.083
.125
.157
.118
.279
1138
1138
1138
1138
1138
1138
1138
1138
.097
.121
.052
.078
.123
.149
.109
.272
a Nippon (1968–1982); Toyo keizai (1975); Keizai (1975); Daiwa (1968–1982).
r Profitability: The ratio of a firm’s operating income (#95) to total assets
(#89).
r Sales: A firm’s net sales (#90).
r Financial Shares: The fraction
of a firm’s shares held by financial
institutions (#202/#226).
r Industry Dummies: See summary statistics. In general, we rely on
regressions over all industries. To check whether manufacturing firms
are relevantly different from others, in Table VII we also report
regressions on the manufacturing firms only.
The Data. We assemble our basic financial data from the
Nikkei NEEDS database. We use data on all TSE Section 1 firms (the
largest firms) from 1968 to 1982. After dropping firms with two years
of data or less and extreme outliers, we obtain the sample described in
4.1.3
188
Journal of Economics & Management Strategy
Table VII.
Determinants of Interest Rates
A. Basic Results (using Profitability, 1968–75 and 1976–82 means, 2SLS)a
All Industries
Volatility
Mortgageable Assets
Leverage
Leverage Squared
Financial Shares
Bank Debt
n
Manufacturing Only
1968–75
1976–82
1968–75
1976–82
1.423 (3.25)
−.015 (2.58)
−.305 (5.78)
.174 (4.67)
−.014 (2.20)
.246 (1.37)
948
2.028 (3.87)
.001 (0.12)
−.154 (3.01)
.055 (1.53)
.004 (0.42)
.142 (1.17)
1026
1.618 (2.84)
−.009 (0.80)
−.443 (5.13)
.265 (4.36)
−.017 (1.96)
1.39 (4.93)
649
1.378 (2.36)
.018 (1.16)
−.156 (2.62)
.056 (1.35)
−.005 (0.43)
.690 (3.02)
684
B. Variations
1. Using Surrogate Q (1968–75 and 1976–82 means, 2SLS)b
All Industries
Volatility
Mortgageable Assets
Leverage
Leverage Squared
Financial Shares
Bank Debt (/107 )
n
Manufacturing Only
1968–75
1976–82
1968–75
1976–82
2.319 (2.15)
−.080 (3.24)
−.260 (1.72)
.048 (0.43)
−.060 (2.90)
9.30 (5.32)
258
1.368 (2.19)
.001 (0.14)
−.181 (2.48)
.078 (1.48)
.003 (0.25)
.079 (0.67)
728
2.125 (1.65)
−.114 (2.98)
−.520 (1.73)
.222 (1.02)
−.079 (2.75)
11.1 (4.83)
196
1.352 (1.85)
.021 (1.16)
−.134 (1.38)
.041 (0.59)
−.004 (0.29)
.565 (2.43)
505
2. Using 1968–82 Means (2SLS)c
Volatility
Mortgageable Assets
Leverage
Leverage Squared
Financial Shares
Bank Debt (/107 )
n
All Industries
Manufacturing Only
2.637 (5.44)
−.013 (1.64)
−.217 (3.21)
.095 (1.99)
−.004 (0.44)
.206 (1.24)
1044
2.581 (4.37)
.001 (0.05)
−.342 (3.67)
.178 (2.68)
−.004 (0.32)
1.07 (3.52)
696
Continued
Table VI. We use stock price data from the Daiwa (volumes for 1968–82)
securities firm.
4.1.4 Results
4.1.4.1 Risk. The results are consistent with our hypothesis that the
capital market cleared: Interest rates reflect risk and show no evidence
189
The Loan Market in High-Growth Japan
Table VII.
Continued
3. Using OLS (1968–75 and 1976–82 means)d
All Industries
1968–75
Volatility
Mortgageable Assets
Leverage
Leverage Squared
Financial Shares
Bank Debt (/107 )
Profitability
Sales (/107 )
n
Adjusted R2
1.088 (2.45)
−.001 (0.08)
−.353 (6.58)
.213 (5.62)
−.007 (1.12)
−.122 (0.88)
948
.12
1976–82
.603 (1.33)
.003 (0.36)
−.313 (5.84)
.206 (5.49)
−.005 (0.77)
−.689 (3.23)
.115 (3.86)
.161 (3.46)
948
.15
1.879 (3.51)
.011 (1.01)
−.162 (3.14)
.063 (1.77)
.010 (1.09)
.006 (0.06)
1026
.08
1.901 (3.58)
.011 (1.04)
−.131 (2.55)
.065 (1.83)
.015 (1.66)
−.261 (1.83)
.173 (4.10)
.064 (2.22)
1026
.10
Manufacturing Onlye
1968–75
Volatility
Mortgageable Assets
Leverage
Leverage Squared
Financial Shares
Bank Debt (/107 )
Profitability
Sales (/107 )
n
Adjusted R2
1.348 (2.47)
−.001 (0.07)
−.481 (5.79)
.305 (5.22)
−.002 (0.27)
.132 (0.62)
649
.09
.962 (1.84)
−.006 (0.63)
−.485 (6.24)
.333 (6.14)
−.010 (1.33)
−2.57 (8.11)
.096 (2.91)
1.45 (9.82)
649
.22
1976–82
1.213 (2.10)
.026 (1.73)
−.160 (2.71)
.067 (1.62)
.005 (0.44)
.092 (0.54)
684
.06
1.363 (2.38)
.024 (1.59)
−.133 (2.24)
.066 (1.63)
.002 (0.25)
−.637 (2.56)
.128 (2.57)
.388 (3.88)
684
.09
Notes: The 2SLS regressions treat bank debt as endogenous. Coefficients followed by the absolute value of the
t-statistics. All regressions include a constant term not reported here. The OLS regressions include industry dummies,
which are not reported here. For sources, see Table VI.
a Dependent variable: interest rate; Instruments of bank debt: profitability, volatility, sales, industry dummies.
b Dependent variable: interest rate; Instruments of bank debt: surrogate Q, volatility, sales, industry dummies.
c Dependent variable: interest rate; Instruments of bank debt: profitability, volatility, sales, industry. dummies.
d Dependent variable: interest rate.
e Dependent variable: interest rate.
of rationing (see Table VII, Panel A.). According to our basic results,
firm volatility strongly predicts the interest rate banks charge for credit.
Whether we look at the early period (1968–75) or later (1976–82); whether
we look at all industries or only manufacturing firms; whether we
use two-stage least-squares (2SLS) or OLS (Panel B.3.); whether we
instrument the demand for loans with our proxy for Q (Panel B.1.) or firm
profitability (Panel A.)—whatever set of measures we use, the coefficient
190
Journal of Economics & Management Strategy
on firm volatility consistently is positive and generally statistically is
significant.
4.1.4.2 Mortgageable Assets. As basic logic predicts, firms with large
stocks of mortgageable assets borrow at lower rates. This result is less
robust than the results for volatility, however, and appears only among
the 1968–75 estimates.
4.1.4.3 Equity Cushion. The impact of a firm’s equity cushion (inversely proxied by leverage) on interest rates is nonlinear. Initially, an
increase in leverage is associated with lower interest rates, but at high
leverage levels interest rates again rise. We offer no explanation for why
interest rates would fall with leverage at lower levels. That they would
increase at higher leverage levels, however, is exactly what one would
expect. The most plausible estimates are those for 1968–75 for the largest
sample (Table VII, Panel A): interest rates rise as leverage climbs beyond
89% (for all firms) or 84% (for manufacturing firms).
Potentially, of course, the level of leverage at a firm will depend
on the interest rate the firm must pay to borrow. To address this endogeneity, we re-run our Table VII Panel A regressions without leverage
or leverage squared. The results (available upon request) remain largely
unchanged: in the 1968–75 2SLS estimates, the coefficient on volatility
ranges from 2.1 to 2.5 with a t-statistic between 1.49 and 4.25; in the
1976-82 estimates, it ranges from 0.928 to 1.769 with a t-statistic between
1.41 and 3.43.
4.1.4.4 Q. By basic finance theory, Tobin’s Q is the appropriate proxy
for a firm’s business opportunities, and the 2SLS estimates in Table VII,
Panel B.1. use our surrogate Q. This presents two problems. First, we
have neither the market value of a firm’s debt nor the replacement costs
of its assets. As a result, our surrogate Q simply measures the ratio of a
firm’s market capitalization to accounting capital. Second, for the first
period (1968–75), we have stock price data only for a minority of the
firms. Accordingly, the use of Q dramatically lowers sample size. Given
these problems with our surrogate Q, in Panel A we instrument bank
debt with accounting profitability. Largely, this yields results consistent
with regressions using surrogate Q (Panel B.1.).
4.1.4.5 Bank Dominance. Several scholars argue that Japanese banks
use their influence over the firms most dependent on them to extract
rents through high interest charges (e.g., Weinstein and Yafeh, 1995). In
fact, Table VII shows nothing of the sort. For 1968–75, the more heavily
financial institutions have invested in a firm, the lower the interest rate
a firm pays; for 1976–82, the results are insignificantly different from
zero. We have no theory about why the interest rate would fall with
The Loan Market in High-Growth Japan
191
bank shareholdings. Suffice it to say that the evidence is inconsistent
with any claim that banks raise rates at the firms they most dominate.
4.1.4.6 Implications. We do not claim that these results prove that
banks charged market-clearing rates. We do claim that the results
suggest that banks charged interest rates that reflected borrower risk.
Obviously, that result is exactly what one would expect if the interest
rate caps did not bind. More generally, it is exactly what one would
expect if the loan market had cleared.
4.2 Determinants of Bank Deposits
Turn then to the factors that best predict the amount of deposits a firm
will keep. As noted earlier, many observers claim that Japanese banks
routinely demanded deposits to raise the effective interest rate they
charged on their loans. According to the very documents on which these
observers rely, however, banks never demanded deposits from large
firms—and our sample (TSE Section 1 firms) includes only large firms.
Suppose, however, that banks did demand deposits to raise interest rates. All else equal, deposits should be associated positively
with three sets of variables: (1) loan interest rates; (2) the amount of
bank debt; and (3) bank shareholdings. First, deposits would be associated positively with interest rates because banks would demand them
disproportionately from firms facing the regulatory interest cap. By
contrast, we hypothesize (1) that deposits help a bank monitor a debtor’s
performance; (2) that deposits partially substitute for mortgages; and (3)
that higher interest rates raise the cost of holding large deposits. For all
these reasons, we predict that deposits will be associated negatively
with observed interest rates.
Second, if banks demand deposits to raise interest rates, deposits
also would be associated positively with debt levels. After all, a bank
would need a bigger deposit to raise the effective interest rate on a bigger
loan. By contrast, we predict no relation between deposits and loans.
Third, if (as the conventional wisdom asserts) banks exploit those
firms most dependent on them, they would exploit firms in which they
have large equity stakes. If so, then financial shareholdings would be
associated positively with deposit levels. By contrast, for the reasons
previously discussed, we expect no relation between deposits and this
variable.
In addition, observed interest rate held constant, we predict that
the size of deposits will be associated (1) positively with sales, since sales
amounts reflect the transactional demand for cash; (2) negatively with
the supply of mortgageable assets, since deposits partially substitute for
mortgages; and (3) positively with the volatility of firm performance,
192
Journal of Economics & Management Strategy
again since deposits partially substitute for mortgages (and volatile
firms stand to gain the most from providing mortgages).
4.2.1 The Model, Data, and Variables. We employ a simple
OLS model:
Deposits = h(interest rate, bank debt, mortgageable assets, sales,
volatility, financial shareholdings)
We use the same datasets we used previously but normalize the relevant
variables by firm size. We define the following variables:
r Deposits/Assets: The ratio of a firm’s cash and deposits (NEEDS, #3)
to total assets (#89).
r Bank Debt/Assets: The ratio of a firm’s bank debt (#s 46, 47, 63) to total
assets (#89).
r Mortgageable Assets/Assets: The ratio of a firm’s tangible assets (#21)
to total assets (#89).
r Sales/Assets: The ratio of a firm’s net sales (#90) to total assets (#89).
r Volatility, Interest Rate, Financial Shares, and Industry Dummies: As
previously defined.
4.2.2 Results. As we expected, the results in Table VIII indicate
that banks did not use compensating deposits to raise effective interest
rates. Instead, deposit levels are associated negatively and significantly
with observed interest rates in all specifications and are associated
only insignificantly with bank loan levels. The results also suggest
that deposits substitute for mortgageable assets: The coefficients on
mortgageable assets are consistently and strongly negative, while the
coefficients on volatility are positive (though not statistically significant). The coefficients on sales remain a puzzle: positive for 1968–75
as predicted but negative for 1976–82 (and statistically significantly for
manufacturing firms).15 The coefficients on shareholdings by financial
institutions are significantly negative—refuting the claim that banks
used their power to extract rents.
5. Credit Rationing in Ocean Shipping
5.1 The Programs
To give context to the institutional detail and econometric results,
consider what happened in one of the industries where the Japanese
15. If sales proxy for profitability and banks see deposits as a substitute for mortgages,
then they may demand greater deposits when sales fall. This effect could offset the positive
correlation predicted by the expectation that sales reflect the transactional demand for
cash—yielding the indeterminate effect we observe.
193
The Loan Market in High-Growth Japan
Table VIII.
Determinants of Bank Deposits
All Industries
1968–75
Interest Rate
Bank Debt/Assets
Mortgage/Assets
Sales/Assets
Volatility
Financial Shares
n
Adjusted R2
−.197 (3.11)
.001 (0.10)
−.146 (12.99)
.004 (1.23)
1.236 (1.50)
−.082 (7.43)
948
.22
Manufacturing Only
1976–82
1968–75
1976–82
−.136 (3.13)
.007 (0.68)
−.141 (12.19)
−.005 (1.63)
.415 (0.60)
−.078 (6.94)
1026
.17
−.272 (4.04)
.003 (0.21)
−.168 (9.27)
.013 (2.26)
1.428 (1.56)
−.075 (5.75)
649
.17
−.215 (4.60)
.010 (0.82)
−.140 (7.68)
−.011 (2.44)
.229 (0.35)
−.075 (6.04)
684
.16
Notes: Dependent variable: deposits/assets. The regressions are OLS. Coefficients followed by the absolute value of
the t-statistics. All regressions include a constant term; regressions on all industries include industry dummies not
reported here. For sources, see Table VI.
government tried hardest to ration credit. More specifically, consider
finance patterns among the ocean shipping firms during the heyday
of Japan’s rapid economic growth. The war had left shipping firms
decimated, and the government apparently decided to help them rebuild. Toward that end, it aggressively tried to manipulate investment.
Although it started during the occupation, it tried even harder during
the mid-1960s.
The shipping industry did grow. Given the custom of registering
ships in nominal places like Panama or Liberia, national capacity can
be hard to estimate. Still, from 1955 to 1975, Japanese-flag capacity
increased from 4 million tons to 40 million. By 1975, no country except
Liberia with its 66 million tons could claim more.16
To the shipping and shipbuilding firms the government did offer
massive subsidies. From 1961 to 1970, it routed the shipping firms
33.7% of all Japan Development Bank (JDB) loans. To shipbuilding firms
preparing vessels for export, it routed 48.3% of all Export-Import Bank
(Ex-Im Bank) loans.17 According to Horiuchi and Otaki (1987), during
the early years of the decade, shipping firms obtained over half of their
16. Nihon senshu (various years). Other countries in 1975—UK: 33 million tons;
Norway: 26 million tons; Greece: 23 million tons; USSR: 19 million tons; US: 15 million
tons.
17. Nihon zosen (1980). The annual ratios of shipping firm loans from JDB ranged from
17.8% (1962) to 44.7% (1965). The mean of the annual ratios was 31.8%. Over the decade,
annual total loans by JDB averaged 204 billion yen. The annual ratios of shipbuilding firm
loans from the Ex-Im Bank ranged from 40.8% (1961) to 58.4% (1964). The mean of the
annual ratios was 48.6%. Over the decade, annual total loans by the Ex-Im Bank averaged
247 billion yen.
194
Journal of Economics & Management Strategy
capital investment from the government. Only coal-mining firms kept
their dependence levels higher.
By these loans, the government transferred enormous wealth. The
JDB raised its funds (through the Ministry of Finance’s so-called “Fiscal
Investment & Loan Program”) from government-run financial institutions like the postal savings system. It then lent the funds to private firms
at 6.5% (Nihon senshu, 1970). To shipping firms on approved projects,
it lent 60–80% of the cost of a ship and spread repayment over 11–
13 years.18
The subsidies did not stop there. First, the government used the
general budget to cut the cost of JDB loans further. For much of the 1960s,
it forgave 2.5% of the 6.5% interest and charged shipping firms only the
remaining 4% (Nihon senshu, 1970). Second, it informally guaranteed
private-sector loans for the rest of the cost of a ship. In exchange,
the banks loaned shipping firms the necessary funds at 8.4–9.1% and
extended repayment over eight years.19
Third, usually the government paid 2–3% of the stated interest
on the private-sector loans. As a result, shipping firms borrowed from
private banks at 6–7.1% net.20 In effect, those who complied with the
government program borrowed most of the cost of a ship from the JDB
at 4% and the remainder from private banks at 6–7%.
Last, if financially troubled, the government let shipping firms
defer repayment even beyond the (already generous) contractual terms.
When shipping firms found themselves in distress after the Suez Canal
reopened in the late 1950s, for example, it deferred their JDB obligations.
It then induced private banks to allow similar deferrals.
5.2 Sanko
5.2.1 The Issue. If the program’s structure is clear, its effect is
less so. Reflecting the standard wisdom, Hoshi and Kashyap (2001)
argue that the program let the government “tightly regulate[] the
number of new vessels that could be produced each year” (p. 159).
Yet to do so, the government would have needed both (1) to stop
firms from borrowing at market rates on unapproved projects; and
18. Nihon senshu (1970). In general, the expected life of a ship still would have
exceeded this repayment period.
19. Nihon senshu (1970). Eight years from delivery of the ship. The terms were
determined through negotiations between the bank trade association and the government.
For details of the negotiations, see Ginko kyokai (1965, pp. 347–67).
20. The subsidies called for partial repayment if profits/capital exceeded 10% (Nihon
senshu, 1970).
The Loan Market in High-Growth Japan
195
(2) to stop them from arbitraging funds from approved projects to the
unapproved.
If the government could stop both unapproved loans and arbitrage, it had at least a shot at regulating investment. If not, it seldom
would have affected a firm’s returns on its marginal investments (Miwa
and Ramseyer, 2001b). Not changing marginal calculations, it would
seldom have affected either the level or direction of investments. Not
affecting output, for better or worse it would not have implemented any
“industrial policy.” So—could the government stop unapproved loans
and arbitrage?
5.2.2 The Strategy. Among the shipping firms, none was more
outspoken than the Sanko steamship firm. Before the early 1950s, Sanko
had taken government subsidies and had complied with government
mandates. In the mid-1950s, however, it decided to go it alone. Rather
than take and comply, it would raise its own funds and would follow
its own plans. While its rivals stayed within the government’s orbit, it
repaid its JDB loans and turned exclusively to private capital (Table IX,
Panel D).
In Sanko’s eyes, the subsidies brought too much control. Loans
always come with terms, of course—whether in Japan or in the United
States, whether from the government or from private banks. Sanko was
willing to accept the terms private creditors and investors imposed. It
was not willing to accede to the government’s.
For Sanko, the government loans presented several problems.
First, the government pushed obsolete services. The government
claimed to be rebuilding the industry to its prewar glory. Because
regular, scheduled freightliners had been central to the industry prewar,
it focused the post-war program on liners as well. Yet as Sanko saw it, the
industry had shifted. The future lay not with standard liners but with
industry-specific ships like oil tankers, operating on shipper-specific
schedules. Were it to accept the subsidies, it would need to focus on
services.
Second, the government imposed a long and cumbersome loan
application process but promised funding only year by year. Again, as
Sanko saw it, it had to be able to plan over several years in order to offer
its clients what they needed. Rather than apply annually for funds it
might or might not obtain, it had to be able to work with its clients long
term.
Third, through its loan covenants the government demanded a
veto over new projects. To Sanko, this posed trouble on two fronts. On
the one hand, in order to help the industry earn monopoly rents, the
196
Journal of Economics & Management Strategy
Table IX.
Relative Performance of Shipping Firms
A. Stock Market Capitalization (in million yen)
1964
1968
1973
Sanko
a NYK
a Mitsui OSK
a Japan Lines
a Kawasaki
a Yamashita
a Showa
Shinwa
Daiichi
Iino
3,591
18,104
11,004
11,099
7,740
2,740
2,160
1,427
1,000
2,924
6,615
28,908
26,800
14,406
17,280
7,905
6,750
3,600
2,320
3,456
1978
514,062
146,439
112,800
153,731
101,655
45,540
29,160
15,012
5,400
14,688
163,697
162,383
109,435
83,580
45,098
34,425
24,300
22,194
12,240
13,920
B. Capacity (Number of Ships, 1,000 tons)b
1964
1968
1973
Sanko
NYK
Mitsui OSK
Japan Lines
Kawasaki
Yamashita
Showa
Shinwa
Daiichi
Iino
13
172
245
102
135
127
53
84
64
52
280
1,925
2,548
1,493
1,684
1,143
844
555
436
558
53
261
289
129
189
130
70
122
97
55
1,858
4,993
4,738
3,123
4,246
2,971
1,754
1,377
1,346
1,119
173
299
291
259
206
190
123
131
129
58
12,255
10,867
10,372
15,673
7,775
8,688
4,756
4,376
3,800
2,447
C. Debt (Long Term, Short Term; in million yen)
1964
1968
1973
Sanko
NYK
Mitsui OSK
Japan Lines
Kawasaki
Yamashita
Showa
Shinwa
Daiichi
Iino
3,247
31,114
34,767
21,078
27,244
15,605
9,458
8,936
1,264
7,307
1,594
10,661
23,521
9,661
14,260
16,671
9,379
3,078
2,223
8,178
19,719
92,241
71,864
56,779
67,898
47,173
31,238
17,397
18,790
1,659
7,282
13,660
11,527
9,332
10,064
6,412
3,077
2,096
1,204
1,360
109,368
175,963
143,881
106,810
91,300
70,024
59,726
32,277
30,590
20,349
43,820
49,604
50,511
26,581
25,737
18,682
9,346
5,207
4,613
3,551
1978
305
305
282
238
199
178
138
155
133
74
24,637
12,321
10,058
19,745
9,970
8,795
5,905
4,979
3,635
2,564
1978
183,628
119,104
102,509
58,494
97,830
43,793
45,255
23,916
19,079
23,219
64,156
118,281
90,938
56,238
41,901
26,549
19,729
11,088
9,298
8,331
Continued
government often wanted to block construction just when Sanko—as
industry renegade—wanted to expand. On the other hand, the government wanted Sanko under the control of (and perhaps as a mere division
of) a larger, more compliant shipping firm. To reduce competition, it had
decided to consolidate the industry into six firms (or firms under the
197
The Loan Market in High-Growth Japan
Table IX.
Continued
D. Composition of Bank Debt (% JDB, Name of and % Borrowed from
Other Principal Creditor)
1965
1970
1975
Sanko
NYK
Mitsui OSK
Japan Lines
Kawasaki
Yamashita
Showa
Shinwa
Daiichi
Iino
0
73
65
66
65
68
61
58
57
45
Dwc
Mbd
NAe
IBJf
NA
Swg
IBJ
IBJ
Smth
∗j
34
5
9
11
8
17
16
19
0
76
75
77
67
78
77
77
80
80
LTCBi
IBJ
IBJ
IBJ
NA
IBJ
LTCB
IBJ
IBJ
IBJ
13
4
5
5
0
59
61
53
61
68
67
74
76
70
4
6
8
5
5
E. Annual Shareholder Rate of Return
1955–60
1960–65
1965–70
Sanko
NYK
Mitsui OSK
Japan Lines
Kawasaki
Yamashita
Showa
Shinwa
Daiichi
Iino
6.1
1.8
−0.9
8.6
3.5
5.4
5.2
−2.6
N.A.
2.0
10.5
7.5
4.5
5.3
10.5
−3.5
0.6
−2.7
N.A.
−24.9
32.0
11.9
8.6
9.4
12.4
12.9
16.6
13.8
6.9
6.9
LTCB
Mb
IBJ
∗∗k
IBJ
Sw
LTCB
IBJ
IBJ
IBJ
9
10
6
7
8
7
7
8
5
7
1970–75
61.5
25.1
25.9
49.6
27.4
33.9
31.9
38.9
17.8
33.0
Notes: In Panels A–C, we use 1964 data rather than 1963 because of mergers in 1964 between listed and unlisted firms
(about which data are not available). Sources: Nihon (1955–1975); Kyoiku sha (1980); Keizai chosa kai (1975).
a Firms designated as “core” firms under government policy. The government hoped to consolidate the industry
around these six.
b For reasons of data availability, the 1964 data in Panel B reflect mid-year data rather than end-of-year data as
elsewhere.
c Daiwa.
d Mitsubishi Bank.
e Information not available.
f Industrial Bank of Japan.
g Sanwa Bank.
h Sumitomo Bank.
i Long-term Credit Bank.
j Kawasaki Shipbuilding.
k Bank of America.
control of the six), and Sanko was not to be one of the six. Fundamentally,
Sanko and the government were now on adversarial terms.21
21. The government also pressured private banks not to fund Sanko. Through their
trade association, the banks had agreed to report any new loans they made to shipping
firms to the association. Now the government could urge them not to fund nonconformist
firms like Sanko.
198
Journal of Economics & Management Strategy
5.2.3 The Results. Sanko jettisoned the government subsidies
all the way to the bank. It had opened the 1950s with virtually nothing.
It closed the 1960s as the most profitable firm in the industry. During
the last half of the decade, it earned shareholder returns of 32% a year,
and by the early 1970s, 62% (Table IX, Panel E). Its closest rival during
the late 1960s was Showa, but it earned only 17% and in the early 1970s
only 32%. Its closest rival during the early 1970s was Japan Lines, but it
earned only 50% and in the late 1960s only 9%.
The Organization of the Petroleum Exporting Countries (OPEC)
transformed the industry in 1973 (more on this in section 5.4), but
by then Sanko had grown from the sixth-ranked firm (in 1964) to the
largest. From a stock-market capitalization of 3.59 billion yen in 1964,
it had boomed to 514 billion yen by 1973, three times its nearest rival
(Table IX, Panel A). Despite making no “approved” vessels, it commanded a shipping capacity second only to Japan Lines (Table IX, Panel
B). Despite refusing to participate in the government’s loan program, it
serviced the third largest debt in the industry (Table IX, Panel C).22
Flout as it did government policy, Sanko raised funds straightforwardly. First, it sold stock and retained its earnings. In 1952, it had
paid-in capital of only 420 million yen. By 1956, it had 1,300 million
yen; by 1964, 4,700 million yen; and by 1974, 31,000 million yen (Kyoiku
sha, 1980; Sanko, 1968). Second, it leased. From 1963 to 1971, Sanko
increased the number of ships it controlled from 13 to 108; instead of
buying all of them, it leased about half (Sanko 1961, 1966). Depending
on the contractual terms, leasing can have identical economic effects
to borrowing. For Sanko, the identity presented a standard financing
strategy. Third, it borrowed. On the one hand, Sanko borrowed from
banks—sometimes from a single bank and sometimes from multiple
banks (Sanko, 1961). Generally, it arranged for its client—the firm on
whose behalf it would eventually operate the ship—to guarantee it
business (Sanko, 1968). On the other hand, Sanko negotiated deferred
payments to the builders from which it bought its ships (Sanko, 1966).
The trick involved arbitrage. Even if the government could discourage
banks from lending directly to Sanko, it did not try to prevent banks
from lending to shipbuilding firms that sold to Sanko. If those firms
then let Sanko defer its payments to them, they effectively arbitraged
their own credit. Suppose a firm obtained a subsidized loan through the
Export-Import Bank to sell ship A abroad. If it then deferred payment
on the ship B it sold to Sanko, it even arbitraged the government loan
on the exported ship.
22. Nor was it just quantity—it was quality too. Throughout the period, Sanko
relentlessly upgraded its fleet, selling unprofitable ships and modernizing and automating
what it kept (Kyoiku sha, 1980, pp. 32–33).
The Loan Market in High-Growth Japan
199
The moral is simple. During the 1960s, the government intervened
heavily in the ocean shipping industry. It intervened for a specific
purpose (rebuild the industry) and offered compliant firms massive
subsidies (low-cost loans). Throughout the decade, it did what it could to
run Sanko out of business: It paid competitors subsidies but not Sanko;
it lent competitors money but not Sanko; it encouraged private banks
to lend to competitors but not to Sanko. Ever the nonconformist, Sanko
flouted all this to spectacular success. By the early 1970s it had raised
enough funds to catapult itself into preeminence and generated high
enough profits to earn its investors huge returns.
5.3 Tanker Firms
Nonconformity neither started nor stopped with idiosyncratic Sanko.
Sanko may have been the most visible shipping firm to buck national
shipping policy, but it hardly was alone. If it questioned the government’s unwillingness to promote tankers, so did many petroleumrefining firms. Rather than defer to national policy, some bought their
own tankers or formed transportation subsidiaries that did.23
Through such policies, the refining firms integrated vertically into
transportation. Sometimes, they bought the tankers in conjunction with
foreign firms. These foreign firms could borrow abroad, of course.
Arbitrage being what it is, Japanese firms that entered joint ventures
with them could finance both their refineries and tankers abroad.
Consider Tokyo tankaa [Tanker], the first of the specialized firms
(Nihon sekiyu, 1982). The Nisseki refining firm had formed a joint
venture with Caltex in 1951. That venture (it bought a 64% interest)
then teamed up with Nisseki (4%) and Koa Petroleum (32%) to form
Tokyo Tanker in 1952. Initially, Tokyo Tanker leased a ship from the
joint venture itself. By 1956, it began building its own fleet. For its first
vessel, it borrowed $4 million from the First National City Bank of New
York. It borrowed another $5 million from First National in 1958 and $5
million from Chase Manhattan Bank in 1959.
Internationally, vertically integrated tanker operations were the
norm (Okaniwa, 1981), and Nisseki had close ties to Caltex. Yet within
Japan even petroleum firms not tied to foreign firms borrowed abroad
and bought their own tankers. From 1955 to 1963, the independent
Idemitsu kosan firm borrowed $56 million from Bank of America and
Esso ($6 million of that for tanker capacity). Independent Maruzen
sekiyu borrowed $61 million from the Bank of America, Unoco, and
23. Refineries affiliated with foreign firms used tanker subsidiaries because Japanese
law prohibited the ownership of Japanese flag vessels by firms with foreign national
directors. See Senpaku ho [Ship Act] of 1899, Law No. 46, § 1.
200
Journal of Economics & Management Strategy
Continental Illinois (also $6 million for tankers). And independent
Daikyo sekiyu, Nihon kogyo, and Shin Ajia sekiyu each borrowed lesser
amounts abroad (Sangyo keikaku, 1965). Indeed, from 1960 to 1963, only
16 of the 41 tankers built were funded by the government; 11 others were
funded abroad (Tonen, 1979).
The petroleum-refining firms built considerable transportation
capacity. By 1978, Tokyo Tanker had eight tankers carrying 749 thousand
tons. Idemitsu (with its own tanker subsidiary) owned 10 tankers (1,200
thousand tons); Daikyo (also with a tanker subsidiary) owned three (189
thousand tons); and Maruzen owned two (46 thousand tons). The firm
with the most tanker capacity, however, remained the Sanko shipping
firm: 23 tankers carrying 2.6 million tons (Nihon tankaa, 1980).
In Table X, we detail the shipping capacity firms developed outside
government policy. Of all new ships in 1965, they produced 36% (18% of
total capacity) beyond the official programs. Like Sanko, they apparently
found few financing barriers they could not circumvent. To fund these
Table X.
Government-Approved and Independently
Produced Vessels, 1961–73
Government-Approved Ships
Production
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
Independent Ships
Financing
Production
Noa
Capb
Govc
PBd
Eqe
No
Cap
Gov/Totf
27
13
18
41
65
75
56
57
57
45
41
37
25
498
393
567
1,209
1,825
1,909
2,033
2,308
2,747
2,624
3,218
3,304
1,985
.54
.71
.77
.78
.77
.76
.76
.77
.62
.61
.53
.52
.49
.56
.29
.23
.22
.23
.24
.24
.23
.28
.30
.30
.30
.32
0
0
0
0
0
0
0
0
42
15
15
32
36
69
72
128
161
118
192
115
115
581
209
170
424
405
483
809
841
1,191
1,162
3,706
1,307
4,396
.46
.65
.77
.74
.82
.80
.72
.73
.70
.69
.46
.72
.31
.10
.09
.17
.18
.19
Notes: Slightly different figures for independently produced ships appear in Nihon zosen (1980, p. 139, tab. 1).
Sources: Nihon senshu kyokai (1970 ed., pp. 172-73, and 1981 ed., pp. 284-85); Nihon zosen kogyo kai (1980, p. 391);
Ginko kyokai (1965, p. 361).
a Number of ships built.
b Capacity of ships built, in 1000 tons.
c Fraction of shipbuilding costs funded by government loans.
d Fraction of shipbuilding costs funded by loans from private banks.
e Fraction of shipbuilding costs funded out of shipping firm equity.
f Total capacity of ships produced with government approval, divided by total capacity of all ships produced.
The Loan Market in High-Growth Japan
201
ships, they borrowed 15% from banks, 41% from trading partners (such
as shipbuilding firms), and 26% abroad (Nihon senshu, 1970; Ginko
kyokai, 1965).
5.4 The Oil Embargo
By the time OPEC was organized, all of this changed. Facing radically
higher oil prices, western firms now cut the amount of oil they consumed
and, hence, shipped. They also began looking harder for oil outside of
the Middle East—and further cut the amount they shipped. Firms that
had invested heavily in tankers suffered, and Sanko suffered as much
as any. By 1985, it filed for bankruptcy. Ex post, it had gambled and lost.
Gambles that go bad ex post can still be good ex ante, of course. At least
investors seem to have thought Sanko a good gamble ex ante. And in
losing ex post, Sanko was also in good company. Even the governmentfavored firms lost heavily. Among the government’s anointed six, Japan
Lines had failed and had been acquired by Yamashita by 1988.
Good or bad ex ante, Sanko’s gamble was one in which it had been
able to indulge. It was not a gamble the capital market stopped. Neither
was it a gamble the government prevented. Competitive capital markets usually will route entrepreneurs the funds they need for sensible
gambles. Often, they will route them the funds for foolish gambles too.
Sensible or foolish, Sanko’s gamble was one the government opposed—
and the market funded.
6. Conclusions
By most accounts, in the 1960s and 1970s the Japanese government
barred domestic competitors to bank loans. It banned foreigners from
investing in Japan. It capped loan interest rates. Thereupon, it rationed
credit to its favored firms.
In fact, our data—comprehensive financial data on all TSE-Section
1 firms—and examination of the historical record suggest that the
government did nothing of the sort. It did not shut down domestic
competitors, did not stop foreign investors, and did not relevantly cap
loan interest rates. It did not ration credit, for the credit market cleared.
Because banks could set nominal interest rates at market levels, for
large firms they did not use compensating deposits to hike effective
rates.
The Japanese story of the 1960s and 1970s is not a story about a
government that directed credit to promote industrial policy. Fundamentally, governments in modern advanced economies lack the means
202
Journal of Economics & Management Strategy
to direct credit. Instead, the Japanese story is a more mundane story
that tracks the contours of standard economic theory. At its root, it
is a story about decentralized financial markets that allocated funds
competitively.
Although the Japanese government did claim to regulate access to
credit, like most other governments at most times it lacked the means
to do so. After all, money is fungible, durable, readily concealable, and
widely available. Because it is all of these things, it also easily is tradable
and effectively is arbritrable. Faced with government controls that did
not bind, Japanese firms raised their funds in markets that cleared.
Faced with controls that did not bind, they raised their funds where
they wanted for what they wanted.
References
Ackley, G. and H. Ishi, 1976, “Fiscal, Monetary, and Related Policies,” in H. Patrick and
H. Rosovsky, eds., Asia’s New Giant: How the Japanese Economy Works. Washington, DC:
The Brookings Institution, 153–247.
Calder, K.E., 1993, Strategic Capitalism: Private Business and Public Purpose in Japanese
Industrial Finance, Princeton, NJ: Princeton University Press.
Cargill, T.F., 2000, “What Caused Japan’s Banking Crisis?” in Hoshi and Patrick, eds., 2000,
Crisis and Change in the Japanese Financial System. Boston, MA: Kluwer Academic,
37–58.
—— and N. Yoshino, 2000, “The Postal Savings System, Fiscal Investment and Loan
Program, and Modernization of Japan’s Financial System,” in Hoshi and Patrick, 2000,
Crisis and Change in the Japanese Financial System. Boston, MA: Kluwer Academic,
201–230.
Caves, R. and M. Uekusa, 1976, Industrial Organization in Japan, Washington, DC: The
Brookings Institution.
Daiwa shoken keizai kenkyujo [Daiwa Securities Research Institute], ed., Volumes for
1968–82, Anarisuto gaido [Analyst Guide], Tokyo, Japan: Daiwa shoken keizai kenkyujo.
Flath, D., 2000, The Japanese Economy, Oxford, UK: Oxford University Press.
Ginko kyokai [Banking Association], ed., 1965, Ginko Kyokai 20-Nen Shi [A 20-Year History
of the Banking Association], Tokyo, Japan: Ginko kyokai.
Hamada, K. and A. Horiuchi, 1987, “The Political Economy of the Financial Market,” in
K. Yamamura and Y. Yasuba, eds., The Political Economy of Japan, Volume I: The Domesic
Transformation, Stanford, CA: Stanford University Press, 223–262.
Henderson, D.F., 1986, “Access to the Japanese Market: Some Aspects of Foreign Exchange
Controls and Banking Law,” in G.R. Saxonhouse and K. Yamamura, Law and Trade
Issues of the Japanese Economy: American and Japanese Perspectives, Seattle: University of
Washington Press, 131–156.
Horiuchi, A. and M. Otaki, 1987, “Kin’yu: Seifu kainyu to ginko kashidashi no juyosei
[Finance: The Importance of Government Intervention and Bank Lending],” in K.
Hamada, M. Kuroda and A. Horiuchi, eds., Nihon keizai no makuro bunseki [The Macro
Analysis of the Japanese Economy], Tokyo, Japan: Tokyo daigaku shuppan kai, 123–148.
Hoshi, T. and A. Kashyap, 2001, Corporate Financing and Governance in Japan, Cambridge,
MA: MIT Press.
——, —— and D. Scharfstein, 1990, “The Role of Banks in Reducing the Costs of Financial
Distress in Japan, Journal of Financial Economics, 27, 67–88.
The Loan Market in High-Growth Japan
203
——, —— and ——, 1991, “Corporate Structure, Liquidity, and Investment: Evidence from
Japanese Industrial Groups,” Quarterly Journal of Economics, 106, 33–60.
—— and H. Patrick, eds., 2000, Crisis and Change in the Japanese Financial System. Boston,
MA: Kluwer Academic.
Ito, T., 1992, The Japanese Economy, Cambridge, MA: MIT Press.
——, 2000, “The Stagnant Japanese Economy in the 1990s: The Need for Financial
Supervision to Restore Sustained Growth,” in Hoshi and Patrick, eds., 2000, Crisis
and Change in the Japanese Financial System. Boston, MA: Kluwer Academic, 85–107.
Itoh, M. and K. Kiyono, 1988, “Foreign Trade and Direct Investment,” in R. Komiya, M.
Okuno and K. Suzumura, eds., Industrial Policy of Japan, Tokyo, Japan: Academic Press,
155–181.
Keizai chosa kai [Economic Research Institute], ed., 1975, Keiretsu no kenkyu [Research on
the Keiretsu], Tokyo, Japan: Keizai chosa kai.
Kitahara, M., 1970, Zukei Nihon no Kin’yu [Illustrated: Japanese Finance], Tokyo, Japan: Zaikei
shoho sha.
Kosei torihiki iinkai [Fair Trade Commission], ed., 1979. Kosei Yokin no Jittai [The Reality of
Compulsory Deposits], Tokyo, Japan: Kosei torihiki iinkai.
Krugman, P., 1994, The Age of Dimishing Expectations: US Economic Policy in the 1990s,
Cambridge, MA: MIT Press.
Kuroda, I. and Y. Oritani, 1979, Waga Kuni no “Kin’yu Kozo Tokucho” no Saikento [A
Reexamination of “The Peculiarity of Financial Structure” in Our Country], [Nichigin]
Kin’yu kenkyu shiryo, 2, 1–22.
Kyoiku sha, ed., 1980, Kaiun Gyokai: Joi 10sha no Keiei Hikaku [The Ocean Shipping Industry:
The Comparative Management of the Top 10 Firms], Tokyo, Japan: Kyoiku sha.
Lincoln, J.R., M. Gerlach and C. Ahmadjian, 1998, “Evolving Patterns of Keiretsu Organization and Action in Japan,” Research in Organizational Behavior, 20, 303–345.
Macey, J.R. and G.P. Miller, 1995, “Corporate Governace and Commercial Banking: A
Comparative Examination of Germany, Japan, and the United States,” Stanford Law
Review, 48, 73.
Meerschwam, D.M., 1991, “The Japanese Financial System and the Cost of Capital,” in P.
Krugman, ed., Trade with Japan: Has the Door Opened Wider? Chicago, IL: University of
Chicago Press, 191–220.
Milgrom, P. and J. Roberts, 1994, “Complementarities and Systems: Understanding
Japanese Economic Organization,” Estudios Economicos 9, 3.
Miwa, Y. and J.M. Ramseyer, 2000, “Corporate Governance in Transitional Economies:
Lessons from the Prewar Japanese Cotton Textile Industry,” Journal of Legal Studies, 29,
171–203.
—— and ——, 2001a, “Nihon no Keizai Seisaku to Seisaku Kenkyu [Japanese Economic
Policy and Policy Research],” Keizai Kenkyu, 52, 193.
—— and ——, 2001b, “Directed Credit? Capital Market Competition in High-Growth
Japan,” Harvard John M. Olin Discussion Paper, No. 334, September.
—— and ——, 2002a, “The Myth of the Main Bank: Japan and Comparative Corporate
Governance,” Law & Social Inquiry, 27, 401–424.
—— and ——, 2002b, “Banks and Economic Growth: Implications from Japanese History,
Journal of Law & Economics, 45, 127–164.
—— and ——, 2002c, “The Fable of the Keiretsu,” Journal of Economics & Management
Strategy, 11, 169–224.
—— and ——, 2002d, “The Value of Prominent Directors: Corporate Governance and
Bank Access in Transitional Japan,” Journal of Legal Studies, 31, 273–301.
Morck, R. and M. Nakamura, 1999, “Banks and Corporate Control in Japan,” Journal of
Finance, 54, 319–339.
204
Journal of Economics & Management Strategy
Myers, S.C., 2001, “Capital Structure,” Journal of Economic Perspectives, 15, 81–102.
Nakabayashi, T., 1968, Shin ginko jitsumu koza [New Lectures on Banking Practice], vol. 15.
Tokyo, Japan: Yuhikaku.
Nihon ginko [Bank of Japan], ed., 1974, Keizai Tokei Nempo [Economic Statistics Annual],
Tokyo, Japan: Nihon ginko.
——, 1977, Waga Kuni no Kin’yu Seido, dai 8 ban [The Financial System in Our Country, 8th
ed.], Tokyo, Japan: Nihon ginko.
Nihon sekiyu seisei [Japan Petroleum Refining], ed., 1982, Nihon Sekiyu Seisei 30 Nen Shi
[A 30-Year History of Nisseki], Tokyo, Japan: Nihon sekiyu seisei.
Nihon senshu kyokai [Japan Shippers Association], ed., 1970, Kaiun Tokei Yoran [Survey of
Ocean Shipping Statistics], Tokyo, Japan: Nihon senshu kyokai.
Nihon shoken keizai kenkyu jo [Japan Securities Research Institute], ed., Volumes for
1955–75, Kabushiki Toshi Shueki Ritsu [Rates of Return on Common Stock], Tokyo, Japan:
Nihon shoken keizai kenkyu jo.
Nihon tankaa kyokai [Japan Tankers Association], ed., 1980, Nihon Tankaa 50-Nen no Ayumi
[The 50-Year History of the Japan Tanker Association], Tokyo, Japan: Nihon tankaa kyokai.
Nihon zosen kogyo kai [Japan Shipbuilders Association], ed., 1980, Nihon Zosen Kogyo
Kai 30-Nen Shi [A 30-Year History of the Japan Shipbuilding Industry Association], Tokyo,
Japan: Nihon zosen kogyo kai.
Nippon keizai shimbun sha, ed., 2001. “Nikkei zaimu deeta [Nikkei Financial Data]”
(NEEDS). Available online for faculty at University of Tokyo Economics Department..
Okaniwa, H., 1981, Shintei: Kaiun no Gaiyo [Revised Edition: An Outline of Ocean Shipping]
Tokyo, Japan: Seisan do.
Okura sho [Ministry of Finance], ed., 1965, 1966, Ginko Kyoku Nempo [Annual Report of the
Banking Bureau], Tokyo, Japan: Okura sho.
——, Volumes for 1965, 1970, 1975, and 1995, Hojin Kigyo Tokei Nempo [Corporate Firm
Statistics Annual], Tokyo, Japan: Okura sho.
Patrick, H.T., 1972, “Finance, Capital Markets and Economic Growth in Japan,” in A.W.
Sametz, ed., Financial Development and Economic Growth: The Economic Consequences of
Underdeveloped Capital Markets, New York: New York University Press, 109–139.
——, 2001, “From Cozy Regulation to Competitive Markets: The Regime Shift of Japan’s
Financial System,” Columbia Business School Center on Japanese Economy and
Business, Working Paper No. 186, April.
Pempel, T.J., 1978, “Japanese Foreign Economic Policy: The Domestic Bases for International Behavior,” in P.J. Katzenstein, ed., Between Power and Plenty: Foreign Economic
Policies of Advanced Industrial States, Madison: University of Wisconsin Press, 139–190.
Rajan, R.G. and L. Zingales, 1995, “What Do We Kow about Capital Structure? Some
Evidence from International Data,” Journal of Finance, 50, 1421–1460.
Sangyo keikaku kaigi [Industrial Planning Conference], ed., 1965, Kokusai Sekiyu Josei to
Nihon no Sekiyu Seisaku [The International Petroleum State of Affairs and Japanese Petroleum
Policy], Tokyo, Japan: Heibon sha.
Sanko kisen, K. K., 1961, Securities Filings, Tokyo, Japan: Sanko kisen, K.K.
Sanko kisen, K.K., 1966, Securities Filings, Tokyo, Japan: Sanko kisen, K.K.
Sanko kisen, K. K. [Sanko Shipping], ed., 1968, Shagyo Kaiko Roku [Records of the Company
Business], Tokyo, Japan: Sanko kisen.
Senpaku ho [Ship Act] of 1899, Law No. 46.
Shukan toyo keizai [Toyo keizai Weekly], ed., 1975, Kigyo Keiretsu Soran [Overview of Firm
Keiretsus], Tokyo, Japan: Toyo keizai shimpo sha.
Smith, A., 1984, “The Japanese Foreign Exchange and Foreign Trade Control Law and
Administrative Guidance: The Labyrinth and the Castle,” Law and Policy in International
Business, 16, 417.
The Loan Market in High-Growth Japan
205
Tokyo shoken torihikijo, ed., 1985, Tosho Tokei Nempo [TSE Statistical Annual], Tokyo, Japan:
Tokyo shoken torihikijo.
Tonen tankaa [Tonen Tanker], ed., 1979, Tonen Tankaa 20-Yen Shi [A 20-Year History of Tonen
Tanker], Tokyo, Japan: Tonen tankaa.
Tsusho sangyo sho [Ministry of International Trade & Industry], ed., 1990, Tsusho Sangyo
Seisaku Shi, Dai 6 Kan [A History of Trade and Industry Policy, Vol. 6], Tokyo, Japan: Tsusho
sangyo sho.
Ueda, K., 1994, “Institutional and Regulatory Frameworks for the Main Bank System,”
in M. Aoki and H.T. Patrick, eds., The Japanese Main Bank System: Its Relevance for
Developing and Transforming Economies, Oxford, UK: Oxford University Press, 89–108.
Un’yu sho [Ministry of Transportation], ed., 1980, Nihon Kaiun no Genkyo [The Current
Status of Japanese Ocean Shipping], Tokyo, Japan: Nihon kaiji koho kyokai.
Un’yu sho, ed., 1986, Gaiko Kaiun no Genkyo [The Current Status of Overseas Ocean Shipping],
Tokyo, Japan: Nihon kaiji koho kyokai.
Wallich, H.C. and M.I. Wallich, 1976, “Banking and Finance,” in H. Patrick and H.
Rosovsky, eds., Asia’s New Giant: How the Japanese Economy Works, Washington, DC:
The Brookings Institution, 249–315.
Weinstein, D.E. and Y. Yafeh, 1998, “On the Costs of a Bank-Centered Financial System:
Evidence from the Changing Main Bank Relations in Japan,” Journal of Finance, 53,
635–672.
Zenkoku chiho ginko kyokai [National Association of Regional Banks], ed., 1960–1975,
Kin’yu Ginko Shotokei [Statistics for Finance and Banking], Tokyo, Japan: Zenkoku chiho
ginko kyokai.