Question

Describe the Main Bank System and its contribution to the Post-WWII economy of Japan.

Answer 1

The Main Bank System and its contribution to the Post-WWII economy of Japan:

There is no disputing that Japan had a tremendous run of industrialization and economic growth in the postwar era, up through the peak of the bubble economy in 1989. During at least the latter part of this period, it was felt by many observers in other countries that Japan had discovered a unique and culturally Japanese model of economic and financial operation which was superior to those of traditional Western market-based economies, and the structure and organization of the banking system—in particular its powerful and prestigious large city banks and their “main bank system”—was cited as a key advantage of this model.

Obviously with hindsight the Japanese model as it was then understood was far from the perpetual-motion machine that it appeared to be, and Japan’s economy, having gone (in the words of Gordon) from “revival to miracle to mess,” has been largely stalled for over two decades. But the main bank system continues as a key part of Japan’s economic and financial landscape. What is this system, and what were its origins? What if anything did the main bank system contribute to Japan’s rapid growth during the years of the economic miracle, and was it ultimately responsible for the end of the bubble economy?

Characteristics of a Main Bank

Almost all mid-sized and larger Japanese corporations are said to have a "main bank," but the systematic definition of exactly what a main bank is neither clear-cut nor consistent among authorities. Nevertheless, a Japanese company can immediately self-identify its own main bank (and frequently those of its competitors and suppliers as well), and the term is widely recognized and used by bankers, corporates, and regulators. What is a main bank?

General indicia of a main bank relationship may include any or all of the following characteristics:

• The main bank is a bank of long-standing relationship with the corporate, potentially its initial banking relationship
• The main bank is a part of the corporate's historical or current group (either zaibatsu or later kigyo shudan)
• The main bank has the largest share of lending to the company, or in some cases has the largest share of lending excluding long-term loans from the Japan Development Bank (JDB) or the long-term credit banks
• The main bank has the largest share of deposit balances, especially non-interest- bearing deposits) from the subject company, and clears and processes the company's transactions with other firms.
• The main bank guarantees the foreign bonds of the company, or serves as a trustee for its domestic or foreign bonds. During the miracle years we discuss, city banks could not underwrite bond issues; since 1993 when this was changed main banks through their securities arms have taken over the top rank in bond underwriting as well.
• The main bank is an important or principal (but not controlling) shareholder in the firm.
• The main bank provides management support to the company, via temporarily seconded staff, advisory services, or by placing retiring employees of the bank in executive positions or directorships of the company.

The main bank relationship, while not a legal one, is an intentional mutual relationship explicitly entered into on both sides, and not simply an evaluation of which of a variety of corporate counterparts best meets the criteria above. However, as we describe below the operations and duties of the parties under this system are not explicit and their interpretation varies.

Structure of the Japanese Banking System

As we are primarily concerned with the period of the Japanese "economic miracle," roughly defined as between the end of post-WWII occupation in 1952 and the bursting of the "bubble economy" in 1990, we will focus on the structure of the banking system during that time, including as necessary some discussion of origins of the system as well as changes post the bubble era.

The Japanese banking system during the miracle years was primarily a three-tiered one.

Long-term Credit Banks and Trust Banks

Long term credit (at various times determined as over two, three, or five years) was extended by the long-term credit banks (Industrial Bank of Japan [IBJ], Long-term Credit Bank [LTCB], Nippon Credit Bank [NCB], and the state-owned Japan Development Bank [JDB]) and to a lesser extent by the trust banks (many of which were themselves affiliates of city banks or kigyo shudan). The long-term credit banks were not permitted to take deposits, and in compensation during the miracle years had a monopoly on the issuance of financial debentures, and much more freedom to issue bonds. Note that the long-term credit banks were more important to the development of the economy than were the trust banks not simply by virtue of their greater size, but mainly because they had the ability to perform their own credit analysis and to some extent monitoring, while the trust banks were known to be individually weak in this area, and so relied upon others to perform these functions, after which they would lend individually or take part in a syndication

Post-bubble economy, the long-term credit banks are almost irrelevant to Japan's financial sector: IBJ has been integrated into Mizuho's city bank unit, and LTCB and NCB are small and moribund. Whether by cause or consequence of their slide into desuetude, the regulatory regime for city banks and long-term credit and trust banks is now virtually identical.

City Banks

Short term loans for working capital purposes, deposits, and transaction and other fee- based services for most companies of significant size (e.g., listed companies, companies with international operations, companies which are part of major kigyo shudan) were handled by the city banks (Mitsubishi Bank, Sumitomo Bank, Mitsui Bank, Bank of Tokyo, Taiyo-Kobe Bank, Sanwa Bank, Tokai Bank, Fuji Bank, Dai-ichi Bank , Nippon Kangyo Bank, Daiwa Bank, Kyowa Bank, and Saitama Bank being the major ones), sometimes also termed the commercial banks. City banks were headquartered in the largest cities (primarily Tokyo, Osaka, Nagoya, and Kobe), but could have national branch networks.

The city banks, unlike the long-term credit and trust banks, were funded mainly by both retail and corporate deposits, with an emphasis on demand deposits and short-term time deposits. During the period under scrutiny, city banks' deposit rates were strictly regulated, and they were unable to take time deposits for periods in excess of one or two years.

The city banks were (and continue to be) the main banks of choice for most sizable corporates. Unlike the trust and regional banks, they were able to engage in the extensive monitoring which being a prime lender requires (although later experience indicates that even the city banks were less than adequate at this function), and unlike the long-term credit banks the city banks were able to offer a full suite of deposit, international, and bond guarantee products to corporates.

As the spiritual heirs to the zaibatsu holding companies, the city banks also had a large historical advantage, given that "it is improbable that the main bank will be displaced from its lead position by another bank [and], it is virtually impossible if the client corporation and its main bank are members of the same kigyo shudan."This large and dependent "installed base" has helped to keep new city banks from forming and taking away business, except through the merger of smaller or failing city banks.

However, it should be noted that although the overwhelming majority of main bank relationships were with city banks during the miracle years, IBJ was a lone exception in that it had numerous clients for which it was the main bank. Some of these clients were formerly government-owned enterprises such as Nippon Steel, and some according to Packer were legacies of IBJ's championing of the "new zaibatsu" groups during the pre-war era, which were independent of the old-line combinations and which did not have their own captive banks inside the group.

While the city banks were at a disadvantage in extending long-term loans to their clients, the debenture-issuing activities of long-term credit banks gave rise to a classic example of Japanese cooperation, in which the city banks became the major buyers of long-term credit bank debentures (an average of 45% between 1956 and 1965), replacing the former primary buyer the Ministry of Finance, in return for which an unwritten rule specified that the long-term credit banks would extend credit to each bank's customers of approximately double the value.

Regional Banks

Almost all of the remaining banks in Japan were strictly regional banks, with offices limited to one municipality or prefecture and a predominantly retail and municipal government deposit base. Regional banks would rarely be main banks except to the smallest of companies, and so had little access to opportunities to originate loans to substantial corporations.

Instead, they would join the city banks in syndicates, relying on the credit analysis and monitoring capabilities of the main bank, and sharing in the interest income — but not the more coveted fee income or large compensating balances. In addition, the regional banks were encouraged to buy long-term credit bank debentures, in return for which "they would be made the agents of long-term credit bank loans to customers in their region. “This delegated origination has been described as the "convoy system."

Origins of the Main Bank System

The main bank system became (and continues to be in many ways) of huge importance to Japan during the economic boom, as bank lending was the major source of corporate funds, and the main banks were the coordinators of its origination, the agents of its monitoring, and the parties responsible for restructuring or otherwise resolving firms which borrowed too much. This is considered to be such a distinguishing characteristic of Japanese finance that Sheard states simply that: “Traditionally large firms in Japan have relied heavily on direct bank finance, rather than equity or bond issues..."

Has this always been so? And if not, how did the main bank system come about? What are its origins?

Despite the perception that dependence on bank lending and the main bank system is a "traditional" feature of Japanese finance, up until the militarization of Japan in the mid-1930s, Hoshi and Kashyap analyze multiple studies which conclude that in fact Japanese companies were primarily funded by equity issuance and bonds, with bank loans less important. The commercial bank loans which were originated during these years were mainly to smaller, unincorporated businesses or to individuals engaged in commerce, with only a small minority of loans made to corporations.

Indeed, the development of Japan as what Ozawa Terutomo terms a society of "bank-loan capitalism" is a phenomenon of the mid 1950s and after, not a consistent feature of the modern Japanese economy.

As for origins, there are several explanations put forth for the genesis of the main bank system:

Tradition and Culture

The traditional explanation highlights the origin of the main banks (ex-IBJ, which is to some extent sui generis) as instruments or captives of the zaibatsu. In this model of the development of the system, the period encompassing the immediate pre-war years through the occupation are treated as an aberration, after which the zaibatsu are re-formed as either keiretsu or kigyo shudan, and the banks resume their former roles.

Scher directly connects the early banking and post-occupation periods: "...many aspects of the main banks' relationships with their client firms predated the introduction of modern banking practices in Japan during the Meiji period..."while Tamaki elides the pre-war years altogether and sees a break in the system only during the occupation, which ends with the symbolic reversion to their original names of Mitsubishi (temporarily Chiyoda), Sumitomo (Osaka), and Mitsui (Imperial) banks in 1952. He further notes that with the dissolution of the zaibatsu honsha (holding companies) the banks were freed from oversight by the controlling families and able to "claim the pivotal positions of former zaibatsu industrial groups."

In this theory of origin, the Japanese business culture has a powerful tropism towards related groupings of companies offering mutual support, which can be disrupted by exogenous shocks but which will then tend to pull back towards a group structure. Bhappu and others have theorized that this leaning is a manifestation of the Japanese family as an organizing principle for society and corporate management. Their model uses the ie (household) of merchant families in the feudal era as a guide to the interlinks of corporate group structures during the expansion era.

Scher also cites the cultural theory as one potential origin of the main bank system, noting from his interviews with bankers that as with other forms of Japanese social organization the main bank system is dependent on "traditional hierarchical and communal relationships, such as giri (obligation), onjoshugi (paternalism), and ie ishiki (family consciousness)" Based on these cultural factors, he distinguishes the Japanese firm from the Western concept of a bundle of discrete legal rights and responsibilities, terming it in contrast a "'nexus of implicit relational contracts,' indicative of its character as a high-context communal society.

Certainly this origin theory is attractive in that it fits well with what we know of Japanese society in other fields; however, it fails to explain why these exceptionalist cultural factors were not controlling during the pre-militarization period, and why they reverted so abruptly to type after the years of occupation.

Also, although the city banks were mostly originally affiliated with the zaibatsu, it is not clear that they took on the same role in financing the zaibatsu companies that the main banks later did with the kigyo shudan. Ogura Shinji, in his explanation of the history of the Mitsui bank, highlights the work of Morikawa Hidemasa which quotes the early twentieth century executive director of the Mitsui Bank head office to show the opposite:

'Deposits should be kept from growing, insofar as possible without offending customers. Funds should be safely tied up in loans without becoming uncollectable, even at low interest rates.' The policy had been adopted, he explained, in response to outside criticism that 'Mitsui is accepting deposits from others and operating a variety of its own enterprises. It is, however, extremely unsound and improper for it to invest in its own enterprises funds received on deposits from others.'

Morikawa concludes, and Ogura concurs, that:

The zaibatsu with banks were cautious...in using their own banks. For a zaibatsu to sink its own deposits into its enterprises and then not to be able to recover them would do devastating damage not only to the reputation of the bank but also to the family's honor. Reflecting the zaibatsu families' fear of this outcome, the zaibatsu with banks avoided as much as possible investing their bank's funds in their own enterprises, a restraint that became particularly marked after 1900.

The Wartime Economy

Although the Japanese banking system prior to the militarization of society during the immediate pre-war period was relatively autonomous—at least within the confines of the zaibatsu system for banks with such affiliations—it was also relatively undeveloped. Banks generally lent against marketable securities or other ready collateral, and had little ability to do independent credit analysis or engage in ongoing monitoring of borrowers. The fragmented banking market worked against scale and the accumulation and channeling of large capital sums.

With the advent of the military buildup and accompanying aggressive industrial modernization policy, commensurately large-scale lending was required. A combination of bank failures following the great crash of 1927, higher capital requirements imposed by the government, and government-encouraged mergers reduced the number of Japanese banks from 1422 in March, 1927 to 194 at the end of 1941 and only 61 at the end of 1945. Under the Konoe administration, these remaining, larger banks were subject to the Ordinance for the Funds Management Bank and Other Institutions, the Outlines of Basic Financial and Monetary Policies, and the new Bank of Japan Act, which grouped them together into the Ordinary Bank Control Association and essentially placed their lending operations under direct government control.

As part of the drive to efficiently meet the needs of companies in munitions and other essential industries, in January 1944 the Designated Financial Institution System for Munition Company Financing was introduced. Under this system, each company was required to explicitly identify (or be assigned) a single main (manager) bank, which would henceforth be responsible for meeting its credit needs. These main banks were in the vast majority of cases (202 of 217 in total) either one of nine large city banks or IBJ. The main banks were expected to form syndicates in order to quickly meet any credit requests, and the funds were required to be provided in the form of bank loans rather than bonds. Thus this theory of origin accounts for most of the salient points of the main bank system.

Structural Changes and Phases of Development

A third potential explanation for the rise of the main bank system is a structural one. Under this theory, it was inevitable that Japan's rapid industrial expansion would outpace the clubby and limited pre-militarization banking system, and force it also to go through consolidation and rationalization. Japan's rather boutique family and group-dominated banking system was simply not robust enough to finance and control a modern industrial sector.

Ozawa presents Japan's viable options for financing its rapid growth as either borrowing from abroad (i.e., running a current-account deficit) or using the central bank and commercial banking systems to create additional credit internal to the country. In his analysis, Japan was pressed into the second model by a limited appetite on both sides for Japan to borrow extensively abroad, and by the inability of the primary equity capital market to keep up with the demand for capital.

Scher in his analysis also lays out (but does not necessarily favor) this argument, which he terms a "historical view." He notes that due to the large scale of wartime and postwar industrial projects, and the limitations on how much of their capital (rather than assets) banks could devote to a single company or project, firms came to have relationships with 20 or 30 banks. When coupled with the greater prominence of the long-term credit and trust banks in lending, and the government's active role in selecting large new industries for rapid development, Scher's development of this theory of origin argues that the advent of a main bank system whereby one party would take the lead in coordinating the required large lending syndicates and monitoring the performance of the firm was 11 simply a natural consequence of the need for efficiency and coordination, and not an innately Japanese development based on culture.

There appear based on the arguments to be solid grounds to believe that the structural origin theory of the main banks has at least some contributing merit; however, it does not explain why the postwar main bank system looks—at least in terms of its major players and their relationship to clients—very similar to that of the early zaibatsu era. Nor does it explain why the system changed to be securities capital-markets dominated during the initial period of pre-war rapid industrial growth.

As none of the three major theories of origin can completely explain the emergence of the main bank system in Japanese finance, and as they are not mutually exclusive, it seems probable that an accurate assessment of the genesis of the system will contain elements of all three: The structural growth of Japan's corporate sector and the outsized financing required during the war and reconstruction years demanded a larger and more professionalized banking system, but one which could work in harmony with both the government-driven system of administrative guidance and the group-centric organization of Japanese corporations and the consensus-seeking behavior of their managements. The explicit

assignment of main banks during the war, and subsequently the abrupt vanishing of the coordinating zaibatsu honsha during the occupation, left both a vacuum of control over the investments and activities of the far-flung former zaibatsu companies, and a need for intra-group and government/commercial coordination to replace the dismantled mechanisms of direct control, both of which roles the main banks eagerly filled.

The Main Bank System in Operation

The operation of the main bank system is complex and somewhat opaque, given that almost none of the major functions which separate a main bank from an ordinary bank are contractual and thus easily pinned down.

• Lending and Deposit Taking
• Equity Investment and Cross-shareholdings
• Management Support
• Monitoring
• Restructuring and Distress

The Impact of the Main Bank System on Japanese Economic Growth during the Miracle Years

The Japanese government made a clear decision not to pursue the bond-market focused financial sector that the US occupation forces wished them to implement, and instead turned to a bank-led system. The easiest explanation for this choice is that policy-makers were overridingly concerned with a stable and predictable macroeconomic environment minimal reliance on external funding and an increased current account deficit and the ability to direct funds to targeted industries seen as essential for economic development.

As previously discussed, insulating the market from international funding allowed Japan to create sufficient credit domestically through the banking system, and allowed for a more predictable availability of credit not subject to the vagaries of world markets, albeit one requiring a certain measure of domestic financial repression. Parker sees this decision as a quid pro quo generally understood by all parties in Japan, “an implicit social contract until late in the [1970s] whereby households accepted low yields on savings deposits and poor housing quality in exchange for a system that generated rapid growth in GNP and household incomes.” This is a feature of the banking system in general, but not specific to the main banks.

However, the goal of promotion of targeted industries in a consistent manner calculated to promote maximum macroeconomic stability was made considerably easier by the creation of the main bank system. Using the mechanism of administrative guidance, the government was able to select favored industries and projects for banks to fund without in general making direct credit allocation decisions. Such a system required the analytical capacity and monitoring ability of a main bank, which would be responsible for performing due diligence on the specific borrowers and projects, assembling the lending syndicate, and restructuring the company in the event of distress.

Of the four initially-targeted sectors in the 1950s—iron and steel, electric power, marine shipping, and coal mining—the latter two were (correctly) considered poor credit risks by the main banks by the 1960s, and private lending to them fell off sharply. This shows the main bank role as an information intermediary acting for the benefit of the government’s economic goals.

From the corporate perspective, Japanese firms did enjoy preferential and lower cost access to capital in the 1970s and 80s, but by 1990 this disparity had disappeare — coterminous with the end of the bubble economy. While it is difficult to quantify the impact that the lower cost of capital of firms had on the Japanese economic miracle, given the vociferous envious complaints made by their international competitors it is reasonable to think that it was material.

Interestingly, the multiple studies reviewed by Weinstein and Yafeh show that firms with a main banking relationship, although they do have better access to capital, were not any more profitable—nor did they grow any faster—than non-main-bank clients, and their cost of capital was not provably lower. This implies that whatever impact the main bank system had on the cost of bank loans in Japan was consistent across the board, and not enjoyed only by their clients.

Did the Main Bank System Cause The Bubble?

Japan’s asset bubble was certainly facilitated by banks, and overlanding with poor credit analysis resulted in the forced merger of many of the oldest and most prestigious names in the industry. But was the main bank system to blame?

Both Ogura and Hoshi/Kashyap make an interesting commentary on the bubble financing. In essence their collective observation is that the government’s loosening of offshore bond market access in 1980 after the oil price shock had the effect of driving down the profits to be had from relationship banking to large corporate groups, as the borrowers now had access to large pools of money elsewhere. This forced banks to look for other borrowers, and in the run-up of property and stock prices secured lending looked like an easy and profitable diversification which would free the banks from having to take thinner and thinner spreads from lending to the major corporate groups. Sadly, the banks did not have the same degree of information about their new borrowers that they did about the old ones, nor did they have the same government and group support to help them work out loans in distress.

In the end, it would appear that both borrowers and lenders tried to renegotiate the terms of their implicit main bank system understandings so as to do better elsewhere, and both wound up worse off for it. It may be rather this attempt to abandon the main bank system which produced the bubble, and its aftermath.