Deregulation and competition in Japanese banking.

Allen B. Frankel and Paul B. Morgan, of the Board's Division of International Finance, prepared this article. An earlier version was presented at the Conference on Bank Structure and Competition sponsored by the Federal Reserve Bank of Chicago, May 6-8, 1992.

In the past fifteen years, the Japanese financial system has been the focus of a series of liberalization measures aimed at modernizing the intermediation process and improving the efficiency of Japanese corporate finance. These policy developments have stemmed largely from pressures external to the domestic banking sector itself, such as the substantial increase in government debt as a result of changes in the flow of funds in Japan after the OPEC oil shocks; increased competition in international financial markets; and a new emphasis on bank capital management. These influences are engendering fundamental changes in the system of finance in Japan, the objectives on which Japanese bankers place importance, and the competitive position of the banking system vis-a-vis the international sector and the domestic nonbank financial sector. This article provides an overview of the forces that have induced changes in the Japanese banking system and attempts to discuss these changes in the context of the ongoing financial reform effort in Japan. (For an overview of the process of change in the Japanese banking system, see chart 1 .)

EVOLUTION OF JAPAN'S FINANCIAL SYSTEM

For more than three decades after World War II, the Japanese financial system was highly regulated and remained largely isolated from the rest of the world. First, Japanese monetary authorities administratively determined all interest rates, including those on bank deposits and loans as well as coupon rates on government bonds and bank debentures. Second, various types of banking firms and other financial service firms were legally and administratively confined to a specified range of activities (see table 1 for an overview of the current limitations). Third, capital markets were repressed by guidelines, such as strict collateral requirements for the issuance of corporate bonds. Consequently, few alternatives to bank financing existed for even the largest firms.

Internationally, exchange controls and limits on foreign activities restricted the access of Japanese financial firms to foreign financial markets. Until the 1970s, only a limited number of financial firms, including foreign banks granted special concessions, were licensed to engage in foreign exchange transactions. This system of foreign exchange controls and licenses effectively separated Japanese domestic markets from financial markets abroad.

Domestic

The OPEC oil shock in 1973 signaled a turning point in the operation of the domestic Japanese financial system. The contractionary effect of the oil price increases ushered in a period of sizable government deficits, which resulted in a buildup in the outstanding stock of Japanese government bonds. Outstanding government debt rose as a percentage of GNP from 5 percent in the early 1970s to 40 percent in the early 1980s. [1]

The buildup of Japanese government debt forced changes in the relationship between the Bank of Japan and members of the underwriting syndicate for government bonds. Traditionally, syndicate members purchased newly issued government bonds with the understanding that the debt could later be sold to the Bank of Japan at a price that would ensure the avoidance of losses on the original bond purchase. This arrangement involved ex post compensation of losses by the Bank of Japan as fiscal agent for the Japanese government.[2]

With the surge in government bond flotation, Japanese authorities concluded that the Bank of Japan could no longer guarantee the repurchase of the growing volume of government debt without jeopardizing its ability to exercise monetary control. [3] The elimination of the agreement between the Bank of Japan and the underwriting syndicate is evocative of the 1951 accord between the Federal Reserve and U.S. Treasury from which the Federal Reserve received discretionary authority for monetary policy.[4] Facing financial strains in the absence of the Bank of Japan's backstop facility, the syndicate successfully lobbied in 1977 for permission to develop a secondary market. Syndicate members viewed the secondary market as a mechanism for off-loading their holdings of seasoned bonds when they were called upon to purchase new debt. The Japanese authorities allowed the secondary market to develop gradually, but the syndicate process did not function smoothly during the period of large net issuance of government debt. As Suzuki notes, over two periods in 1981 and 1982, in fact, syndicate participants publicly refused to carry out their underwriting responsibilities.

The introduction of secondary market trading marked the end of the postwar era of absolute administrative control of Japanese interest rates. Japanese banks and securities companies were given formal authorization for a market-rate funding mechanism for their bond purchases through the use of short-term repurchase agreements; and as a result, the gensaki market came into being. The primary investors in the gensaki market were Japanese nonfinancial companies, which often used the market to exploit arbitrage opportunities to borrow at administratively determined rates and invest at higher, market-based rates. Japanese securities companies used the gensaki market to finance their inventories of bonds and, in the process, greatly expanded their overall share of the underwriting of Japanese government bonds. Overall, in terms of adapting their financial structures to accommodate large government debt issues, the Japanese made choices similar to those made in other countries: that is, to tolerate limited disintermediation in the interest of buttressing the arrangements used in government debt underwriting.

In the 1970s, however, the Japanese authorities were reluctant reformers of their highly segmented financial system. For example, until 1979, Japanese banks were not permitted to issue yen-denominated certificates of deposit (CDs) as a source of market-based funding. The post-1984 surge in the growth of the Japanese domestic money markets paralleled changes in the regulation of the access of Japanese banks and nonfinancial corporations to international markets (chart 2). These changes liberalized the regulation of Japanese banks' access to international markets and thus had an important influence on domestic market reform.

Through most of the postwar period, the Bank of Japan relied heavily on the administration of its credit facilities for bank borrowers as a mechanism enabling it to fulfill its monetary policy responsibilities. The Bank provided a continuing source of credit to Japanese commercial banks for funding bank asset expansion--a so-called overborrowed position. In return, the Bank conditioned the availability of such funds on an adherence by the banks to guidelines concerning their individual lending behaviors. This practice of "window guidance" of domestic credit activities persisted until 1991, but it is believed to have become more consultative than directive in the later years.

Besides giving administrative guidance regarding credit policies, the regulatory authorities restricted the city banks' own efforts to raise capital. [5] In 1983, the Japanese authorities first began to encourage city banks to take advantage of market-determined share prices. The behavior of share prices of Sumitomo Bank since 1977 provides an example of the effects of this policy change (chart 3). Before 1983, all city bank shares traded in the same narrow range shown for Sumitomo Bank. After 1983, the share price of Sumitomo Bank rose, as did that of other city banks. The city banks responded to the surge in their share prices by making large primary stock offerings at the higher market prices. The proceeds of these share offerings were earmarked to finance investments in computer facilities and overseas branch networks. Gradually, however, share prices of city banks have stopped moving in lockstep.

International

The Foreign Exchange Law of 1980 marked a watershed in Japanese financial policy: It reversed the presumption that all international financial transactions by Japanese residents were subject to government control unless explicitly authorized. In particular, it ushered in a period in which Japanese commercial banks ceased operating overseas solely to finance the growing share of world trade accounted for by Japan's exports of finished goods and imports of raw materials. That is, Japanese authorities accepted the need for flexibility in the overseas activities of Japanese banks by enabling the banks to respond to the increasingly sophisticated financing requirements of their internationally active corporate customers.

Two measures contained in the Foreign Exchange Law of 1980 proved to be of particular importance in integrating Japanese domestic money markets with international markets. These were the authorizations (1) for Japanese banks to borrow and lend foreign currencies freely (both at home and abroad), subject only to prudential guidelines, and (2) for Japanese companies to finance themselves abroad through borrowing denominated in foreign currency.

Throughout the early 1980s, the Japanese authorities further reformed their regulation of Japanese residents' participation in international markets. The cumulative effect of these liberalization measures was the opening up of important channels of intermediation through which Japanese interest rate conventions could be circumvented by transactions routed through offshore financial markets. As a result of the arbitrage opportunities generated by these liberalizations, yen interest rates in domestic and international markets became much more tightly linked: The standard deviation of the differential between three-month Euroyen and domestic gensaki interest rates fell 88 percent between the 1975-80 period and the 1981-85 period (from 190 basis points to 23 basis points). Further, reforms such as the 1984 abolition of "swap limits" for spot transactions had the effect of repatriating yen money markets from foreign locations, as evidenced by the post-1984 surge in the volume of domestic money market transactions, which was discussed earlier.

Throughout the late 1980s, the Euroyen market expanded sharply in response to a strong surge in cross-border lending of yen to the Japanese nonbank sector by the offshore offices of Japanese. banks. This form of bank lending had the advantage of not being covered by Bank of Japan window guidance, although activities in the Euroyen market were monitored. Japanese commentary suggests that the accommodation of such borrowing was a component of a strategy for financial liberalization in which offshore experience was used to inform the implementation of domestic financial reform.[6]

ELEMENTS OF FINANCIAL LIBERALIZATION

Two of the most important developments in the process of financial liberalization were the gradual deregulation of interest rates beginning in May 1979 and the change in the corporate client base of Japanese banks as a result of interest rate deregulation.

Interest Rate Reform

The growth of the gensaki market, along with the introduction in 1980 of the chnkoku (government bond mutual fund) market, caused a disintermediation of funds from the banking sector as corporations rapidly sought to capture the higher yields available in markets offering unregulated interest rates. The consequent funding pressure on the Japanese banks caused by disintermediation led to the introduction of negotiable CDs offered by commercial banks at liberalized interest rates. At the outset, the restrictions set for CD issuance narrowly limited the maturity, minimum denomination, and total funding ceiling for each bank. These restrictions have been cased over time but not eliminated; for example, Japanese banks still cannot issue floating-rate deposits--a potential source of funding that would more closely mirror standard loan-pricing terms in Japan. A chronology of the liberalization of interest rates on Japanese deposits is shown in table 2.

The succession of regulatory reforms that followed the authorization of CDs in 1979 has caused the interest costs of banks to become increasingly sensitive to movements in market interest rates. One change that was especially important in this process was the introduction of money market certificates (MMCs) in 1985, with interest rates linked by formula to open market rates on designated instruments. A steady increase has occurred in the percentage of Japanese bank deposits that have liberalized, market-based interest rates (chart 4). The increases reflect the relaxation of restrictions regarding the minimum denomination and maturity of deposits as well as the attraction of liberalized deposit rates during the period of rising yen interest rates between 1988 and 1990. In 1991, the share of liberalized deposits for the city banks fell in response to a decline in the attractiveness of such deposits compared with that of assets with administratively determined interest rates, such as postal savings accounts.

As a result of their increased reliance on market-rate funding during the period of sharply rising short-term interest rates, the banks' pretax profits declined sharply (chart 4). In response, the banks adjusted their methods of determining their prime rates for short- and long-term loans (in 1989 and 1991 respectively), so that the rates would more closely track actual funding costs. Formerly, rates on bank loans had been based on the official discount rate of the Bank of Japan (for short-term loans) or the rate paid on debentures issued by long-term credit banks (for long-term loans). Not unexpectedly, the banks encountered resistance to their revised loan-pricing formulas from large corporate customers. In fact, some of the deterioration in corporate liquidity since the adoption of the new lending rates could be associated with corporate customers' choosing to retire bank loans rather than to roll over credits priced under the revised formulas (see chart 5).

By regulation, there are restrictions on the maturities of deposits that Japanese banks can issue. City banks and other ordinary commercial banks had been limited to two-year time deposits until November 1991, when the limit was raised to three years. As noted previously, the city banks cannot issue floating-rate deposits. However, despite the short-term nature of their liabilities, the city banks steadily extended the maturity of their loan portfolios: The percentage of loans with terms longer than one year grew from 33 percent in 1980 to 57 percent in 1991. Such loans are made over-whelmingly on a floating-rate basis, with the rate reset periodically based on banks' posted long-term interest rates.

The interest rate risk associated with this maturity mismatch became especially harmful to the banks between 1989 and 1991, when increases in short-term interest rates resulted in an inverted yield curve in the yen market. Over this period, Japanese banks made limited use of interest rate derivative products to manage their individual interest rate exposures. Recently, however, the banks have begun to increase their use of instruments such as interest rate swaps and futures; for example, the reported volume of yen-interest-rate swap transactions exhibited particularly strong growth in 1991.

Historically, Japanese banks have been able to raise effective loan yields above posted lending rates through the maintenance of compensating deposit balances by loan customers (chart 6). Under compensating-balance arrangements, corporate borrowers hold interest free (or low interest) deposits, either as a condition of the formal loan contract or to maintain a "healthy, stable relationship" with their banks. According to data from an annual survey by the Japan Fair Trade Commission, the reliance on these compensating balances for loans to small businesses declined steadily throughout the 1980s, from 45 percent of surveyed loan contracts in 1980 to 26 percent in 1990. The survey covers only companies with less than 100 million in capital, which now account for approximately 70 percent of total city bank lending. In view of the rising share of small business lending as a portion of the city banks' total loans, the overall reduction in loans with compensating balances as a percentage of banks' corporate loans could be considerably smaller. For those loans with compensating-balance requirements, the average rate maintained in 1990 was approximately 20 percent.

Despite the financial benefits provided by compensating balances, these arrangements have become less attractive for large Japanese banks. The deposit balances held under these agreements artificially inflate asset levels by increasing the book value of loan portfolios to a level above that of the funds actually extended. During the mid1980s, a period of capital abundance, the consequent asset inflation was not problematic. However, since the late 1980s, banks have begun to view capital as an increasingly scarce commodity and have tended to turn away from practices that increase the leveraging of their capital positions.

The process of interest rate deregulation in Japan has caused a rationalization of the banking business in terms of raising funds and extending credit. As the importance of the administratively determined interest rate structure fades, banks will face an increasingly competitive environment. The fallout from the increase in bank competition has been staggered throughout the period of gradual elimination of interest rate restrictions. As discussed earlier, the Japanese city banks have remained ahead of the smaller regional and cooperative banks in their share of deposits paying liberalized interest rates (chart 4). Yet, as the minimum denomination of money-market-related deposit instruments continues to decline while deregulation proceeds, the effect on the smaller institutions, whose small business clients and retail customers will then be able to gain access to the instruments, will become stronger (see table 2).

Corporate Financing Developments

Historically, Japanese city banks maintained close relationships with the largest Japanese corporalions. As a byproduct of numerous financial developments, including the introduction of alternative sources of corporate finance, these firms have become less dependent on banks for theft financing needs. City banks have responded to these changes by placing a greater emphasis on developing relationships with small and medium-size businesses, which previously had been financed primarily by regional banks and smaller credit cooperatives (shinkin banks) as well as by extensions of interfirm trade credit by larger Japanese companies. As a result, over the 1986-90 period, the share of city bank loans made to large Japanese corporations-- those with more than Y100 million in capital-- declined nearly 20 percentage points, to only 30 percent.

A comparison of the overall sources of finance for Japanese corporations over the two periods 1981-85 and 1986-90 shows that trade credit accounted for a larger share of corporate finance in the earlier period (18 percent compared with only 5 percent in the later period) (see chart 7). The shift in bank relationships toward smaller firms accounts for much of the reduction in trade credit extended, while the increased reliance on direct corporate financing by large corporations explains the reduction in the overall share of bank credit as a source of financing.

The increased importance of domestic and international securities markets as sources of funds for large Japanese companies is reflected in the nearly 30 percent share of corporate funds raised between 1986 and 1990 through the issuance of domestic securities (including commercial paper) and borrowings in international markets. Borrowings in international markets included large amounts of equity-linked bonds issued by Japanese banks and nonfinancial companies in two forms: convertible bonds and straight bonds with detachable warrants. Japanese corporations (inclusive of banks) issued more than $30 billion of convertible bonds denominated in Swiss francs between 1987 and 1990. According to market observers, few issuers hedged the currency exposure of these issues given their view that the securities would inevitably be converted into stock. Japanese corporations believed that convertible issues offered lower costs of equity issuance than direct issuance of equity, perhaps reflecting the Japanese practice of offering new issues at a discount from market price to existing shareholders. The choice of denominating a convertible issue in a foreign currency, such as the Swiss franc, was strongly influenced by the fact that the lower nominal cost of such an issue would result in higher reported current earnings than if the issue were denominated in yen. There is little evidence that the "speculative" nature of such financing choices by various Japanese firms, before the sharp 1990 decline in the Japanese stock market, was factored into market assessments.

Between 1985 and 1991, Japanese private nonbanks increased their outstanding debt issued in international markets nearly six-fold, to a level of $350 billion. For the most part, this choice of financing alternatives was in response to large differentials in the cost of financing that favored international over domestic Japanese markets. In turn, the cost differentials can be traced to particular characteristics of Japanese financial regulations and domestic financing practices. For example, the gap between the rates charged for prime loans by the city banks and those available in the Euromarket have provided an incentive for internationally recognized Japanese firms to search actively for less costly alternatives to domestic bank funding. [7]

The low level of domestic bond issuance by Japanese corporations has been attributed in part to the costs imposed by the "commissioned" bank system in Japan. Under Japan's Securities and Exchange Law, only securities firms are licensed to underwrite corporate bonds; however, banks perform the role of trustees. The fees received by commissioned banks include a charge based on the presumed responsibility of the trustee bank for repurchasing the secured bonds of a defaulting corporation. As Brian Sernkow notes, the Bond Underwriters Association of Japan has estimated that for a Y10 billion bond, the commission fees in the domestic bond market are Y53 million, whereas, in the Euromarket, they are only Y3.5 million.

Japan's "main bank" system for bank finance parallels the character of the commission bank system for corporate debt underwriting. One responsibility of a main bank involves ensuring that a client in financial distress makes debt service payments to other bank creditors. It has been argued that the costs of main bank financing to creditworthy Japanese corporations often exceed the benefits and therefore increase the relative attractiveness of alternative forms of financing. Among these costs are those related to financial monitoring as well as those associated with the financial support of the distressed member firms of a main bank's keiretsu, a grouping of financial and non financial companies.

IMPLICATIONS OF THE BASLE CAPITAL ACCORD

In the early 1980s, central banks and regulatory authorities became increasingly sensitive to the absence of mutually agreed-upon rules for conduct in the international banking business. The international debt crisis raised additional concerns regarding the fragility of the international banking system, in view of the potential consequences of debtor country actions on the financial situations of a large number of internationally active banks. This concern led national authorities in the United States and other industrial countries to press banks to bolster their capital positions relative to the risk exposures they assumed.

In the second half of the 1980s, the international assets of Japanese banks surged dramatically. From 1984 to 1988, the Japanese bank share of international bank assets rose more than 10 percentage points to 38 percent.[8] This increase raised further questions as to whether national banking regulators could successfully induce banks to improve their capital ratios in the absence of barriers against further market penetration by Japanese banks. In turn, the possibility of such protectionist responses was one factor motivating efforts to move to a level playing field for internationally active banks through the adoption of international capital standards.

The task of setting out a framework for the capital standard was assigned to the Basle Committee on Banking Supervision, a group of central banks and bank regulators from the G-I0 countries, whose secretariat is furnished by the BIS. By the end of 1987, the committee had agreed on a framework calling for a common capital definition and a risk-asset weighting scheme rather than a simple leverage ratio. The simplicity of the negotiated framework facilitates comparisons among banking systems. In particular, unlike pre-existing national capital definitions with multiple tiers of capital, the new framework has only two (tier 1 and tier 2 capital). Tier 1 capital consists of only the core constituents of the capital base, namely, equity and disclosed reserves. Tier 2 capital includes supplementary elements, such as subordinated debt and revaluation reserves. While the specific composition of tier 2 was left to national discretion, the committee specified several binding limitations on the inclusion of instruments in tier 2, including a 55 percent discount on uurealized gains on securities holdings and a limit on includable subordinated debt at a level of 50 percent of tier 1 capital.

The Basle Committee also agreed on a timetable for establishing transitional capital adequacy guidelines during the implementation period. Under the Basle framework, internationally active banks must meet an 8 percent minimum standard by the end of fiscal year 1992, of which at least half must constitute tier 1 capital. The Basle framework is a negotiated document, which mirrors the situations of individual banking systems. For example, the inclusion of subordinated debt in tier 2 capital had no operational significance for Japanese city banks during the late-1980s negotiations because Japanese banks were then prohibited from issuing subordinated debt. In contrast, U.S. money center banks had relatively large amounts of such debt outstanding as a result of previous efforts to improve their U.S. regulatory capital ratios. As another example, Japanese city banks possessed substantial revaluation reserves, reflecting unreal- ized gains on their cross-shareholdings in other Japanese companies. In contrast, U.S. banks did not.

The requirements in the capital agreement could have been expected, among their other effects, to induce Japanese banks to slow the growth of their balance sheets. From this perspective, one view of the agreement suggests that Japanese banking authorities considered their situation equivalent to that of Japanese trade negotiators who were pressured to accept orderly marketing agreements as a cost of maintaining continued access for Japanese goods to specific foreign markets. A separate line of reasoning is that Japanese authorities actually sought the agreement as a means of leveraging their own efforts to encourage the financial reform process in Japan. While the two explanations are not mutually exclusive, the latter appears more supportable from a historical perspective because important structural reforms in Japan have historically been portrayed publicly as the undertaking of a sacrifice that was necessary to placate hostile foreigners.[9] However, the expectations of the framers of the Basle agreement were not immediately fulfilled in the sense that the asset growth of Japanese banks was unrestrained. Rather, the severe fluctuations in the Japanese stock market in the late 1980s had important consequences for the capital situations of Japanese banks.

Capital-Raising Activities of Japanese Banks

The capital-raising activities of Japanese banks after the adoption of the new capital framework were phenomenal: From 1986 to 1990, the tier 1 capital of the Japanese city banks increased at an average annual rate of 21 percent. There was a strong positive correlation of the capital growth of the city banks with the upward movement of Japanese stock prices in the late 1980s (chart 8). The correlation is largely accounted for by Japanese banks' capitalizing on the opportunities presented by the favorable terms available for the issuance of new equity and convertible bonds as well as from the realization, through sales, of latent gains on their stock holdings. [10] Since the end of 1989, however, the Japanese stock market has fallen sharply. This decline has not only made it more difficult for Japanese banks to raise additional equity but has also lessened the attractiveness of boosting tier 1 capital through the realization of gains on stock holdings.

Before the negotiation of the Basle Accord, no clear relation existed between the growth rate of Japanese city banks' capital and their domestic assets (chart 9). However, the data suggest a linkage between the two growth rates since the announcement of the accord. The strong upward movement in the ratio of city banks' capital to their domestic assets in the late 1980s and the leveling off of this ratio in the past few years lend support to the view that the relationship reflects a new emphasis by the management of Japanese banks on the role of capital (see chart 10). [11]

The fall in Japanese stock prices sharply curtailed the banks' ability to support further asset growth at their historical rates, as documented by the capital positions of the city banks before and after the 1990 stock market decline (chart 11). Before 1990, the banks were constrained by the Basle Accord limits on the amount of allowable tier 2 capital (represented by the distance AB). Under the final Basle Accord guidelines, tier 2 capital could only be included in total bank capital up to the level of existing tier 1 capital (that is, tier 2 must constitute 50 percent or less of the total). While banks were operating under this constraint, any additions to tier 1 capital through retained earnings or equity issues also raised the ceiling on tier 2 capital by an equal amount (segment BC); therefore, every additional unit of retained earnings (including realization of gains on sales of equity) pulled in another unit of surplus unrealized gains (for a total of two new units of capital) to fund asset expansion.

As a result of the stock market decline over 1990-92, however, the drop in unrealized capital gains has caused tier 2 capital to fall below the Basle Accord ceiling of 100 percent of tier 1 capital (chart 11). After the decline in stock prices, increases in capital through retained earnings or equity financing (from AB to AC) now fund only half as much asset expansion because there is no longer any surplus tier 2 capital. Additions to banks' risk-based asset totals now face much higher effective capital charges compared with those that prevailed before the drop in the stock market. The capital gains that the banks realized from the sale of equity securities and the level of equity financing over 1985-90 were sharply affected by the weakness in the Japanese stock market after 1989 (see charts 12 and 13). For example, the city banks in 1989 issued approximately Y1.8 trillion ($12.5 billion) in common stock but have since refrained entirely from new issuance of common stock.

The capital ratios of the city banks are also sensitive to exchange rate movements. Japanese supervisory rules operate to immunize a bank's total capital but not its capital ratio from exchange rate movements. It is our estimated rule of thumb that an increase (decrease) in the yen-dollar exchange rate of Y10 results in a decrease (increase) of 18 basis points in the aggregate Basle risk-weighted ratio of the Japanese city banks. Nevertheless, the limited movements of the yen-dollar exchange rate in recent years have meant that exchange rate changes have not been an important influence on bank capital ratios.

In response to the erosion of the banks' capital bases caused by the stock market decline, Japanese authorities supported the banks by authorizing the issuance of subordinated debt (counted as tier 2 capital) and perpetual preferred stock (counted as tier 1 capital). As of March 1992, the city banks had issued nearly Y6 trillion of subordinated debt (approximately $45 billion), or about 50 percent of their aggregate tier 2 capital at the end of fiscal year 1991). These innovations offset some of the negative effect of the stock market decline on the capital positions of the banks. However, the replacement in bank capital of "costless" unrealized gains on share holdings by costly subordinated debt and preferred stock had financial consequences. These consequences are associated with the need to raise the share of bank funding met by capital. Overall, the substitution of subordinated debt and preferred stock for deposits is estimated to have reduced the aggregate pretax return on equity for the city banks by 81 basis points, to 9.16 percent in fiscal year 1991.

City bank profits were also negatively affected by the mounting expenses associated with the ongoing support of affiliated nonbank financial institutions that are among the major creditors of so-called bubble companies-firms that have been heavily involved in speculative investments, mainly in real estate. The ongoing financial difficulties of these firms have not been reflected in the increased provisions for loan losses by city banks, partly as a result of Japanese accounting rules that do not provide for the disclosure of probable loan losses. Thus, over the next few years, the earnings of the city banks will continue to be adversely affected by losses on loans to bubble firms in financial distress.

Capital Adequacy and Bank Asset Growth

The aggregate worldwide assets of the Japanese city banks declined 5 percent in 1991--the first yearly asset decline since before World War II. A reduction in interbank placements and deposits more than accounted for this decline in assets. Activity in international markets decreased sharply over 1991 in response to the capital pressures caused by the effect of the fall in Japanese stock prices on the unrealized stock gains of the banks. Total assets of the overseas branches of Japanese city banks fell 13 percent in 1991.

The relationship between the Japanese stock market and the average Basle capital ratio of the Japanese city banks indicates the strength of the financial pressures on the banks caused by changes in Japanese stock prices. This relationship can be shown by using the estimated level of risk-weighted assets as of March 1992 (see chart 14). As the Nikkei Stock Average rises above 20,000 approximately, a declining share of the addition to unrealized gains caused by the stock price appreciations may be included in tier 2 capital (that is, the slope of the line decreases). Point B in the chart reflects a Nikkei level of 22,000. At this point, without any change in asset levels, the average capital ratio of the banks will increase to 8.8 percent. Alternatively, this appreciation will create a capital cushion sufficient for the city banks to increase their aggregate risk-weighted assets 13 percent and still meet the 8 percent capital standard (represented by the shift from B to B*). Similarly, point C reflects the effects of a fall of the Nikkei to 14,000. To compensate for this decline and mainrain an 8 percent capital ratio, the banks have to reduce their aggregate risk-weighted assets 14 percent (movement from C to C*). Alternative means of adjustment available to the banks include increases (or decreases) in various instruments recognized as tier 1 or tier 2 capital, such as subordinated debt or perpetual preferred stock.

Overall, available data and anecdotal evidence suggest that the response of the Japanese city banks to binding capital requirements has been the following. First, the banks cut back on the allocation of capital to support money market activities. Second, pricing objectives for domestic and international credits were increased to improve returns on equity and assets. Third, the banks began to reduce their lending to, and sell off theft stock holdings of, Japanese firms with which they did not expect to have sufficiently profitable long-term banking relationships. Each of these responses emphasizes the heightened awareness of the banks to the importance of maintaining a sufficient risk-adjusted return on capital.

THE OUTLOOK FOR FINANCIAL REFORM

The segmentation currently present in the Japanese financial system is much greater than that in the United States. The Japanese system of specialized banking and credit intermediaries has remained largely unchanged since its reconfiguration after World War II. Traditional Japanese decisionmaking has operated to create a reform process in which the effect of various liberaiization measures must be deftly balanced among all constituencies, thereby demanding that every attempt be made to minimize the costs inflicted on any one sector of the financial system. Such loss-sharing arrangements have operated to preserve a segmented system by necessitating a gradual approach to deregulation in which adequate time must be given to assess accurately the effect of each liberalization measure before the undertaking of further reform. An example of this approach is the fifteen-year process of interest rate deregulation shown in table 2.

Within this framework, regulatory barriers to entry have been claimed as the right of protected firms. [12] The negotiation of compensation for the removal of various restrictions on intersectoral competition has added significantly to the difficulties of reforming the Japanese financial structure. [13]

For example, the financial reform package adopted by the Japanese government in 1992 does not call for the provision of stock brokerage services by Japanese banks, in contrast to the current U.S. regulatory regime, under which U.S. banks are permitted to control firms that provide stock brokerage services. This omission in the Japanese reform package was strongly influenced by concerns about the adverse consequences of bank entry on the competitive positions of the smaller securities firms, which have been undergoing a period of financial weakness. Several of the measures that are contained in the Japanese government's financial reform package are listed in table 3. In most cases, Japanese financial companies currently Operating in various sectors would be allowed to enter into new financial activities only through separate de novo special-purpose subsidiaries. However, the package contains an exception that would permit a bank to acquire a failing securities firm and continue its full-brokerage operations.

In the past, Japanese banks have been encouraged to compete for regulatory privileges. In the 1980s, such competition among Japanese financial institutions appears to have been channeled to international markets by regulatory actions designed to accommodate and encourage the internationalization of Japanese finance. In the late 1980s, the character of such competition was influenced by the substantial increase in financial wealth controlled by the banks in the form of unrealized capital gains on stock holdings. In our view, important spillover effects from such competition among Japanese banks have been observed in various financial markets. For example, the low returns available on traditional bank lending associated with the expansion of Japanese banks provided incentives for U.S. banks to specialize in the processes of credit origination and financial engineering.

In summary, the changeover to capital-based regulation of Japanese banks should, in itself, encourage important changes in the structure of domestic and international banking markets. Under this regime, unless the Japanese stock market were to rise significantly over the next few years, the capacity of the Japanese banks to continue to implement change in a deliberate and considered fashion will be reduced. In particular, in such circumstances it is likely that Japanese banks will place more emphasis on reviewing theft roles as financial monitors of Japanese nonfinancial firms. [14]

NOTE. The authors especially wish to thank Hiroshi Nakaso, Alicia Ogawa, and Larry Promisel for their comments and assistance.

1. The increase in government debt was reversed later in the 1980s, as government budget surpluses facilitated the retirement of outstanding government bonds.

2. Yoshio Suzuki concludes that underwriting of government debt during the period of low administered rates before 1975 was not a source of adverse pressure from the viewpoint of either liquidity or profitability for syndicate members. See Yoshio Suzuki, Money, Finance, and Macroeconomic Performance in Japan (Yale University Press, 1986).

3. Kumiharu Shigehara describes the Bank of Japan's policy before 1975 as one in which the outright purchase of Japanese government bonds was generally kept in line with the trend increase in the monetary base. Also, Japanese law prohibits the Bank of Japan's direct subscription to new government bond issues. See Kumiharu Shigehara, "Japan's Experience with Use of Monetary Policy and the Process of Liberalization" (paper presented to the Pacific Region Central Banks' Conference on Domestic Monetary Policy sponsored by the Reserve Bank of Australia, October 12-13, 1990).

4. Paul M. Horvitz notes that, soon after the accord, long-term U.S. government bond prices dropped below par for the first time since 1937. Paul M. Horvitz, Monetary Policy and the Financial System, 3d. ed. (Prentice-Hall, 1983).

5. There are two types of ordinary commercial banks, city and regional. The distinction is not a legal one but has become embedded in practice. In contrast to the regional banks, each of the eleven city banks is headquartered in a major city and has a national branch network.

6. This interpretation is bolstered by Kazuhito Osugi's commentary on the role of cross-border lending in reducing the importance of window guidance by the Bank of Japan in restraining competition among Japanese banks. See Kazuhito Osugi, "Japan's Experience of Financial Deregulation since 1984 in an International Perspective," Economic Papers 26 (Bank for International Settlements, Basle, January 1990).

7. "Overseas Markets Beckon," International Financing Review (April 27, 1991), p. 6, and "IADB Readies First Floating-Rate Note," International Financing Review (May 11, 1991), p. 4. The two articles identify the relationship between corporate lending rates and the Japanese long-term prime rate and describe how this led to the first floating-rate bond issued in Japan.

8. The reported share is based on data on international banking assets reported to the Bank for International Settlements (BIS) by the G-10 and other reporting countries. The data include bank claims on nonlocal customers in foreign and domestic currencies and claims on local residents in foreign currencies.

9. For example, students of the Japanese system continue to debate the role of Admiral Perry's "black ships" in prompting the structural reforms undertaken by the Meiji state.

10. In the late 1980s, Japanese authorities effectively discouraged Japanese banks from reducing theft exposures to heavily indebted middle income countries through secondary market sales at a discount. However, the banks were permitted to reduce their exposures in connection with their participation in restincturing agreements for Argentina, Brazil, and Mexico, and, subsequently, in connection with their involvement in the Brady debt reduction agreement for Mexico. The resulting tax losses from such transactions were offset by the capital gains from stock sales that were routinely reversed so as to pernut the bank to meet its obligations as a reliable shareholder of client firms.

11. Japanese accounting practices require banks to value securities at the lower of cost or market. This accounting convention means that the stock market decline had litfie direct effect on banks' reported tier 1 positions because most stocks continued to he reported on a cost-of-acquisition basis.

12. Ronald Dore has observed that a general feature of Japanese industry policy has been that it is not ruthless. He cites as an example the prolonged and fruitless efforts to end a loom registration system implemented in the 1950s to control capacity in the Japanese weaving industry. See Ronald Dore, Taking Japan Seriously: A Confucian Perspective on Leading Economic Issues (Stanford University Press, 1987), p. 202.

13. James Home has examined how Japanese financial regulatory policy is formulated and implemented. Through case studies, he illustrates a set of relevant idiosyncrasies, which stem from the Japanese institutional and political framework. Nevertheless, he admonishes the reader to be prepared to recognize "that there is much in the process of regulatory policy-making in Japan's financial markets which policy-makers and partcipants in other countries will recognize." See James Home, Japan's Financial Markets: Conflict and Consensus in Policy-making (George Allen & Unwin, Sydney, 1985).

14. For a more extended discussion of the possible significance of differences in national financial structures, see Alien B. Frankel and John D. Montgomery, "Financial Structure: An International Perspective," Brookings Papers on Economic Activity, 1:1991, pp. 257-97.

SELECTED REFERENCES

Bank for International Settlements. Changes in the Organisation and Regulation of Capital Markets. Basle: Monetary and Economics Department of the Bank for International Settlements, 1987.

Dore, Ronald. Taking Japan Seriously: A Confucian Perspective on Leading Economic Issues. Stanford, Calif.: Stanford University Press, 1987.

Federation of Bankers Associations of Japan (Zenginkyo). The Banking System in Japan. Tokyo: Zenginkyo, 1989.

Frankel, Allen B., and John D. Montgomery, "Financial Structure: An International Perspecfive," Brookings Papers on Economic Activity 1:1991, pp. 257-97.

Home, James. Japan's Financial Markets: Conflict and Consensus in Policy-making. Sydney: George Allen & Unwin, 1985.

Horvitz, Paul M. Monetary Policy and the Financial System. 3d ed. Englewood Cliffs, N.J.: Prentice-Hall, 1983.

"IADB Readies First Floating-Rate Note," International Financing Review, issue 877 (May 11, 1991), p. 4.

Morimoto, Tetsuya, and William J. Seiter. "Japan's Financial System Reform: Stalled on the Tracks," International Financial Law Review, September 1991, pp. 30-33.

Osugi, Kazuhito. "Japan's Experience of Financial Deregulation since 1984 in an International Perspective." Economic Papers 26. Basle: Bank for International Settlements, January 1990.

"Overseas Markets Beckon," International Financing Review, issue 875 (April 27, 1991), p. 6.

Semkow, Brian Wallace. "Japan's 1992 'Big Bang' and Other Financial Reform," Journal of International Banking Law, March 1992, pp. 89-96.

Shigehara, Kumiharu. "Japan's Experience with Use of Monetary Policy and the Process of Liberalization." Paper presented to the Pacific Region Central Banks' Conference on Domestic Monetary Policy sponsored by the Reserve Bank of Australia, October 12-13, 1990.

Suzuki, Yoshio. Money, Finance, and Macroeconomic Performance in Japan. New Haven, Conn.: Yale University Press, 1986.

Viner, Aron. Inside Japanese Financial Markets. Homewood, Ill.: Dow Jones-Irwin, 1988.