This Selected Issues paper analyzes macroeconomic developments and prospects for Japan during the 1990s. Following a surge in activity during 1996 and early 1997, the economy fell into recession in the second quarter of 1997. Real GDP fell by 3¾ percent during the four quarters ended March 1998; the unemployment rate reached historical highs; and deflationary pressures reemerged. The downturn was largely unexpected, and most forecasters had projected growth of about 2 percent in 1997. This paper also examines fiscal policy issues for Japan.


This Selected Issues paper analyzes macroeconomic developments and prospects for Japan during the 1990s. Following a surge in activity during 1996 and early 1997, the economy fell into recession in the second quarter of 1997. Real GDP fell by 3¾ percent during the four quarters ended March 1998; the unemployment rate reached historical highs; and deflationary pressures reemerged. The downturn was largely unexpected, and most forecasters had projected growth of about 2 percent in 1997. This paper also examines fiscal policy issues for Japan.

IV. Resolving Japan’s Banking System Problems1

A. Origins of Banking System Weakness and the Emerging Policy Response

1. The roots of Japan’s banking problems lie in the asset price inflation of the late 1980s. Initial steps towards financial liberalization starting in the mid-1980s allowed small unsophisticated financial institutions (such as credit cooperatives) to venture into new areas, particularly funding housing finance companies affiliated to banks (the jusen) and other real estate investments. At the same time, with large corporations increasingly obtaining funds directly from capital markets, lending by the major banks also was increasingly directed toward real estate projects and small- and medium-sized enterprises, often collateralized by real estate. Rapid credit growth was accompanied by a doubling of stock prices over the second half of the 1980s and a massive rise in commercial real estate prices, which increased by 400 percent. A sharp increase in interest rates in 1990 burst this asset price bubble. Stock prices fell abruptly in 1990–91, and land prices continued to decline over the following seven-year period.

2. Bank provisioning following the asset price collapse has been gradual. In the absence of public funds and with limited scope for raising funds in equity markets, banks limited the pace of provisioning to that afforded by the resources from operating income and the realization of latent gains on stock holdings.2 A departure from this pattern occurred at the end of FY 1995, when banks made special provisions with a modest financial assistance from the government, in order to address the problems of the jusen. In the seven years through March 1998, the 19 major banks have set aside ¥37 trillion in provisions and write offs (Figure IV.1), which were only partially offset by the realization of ¥13 trillion in capital gains from stock holdings. For most of the period, the stock of disclosed nonperforming loans, net of reserves, remained at levels corresponding to a couple of years’ operational income, with a continuous flow of new problem loans requiring new provisioning efforts. The persistent deterioration of banks’ profitability (the banks cumulated ¥6 trillion in losses over the period) led to a continuing decline in their stock prices and credit ratings (Chart IV.1).



Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A004

Source: Fitch IBCA


Citation: IMF Staff Country Reports 1998, 113; 10.5089/9781451820447.002.A004

Sources: WEFA, Bloomberg, and Moody’s.

3. The authorities’ approach to the banking sector was predicated on the expectation that a resumption of economic growth would permit banks’ and borrowers to recover their financial strength. Over the last three years, however, this orientation has gradually changed, and the authorities have introduced a broad range of measures to address aspects of banks’ problems more forcefully. In early 1996, the decision was taken to overhaul the regulatory and supervisory framework, after the large public outcry associated with the collapse and bailout of the jusen, and to extend the deposit insurance system to provide a full guarantee of bank deposits through 2001. Moreover, in late 1996, the government announced the “big bang” reforms, a blueprint to phase in free and open competition and permit market incentives to allocate capital within Japan (discussed in Chapter V). Key aspects of these measures started to be implemented in late 1997, while the authorities have responded to increasing banking strains with a series of additional steps, amid growing reluctance of banks themselves to participate in traditional, orchestrated “convoy-style” rescue operations. Notable among the new measures was the decision to provide public funds to strengthen the deposit insurance system and to facilitate the recapitalization, restructuring, and consolidation of Japanese banks. Since the spring of 1998, further initiatives have been announced, including mechanisms to catalyze the workout of real-estate backed loans and a “bridge bank” facility to reduce the transition costs associated with the resolution of failed banks.

4. The overhaul of the regulatory and supervisory framework is based on three components. The first component gives banks the main responsibility for assessing the asset quality of their portfolio and provisioning accordingly. It requires banks to periodically carry out a self-assessment of the quality of their assets, and to set loan loss provisions to reflect loss rates on different categories of loans, subject to external audits and approval by bank examiners. The second component involved the introduction of a framework for Prompt Corrective Action (PCA), which establishes a set of structured early intervention and resolution rules to be applied in response to a deterioration of regulatory capital below capital adequacy standards (8 percent for banks with international business, and 4 percent for banks with only domestic operations). The third component is the establishment of an independent Financial Supervisory Agency (FSA), separate from the Ministry of Finance (MOF), and in charge of carrying out the supervisory responsibilities previously carried out by the MOF. The self-assessment and PCA were applied from April 1998, while the FSA started its operations in June 1998.

5. Notwithstanding the progress being made, Japan’s banking system problems continue to be a major source of downside risk for economic performance. Weakness can be seen in four key areas:

  • Problem loans remain to be fully recognized on banks’ balance sheets. The scale of uncovered losses remains a major source of uncertainty.

  • Banks’ capital base is weak, especially once securities’ holdings are marked to market and the full extent of deterioration in loan quality is recognized. Moreover, exposure of bank capital to market risk related to equity holdings remains a major source of vulnerability.

  • The process of loan recovery continues to be cumbersome and slow, and is not well suited to moving forward the necessary corporate debt restructuring.

  • Bank franchise value is being eroded by financial liberalization, with the situation of long-term credit banks and trust banks becoming increasingly difficult.3

B. Recent Banking Developments

6. Banking system strains became increasingly apparent during the course of 1997.4 In April 1997, Hokkaido Takushoku Bank (HTB), one of the major commercial (city) banks announced a plan to merge with the smaller regional Hokkaido Bank, while one of the three long-term credit banks (Nippon Credit Bank) undertook a major restructuring with the financial support of private banks and the Bank of Japan. Over the summer, however, concern grew regarding the success of the merger of HTB, which finally stalled over the magnitude of HTB’s bad assets and was postponed sine die in September (Box IV.1). The failure of Sanyo Securities on November 3 heightened concerns among market participants about the ability of Japanese financial institutions to honor their interbank obligations and led to a sharp decrease in the liquidity of the domestic interbank market and a substantial rise in the Japan premium (discussed in Chapter III). As a result of these pressures, HTB found itself unable to raise funds in the market. On November 17, it applied—with the support of its supervisors—to transfer its problem loans to the Deposit Insurance Corporation (DIC) and its healthy assets and liabilities to Hokuyo Bank, a second-tier regional bank also based in Hokkaido. The collapse of HTB was followed shortly thereafter by that of Yamaichi Securities, the fourth largest securities house in Japan.

Table IV.1.

Japan: Structure of the Banking System, March 1998

(In trillions of yen, unless otherwise noted)

article image
Sources: Economic Statistics Monthly Bulletin; Bank of Japan; Federal Bankers Association of Japan.

Includes balances in principal accounts of overseas branches.

Mainly subsidiaries of city banks and other financial institutions.

Principal accounts only.

March 1997; deposit data are for Postal Saving System.

7. Prompt intervention by the Bank of Japan after the collapse of HTB and Yamaichi avoided the repetition of the financial market disruptions that followed the collapse of Sanyo Securities, but overall market conditions continued to deteriorate in December. Fears of additional bank closures were reflected in large deposit withdrawals from weak banks, and contributed to the decision by credit rating agencies to consider downgrades for several banks. Market discipline led to a tiering in stock markets, with marked declines in stock prices of weaker banks. Market tension intensified with a spate of bad economic news in early December, which sent Japan’s stock market to a six-year low, and raised pressures on banks whose capital bases were most vulnerable to changes in stock prices. The imminent implementation of the new supervisory framework, which requires supervisors to take prompt corrective action whenever a bank’s capital-to-risk-weighed-asset ratio falls below a certain level, increased pressures on banks, and prompted banks’ efforts to reduce their risk-weighted assets, contributing to the “credit crunch.”

8. Responding to this situation, in late December the Government announced a set of emergency measures to stabilize financial markets and bolster depositors’ confidence. The measures (discussed in more detail in Section D) centered on strengthening the financial condition of the Deposit Insurance Corporation (DIC) and providing public funds for bank recapitalization, and were accompanied by several regulatory changes that assisted the major banks to meet capital adequacy requirements—most notably permission to value securities at cost instead of the minimum of cost or market price, and to include unrealized gains on land holding in Tier 2 capital.

9. This package represented the first time that public funds were made available to the banks on a massive scale, and, together with provision of liquidity by the Bank of Japan, helped to reduce market pressures. The corresponding legislation was passed in late February, and in late March (shortly before the end of the financial year), all but one of the major banks and three large regional banks received capital injections. In order to address the concerns of weak banks that application for these funds could be interpreted as a signal of fragility, these injections were of similar magnitude irrespective of the size or needs of individual banks. The injections complemented banks’ efforts to improve capital ratios by reducing risk-weighed assets, including through restrained lending, the sale or securitization of about ¥4 trillion in existing assets, and the issuance of non-voting preferred stock in international markets (at a significant premium over U.S. Treasury bonds).5

The Collapse of Hokkaido Takushoku Bank

Hokkaido Takushoku Bank (HTB), also known as Takugin, was founded in 1900 as a state-supported bank aimed at providing funds for the development of the northern island of Hokkaido (Takushoku can be loosely translated as “colonial”). After World War II, it became the smallest national commercial “city” bank, with its activities remaining concentrated in Hokkaido. The island’s economy is still less diversified than that of the south-central regions of Japan and was strongly affected by the collapse of the real estate market after the bursting of the asset price “bubble.” HTB’s 70-odd branches outside Hokkaido generated little income to the bank.

At the end of FY 1996, HTB disclosed ¥935 billion in problem loans (equivalent to 10 percent of its assets), and announced a plan to merge with Hokkaido Bank—the 23rd largest regional bank—by the end of the following fiscal year and to withdraw from international operations. The main advantage seen by markets in the merger with the relatively small Hokkaido Bank (whose assets were about a third of those of HTB) was the possibility of a drastic consolidation of the two banks, whose geographic base overlapped (each had 133 branches on the island). Neither bank was particularly well capitalized and both had asset-quality problems (Hokkaido Bank had ¥166 billion in bad debts against a capitalization of ¥44 billion).

In early summer, the two banks agreed on standards to evaluate each bank’s debts and provisioning requirements, but Hokkaido Bank started to express doubts about HTB’s ability to absorb losses from its bad loans. In early July, HTB cut off credit to Tokai Kogyo Company (the 46th largest construction company in Japan). The decision of HTB, as a main bank, to let Tokai Kogyo file for bankruptcy contrasted with the support Fuji Bank, as main bank to the much larger Tobishima company, had offered in a similar situation and—although permitting the bank to clear the decks of a problem ahead of the merger—forced HTB to write off nearly ¥47 billion in loans. Concerns about the soundness of HTB’s other loans, including to several subsidiaries, finally led the merger to be postponed sine die in early September.

In the following months, the HTB faced increasing difficulties in raising funds, and it finally collapsed after liquidity in the interbank market dried up following the failure of Sanyo Securities in early November.

An inspection carried out by the MOF in the weeks after HTB’s collapse indicated that the bank’s liabilities exceeded its assets by ¥840 billion, before accounting for ¥200 billion of bad loans to affiliated companies. The bank had ¥940 billion in bad loans and an additional ¥1,350 billion in questionable loans.

The resolution of HTB is expected to be completed by late 1998, with Hokuyo Bank and Chuo Trust Bank acting as receiving banks for the bank’s good assets in the Hokkaido region and the main island (Honshu), respectively, and the RCB taking over the doubtful and uncollectible loans. The transfer of substandard loans to the receiving banks has been delayed by disagreements about the price of these transfers. Chuo Trust Bank, for its part, plans to apply for a recapitalization with DIC funds, as its capital was insufficient to support the stock of good loans received from HTB.

The collapse of HTB, in addition to representing a movement away from the “convoy” system in which strong banks are called to support the weak ones, underscored serious shortcomings in financial disclosure practices in Japan. The amount of bad loans uncovered by the supervisors after the collapse of the bank was much higher than that disclosed by the bank at the end of FY 1996, and the asset quality was much worse than suggested by the provisions made by the bank. Notably, loans to Tokai Kogyo—which had been heavily restructured in 1993—were not included among problem loans disclosed by HTB, or by any of the other banks that held the company’s remaining ¥80 billion in debt.

10. At the end of the 1997 fiscal year, banks took up the room provided by access to public funds and changes in accounting methods to increase their provisions and writeoffs, while succeeding in most cases in boosting their reported regulatory capital ratio (Table IV.2 and IV.3). A number of the strongest banks among the 19 major banks increased their loan loss provisions and charge offs by a factor of three to five vis-à-vis the previous fiscal year. Provisions and writeoffs for the major banks as a whole doubled to ¥10.6 trillion, while the total for all deposit-taking institutions amounted to ¥15 trillion. Because banks’ net operating profits (gyomu-juneki) contracted sharply, especially among trust banks whose funding costs increased, the boost in provisions implied large negative pre-tax profits (keijo rieki) for most major banks, including all the city banks.

Table IV.2.

Japan: Capital Adequacy Ratios and Selected Hidden Reserves for the Major 19 Banks

(In percent of risk-weighted assets, except where noted otherwise; end-March)

article image
1/ Unrealized gains measured in relation to purchase values. Under the old standard, holdings of listed equities were accounted at the lowest of purchase (book) and market price for the effect of computing Tier-1 capital (i.e., a drop of market values below book values would reduce Tier-1 capital), and up to 45 percent of the excess of market value over book value could be included in Tier-2 capital. Under the new standard, banks can choose to use only book values (i.e., unrealized loss do not reduce Tier-1 capital, and unrealized gains cannot be included in Tier-2 capital).2/ Under the new standard, up to 45 percent of unrealized capital gains on real estate holdings can be included in Tier-2 capital.

These banks did not adopt the new (optional) standard.

The risk weight of loans linked to trust accounts is less than the ordinary weight for loans.

Table IV.3.

Japan: Profit and Loss Accounts of the Major Banks, end-FY1997 1/

(In billions of Japanese yen unless otherwise noted)

article image
Source: Fitch-IBCA Ltd.

Fiscal year ending March 31 of the year shown.

Net operating profits before specific loan-loss charges and gains on the investment portfolio (source indicates that due to adjustments, this measure cannot be precisely calculated from public data).

Pre-tax, pre-special-item current profits that include those from securities holdings.

11. Despite the increase in provisions, the credit ratings of several major banks continued to be lowered, on concerns about profitability and asset quality. On June 26, 1998, one of the major Japanese banks, the Long-Term Credit Bank (LTCB) announced its intention to merge with Sumitomo Trust Bank (STB) with the assistance of the authorities. This came after a period of heavy market pressure on LTCB in the second half of June.6 The STB has indicated that a condition of the merger would be that the bad loans of LTCB would need to be shifted elsewhere, but the terms of the merger are still under discussion.

12. Since the end of the fiscal year, the Japanese authorities have taken further steps to address banking sector problems. Under the framework provided by the emergency measures approved at the beginning of the year, some 45 credit cooperatives and four small regional banks have been closed, thus accelerating the consolidation process that has reduced the number of Japanese deposit-taking institutions by some 15 percent since FY 1992. In May 1998, plans were announced to establish a new mechanism for debt workouts, supported by adjustments in the tax treatment of debt forgiveness. This was followed by the announcement in early July 1998 of a “bridge bank” scheme to continue the operation of failed banks—including the extension of credit to “solvent borrowers in good faith”—thus reducing the transitional costs associated with bank closures. In the weeks following the inauguration of the FSA, the authorities also announced a plan to immediately inspect the 19 major banks. Such a plan came on the heels of calls from Bank of Japan officials (including the Governor) encouraging banks to disclose the results from the self assessment of their individual portfolios conducted at the end FY1997.

C. The Bad Loan Problem and Impediments to its Resolution

13. A variety of approaches have been used to estimate banks’ bad loans and uncovered losses:

  • Disclosed non performing loans (NPL) of all deposit-taking institutions amounted to ¥35 trillion at end-March 1998, based on new, more stringent, standards intended to be broadly in line with those used in the United States. On the old standard, banks’ disclosed NPLs would have been ¥25 trillion at end-March 1998, down from ¥28 trillion at end-September 1997.7

  • As an alternative to relying on disclosed nonperforming loan figures, rating agencies and other market participants have often estimated losses taking into account the flow of real estate lending during the “bubble” period, which amounted to about ¥80 trillion. Agencies estimate that most of the resulting stock is impaired, against a residual value of collateral estimated at only ¥20–30 trillion.8 This figure would include loans to the jusen, loans sold to the Cooperative Credit Purchasing Corporation (CCPC) special vehicle set up in 1993 (Box IV.2) and problem loans of banks’ affiliates, as well as banks’ own problem loans. Taking account of estimates of recoveries from collateral and accumulated loan loss reserves, total uncovered losses were estimated by Fitch-IBCA (a bank rating agency) at around ¥16 trillion at end-March-1998, mainly concentrated in the smaller regional banks and cooperatives.

  • The self-assessment of asset quality conducted by banks showed total problem loans of ¥71 trillion net of specific loan-loss reserves at end-March 1998, although this estimate has not yet been subject to supervisory scrutiny. Problem loans under this definition include loans classified as substandard (category II), doubtful (category III), and loss (category IV). The equivalent figure for all deposit-taking institutions (including the credit cooperatives) is ¥88 trillion. These figures do not include loans sold to the CCPC and loans to the jusen.

14. The self-assessment results can be used to estimate likely uncovered losses. Using loss rates for the various categories derived from a Bank of Japan study of the loss experience over 1993–96, the self-assessment figures for end-March 1998 would imply a total uncovered loss in all deposit-taking institutions of ¥19 trillion (around 4 percent of GDP, see tabulation below). Nevertheless, many market analysts believe that the remaining uncovered losses are likely to be considerably higher because (i) banks may well have been over-optimistic in loan classification, placing too high a proportion of impaired loans in category II rather than categories III and IV and providing excessive valuation for loan collateral; (ii) loss rates for category II loans may be higher than during the 1990s when banks were not actively disposing of bad loans (even taking into account the special provisions made in 1995 and captured in some of the figures used in the Bank of Japan study); and (iii) the exposure to Asia may not be fully captured in category II loans, and may result in losses greater than the 2–3 percent currently provisioned (Table IV.4).

Resolution Agencies in Japan

The Cooperative Credit Purchase Company (CCPC) was created in 1993. The CCPC provided a mechanism to allow banks to transfer loans at a discount, thus satisfying tax requirements in the tax law, while avoiding bankruptcy of debtors (loan-loss provisions are automatically tax-deductible only when they follow the foreclosure of collateral or the sale of the loan at a loss). Banks remain responsible for covering the difference between the transfer price to CCPC and the final disposal price, and generally for managing the loan. Reflecting its main mandate, the CCPC does not actively seek to resolve the assets under its care, and at its current pace, it will take another 5–8 years to dispose of its inventory. Collections on an original portfolio of ¥15 trillion (purchased at a price of ¥5.7 trillion) have amounted to ¥1.1 trillion. Sales, which are most often arranged by debtors themselves, picked up in 1997, but are still low; moreover the disposal of the asset does not automatically entail a reduction in the debtor’s liability, which occurs only after the three parties have received an agreement from the courts.

The Housing Loans Administration Corporation (HLAC) was created in 1996 to resolve within a 10-year period some 300,000 loans left by the seven failed housing financing companies affiliated with banks (the jusen), and received an endowment of ¥0.6 trillion for this purpose. The HLAC has liquidated about one-fifth of its original ¥4.6 trillion portfolio. Claims that banks have knowingly transferred to jusen their worst assets, and that as much as 10 percent of the ¥1 trillion corporate loan book the agency built up may be tied up to criminal (yakuza) concerns—suggest that loan recovery will continue to be slow.

The Resolution and Collection Bank (RCB) was, until recently, in charge of the assets of failed credit cooperatives only. RCB is the successor of the Tokyo Kyodo Bank created in 1995 to deal with assets left by the failure of credit unions in the Tokyo region. As of end-FY1997, the RCB had received loans with a face value of ¥1.5 trillion, at an average discount of about 70 percent. Although it is a bona fide resolution bank, the RCB has also been slow in selling assets. Despite the relatively high discount at which it received most of its assets, the RCB had sold only 19 percent of its inventory by end-FY1997.

Table IV.4.

Japanese Bank Exposure to Asia

article image

Source: Bank’s financial statements, reported by the Nihon Keizay Shinbun-Nikkei.

Source: Bank of International Settlements.

Estimates of Loan Losses Based on the Self-Assessment Exercise 1/

(In trillions of yen, unless otherwise indicated)

article image
Source: Ministry of Finance; Bank of Japan Quarterly Bulletin, “Utilization of Financial Institutions Self Assessments in Enhancing Credit Risk Management”, February 1998; and staff estimates.

Net of specific provisions. Classified loans at end March 1998.

Derived from Bank of Japan study.

15. The persistence of large amounts of bad loans reflects the slow progress in dealing with the bad loan problem observed since the early 1990s. Several factors have contributed to this outcome:

  • Japanese banks and officials did not foresee the extended period of slow growth experienced in the 1990s and the continued slump in real estate prices. They believed that over time current earnings would be sufficient to build provisions and cover losses without public assistance.

  • The complex web of relationships between bank shareholders and main customers reduced the incentives for banks to improve their profitability at the risk of the viability of borrowers.

  • The legal infrastructure in Japan is not suited for dealing with recovery and debt restructuring in an expeditious fashion. Since there is no “project” financing, all loans to corporations are on a full recourse basis, implying that foreclosure of one loan would threaten the whole company. Moreover, foreclosure and bankruptcy procedures are cumbersome and can be time consuming, while procedures for out-of-court settlements are inflexible.

  • With prospects for loan recovery limited, banks have little incentive to liquidate the questionable parts of their loan portfolios, as the carrying cost of bad loans (after provisioning) is very low, and the potential for upswing in the value of the asset is greater than that of a further loss.

  • Until recently, the process of loan securitization was difficult and costly, reducing the scope for sales of problem loans to private investors.9

  • Provisions and debt forgiveness have not been automatically tax deductible, reducing banks incentives to set realistic levels of provisions or write off debts.

  • Multiple liens attached to most real estate used as collateral, which have proved to be difficult to rank, consolidate, and clear (the CCPC, for instance, only accepts assets with single or well defined liens, held by parties who have reached agreement on the way to deal with the asset).

D. Use of Public Funds

16. In December 1997, the authorities decided to provide ¥30 trillion in public funds for the purpose of strengthening the Deposit Insurance Corporation (DIC) and to create a financial crisis management fund. In contrast to the vocal public opposition against providing public funds to resolve the jusen, this most recent initiative to provide public monies was broadly supported, in part because, after the failures in November, it was widely recognized that the resources of the DIC were inadequate.10 These measures were followed in early July 1998 by the announcement of the “bridge bank” scheme.

17. Legislation was adopted in February 1998 to provide the DIC with adequate financial resources to offer the full protection of banks’ deposits and most credits until March 2001 and to ensure the efficient management of assets received from failed banks.11 The legislation also provided a mechanism for the DIC to play a role in the consolidation of the banking sector. The following measures were taken for these purposes:

  • To strengthen the deposit insurance system, the DIC was provided with ¥7 trillion in the form of government bonds, plus the authority to issue an additional ¥10 trillion of government guaranteed bonds, if needed, to meet liquidity needs in purchasing assets from failed institutions.

  • The Resolution and Collection Bank (RCB)—a unit of the DIC—had its authority expanded to permit it to take over assets from financial institutions in addition to credit cooperatives, and had its investigative powers and collection ability expanded.

  • In addition to protecting depositors, the DIC was allowed to purchase doubtful and other nonperforming loans from insolvent institutions to facilitate mergers with healthy institutions or to create a new institution by combining two or more failed institutions.

  • A new Financial Management Account was created, funded with ¥3 trillion in government bonds to be transferred to the DIC, and the authorization to the DIC to issue up to ¥10 trillion in government-guaranteed bonds, for the purpose of recapitalizing banks.

18. The terms under which the DIC purchases problem loans remains a difficult issue, since no comprehensive valuation framework (e.g., analysis of future cash flows under generally applied assumptions and specific parameters of individual loans) has been adopted.12 In the first operation using the new framework announced in May (involving two banks in the Osaka region), these prices were not disclosed, but the recapitalization effort required from the original shareholders was small in proportion to the stock of substandard loans expected to be bought by the DIC. A related issue is that of the price paid by receiving banks for substandard loans. In the past, banks have received these loans at face value. As the quality of these assets deteriorated, however, the receiving bank faced growing problems. In the extreme case of Midori Bank, which itself failed three years after being set up as a receiving bank, the government felt compelled to recapitalize the bank without penalizing its shareholders.13 In order to avoid the repetition of this experience, the authorities have recognized the need for transferring substandard loans at a realistic discount.

19. The objective of the financial management crisis account is to permit the DIC to increase the capital base of banks for any of the following purposes: (i) to support the merger of a failed bank (the receiving bank may need additional capital to support the received assets, independent of the quality or transfer price of these assets); (ii) to avert systemic risks; and (iii) to protect a region from the consequence of a liquidity crisis. Banks can apply to use this facility on a voluntary basis, and purchases are to be approved by a high-level committee, based on the submission of a program for improving banks’ operations and management and criteria supporting the requirement in the law that the applying financial institutions are solvent. The facility entailed the establishment of a new account at the DIC to be used for the purchase of preferred stocks and subordinated loans or bonds issued by financial institutions until March 2001. The law required these purchases to be made under conditions that would not make future sales by the DIC of these instruments difficult, but did not establish an obligation of, or a time for, proceeding with such sales.

20. All major banks (except for Nippon Trust, now a subsidiary of Bank of Tokyo-Mitsubishi) and three large regional banks qualified for a first round of recapitalization in March on the grounds of reducing the systemic risks and after submitting plans to improve their operations. These plans were built around a reduction in personnel expenses, the closure of branches, and a decrease in the number of directors. Although most banks received about ¥100 billion, the terms under which the funds were provided varied among banks, reflecting the committee’s judgement about the soundness of individual banks. These terms were based on the examination of banks’ self-assessments and other documents provided by banks to the committee and the MOF (Table IV.5)

Table IV.5.

Japan: Conditions for the Subscription of Capital Using Public Funds, March 1998 1/

(In billions of Japanese yen unless otherwise stated)

article image
Source: Deposit Insurance Corporation (Japan), Moody’s, and Bloomberg.

Spreads measured in basis points (bp) vis-à-vis yen rates in the London market

A= Spread during the first five years, B= Spread after the first five years, if the bond is not called or the loan repaid.

Percentage change between July 1, 1997 and January 1, 1998.

Basle Committee capital adequacy ratio.

21. In early July, the authorities announced a bridge bank facility to take over the operations of failed institutions (Box IV.3). The aim would be to arrange a merger for a failed bank within two years, but three one-year extensions would be provided if necessary. A public holding company would manage all of the bridge banks. Necessary legislation is expected to be submitted to an extraordinary session of the Diet in late July.

22. This new scheme shares some features with one of the approaches used in the United States by the Federal Deposit Insurance Corporation (FDIC) since the late 1980s, and should facilitate the process of dealing with failed institutions without disrupting the credit mechanism. A number of details on its implementation, however, remain unclear, most notably the parameters guiding the choice of category II loans that will be deemed eligible to be transferred to a bridge bank (i.e., which loans would be allowed to be renewed), and which ones would be transferred to the RCB to be resolved. Also, new loans would need to be approved by a committee set up by the DIC following still-to-be established guidelines, but it is not clear how stringent the screening process will be. Finally, although the plan sets an initial two-year horizon for the resolution of bridge banks (with the possibility of three additional one-year extensions), the approach to be used to handle the eventual sale or liquidation of these institutions has not been disclosed. In the United States, in most cases bridge banks have been resolved quickly, with the FDIC prepared to accept low prices to find a buyer.14

The Design and Operation of the Bridge Bank Scheme

The bridge bank scheme would operate in two stages—a first phase in which the failed bank would continue to operate under a financial administrator and a second phase—relevant if a receiving bank cannot be quickly found—in which a public bridge bank would be set up temporarily to continue the bank’s operations.

  • When a bank fails, the Financial Supervision Agency (FSA) would promptly appoint a financial administrator (receiver) to administer the business and manage the assets of the bank. The failed bank would continue to provide loans to sound borrowers, upon approval by the administrator. Non-performing loans would be transferred to the Resolution and Collection Bank (RCB), with funding from the DIC.

  • The financial administrator would seek to identify a suitable receiver bank to take over the failed bank’s operations as soon as possible. However, if a suitable bank could not be identified quickly, then the bank’s operations would be transferred to a public bridge bank. Public bridge banks would be established as subsidiaries of a holding company established by the Deposit Insurance Corporation (DIC).

  • The high-level Financial Crisis Management Board (FCMB) would establish a committee to scrutinize the assets of the failed bank, and divide them between sound loans that would be kept with the bridge bank and bad loans to be transferred to the RCB. Guidelines for the classification are to be established by the FCMB. A critical issue will be the treatment of questionable (Category II) bank loans.

  • The DIC would provide financial support to the public bridge banks as necessary, utilizing the ¥13 trillion fund set up earlier this year for bank recapitalization, as well as covering losses of the bank.

  • The bridge bank would continue providing loans to sound borrowers taken over from the failed bank, upon approval by a special committee set up by the holding company. Loans to sound borrowers could also be provided by government financial institutions on the recommendation of the bridge bank. (The recent fiscal stimulus package contained additional funds for such additional lending.)

  • The bridge banks are to be set up for an initial period of two years with a view to finding suitable bank or banks to take over their assets and liabilities. Up to three one-year extensions could be provided. How the exit process would be managed, including procedures to follow if a buyer is not found within five years, are not specified.

E. Recent Initiatives to Help Debt Workouts

23. The authorities have recently announced three sets of measures to help the workout of real estate loans: the easing of regulations covering asset-backed securities and special purpose corporations (SPCs) for holding securitized assets, the use of public money to buy and consolidate odd plots of land and changes in zoning regulations for certain areas, and the creation of arbitration panels to mediate the resolution of bad loans.

24. In June 1998, legislation aimed at stimulating the securitization of assets, in particular bad loans, was approved by the Diet. The law—first announced in early 1997—will regulate securities backed by loans collateralized by real estate. It will also facilitate the creation of SPCs with the ability to secure claims on specific assets, supported by a centralized system for registering secured interest in (or ownership of) specified financial assets. Under the new regulations, the original borrowers will no longer need to be informed about the sale of their loans.15 Favorable tax treatment will also be granted to these entities and the related transactions, reducing their cost. Although the basic framework for the establishment of the SPCs is well advanced, measures are still being formulated to ensure the full disclosure of the quality of the assets to be securitized (which was previously side stepped by allowing issuers to wrap the securities with enhancements from insurers).

25. In May 1998, a plan was announced to establish an arbitration panel, with the pertinent legislation expected to be submitted to the Diet shortly. The proposed panel would clarify and consolidate the liens on real estate collateral and mediate the terms of agreements between debtors and creditors. To support such debt workouts, the tax code has been amended to permit banks to deduct from their taxable income the losses incurred as consequence of these agreements, and to allow debtors to offset the corresponding windfall gains against past and future losses.

F. Changes in Bank Regulation and Supervision

26. In early 1998, several regulatory changes were made that made it easier for banks to meet their regulatory capital requirements. These changes included both modifications in the accounting of banks’ stock and land holdings, as well as in the weights assigned to several classes of assets held by banks:

  • Banks were allowed to shift the valuation of securities in their portfolios from a “lower of cost or market” to a “cost” basis.16 All but three of the 19 major banks switched to the new regime at the end of FY1997, with seven of them carrying unrealized losses in their stock portfolios totaling ¥1 trillion.17

  • Banks were allowed to include 45 percent of unrealized valuation gains on real estate holdings in Tier-2 capital. The value of these gains for the largest 19 banks amounted to ¥1.2 trillion.

  • Banks were allowed to fully account the formerly mandatory reserves for losses on trading account securities and the “Government Bond Price Fluctuation Reserves” as part of Tier-1 capital (the estimated contribution of this measure to banks’ capital was estimated around at ¥0.1 trillion).

  • The application of the 4 percent capital adequacy ratio for banks without overseas operations was postponed to March 31, 1999. This grace period was granted to banks that had submitted restructuring plans. Of the 80 international banks in early 1997, around 35 (including one of the major banks, the long-term credit bank Nippon Credit Bank) have withdrawn from international operations, thus potentially obtaining a grace period as well as halving their eventual capital adequacy requirement in relation to the 8 percent ratio required from other banks.

27. Several changes in the weights applied to specific classes of assets were also implemented, following the Basle accord. These changes, in addition to their immediate effect on attenuating the risks of a credit crunch by easing bank’s capital requirements, in some cases improved incentives for new financial intermediation mechanisms.

  • The weight applied to loans to securities houses was reduced from 100 percent to 20 percent.

  • Where the amount of recourse liability retained by a bank is less than 8 percent, the bank shall maintain capital for the recourse liability equal to the amount of credit retained.

  • The weight on loans insured by credit guarantee associations was reduced to 10 percent, thus bolstering the effects of the resources made available by the government in early 1998 to guarantee or otherwise insure a ¥13 trillion flow of new loans during fiscal years 1997 and 1998.

  • Banks with only domestic activities were allowed to net out compatible loans and deposits of same borrowers (the average cash and deposit to borrowing ratio of Japanese firms, although less than half of its value in the early 1990s, is still large, standing at around 15 percent).18

28. The major changes in bank supervision were the application of the prompt corrective action (PCA) framework for banks with overseas operations as of April 1, 1998, and the start of operations of the new supervisory agency (the FSA) on June 22, 1998 (see Chapter V for a description of the structure, responsibilities, and operation of the FSA). The Japanese PCA framework is similar to, but somewhat less demanding than PCA as applied in the United States. For example, banks operations are only suspended when capital falls below zero (or there are strong indications that this will happen shortly), while in the United States, supervisors can intervene when the capital ratio falls below 2 percent (Table IV.6).

Table IV.6.

Japan and the United States: Summary of Prompt Corrective Action

article image
Sources: Japan, Ministry of Finance; and United States, Federal Deposit Insurance Corporation;

The international capital standards (Basle capital adequacy standards) apply to banks with international operations. The adjusted national capital standards apply to banks with purely domestic operations.

The total capital ratio cited is the total risk-weighted capital; the leverage ratio is the ratio of Tier-1 capital to total assets.

29. Two weeks after the inauguration of the FSA, the authorities announced an imminent and intensive inspection of the 19 major banks, to be carried out in collaboration with the Bank of Japan. At the occasion, the authorities noted that, based on the results of that inspection, strict measures would be taken, if necessary, depending on the supervisors’ assessment of those banks’ capital adequacy ratios and PCA rules. The measures mentioned ranged from the requirement of having banks drawing up and implementing special management improvement plans, to the suspension of the operations of severely undercapitalized banks.

G. Remaining Challenges and Risks

30. In seeking to resolve Japan’s banking problems, the greatest challenge will be to strike the proper balance between short-term macroeconomic objectives (of avoiding deflationary pressures and restoring growth) and the medium-term structural objectives of promoting and ensuring a market-based restructuring of Japan’s banking sector. The provision of ¥30 trillion in public funds to the DIC has provided scope for restoring the soundness of the banking system and supporting the resumption of economic growth on a sustained basis, by facilitating the resolution of insolvent banks and the recapitalization and deep restructuring of the core banking system. However, there are several key issues that will need to be addressed in the implementation of an effective restructuring strategy:

  • First, mechanisms are needed to encourage weak and undercapitalized banks to seek access to the recapitalization window, in order that they can start lending again. These mechanisms also need to be compatible with a strategic reorientation of banks’ activities, organization, and governance rules, toward a focus on shareholder value and return on equity. Such a reorientation will require much more than a reduction in personnel costs on the part of the major banks.

  • Second, it is important to ensure that access to the financial crisis management facility does not result in flows of low cost capital to inefficient sectors of the economy, including the construction sector. Rather, these public monies need to contribute to the fundamental bank restructuring and resolution of the debt overhang that now exists in parts of the corporate sector.

  • Third, the potential financial gains from successful and profitable restructurings supported by capital injections with public funds should be shared with the DIC (and hence the taxpayer).

31. To address these issues, the implementation of recapitalization and restructuring plans with public funds should adhere to some guiding principles, such as:

  • Public funds should be targeted to create a stronger, more profitable banking system.

  • Publicly funded asset acquisitions should be based on transparent, cash-flow based loan-valuation methods.

  • Private market solutions should be encouraged to the extent possible.

  • Shareholders and management should bear responsibility for losses and poor performance.

  • The terms of recapitalization should provide clear and strong incentives for the eventual replacement of public sector funding with private market of capital.

32. The use of public money has heightened the need for further improvements in accounting and disclosure standards, and in internal risk control mechanisms and corporate governance. Although the introduction of bank self-assessment constitutes an improvement in this area, its effective implementation will depend on a continuous and close evaluation of its results by a core of well trained supervisors and on a clear translation of these results into banks’ financial statements. Raising disclosure standards to the high end of the spectrum of international practice would also help restore market confidence. For instance, the reporting of asset quality to supervisors and the release of financial statements is currently done on a semiannual basis in Japan; in contrast, in a number of other advanced countries, such reporting is on a quarterly basis, thus providing more frequent opportunities for supervisors and markets to evaluate balance sheets.

33. The introduction of the prompt correction action framework was a major step in establishing an effective structured mechanism for early intervention and resolution. However, the application of this framework in Japan also falls below the most stringent international practices. For example, as mentioned above, in the United States, for example, the trigger points for determining regulatory action, such as whether a bank would be required to formulate a recapitalization and restructuring plan, or whether public intervention is required, are higher than in Japan. Also, the recent change in the accounting rules governing the valuation of securities holdings of Japanese banks has diluted the scope for the PCA framework to respond to an erosion of banks’ true capital position. Moreover, the distinction for capital adequacy purposes between banks with and without overseas activities needs to be phased out over time, since domestic banks also need to be strongly capitalized.

34. Effective banking supervision is the last line of defense for ensuring accurate recognition of asset quality problems and their prompt correction. Experience in other advanced countries suggests that the challenges in fulfilling this task in Japan are likely to increase following the introduction of new financial instruments, the liberalization of entry in several financial activities, and the more complex organizational structures allowed by the “big bang.” International experience also indicates that the success of the FSA in ensuring effective supervision will depend on providing the agency with a clear mandate, supported by operational autonomy, and balanced by public accountability.

35. The effectiveness of recent initiatives to accelerate loan disposal will hinge on ensuring a transparent framework for the arbitration panels to decide on the efficient valuation of loans and sharing of losses. An active secondary market for asset-backed securities would facilitate, the work of those arbitrations panel by making the price-discovery process easier. The implementation of the securitization law next September should give a boost to this market, if accompanied by the establishment of disclosure rules regarding the asset quality underlying the new securities. An acceleration of current plans to reform Japan’s commercial code and bankruptcy laws—as is now being considered by the government—would help to establish a more enduring and comprehensive framework for the resolution of corporate distress, and would help to foster a more rapid resolution of bad debt problems. The key challenge will be to balance the legal principles of attempting to maximize the amount of funds recovered from debtors with the desire to preserve the economic value of collateral.

36. The prompt resolution of banks’ bad loan problems will be essential to prepare banks to face the increased in competition for the provision of financial services that is an inevitable consequence of the “big bang” reforms. These reforms are also likely initially to put further strains on banks’ balance sheets. On the liability side, the proliferation of mutual funds may reduce banks’ deposits, and increase banks’ cost of funding. Experience in other advanced countries indicates that banks often succeeded in raising funds in money markets to replace deposits, inter alia, by selling their liabilities to affiliated short-term mutual funds. Such a strategy, and a growing acceptance by Japanese investors of mutual funds sponsored by foreign firms or geared toward investments abroad, would raise banks’ funding costs.

37. The counterpart to these changes in the liability side of banks’ balance sheets will probably involve a rebalancing of banks’ portfolios of stocks and loans, through a gradual reduction in their stock holdings as well as the securitization of loans. Some of the recent regulatory changes that have reduced the capital requirements associated with the holding of securitized loans and commercial papers can contribute to this process (thereby reducing the risk of a credit crunch during the liberalization process and favoring an increase in the return on banks’ equity).

38. In sum, a shift away from high-volume, low-margin lending towards greater reliance on fee income and capital market operations would support a more profitable and efficient use of banks’ capital, if adequately regulated and supervised. These transformations in response to a changing environment would, however, remain difficult as long as doubts linger over the quality of banks’ assets and the need for further provisioning. The problems of managing risk in such circumstances are illustrated by recent difficulties in securitizing domestic corporate debt in the absence of assurances about the quality of the assets to be securitized, and the high premium charged by markets to hold uninsured liabilities of Japanese banks.


Prepared by Joaquim Levy.


Japanese banks have traditionally held a large portfolio of stocks. Banks1 stockholdings corresponded to about 17 percent of the capitalization of the Japanese stock market in the mid-1970s, a share that rose to more than 25 percent by 1988, reflecting the expansion in cross-shareholdings among financial as well as nonfinancial corporations. Owing to the historical appreciation of stock prices, capital gains on these stocks, until realized, were translated into sizeable hidden reserves. The Nikkei 225 stock price index increased five fold in 1970–85, and another 350 percent in the following five years. In accordance with the 1988 Basle agreement establishing the capital adequacy requirement of 8 percent of risk-weighed assets, the Japanese authorities have allowed banks to include 45 percent of hidden reserves in Tier-2 capital.


Such banks do not have a retail base, and have been under threat from the change in regulations under the “big bang” that would allow all banks to issue long-term debentures (currently a monopoly of long-term credit banks) and sell investment trusts. In the case of the long-term credit banks, greater competition will compound the loss of their franchise value that started in the 1980s with the changes in the long-term financing of large Japanese corporations (which led these banks to have an exceptionally large exposure to the real estate sector). Trust banks, which receive most of their funds for investment from pension funds and insurance companies (acting inter alia as custodians of these funds), have already been weakened by the liberalization of pension fund allocation rules. The new rules now allow professional asset management companies to handle an increasing share of pension fund resources through diversified portfolios, and have reduced the attractiveness of standard investment trusts. Trust banks (including the specialized subsidiaries of city banks) and long-term credit banks account for about 10 percent of the assets of the Japanese banking system (Table IV.1).


Strains were also felt in other sectors of the financial system, as discussed in Chapter V.


In the final months of FY 1997, four Japanese banks raised about US$5 billion in funds, using these instruments. These so-called Opco Preferred securities use a credit-linked note structure, similar to the “capital securities” or “trust-issued preferred stock” sold by U.S. banks since 1996, and refined by several European banks in 1997. Credit agencies have treated them broadly as debt-like instruments, similar to junk bonds, even if they rate them one notch above “speculative” instruments. These instruments were issued at premia of 225–440 basis points over U.S. Treasury paper, and were lightly traded afterwards (when reported premia widened to more than 600 basis points). Although not convertible (thus avoiding the risk of dilution) and nominally being perpetual securities, these securities can usually be called by the issuer after 10 years subject to supervisory approval; interest payments are tax-deductible under Japanese law. In general, these instruments are designed with U.S. pension funds in mind (i.e., satisfying the requirements of SEC’s Regulation 144a) and are most often issued by a foreign subsidiary of the concerned bank. These interest-bearing instruments have been treated by regulators as Tier-1 capital because interest payments can be missed and are non-cumulative, i.e., payments need not to be made up to investors (but the bank must pay dividends on the instrument before paying dividends on its preferred stock or its common stock). Market observers have noted, however, that failure to make these payments would have severe reflections in the access of Japanese banks to international money markets.


The LTCB is the second largest long-term credit bank, a type of bank that has relied on issuing long-term debentures to fund long-term corporate lending. The new bank would be the seventh largest in Japan.


Nonperforming loans (NPLs) are defined to include loans to borrowers in bankruptcy, loans that have been restructured, and loans with interest arrears. The shift in accounting rules implemented last March requires disclosing a loan as nonperforming when interest arrears exceed three months (previously six months) and all restructured loans with reduced interest rates or extended maturities (previously only when the interest rate was reduced below the official discount rate).


Official estimates agree that the collateral value of most impaired real-estate loans amounts to about thirty percent of the original value of the loan, reflecting broadly the 70–80 percent decline in commercial real estate prices since 1989.


Notwithstanding these difficulties, ¥3 trillion of bad real-estate loans was sold in the past year, mainly through off-shore special-purpose corporations and at discounts reportedly varying between 70 and 95 percent.


The projected income for the DIC in FY1996–2000 (including the surtax to protect deposits over ¥10 million) was ¥2.7 trillion, of which about ¥1.3 trillion was already committed by the time the HTB failed, and most of the remainder would be absorbed by HTB’s net liabilities of around ¥1.1 trillion.


According to statements by the authorities, unguaranteed loan trust and senior debts such as bank debentures are fully protected, but subordinated debt has not been explicitly included in this group. In the past, subordinated creditors (often financial institutions) have been required to bear some losses, even when the “subordination clauses” were not formally triggered by events leading to the collapse of the borrower.


On July 12, the National Land Association announced that, in cooperation with the Japanese Association of Real Estate, it would change its method of assessing the value of land held by financial institutions as collateral to one based on discounted cashflows, instead of the traditional combination of the recent sale price of nearby property and the acquisition cost of the land in question. This could constitute a first step to widespread use of those methods.


Midori Bank was set up after the Hanshin earthquake as the successor to a failed regional bank, with capital contributions from financial institutions in the Kobe region and loans from the Bank of Japan. The authorities have indicated that this was a special case and that the treatment dispensed to this bank should not be seen as setting a precedent.


The FDIC has used the bridge bank method to create 32 bridge banks from 114 separate institutions. Thirty of these institutions were resolved in seven months or less; the other two in less than 2½ years.


Such a system is similar to that provided in the U.S. Uniform Commercial Code. Since 1993 the MOF has dispensed with the need to inform debtors prior to securitizing car loan and lease receipts.


In an associated move, capital requirements on market risks were implemented in line with the amendments to the Basle Capital Accord. Banks can still change the lock-in prices of their stock holdings by selling and buying back such stocks at any time.


The Nikkei 225 index fell 8 percent between end FY1996 and FY1997, which contributed to a ¥6.2 trillion loss on the equity portfolio of the 19 major banks. At the end of FY1997, remaining hidden reserves amounted to ¥1.7 trillion, concentrated in three banks.


On-balance-sheet netting was allowed under the following conditions: (i) netting is legally enforceable; (ii) the maturity of the deposit is at least as long as the corresponding loan; (iii) the positions are denominated in the same currency; and (iv) the bank monitors the relevant exposure on a net basis.

Japan: Selected Issues
Author: International Monetary Fund