III Financial Supervisory and Regulatory Issues
- International Monetary Fund
- Published Date:
- January 1995
Policy Challenges Posed by Derivatives
The dramatic collapse of Barings, a mid-sized, blue-chip investment bank established in 1762, which occurred with little warning and great speed, involved the extensive proprietary use of derivatives in establishing large, highly leveraged Nikkei 225 futures positions on the futures exchanges in Singapore and Osaka. The multinational nature of the firm’s activities obscured its consolidated risk position sufficiently that neither the relevant supervisors nor the firm’s counterparties acted in time to discipline its activities. Hence Barings’ failure, coming on the heels of other financial tremors involving derivatives, for example, Orange County,32 and Metallgesellschaft,33 added new urgency to the initiatives now under way to strengthen the supervisory and regulatory infrastructure governing global derivative markets.
The good news is that the supervisory and regulatory arrangements currently in place proved adequate in containing any possible systemic spillovers from these accidents. The internal risk management systems of the major banks and securities houses had been effective in keeping their financial exposure to the failed entities well within tolerable limits. Indeed, the crisis involving Barings and its subsidiaries was resolved without a lasting negative effect on global markets. Moreover, the risk-management systems of the futures exchanges in Singapore and Osaka, where most of Barings’ positions had been established, prevented a threat to the financial integrity of the exchanges.
The resolution of Barings, which ultimately involved the purchase of a significant share of Barings’ assets by the Dutch bank ING, was helped by favorable circumstances. Markets were at the time not subject to disturbances and uncertainties from other sources, and Barings’ problems were quickly recognized by market participants as firm specific. The sequence of events involving the collapse of Barings and its resolution is a textbook case of the failure of financial institutions: the management of a major global institution fails to control its risk positions; it loses capital; and the central bank is able to close the institution without negative consequences for the stability of the financial system. The resolution of Barings will undoubtedly have a profound effect on market discipline; the swift and resolute reaction of the Bank of England will be a milestone in the containment of moral hazard in financial markets.
Nevertheless, the fact a failure of this size caught supervisors by surprise suggests that renewed efforts to strengthen the existing international supervisory and regulatory infrastructure are called for. In broad terms, these efforts should be aimed at fostering a competitive market environment in which international financial activity can evolve efficiently, where market discipline induces firms to control their risk, and where the failure of major institutions can be managed in such a way as to limit spillovers to other markets. To this end, it will be necessary to introduce better internal risk-management methods to a larger number of the major financial institutions; to set the size of capital requirements governing market risk sufficiently high to offset the benefits flowing from explicit or implicit official financial guarantees for the affected institutions; to improve accounting and disclosure standards to increase the transparency of markets; to reduce risk in wholesale payments systems to limit the spillover of market disturbances; to reform legal codes relating to netting and bankruptcy of multinational financial firms; and to improve international coordination among regulators. Previous capital markets reports have dealt with many of these subjects in detail. Some of the more recent developments are summarized here.
Growth in Derivative Markets
The spectacular growth of derivative markets since the late 1980s and the rapid expansion of banks’ involvement in these activities constitute perhaps the most worrisome aspect of recent developments in this area. Financial history contains many examples of rapid expansions in certain types of financial exposures that result in major losses during periods of consolidation. The expansion of bank lending in particular sectors, such as lending to developing countries in the 1970s and early 1980s, to real estate in the mid 1980s, and for highly leveraged transactions in the late 1980s, are some of these examples.
Derivative markets have continued to grow in recent years. The notional principal34 of all outstanding exchange-traded and over-the-counter derivative contracts increased from less than $2 trillion at the end of 1986 to more than $20 trillion at the end of 1994, an average annual growth rate of 140 percent.35 The replacement value of these products36 is estimated to have remained at around 2.5 percent of notional value. Hence, at the end of 1994, approximately $500 billion of replacement value was outstanding, compared with the capital base of less than $200 billion of the 12 largest dealers, who together are responsible for the vast majority of over-the-counter transactions in derivatives.
The development of derivative markets has been accompanied by a dramatic change in the activities of many of the major international money-center banks, which together are responsible for more than three fourths of all derivatives activities. The financial activities of these banks have moved considerably away from their traditional lending role toward fee-based financial services and proprietary position taking. One of the main activities of banks now is the management of market risk, liquidity risk, and the credit risk associated with derivative activities. These changes, which have occurred in all major countries, are seen most clearly in the behavior of U.S. banks. Among the seven major U.S. money-center banks, interest income declined as a share of total revenue from nearly 70 percent at the end of 1987 to less than 50 percent at the end of 1993. In addition, during the same period, trading income doubled to 14 percent of total revenue. Finally, the volume of over-the-counter transactions in derivatives has also become large relative to total assets. At year-end 1992, the notional principal value of interest rate and currency swaps held off balance sheet by the seven major U.S. money-center banks was 2.3 times the value of their balance sheet assets; the replacement value of the derivatives book of the seven major U.S. money center banks was well in excess of their paid-in equity (Tier I).
Recent Supervisory and Regulatory Initiatives
Regulatory Capital Requirements
The cornerstone of supervisory and regulatory efforts to deal with the developments in international financial markets in recent years has been a rethinking of the role of regulatory capital requirements. It has been recognized for some time that the risk-weighted, ratio-based approach to determining the appropriate amount of regulatory capital is out of step with modern risk-management methodology. The Basle Committee of Banking Supervision, representing supervisory authorities from 12 major countries, and operating under the auspices of the Bank for International Settlements (BIS), is currently developing a very different approach to the determination of regulatory capital required for market risk. It is likely that the new approach to capital requirements is going to affect banking as profoundly as did the 1988 Basle Capital Accord.
As discussed in previous capital markets reports, the focus on credit risk of the 1988 Basle Capital Accord soon came to be viewed by supervisors as too narrow to deal with the market, liquidity, and operational risks inherent in the growth of banks’ trading and derivatives books. In response to this development, the European Union (EU) introduced the Capital Adequacy Directive (CAD) in March 1993, and the Basle Committee introduced a similar proposal for a revised framework for the supervisory treatment of market risk in April 1993. Market developments, however, outpaced regulatory developments, and it became clear during 1994 that the proposed frame-work had sufficiently serious shortcomings to put into question its usefulness. In response, the Basle Committee abandoned the April 1993 proposals in favor of a more radical approach (slated for adoption by the end of 1995 with full implementation by the end of 1997).
Under this new approach, banks would be allowed, if they choose, to use their own internal risk-management models to estimate the bank’s value-at-risk.37 To determine a bank’s new capital adequacy requirement, supervisors would then multiply the bank’s estimate of its value-at-risk by a “safety factor” greater than or equal to three. Supervisors will have the discretion to increase a bank’s safety factor if its internal model is revealed, through time, to be inaccurate. Smaller banks that choose not to use internal models are required to follow a revised version of the standards proposed in April 1993.38
This new approach has the obvious advantage of providing banks with incentives to improve continuously their risk-management systems, and banks will incur no further administrative costs in calculating their regulatory capital requirements. A disadvantage, however, observed during a dry run conducted by the Basle Committee, is that different banks may arrive at different requirements for the same portfolio of derivative instruments, even after controlling for the choice of a confidence interval and the holding period.39 It is expected that validation of the models by the supervisors will lead to a convergence of key aspects of risk-management models, including estimation techniques and possibly common data sets.40 One challenge has been that the model-validation process has required that supervisors retool and become proficient in the use of rather sophisticated modeling techniques in order to be able to engage in this new process of validation. The approach, therefore, is only as effective as supervisors’ ability to judge the adequacy of banks’ risk-management models. It is certain, however, that the calculation of regulatory capital cover for market risk and the process of private risk assessment and risk management have become inextricably bound.
Accounting Standards and Disclosure Requirements
It is widely acknowledged by market participants that current financial accounting systems used for valuing assets and liabilities are not adequately recording the current values of derivatives contracts. This is an important area of concern not only for regulators, but also for market participants that are potential counterparties in derivative contracts. In addition, differences in the accounting treatment of derivatives in national accounting systems are posing significant problems in cross-border counterparty risk assessment. Current financial technology and a high degree of market liquidity, however, allow institutions, and individuals, to change large market positions sufficiently rapidly—in minutes rather than days—to cast doubt on the usefulness of accounting data in providing an up-to-date picture of the financial position of the financial intermediary. Hence, the need for reliable internal position control is not addressed by improvements in disclosure and accounting data.41
A major improvement in accounting standards is under way in the United States. An important feature of the new standards is that derivative positions are disclosed differently if they are traded than if they are used solely for hedging purposes. Further, the standards call for the disclosure of more information about the value of, and the gains and losses from, traded derivatives. The standards also recommend that firms report their value-at-risk. The U.S. accounting standards are rapidly becoming the international reporting norm for those major financial institutions that are active in the dollar markets.
The improvement of accounting standards is also high on the agenda of the BIS. In September 1994, the Euro-Currency Standing Committee of the BIS issued a report on the public disclosure of market and credit risk that emphasizes the importance of quantitative measures of risk. The report recommends that banks disclose enough information for market participants to judge the bank’s average risk exposure. For this purpose, the report suggests that banks disclose summary statistics on their value-at-risk over different holding periods. The report also recommends that ex-ante estimates of value-at-risk be compared with actual outcomes to identify the historical performance of an institution’s risk-management system.
Although globally integrated capital markets are rapidly becoming a reality, the legal, accounting, and, to some extent, the supervisory and regulatory infrastructure have remained national in outlook and application. Capital markets in general, and derivative markets in particular, are global markets. Barings, headquartered in London, established futures and options positions on exchanges in Osaka, Singapore, and Tokyo. Although the resolution of the Barings debacle was made easier by the active cooperation among the regulatory authorities in London, Osaka, Singapore, Tokyo, and other major financial centers, there is much room for improvement. Efforts toward greater global cooperation42 are required in areas such as coordination between the Basle Committee and the International Organization of Securities Commissions (IOSCO) in harmonizing financial regulation for banks and securities houses; the exchange of information among securities regulators; and, to some extent, coordination between the Basle Committee and the European Commission on capital requirements for banks.
In the area of capital requirements, the EU has tried to accommodate the new internal model-based approach proposed by the Basle Committee by allowing top European banks to follow the Basle approach to some extent. In April 1995, the commission’s directorate-general agreed that daily value-at-risk models can be used in conjunction with the CAD provided that the capital requirement according to the firm’s estimate of value-at-risk is no less than the capital requirement according to the CAD. In any event, should the Basle proposal be adopted, the EU has to amend its CAD sometime after its implementation on January 1, 1996. Without full harmonization, EU banks and non-EU banks will be competing on somewhat unequal grounds. Full harmonization between the EU and the Basle Committee is challenging because of their different approaches to banking and banking supervision between European and non-European regulators.
The Basle Committee and IOSCO have attempted to coordinate their efforts in supervising the derivative activities of banks and securities firms, respectively. This coordination effort has proven to be difficult, in part because many European countries have a universal banking system under which banks can perform securities activities directly, while the U.S. and Japan still separate much of the banking and securities business. Furthermore, capital requirements for U.S. securities houses, imposed by the U.S. Securities and Exchange Commission and the U.S. Commodities Futures Trading Commission, focus primarily on liquid capital—defined as capital minus illiquid assets (called haircuts)—with the aim to ensure that registered broker-dealers have adequate liquid assets to meet their obligations to investors and creditors. Little progress has been made in this area to date. The Basle Committee and IOSCO have been able to coordinate their efforts on improving risk management, however, and they have jointly issued guidelines on risk management related to derivatives.
A final concern relates to the coordination among futures and options exchanges and among the authorities with regulatory responsibilities for the exchanges. Most exchanges compete across countries for international business, and how international futures and option exchanges can coordinate the sharing of information regarding clients and market activity without losing their competitive edge is an important challenge in the period ahead (see Box 2). One of the lessons of the Barings collapse is that had there been greater information sharing among regulatory authorities in Singapore and the United Kingdom, and between the futures exchanges in Singapore, Osaka, and Tokyo, the problems experienced by Barings may have surfaced early enough to prevent the collapse. For example, SIMEX, the futures exchange in Singapore, did not have access to information regarding Barings’ positions in Nikkei futures contracts on the futures exchange in Osaka (OSE). Thus, officials at SIMEX could not determine if Barings’ SIMEX position was in fact hedged by offsetting positions on OSE, as was claimed by Barings. Had officials at SIMEX and OSE been able to view Barings’ consolidated position in Nikkei futures contracts in real time, it is likely that both exchanges would have required greater variation margin and prevented Barings’ open position from becoming so large. When the same financial contracts—such as futures and options on the Nikkei 225 index—are cross-listed internationally, as they are on the exchanges in Osaka and Singapore, competition for international business will often put the exchanges in the difficult position of striking the proper balance between gaining a competitive edge, efficiency, and prudential concerns. The importance of this issue in Europe is reflected by the intense competition among the futures exchanges in Frankfurt, London, and Paris.
Strengthening Risk Management at Futures Exchanges
During the past five years, organized futures exchanges have become sufficiently important to the financial system that a disruption in one of the major futures exchanges—as occurred in Hong Kong in 1987—would likely have serious repercussions. As a result, and in the wake of the Barings’ failure, their risk-management practices have come under increased scrutiny.
Under current institutional arrangements, the clearinghouses of futures exchanges are the central counterparties to all trades and guarantee execution of all transactions executed on their exchanges. The daily procedure of marking-to-market and collecting margins (even intraday if necessary) together, reduce the credit exposure of the clearinghouse to not more than one day’s worth of losses on a counterparty’s position. In addition, a guarantee fund maintained by the exchange absorbs potential losses if a counterparty defaults. Exchanges also establish loss-sharing rules that are used to allocate losses among clearinghouse members in the event of a default. Finally, some exchanges impose a number of prudential requirements on their members, including minimum capital requirements and limits on the size of clearing members’ net positions to reduce the risk that any single member can assume and consequently minimize the risk exposure of the clearinghouse.
In the aftermath of the Barings failure, the futures exchanges in Singapore, Osaka, and Tokyo were able to liquidate Barings’ positions quickly enough so that the exposure of clearing members of the exchanges was limited. However, the Barings episode exposed uncertainties in the loss-sharing formulas of the exchanges. These uncertainties are currently being resolved. In addition, the multinational nature of Barings’ business raised questions about the treatment in bankruptcy courts of margins and guarantee funds held by the exchanges.
Continuing Resolution of Banking Problems in Several Industrial Countries
A number of industrial countries have experienced more traditional on-balance-sheet loan losses in their banking sector in recent years. Deregulation of banking in the early 1980s permitted banks in, inter alia, Japan, the Nordic countries, the United States, and the United Kingdom to tap new sources of funds and enter new types of activities—often in areas where they had little prior experience—in an increasingly competitive environment. In addition, rapid asset price inflation and an expansion of lending into real estate tend to reinforce each other. The increased risk taken by banks was exposed by the onset of a worldwide recession in the late 1980s and early 1990s. The usual cyclical decline in asset quality was made worse by a relatively large deflation of commercial and residential real estate values. As a result, the losses incurred by banks in a number of industrial countries were sufficiently large that banks had to slow the growth of their loan portfolios to meet regulatory capital requirements, thus contributing at times to the so-called credit crunch. In some instances, resolution of the nonperforming loans problem required direct intervention of the public sector. In retrospect, it seems that the supervisory infrastructure did not adapt quickly enough to the new competitive environment.
In most of the countries that experienced banking problems in the late 1980s and early 1990s, the situation has improved significantly. While the recovery of profits has been strong in the United Kingdom and the United States, it has been stronger still in Norway and Sweden. In these countries, banks wrote off large portions of their loan books, sold other assets and subsidiaries, sharply reduced noninterest expenses, and, in some cases, achieved consolidation through mergers. Banks in these countries were recapitalized, in some cases by the government and in others by raising private capital. Operating profits increased sharply in 1994, mostly reflecting the decline in provisions. Indeed, the recovery in profits has been so strong that it has raised questions about whether the extent of government assistance to the industry was excessive. The current state of profitability in the industry is such, however, that the authorities in both Sweden and Norway expect to recover a large proportion, perhaps all, of the funds they made available to the banking sectors.
The different resolution strategies employed in dealing with banking losses imply different allocations of losses among the involved parties. Two extreme examples of burden sharing can be identified. The first is a policy of forbearance, under which banks are permitted to carry on with their impaired capital positions in the expectation that they will be able to “earn their way out of trouble” through wider intermediation spreads and improved cash flow generated by the eventual economic recovery. Such an approach broadly characterizes the initial response to problems in the U.S. savings and loan industry and to the problems among major banks in Japan.
The alternative, more radical, approach involves a determined effort to identify problem assets, and devise a work-out program including, in most cases, significant use of public money to recapitalize the banks, while at the same time, strengthening the supervisory and regulatory infrastructure. Norway, Sweden, and Finland followed this approach, with banks in Finland and Sweden creating separately capitalized work-out units (“bad banks”) to which the nonperforming assets were transferred.43 The worst affected banks were made whole by injections of public funds from the government, specialized government agencies, or the central bank. In some cases, these funds were used to facilitate mergers. However, since the rehabilitation efforts covered only banks that were insolvent, this approach raised concerns about competitive equality in the banking industry, and the authorities took care not to give the recapitalized bank an advantage over its competitors. In addition, there were concerns about the moral hazard generated by such blanket rescue plans; however, this event was viewed as unique in the financial history of the countries concerned, and a strengthened supervisory regime was put in place to forestall any further risk taking by depository financial intermediaries.
Box 2.Competition and Cooperation Among Futures Exchanges
Financial centers have emerged throughout the world either to provide financial services to a large and growing economy, such as in London, New York, and Tokyo, or to intermediate financial transactions between a region and other geographic sources of financing, such as in Hong Kong and Singapore in the southeast Asian region. Many of the financial services provided by these financial centers—including foreign exchange transactions, underwriting and syndication services, equity trading, over-the-counter derivative activities—in effect have no natural geographic home. These activities tend to migrate to markets that offer advantages of location, reflecting either a concentration of economic and financial activity or fiscal, regulatory, and cost advantages.1
One area where competition has been fierce and where cooperation is important is on the futures exchanges. The instruments traded on futures markets—stock index futures, currency futures, and bond futures—can be characterized as not having “a natural home,” as their underlying instruments are often “foreign” to those markets. For instance, the London International Financial Futures and Options Exchange (LIFFE) trades German and Italian government bonds futures, and the Singapore and Chicago futures exchanges trade Japanese stock index futures and options. When financial instruments are standardized and cross-listed on several exchanges, the exchanges compete primarily on the basis of transactions costs (commissions, transactions taxes, and margin requirements), and on the basis of liquidity, the efficiency of clearing and settlement systems, and the sophistication of trading mechanisms. To strengthen their competitive position and establish themselves as global players, futures exchanges have been forging alliances to extend their trading activity across different time zones and to enlarge the geographical distribution of their products. For instance, in 1993, the Marché à Terme International de France (MATIF) and the Deutsche Terminbörse (DTB) signed a cooperation agreement that would allow members from one exchange to have direct access to selected products of the respective partner exchange. The exchanges signed the agreement primarily to consolidate their positions within the European time zone, and because of the complementarity of their products, clients, and technologies. A recently announced trading link between LIFFE and the Chicago Board of Trade (CBOT) is another example of how futures exchanges have attempted to increase the global distribution of their products. The trading link would allow LIFFE’s products (e.g., Bunds and Gilts futures and options) to be traded on CBOT’s floor, thereby extending LIFFE’s products to both the U.S. and Asian markets, as the time zones of both regions overlap. Similarly, CBOT’s products (e.g., futures and options on U.S. Treasury bonds) would benefit from the European trading hours. Other exchanges are also considering expanding the distribution of their products through bilateral linkages, including the Hong Kong Futures Exchange and the Philadelphia Stock Exchange, and LIFFE and SIMEX.
|Trading Value in||Trading Value in SIMEX|
|OSE||SIMEX||Trading Value in OSE|
|(In billions of Japanese yen)||(In percent)|
The competition between the Osaka Securities Exchange (OSE) and SIMEX is a particularly revealing example of how derivatives exchanges compete. Both exchanges trade a similar futures contract on the Japanese Nikkei 225 stock index.2 In 1989, the first full year of trading in the OSE, turnover of Nikkei 225 futures on the OSE was ¥ 189 trillion, compared with ¥ 15 trillion on SIMEX—or less than 8 percent of the OSE’s volume (see table on the left). By 1991, trading on SIMEX had declined to less than 2 percent of the OSE’s volume. Since 1991, however, SIMEX’s share of trading in Nikkei 225 futures has increased significantly, reaching about 47 percent of OSE’s trading in 1994.
|Fixed rate fees||2.4||—||0.9||—|
The shift of trading activity from the OSE to SIMEX during this period can be attributed to two factors. First, between the beginning of 1990 and the end of 1991, the OSE tripled its margin requirements, from 9 percent of the futures contract value to 30 percent, to limit volatility spillovers from the futures market to the cash market. The OSE also doubled the size of its commissions in 1993 to protect the fixed commission structure in the Japanese cash markets. Second, SIMEX reduced its margin requirements during this period, to reflect a reduction in price volatility.3 As a result, the transactions cost differential between the two futures exchanges has widened significantly. By early 1995, OSE transactions costs for customers were more than twice those on SIMEX, with commissions accounting for most of the difference (see table above). The difference in transactions costs between the two exchanges also led to a segmentation of the market for stock index futures, whereby Japanese retail investors are largely confined to trading on the OSE, while all other investors, including Japanese institutional investors and securities companies, trade on SIMEX.
Concerns have been expressed that competition between futures exchanges will inevitably lead to a weakening of prudential regulations to the lowest common denominator (“a regulatory race to the bottom”), thereby increasing the potential for systemic problems. Experience, however, suggests that competition between exchanges in the United States, Europe, and Asia has not led to a relaxation of prudential regulations and risk controls to capture financial activity from competing exchanges. SIMEX has been able to gain market share from OSE because it offered more competitive transactions costs without compromising on the quality of its prudential regime. In fact, Singapore has been promoting the quality of its regulatory environment and the related greater security and stability as one of the attractive aspects of its financial center. Similarly, competition among the European futures exchanges has tended to concentrate on offering higher liquidity and greater security of transactions, rather than competing with lower margins or looser risk controls.
Competition between futures exchanges for international business has not, however, encouraged widespread cooperation and sharing of information among the exchanges. Yet, as exchanges have been determined to avoid regulatory arbitrage, there is no reason to believe that there is an inherent contradiction between competing on transactions costs and exchanging information for prudential reasons. The recent Barings collapse is a clear case where greater coordination of information between the OSE and SIMEX might have raised early warning signals that could have averted the collapse of the securities firm. In particular, as officials at SIMEX did not have access to information regarding Barings’ positions in Nikkei futures contracts on the OSE, they were unable to determine whether Barings’ position on the Singapore exchange was in fact hedged by offsetting positions on the OSE, as was claimed by Barings. Had officials at SIMEX and OSE been able to view Barings’ consolidated position in Nikkei futures contracts, both exchanges would have probably required higher margins and forced Barings to reduce its open position.1For instance, London and New York’s stock markets alone account for 90 percent of equities traded outside their home countries; the market for Japanese warrants is more developed in London than in Japan; 90 percent of the world’s outstanding Eurobonds are listed in Luxembourg; and Singapore is the world’s fourth largest foreign exchange center.2The OSE and SIMEX are the most important exchanges for Nikkei 225 futures trading. The Chicago Mercantile Exchange (CME) also trades Nikkei futures, but its share of the total volume is about 3 percent.3Margins were also cut to bring them in line with those of other exchanges, as SIMEX had initially set them well above what standard practice required.
The resolution of the problems in the U.S. savings and loans industry followed the second approach after the initial forbearance approach proved to be ineffective. The Resolution Trust Company, a liquidation agency established to assume the nonperforming assets of institutions that could not meet the regulatory capital standards, required in the end a total of almost $200 billion of public funds to contain the savings and loans problem. In Japan, an asset liquidation company, the Cooperative Credit Purchasing Company (CCPC), was established by banks to take over their bad debts and thereby allow them to take advantage of tax write-offs by realizing losses. This balance-sheet operation overstates the extent to which the banks have resolved their nonperforming loans problems, because the banks remain liable for the losses incurred by the CCPC in the sale of the assets it acquired from them. So far, in Japan, no direct subsidies have been provided by the government (aside from deposit insurance) to the major banks or the CCPC.
The experience with banking problems and their resolution in industrial countries offers a number of generally applicable lessons. First, the bank supervision structure has to evolve quickly in response to financial deregulation. Second, while the goal of bank supervision should be to identify and resolve problems in individual institutions early on, including through bank restructuring, sale or merger, it is highly desirable to set up a clear regulatory rule for closing an institution whose problems cannot be resolved before it becomes insolvent. Alternatively, if it is decided to let a bank with insufficient capital continue to operate in the hope that it may earn its way out of its predicament, then regulatory scrutiny of its activities should be increased. Insolvent or nearly insolvent institutions have strong incentives to undertake high-risk lending strategies with insured deposits in the hope that they might earn their way back into solvency. Third, it is important to establish a clear rule on how to share the nonperforming loans burden among the stake holders without increasing the risk of a systemic disturbance. Once significant losses have accrued on a bank’s balance sheet, they sooner or later have to be shared among the bank’s shareholders, its liability holders,44 its borrowers, and the public sector.
The importance of addressing problems in the banking sector at an early stage is illustrated by two examples, one from Japan and one from France. In Japan,45 the risk in the forbearance approach is evidenced in the most recent failures of two credit cooperatives in Tokyo. The authorities46 were alerted through a special examination in 1993 of a significant nonperforming loans problem but did not close the insolvent credit cooperatives. Instead, a policy of forbearance was adopted. As a consequence of the policy, the two credit cooperatives rapidly expanded their balance sheets—attracting deposits with above-market interest rates and lending to high-risk projects—and with it the volume of their nonperforming loans.
The lack of a clear rule for sharing the burden of recapitalizing weak banks became apparent in the episode of the two failing credit cooperatives. In the past, it had been common to resolve problems through a merger between an insolvent institution and a stronger one, at times supported by an injection of deposit insurance funds.47 However, the condition of the banking system is now such that it is becoming more difficult to persuade the larger banks, which are beset by their own problem loans, to acquire insolvent banks.48 In addition, the failure to deal early on with problems at the smaller institutions may have increased significantly the number of cases where intervention is needed.
The problems among some of the smaller Japanese depository institutions (credit cooperatives) and other specialized financial institutions (especially housing financing companies, commonly known as jusen) also remain a source of concern. In the second half of the 1980s, the housing finance companies (nonbank subsidiaries of banks) greatly expanded their lending to real estate companies with funds borrowed from city banks, agricultural cooperatives, and other financial institutions. A large portion of their loans became nonperforming in the early 1990s.49 All seven jusen negotiated a ten-year plan for interest payment relief in 1993 in order to avoid default on interest payments to their creditors. To date, however, it has not been possible to devise a scheme to reduce the stock of nonperforming assets in the jusen.
The elimination of problem loans from the smaller institutions in Japan has been made more difficult by inadequate disclosure and poor supervision of the smaller institutions. For example, credit cooperatives are formally under the supervision of the local governments, which in some cases are not well equipped to conduct financial examinations, and, unlike the larger banks, they are not required to disclose non-performing loans. Poor disclosure has meant that market discipline could not play its role because depositors and investors were unable to distinguish between the relatively strong institutions and the weaker ones.
The failure to respond quickly to resolve banking problems is also observed in France, on a scale comparable with that of the jusen problem in Japan. Problems with nonperforming assets and weak capital support have continued at the state-owned Crédit Lyonnais. The sources of these problem loans at Europe’s largest bank were its very aggressive expansion into real estate, cross-border lending, and other nontraditional areas, which began in the late 1980s. The subsequent recession in France in the early 1990s led to a significant increase in nonperforming loans. Four features of Crédit Lyonnais’s balance sheet appear to have made it particularly vulnerable to problem loans: the speed with which it expanded its operations exceeded management’s capacity to monitor and control them; lending to the small and medium-sized firms increased significantly as a proportion of total lending; it became heavily exposed to real estate; and its loan portfolio was heavily concentrated in a few borrowers. The magnitude of the recession in France and the crisis in the property market, in which Crédit Lyonnais was one of the main participants, contributed to aggravating the situation. The bank’s reliance on the implicit guarantees of the state, its main shareholder, may also have contributed to its aggressive lending policy.
While the bank’s problems had been recognized as early as mid-1991 by the bank’s supervisor, Commission Bancaire, these problems were not made public until late 1993 and became critical in early 1994 when the government injected capital to maintain the bank’s risk-weighted capital above the Basle minimum of 8 percent. The rescue operation entailed a F 4.9 billion capital injection and the decision to ring-fence, in a Crédit Lyonnais fully owned subsidiary, F 43 billion of nonperforming loans; of this total, F 18.4 billion was guaranteed by the government. Problems at the bank continued to mount in 1994, and at the end of the year, a F 12 billion loss was reported. In March 1995, a second rescue operation was proposed. Crédit Lyonnais will lend F 155 billion to the state-controlled Societé de Participations Bancaires et Industrielles (SPBI), which will use that amount to finance a new special purpose vehicle, Consortium de Réalisation (CDR), and purchase zero-coupon government bonds with a face value of F 35 billion in 20 years. CDR will acquire from Crédit Lyonnais about F 135 billion in assets of which roughly two thirds were nonperforming—including the F 43 billion segregated in 1994—and one third were performing assets, including the bulk of Crédit Lyonnais’s large portfolio of industrial shares. SBBI, which is fully liable for the F 145 billion loan (plus interest), will receive the income from the asset sales carried out by CDR and direct injections from the state. These injections are expected to be repaid by special dividends and a fee of 34 percent of the pretax profits of Crédit Lyonnais if profits are under 4 percent of shareholder funds, or 60 percent of pretax profits if the latter are above the 4 percent return, and capital gains from future privatization of the bank.
The continuing difficulties, and the costly rescue of Crédit Lyonnais, might have been averted had the bank conducted a less expansionary policy and strengthened its internal control procedures, including risk management and controls over its subsidiaries, and had the authorities acted sooner to deal with these problems. A belief that loan quality would improve as the economy recovered seems to have convinced all parties that drastic action was not required. The modalities of the rescue of Crédit Lyonnais have generated serious concerns on the part of its French competitors. This issue is currently under scrutiny by the EU Commission.
Orange County lost $1.5 billion, due to a high-leverage strategy involving reverse repurchase agreements that allowed the county to leverage its assets up from $7.5 billion to $20 billion. When interest rates increased, the value of these instruments dropped significantly. The county was forced into bankruptcy when dealers refused to roll over the reverse repurchase agreements.
MG Corporation, a U.S. subsidiary of Metallgesellschaft AG, Germany, lost almost $1.3 billion through the use of petroleum derivative products. MG had contracted to supply petroleum products to retail distributors at fixed prices, and it had taken a long position in short-term oil futures contracts to hedge its forward commitments. When oil prices dropped significantly in 1993, MG incurred large losses on these hedges, and had to meet margin calls on its outstanding futures contracts. However, while U.S. accounting principles allowed the losses on the futures contracts to be offset by the unrealized gains on the forward contracts, German accounting did not recognize such gains. As a result, and because of doubts about the creditworthiness of the counterparties to the forward contracts, MG’s parent did not use the gains in the forward contracts as collateral for borrowing the necessary margin money. In the event, MG Corp. had to close out the futures positions at significant losses. Culp and Miller (1994) argue that the hedging strategy of MG Corp. was correct, but that the unwillingness of its parent to continue financing margin calls led to a disorderly liquidation of the futures positions, which greatly increased losses on these contracts.
Notional principal amount is the number by which the interest rates or exchange rates in a derivative contract are multiplied to calculate the settlement amount.
This outstanding balance includes interest rate futures and options, currency futures and options, and stock market index futures and options.
The replacement value is the unrealized capital gain or loss of the contract at current market prices; that is, it is the amount that would have to be paid to a third party to induce them to enter into a transaction to replace the contract.
Value-at-risk is an estimate of the maximum loss that a portfolio could generate with a given level of confidence and during a given period into the future. For example, a one-day value-at-risk of $100 million at a 95 percent confidence interval means that there is a 95 percent chance that the loss in portfolio value during the next day will be less than $100 million. In order to estimate value-at-risk it is necessary to estimate the parameters of the probability distribution of the portfolio return.
For further analysis of the capital requirements controversy and this new approach, see the background paper “Capital Adequacy and Internal Risk Management,” pp. 135–49.
In September 1994, the Basle Committee asked a number of banks, active in the over-the-counter derivative markets, to compute risk exposures and capital requirements for four hypothetical portfolios with and without options. There are several reasons why banks might arrive at different estimates of value-at-risk for the same portfolio. First, different parametric and nonparametric approaches to estimate the probability distribution might be used. Second, options risk might be treated differently. Third, different dynamic models of the asset returns might be used. Finally, different sample periods might be used for estimation.
While a bank will be free to choose its own model and its own estimation and simulation procedures, estimates of value-at-risk will be subject to certain quantitative standards and assumptions: a 99 percent confidence interval; a ten-day holding period; a sample period of not less than one year; data set updates at least every three months; and certain aggregation rules. Banks also will have to satisfy some qualitative standards: the maintenance of an independent risk control unit; the integration of its model into day-to-day risk management; the use of stress testing; and the adoption of a periodic review process for its risk measurement and risk-management system.
For proposals to improve transparency and disclosure in derivatives markets see the background paper “Initiatives Relating to Derivatives,” pp. 150–57.
Regional cooperation in the financial area is advancing within the EU under the Single Market legislation.
For details on the restructuring programs in each country see International Monetary Fund (1993c) and (1994). In November 1993, the Finnish government established an asset-management company, Arsenal Ltd., to which the nonperforming assets of the Savings Bank of Finland were transferred at the time of the sale of the bank to a consortium of private banks. The initial capital injection of Fmk 5 billion by the state and the Government Guarantee Fund was augmented by Fmk 6 billion in September 1994.
Attempts to share losses with depositors run the risk that deposits will be withdrawn from the banks perceived as weak, which is desirable from the point of view of market discipline but raises the specter of a general run on deposits in other, including solvent, institutions.
For detailed discussions of the causes of balance sheet fragility in the Japanese banking system see International Monetary Fund (1993c) and (1994), and for a detailed analysis of the asset price cycle see International Monetary Fund (1992) and (1993b). The volume of nonperforming loans in the 21 leading banks decreased slightly in 1994 to around ¥ 13 trillion (about 3 percent of total loans), as banks were more aggressive in disposing of bad assets (mostly to the CCPC) and increased the rate at which they added to loan-loss reserves.
When Tokyo Kyowa and Anzen Credit Unions failed, the Tokyo metropolitan government, along with the Bank of Japan and the Ministry of Finance, arranged a scheme to set up a new bank, which was to acquire all business of these two credit cooperatives—called Tokyo Kyodo Bank—with an injection of capital from the Bank of Japan and private financial institutions, along with financial assistance provided by the Tokyo metropolitan government, the Deposit Insurance Corporation, the Long-Term Credit Bank, the National Federation of Credit Cooperatives, and other private financial institutions. The Japanese authorities announced that the scheme was indispensable for closing the two credit cooperatives without causing a systemic disturbance under the current conditions surrounding the Japanese financial system. They also stressed that they would see to it that the management and shareholders of the two credit cooperatives should be held responsible for the failure, in order to minimize concern for moral hazard.
There have been six cases in which the Deposit Insurance Corporation has provided financial assistance to mergers.
Contributing to this difficulty is the fact that as a result of recent deregulation of financial institutions, banks and securities companies can now move into new lines of business by establishing subsidiaries rather than through mergers.
As of September 1994, the total borrowing of the seven jusen companies was about ¥ 13 trillion, of which ¥ 6 trillion was owed to banks, and about ¥ 5.5 trillion was owed to the financial institutions for agriculture and forestry. The jusen companies are widely reported to hold ¥ 6 trillion in nonperforming loans.