II Deterioration of Bank Balance Sheets
- International Monetary Fund
- Published Date:
- January 1993
Is There a Common Thread?
Although banking cannot lay uncontested claim to being the world’s oldest profession, it is clear that the principles that have helped to define sound banking behavior have a long history. At least five of those principles—namely, avoid an undue concentration of loans to single activities, individuals, or groups; expand cautiously into unfamiliar activities; know your counterparty; control mismatches between assets and liabilities; and beware that your collateral is not vulnerable to the same shocks that weaken the borrower—remain as relevant today as in earlier times. Indeed, behind all of the banking or financial sector crises that have emerged in industrial countries over the past decade—ranging from the developing country debt crisis of the early 1980s, to the saving and loan crisis in the United States, to the bank solvency crisis in Finland, Norway, and Sweden, to the spate of banking strains elsewhere—at least one of those principles has been violated.
There is also little dispute that when such banking and financial crises do occur, their resolution can be costly. For example, it has been estimated that the saving and loan crisis in the United States carried a $180 billion price tag for the taxpayer (equivalent to over 3 percent of GDP) (Table 1). Similarly, the loan losses realized in 1991–92 by banks in Finland, Norway, and Sweden amounted to between 4.2 and 6.7 percent of these countries’ respective GDP and carried with them a substantial degree of government financial support. Moreover, it has by now been amply demonstrated how large actual and projected losses in the banking system can weaken medium-term fiscal consolidation efforts, limit the room for maneuver of monetary and exchange rate policy, and—where banks overreact to past losses by taking an unduly restrictive approach to credit evaluation—restrain the pace of economic recovery.
|Savings and loans||…||…||3.25|
Given all this, a crucial question to ask as a prelude to discussing current banking problems in the industrial countries is why after a long postwar period of stability have banking problems become so widespread and occurred with such frequency. What is it that induces banks in different countries to abandon—seemingly periodically—the principles of sound banking. Although no single answer can do full justice to differing macroeconomic circumstances, institutional structures, and paces of liberalization, at least one common thread running through many recent banking crises merits mention. That thread is the recognition that the competitive pressures unleashed by financial liberalization do not merely increase efficiency: they also carry risks, as banks and other financial institutions alter their behavior to ward off institutional downsizing.
A summary of the forces at work runs along the following lines. Before the financial liberalization of the 1980s, cartelized banking markets, in concert with a host of regulations, served to restrict competition in the financial services industry. Banks and other financial institutions thus enjoyed a financial cushion in the form of excess profits (rents); in fact, banks that operated in these regulated markets were among the most profitable in the world. These rents were essentially obtained from the consumers of financial services who had limited alternatives. An important component of the market value of bank capital under these regimes was the discounted value of the rents that banks had expected to earn in the future. Innovation and liberalization opened many financial markets to intense competition for these rents. Prime borrowers, for example, realized that they could meet a high proportion of their financing needs at lower cost by accessing markets directly via the issuance of commercial paper. Depositors too began more and more to have access to higher-yielding alternatives to bank deposits. This competition in turn increased banks’ own marginal cost of funds. In addition, banks’ cost of equity capital typically rose, as equity markets reacted, inter alia, to banks’ loss of market share in the AAA wholesale market. With both higher funding costs and a lower average quality of client, banks’ required rate of return on assets rose also.
Faced with a potential downsizing of their operations, many banks responded to this new, less hospitable environment by increasing the riskiness of their portfolios—a strategy that was also encouraged by the existence of a net of explicit and implicit government guarantees that both protected depositors and made “failure” a less credible deterrent to excessive risk taking. This increased tolerance for risk shows up in several dimensions, including a particularly rapid expansion of bank balance sheets in the countries with the most pressured banking systems (ratios of aggregate bank balance sheets to GDP jumped by 50 percent or more within several years), in a growing concentration of exposure to single risk classes (often in lending to real estate at the expense of lending to traditional manufacturing customers), and in the mushrooming of banks’ participation in the over-the-counter (OTC) derivative markets (as discussed in more depth in Section III).1
The reduced segmentation in the financial system also allowed intermediaries to seek replication of each others’ profit centers. Later entrants, however, often found that projects with the best risk/return trade-offs had already been taken—forcing them to take on less preferred exposures; in addition, imitators often did not have the same experience and expertise in evaluating new activities as earlier entrants. Bank supervisors frequently found that guidelines appropriate for an environment in which banks were highly profitable and had little incentive to risk their capital by exploring risky areas outside the range of supervisor expertise were less appropriate for this new, more competitive financial landscape—but changes in those supervisory guidelines usually lagged behind developments. In most cases, the increased riskiness of bank portfolios did not translate into an increase in banks’ net operating profits (when scaled by the size of their balance sheets). In the end, the increased risks taken by banks were often exposed and turned into losses by a significant shift in economic conditions—whether a tightening of monetary policy, a large sectoral shock, a decline in asset prices, or a prolonged period of slow economic activity.2 Even loans that were collateralized (such as in real estate) were adversely affected, as the same shock that reduced the borrower’s ability to pay simultaneously reduced the value of the collateral. Where these risks were not adequately protected by increases in banks’ operating profits or a comfortable cushion of capital, the losses subsequently spilled over into the public domain.3
Recent Banking Problems in Some Industrial Countries
Balance sheet problems in banking are widespread among the major industrial countries (Germany is a notable exception).4 Banking sectors in Finland, Norway, and Sweden have been particularly hard hit. Significant problems have also been evident in Japan and in the United States, and to a lesser extent, in France and the United Kingdom.5 This subsection outlines the recent experience and policy approaches in a sample of industrial countries; however, this sample should not imply that banking systems in the countries not discussed in this report are free of problems.
An antecedent to the banking problems in each of the countries examined was a significant change in the pattern of bank lending following an episode of either financial innovation or liberalization or both. These changes included a sharp expansion in bank lending relative to nominal GDP and, in some cases, a growing concentration of loans to particular sectors of the economy.6 The realized and expected losses imposed significant burdens on fiscal, and in some cases monetary, policy as governments made efforts to preserve financial stability.
Chart 1.Nordic Countries: Real Estate Prices
Sources: Central Bureau of Statistics and Den norske Bank—Economics Secretariat, Norway; Enskilda Research, Sweden; Statistics Finland; and Statistics Sweden.
Chart 2.Nordic Countries: Indices of Stock Prices
Sources: Central Bureau of Statistics, Norway; Statistics Finland; and Statistics Sweden.
Banking Crises in Finland, Norway, and Sweden
The banking sectors in three Nordic countries-Finland, Norway, and Sweden—have each been beset by heavy loan losses, necessitating direct government support for a number of banks, including the largest institutions in each of these highly concentrated sectors. As in the United States and Japan, loans to troubled real estate sectors played a prominent role in these developments.7 The losses have not been confined solely to real estate loans, however. In Norway and Finland, credit exposures to the retail, service, and household sectors were also troublesome.
Among these three Nordic countries, the first signs of trouble emerged in Norway in 1987, following a period of several years during which domestic credit rose at an annual rate of more than 20 percent. The removal of credit rationing in 1984 triggered a channeling of bank activity into riskier, more unfamiliar markets, which resulted in strong increases in losses from 1987. Together with rising competition for funds, as well as losses on equity and exchange trade, negative profits resulted.
After a period of rapid credit expansion in 1988–89, banks in Finland began to experience a sharp rise in loan losses. Liberalization of financial markets was followed by an increase in relatively high risk lending by banks and by their increased involvement in securities market speculation. In this case, however, the sharpness of the ensuing banking crisis in 1991 and 1992 was accentuated by the collapse of the real economy. The resulting nosedive in the cash flow of enterprises (that were formerly good credit risks) created balance sheet problems for borrowers, which was exacerbated by reduced bank lending, induced in turn by the decline in banks’ own capital. In addition, the banks suffered from a collapse of the stock market that forced a write-down of bank share holdings. Real estate losses mounted with the collapse of economic activity. Losses from speculation in currency and securities markets also took a serious toll.
Compared with banks in Norway and Finland, banks in Sweden experienced a particularly abrupt increase in loan losses in 1990–92, following a rapid credit expansion in 1986–89. A liquidity crisis in the finance company sector, arising after heavy losses, triggered the crisis.
Although unique factors were at work in each country, the recent banking troubles in Norway, Finland, and Sweden share a number of common elements. Up to the late 1970s, each country had a highly regulated financial system that was dominated by a handful of large banks. The regulatory regimes consisted of controls on interest rates and officially directed credit, and capital controls.
In Norway, the Monetary and Credit Policy Act of 1965 provided for the detailed control of both interest rates and credit volume. Interest rates were maintained at relatively low and stable levels. The state banks received lending quotas as part of the budgetary process, whereas the credit extended by mortgage companies was indirectly controlled through limits on bond issues. Variations in primary and supplemental reserve requirements controlled lending by private banks and finance companies. Foreign exchange regulation was adapted to the objectives of the credit policy and, in particular, to effect the rationing of credit that this policy entailed.
In Finland, bank lending was subject to direct limits on interest rates and indirect limits on volume. Loan rates were constrained by a ceiling on average lending rates, and loan volume was controlled by adjusting the quotas on central bank advances or by altering the spread between the central bank’s borrowing and lending rates. With respect to their liabilities, banks were required to pay a uniform low interest rate on tax-exempt deposits, which facilitated the formation of cartellike arrangements and reduced competition for private funds. The Bank of Finland also made extensive use of exchange controls to limit capital outflows and to direct capital inflows to favored sectors.
An important aspect of economic policy in Sweden was the maintenance of low and stable interest rates. To counter the impact of this policy on credit expansion, the Bank of Sweden relied on moral suasion and, during periods of restrictive policy, on quantitative controls on the volume of commercial bank lending to the public. Liquidity ratios provided a means for directing credit to the Government and to the housing sector on favorable terms, since bonds used to finance these sectors were included in the definition of liquid assets. Foreign exchange controls were used to isolate the domestic market from international developments.
The forces that propelled financial liberalization in the Nordic countries were similar to those in other industrial countries, although they did not emerge until the early to mid-1980s. Chief among these was the emergence of new markets and non-bank financial institutions that escaped existing regulations and the strains imposed by the need to finance increasingly large public sector deficits.
In Norway, the banks in the early 1980s circumvented existing regulations by acting as brokers to mediate loans directly from lenders to borrowers. In some but not all cases, these so-called market loans had a bank guarantee. In the later part of the 1980s, banks sought to avoid limits on credit volumes by selling assets and by booking loans in affiliates that had broad activity powers, such as finance companies (Chart 3). Finland also witnessed the shifting of activity to a so-called gray market as higher inflation resulted in negative real interest rates and a greater excess demand for credit in the early 1980s. The gray market involved the arranging of loans through off-balance-sheet operations at banks’ trust departments and, later, lending by bank-owned finance companies. In Sweden, large public sector deficits in the first half of the 1980s fostered the development of domestic money and securities markets. The tight regulatory policies encouraged the growth of financial intermediaries outside the regulated sector. In particular, finance companies, specializing in leasing and lending to households, expanded rapidly (Chart 3). The expansion in money and securities markets and the rise of nonbank financial intermediaries put pressure on the authorities to undertake financial liberalization measures.
Chart 3.Nordic Countries: Ratio of Lending by Finance Companies to Nonbank Public to Lending by All Banks to Nonbank Public
Sources: Central Bureau of Statistics, Norway; and Sveriges Riksbank.
In Norway, the lifting of quantitative limits on bank credit in January 1984 preceded by more than a year the freeing of interest rates. This led to a surge in the effective demand for bank credit which reached a ratio of 85 percent of nominal GDP in 1986, up from 64 percent in 1983. The liberalization of interest rates in September 1985 and the beginning of a recession in early 1986 slowed the rate of growth somewhat (Chart 4). A tax reform in 1988 also raised the after-tax cost of borrowing by lowering the top marginal tax rate to 48 percent from 66 percent.8
Chart 4.Nordic Countries: Ratio of Bank Credit to GDP
Sources: Central Bureau of Statistics, Norway; Statistics Finland; and Sveriges Riksbank.
Evidence of credit problems started in 1987 in the form of high loan loss provisions by commercial and savings banks (Table A1). Over the next two years, the two industry-operated deposit insurance funds (the Commercial Banks Guarantee Fund (CBGF) and Savings Banks Guarantee Fund (SBGF)) and the Norges Bank came to the assistance of a number of smaller Norwegian banks (frequently by arranging for them to be taken over by stronger banks). In 1990, the banking crisis in Norway deepened, involving for the first time one of the country’s largest banks (Fokus Bank), which received an equity capital guarantee from the CBGF after making large loan loss provisions.
As the demands placed on the two industry-sponsored funds mounted, the Government in March 1991 established a new fund to provide loans to the two industry-sponsored funds. The new Government Bank Insurance Fund (GBIF) received an initial capitalization of NKr 5 billion (0.7 percent of GDP).
In August 1991, the new government fund supplied the CBGF with NKr 2.5 billion in support loans to inject preference capital into the country’s second largest bank, Christiania Bank (NKr 1.8 billion), and Fokus Bank (NKr 0.7 billion). The preference capital was provided after the existing shareholders agreed to substantial write-downs of their shares (to 27 percent of their initial value for Christiania Bank and to 10 percent for Fokus Bank). In October 1991, Christiania Bank alerted the authorities that the bank’s deficit in the third quarter of 1991 would be larger than the bank’s capital and reserves. In addition, losses were also larger than expected in other banks. To provide the funds for recapitalization, the GBIF received additional capital of NKr 6 billion (0.9 percent of GDP) in November 1991. The GBIF was also authorized to lend and provide capital directly to individual banks. A Government Bank Investment Fund was also set up with capital of NKr 4.5 billion (0.7 percent of GDP) to invest in banks on a commercial basis, including through joint ventures with private investors. At the same time, the Government extended to the insurance fund for savings banks a NKr 1 billion (0.2 percent of GDP) grant. In late 1991, a facility was also created at the Norges Bank to provide deposits from the central bank on special terms. Unlike the individual support measures, this arrangement from the central bank was a general measure applied to the banking industry, and not restricted to troubled banks.
In December 1991, Christiania Bank and Fokus Bank received direct equity injections from the GBIF amounting to NKr 5.1 billion and NKr 0.5 billion, respectively. The capital was provided after the values of the banks’ existing shares were written down to zero by Royal Decree. Also in December, Den norske Bank, the country’s largest commercial bank, received NKr 3.3 billion in preference capital from the GBIF as part of an overall recapitalization effort that involved the new government investment fund as well as private shareholders. The former provided NKr 1.7 billion in equity capital and the latter NKr 0.7 billion. At the same time, the existing equity capital was written down to 10 percent of its original value.
As a result of the NKr 13 billion (1.9 percent of GDP) in capital injections (NKr 11.3 billion from the GBIF and NKr 1.7 billion from the Government Bank Investment Fund), the Government eventually held controlling interests in the three largest banks in the country, which together held 85 percent of total commercial bank assets. The Government, however, owns no savings banks, which received NKr 0.5 billion in assistance from the GBIF in late 1991 and early 1992. These capital injections were all made through the SBGF.
The use of interest rate increases to defend the krone during the third quarter of 1992 led to increased losses at the three large commercial banks, which were aggravated by the failure of a large insurance company. To maintain market confidence in banks, the authorities in November 1992 made available further capital. Specifically, at that time, the GBIF made commitments of support totaling NKr 4.6 billion (0.7 percent of GDP). Under these agreements, the fund will provide Den norske Bank with NKr 1.5 billion in preference capital, Christiania Bank with NKr 1.1 billion in equity capital and a NKr 0.9 billion convertible subordinated loan, and Fokus Bank with NKr 0.6 billion in equity capital. In return, these banks on average agreed to cut operating expenses by 15 percent over two years and to lower their risk-weighted assets. To provide the GBIF with sufficient resources to implement these agreements, it was allocated an additional NKr 2.5 billion (0.4 percent of GDP). The GBIF also sold shares in Christiania Bank for an amount of NKr 1.5 billion to the Government Bank Investment Fund.9 Total support by the Government and central bank to commercial banks, savings banks, and the Savings Banks Guarantee Fund reached about 2.8 percent of GDP in 1988–92 (Table 1).
The surge in bank lending in Finland and Sweden took place somewhat later than in Norway, reflecting in part the later implementation of liberalization measures. In Finland, the ratio of bank lending to nominal GDP increased to 76 percent in 1990 from 55 percent in 1986 after most interest rate controls and credit constraints were relaxed (Chart 4). An abrupt fall in economic activity occurred in 1991 following the tightening of economic policies, the collapse of trade with the former U.S.S.R. and Eastern Europe, and the overcapacity in the forestry products sector. At the same time, Finnish banks were caught in an interest rate squeeze, since a significant portion of their loans were tied to a base rate that was set administratively.
As a result, loan losses experienced by Finnish banks increased substantially (Table A1). Accounting for the bulk of these losses are two large institutions, including Skopbank—a commercial bank that serves as the central bank for savings banks. The Bank of Finland took control of Skopbank in September 1991 following an acute liquidity crisis that flowed from the bank’s unusually large portfolio of bad loans.10 In exchange for an equity injection of Fmk 2 billion (0.4 percent of GDP), the Bank of Finland, via a wholly owned holding company, received 53 percent of Skopbank’s shares. At the same time, the Bank of Finland established two other holding companies to remove the bad loans from Skopbank’s balance sheet. Troubled real estate loans and certain equities were transferred to one of the holding companies, while the main industrial credit risks were transferred to the other. These two holding companies had a combined share capital of Fmk 0.7 billion (0.1 percent of GDP). At end-1992, the Bank of Finland had provided Fmk 9.5 billion (1.9 percent of GDP) in loans to fund the holding companies’ acquisition of assets from Skopbank. The ultimate cost of this operation, however, will depend on the price at which the assets acquired from Skopbank are sold.
Responding to declining bank capital positions and fearing a credit crunch, in May 1992 the Government announced a plan of injections of preference capital totaling Fmk 8 billion (1.6 percent of GDP).11 The capital was offered to all banks regardless of their capital ratios and in proportion to their risk-weighted assets (Basle measure). The capital was offered on two dates during August and December of that year. In the first round, 4 commercial banks and 53 savings banks raised Fmk 4.6 billion in capital; in the second, 2 commercial banks, 14 savings banks, and 57 cooperative banks received Fmk 3.3 billion in preference capital.
In April 1992, the Finnish Government established the Government Guarantee Fund (GGF) with Fmk 20 billion (4.1 percent of GDP) to help ensure stability of the banking system and to secure the claims of (domestic and foreign) depositors. In its first operation, the GGF acquired Skopbank from the Bank of Finland for Fmk 1.5 billion in June 1992. However, the central bank continues to own the two holding companies that purchased Skop-bank’s bad assets. Shortly thereafter, the GGF provided Fmk 5.5 billion in capital support to facilitate the merger of 41 smaller savings banks into the newly established Savings Bank of Finland. In December 1992, the GGF provided additional support to both Skopbank and the Savings Bank of Finland totaling Fmk 1 billion and Fmk 6.4 billion, respectively. The total support provided by the GGF in 1992 amounted to Fmk 14.4 billion (2.9 percent of GDP). In February 1993, the Finnish Parliament passed a resolution that requires the Government to guarantee that Finnish banks are able to fulfill their commitments on a timely basis under all circumstances. If necessary, the Parliament committed to provide the Government with sufficient funds and powers to ensure fulfillment of these commitments. In this context, the Parliament also authorized an additional Fmk 20 billion for bank support through a supplementary budget, of which Fmk 5 billion was earmarked for loan guarantees. Total support to date has been huge—about 6 percent of GDP.
In Sweden, the rescission of interest rate controls and quantitative restrictions on credit led to an expansion in bank credit that saw the ratio of bank lending to nominal GDP increase to 68 percent in 1990 from 43 percent in 1986 (Chart 4). The ensuing turnaround in Sweden was occasioned by the worldwide economic slowdown that began in 1990 and by a tax reform that reduced the top marginal tax rate applied to interest deductions to 30 percent from 50 percent and that placed a limit on such deductions.
The banking troubles in Sweden were foreshadowed by developments in the finance company sector in the autumn of 1990. At that time, it was disclosed that a number of finance houses had suffered significant credit losses. These losses caused the market to question the soundness of finance companies as a group. As a result, finance companies that were not affiliated with a bank or owned by an institution that was perceived to be financially sound faced an acute shortage of liquidity as the market for their commercial paper and securities dried up. To weather this crisis, finance houses significantly increased their borrowing from banks and sold their assets.
Troubles in the banking sector came to the fore in the autumn of 1991, when Nordbanken, a majority state-owned bank (71 percent), was compelled to make large loan loss provisions. At that time, the Government subscribed to SKr 4.2 billion (0.3 percent of GDP) of a SKr 5.2 billion new equity issue by the bank, bringing its ownership stake to 77 percent. Following further large loan losses in early 1992, the Government, to facilitate Nordbanken’s restructuring, purchased the shares remaining in public hands for SKr 2 billion (0.1 percent of GDP). The bank was then split into two different components. A new state-owned company, Securum, took over most of the bank’s bad assets, with a book value of SKr 67 billion. The Government also provided Nordbanken with SKr 10 billion (0.7 percent of GDP) in equity capital, and Securum with SKr 24 billion (1.7 percent of GDP) in equity capital and SKr 10 billion in loan guarantees.
The Government has intervened to support two other institutions in addition to Nordbanken.12 In March 1992, Forsta Sparbanken, a savings bank that had been recently merged with Sparbanken Sverige, received support in the form of a SKr 3.8 billion loan at a concessional interest rate and a SKr 3.5 billion loan guarantee. The loan and guarantee were provided to Sparbanken Sverige, which in turn provided the corresponding amounts as equity capital to Forsta Sparbanken. In September 1992, Gota Group, the parent company of the Gota Bank, the country’s fourth largest bank, filed for bankruptcy. At the time, the bank reported large loan losses and the owner of the bank’s holding company, an insurance company, was no longer in a position to support the troubled bank. To bolster confidence in the banking system, the Government issued a broad statement that the obligations of Gota Bank would be honored, but not those of the holding company. The Government also announced its intention to submit legislation to Parliament that would provide a formal safety net for the banking system. In December 1992, the Government took over Gota Bank, although the takeover price has not been finalized. Total government support is estimated to be somewhat above 5 percent of GDP (of which 3 percent has been paid out and the remainder is in various forms of guarantees with varying degrees of risk for the Government).
In December 1992 the Swedish Parliament voted to affirm that the Government would guarantee that banks’ and certain credit institutions’ contractual obligations—other than those to holders of equity and subordinated debt—would be met. The institutions covered by the guarantee include all banks with a Swedish charter and certain mortgage credit institutions such as the Urban Mortgage Bank, General Mortgage Bank, Swedish Ships’ Mortgage Bank, and the National Housing Finance Corporation. The bill also set up a separate agency to administer the Government’s support of the banks and authorized a variety of techniques to implement this support, including direct loans, loan guarantees, and share capital contributions.
In examining common factors in the three Nordic countries, it is apparent that the significant increase in credit relative to nominal GDP experienced in each of these countries before the emergence of banking problems indicates a possible increase in credit risks, as marginal loans are often riskier. However, banks appear not to have been fully compensated for these risks. In Finland, the net interest margin of banks tended to decline during the second half of the 1980s. Banks also pared their net operating expenses, but not by enough to boost their before-tax profits (Table A1). Thus, no additional operating profits were being generated to compensate for the greater lending risks. In Sweden, the net interest margins of banks remained relatively stable during the period of rapid credit expansion, as did net operating expenses. Again, no additional operating profits were being generated at the time bank lending expanded rapidly. Only in Norway did the operating profits of banks increase during the period of rapid credit expansion. However, these gains were subsequently outstripped by the surge in loan loss provisions.
The increased competition faced by banks in these countries could well have increased their tolerance for risk.13 As mentioned above, up to the 1980s, Nordic banks operated in highly regulated markets that inhibited competition and allowed them to earn considerable rents. In fact, during the early part of the 1980s a number of Nordic banks were among the most profitable in the world.14 Financial innovation and liberalization hastened in a new era of increased competition for the rents that had existed under the previous regimes, which in turn altered the incentives for risk taking by bank managers and their shareholders. Since the mid-1980s, bank share prices in Norway, Finland, and Sweden have significantly underperformed those of the markets as a whole (Chart 5); with less private equity at stake, the Nordic banks had less to lose from increasing the risk of their operations.
Chart 5.Nordic Countries: Relative Stock Market Performance1
Sources: Central Bureau of Statistics, Norway; Statistics Finland: and Statistics Sweden.
1 Defined as the ratio of the index of stock prices of banks to the general stock price index.
One key lesson of the Norwegian case is that bank supervision may need to be restructured before financial market liberalization to cope with the new challenges and risks that liberalization entails. In a regulated regime with credit allocation guidelines, compliance with these guidelines may serve as an adequate instrument for limiting excessive risk taking by banks. But once those credit allocations are removed, supervisors may find themselves more at sea—both because of limited experience with monitoring new activities and because of an overstretching of limited personnel.
As in Norway, bank supervision and regulation in Finland proved inadequate to control the risk taking of banks in a liberalized environment. Risk taking (in 1988–89) associated with massive financial market speculation (undertaken in a contest for control of banking markets) proved especially troublesome. Finnish bank supervisors were apparently hampered by, inter alia, the opaque financial status of the large financial/industrial groups that dominate the economy. Supervisors were not empowered to examine subsidiaries and other companies under the control of the banking groups to determine the equity and lending interconnections within the group. In addition, bank supervision lacked sufficient staff resources to provide adequate and timely bank examination. However, it is also important to note that although a better bank supervisory and regulatory environment would have helped to ameliorate the severity of banking problems in Finland, the combination of factors motivating the large expansion of bank credit—expansive impulses stemming from the financial liberalization, the significant gains from trade arising from positive external shocks, and the loose macroeconomic environment—would have made it very difficult to fully avoid banking difficulties.
One option for recapitalizing a troubled banking system is to allow banks to earn above-market interest rate spreads over a sustained period. In the Nordic countries, however, financial deregulation is sufficiently far advanced and financial systems are sufficiently internationalized to limit the feasibility of this task. Hence, they have taken the alternative option of recapitalizing the banking system with public funds coupled with measures aimed at eliminating excess capacity. More specifically, the banks were required to rationalize branch networks, reduce employment, and downsize balance sheets. Nevertheless, the injection of public funds has amounted to a de facto nationalization of these institutions, leaving those governments with the problem of reprivatizing the banks. There is also a risk that availability of cheap public capital to weaker institutions can place them at a competitive advantage relative to the remaining solvent institutions, thereby compounding the latter’s problems.
Strains in the Japanese Banking System
After several years of rapid growth in the exposure of Japanese banks to real estate-related borrowers (Table 2), a decline in real estate prices by up to 30 percent since 1990 has produced—according to market participants—an unprecedented increase in the volume of nonperforming bank loans (Chart 6).15 Suffice it to say that estimates of the size of the bad loan problem vary widely. The 21 major banks disclosed that they held at the end of fiscal year 1992 (end-March 1993) ¥ 12.8 trillion in nonperforming loans (3 percent of all banks’ total loans) in their banking accounts and in trust accounts that benefit from an explicit bank guarantee. The Ministry of Finance reported that nonperforming loans stood at ¥ 8 trillion at the end of fiscal year 1991 and at ¥ 12.3 trillion at end-September 1992. The figure for nonperforming loans at the end of fiscal year 1992 is based on criteria established by the Financial System Research Council. Nonperforming loans are defined as those on which interest has been suspended for at least six months or loans to bankrupt entities. However, the total loans of the 21 major banks comprise less than two-thirds of the total loans of all banks; in addition, the official definition of nonperforming loans excludes restructured loans.16
|Long-term credit banks|
|1992||13.2||26.3||50.4|Chart 6.Japan: Ratio of Urban Land Prices to Nominal GNP
Source: Nikkei Telecom.
A significant proportion of the banks’ bad loans appear to arise from their exposures to nonbank financial institutions, primarily the numerous finance and leasing companies and the eight housing loan companies (jusen).17 According to a survey conducted by the Ministry of Finance, the outstanding loans of the 300 largest nonbank financial institutions totaled ¥ 67 trillion at end-September 1991, of which about 63 percent were secured by real estate. Loans to the real estate and construction sectors amounted to 41 percent of their total loans.18 On the liability side of nonbanks’ balance sheets, approximately 60 percent of their debt is in the form of bank loans. The banks are also significant shareholders in some of the nonbanks.19 Among the major Japanese banks, the trust banks and long-term credit banks appear to have the more significant exposures to nonbank financial institutions; the city banks are reported to have a somewhat smaller exposure.20
In addition, a decline in equity prices of about 50 percent since the end of 1989 has taken a large bite out of the banks’ revaluation reserves, held largely in the form of unrealized capital gains on their long-term equity investments. Near the peak of the stock market boom at the end of fiscal year 1989, the revaluation reserves of the major Japanese banks were sufficiently large to satisfy fully their tier 2 capital requirement (4 percent of risk-weighted assets).21 However, the Nikkei stock index dropped by 38 percent in 1990, and after a 6 percent erosion in 1991, the index lost a further 40 percent through mid-August 1992 (Chart 7). In the month following the announcement of the economic stimulus package in August 1992, it recovered much of the ground it had lost in that year. In early 1993, the stock market recovery gained further momentum, and at the end of fiscal year 1992 the revaluation reserves of the major Japanese banks stood at ¥ 17.8 trillion, up ¥ 0.5 trillion from the end of fiscal year 1991, but ¥ 17.2 trillion below their level at the end of fiscal year 1990 (Table 3).
|Revaluation Reserves1||Subordinated Debt|
|Mar. 91||Mar. 92||Mar. 93||Mar. 91||Mar. 92||Mar. 93|
|Long-term credit banks||6.4||3.4||3.6||0.8||1.0||1.6|
|Total||35.0||17.3||17.8||4.7||7.1||10.6|Chart 7.Japan: Nikkei 225 Stock Index
Sources: International Monetary Fund, International Financial Statistics; and Nikkei Telecom.
Banks have compensated for the decline in their tier 2 capital by issuing subordinated debt and by restraining growth in assets to which the appreciation of the yen contributed.22 At the end of fiscal year 1992, the subordinated debt of the major Japanese banks had increased to ¥ 10.6 trillion (2.3 percent of risk-adjusted assets) from ¥ 7.1 trillion at the end of fiscal year 1991. At the same time, the risk-adjusted assets of the major banks declined to ¥ 466.6 trillion at the end of fiscal year 1992 from ¥ 479.5 trillion at the end of fiscal year 1991, owing largely to an appreciation of the yen and weak demand for loans. Reflecting these developments, together with the upsurge of equity prices, the average ratio of total capital to risk-adjusted assets of the major banks was above 9.2 percent at end-March 1993, while their tier 1 ratio stood at slightly below 4.9 percent.23 Moreover, the reduction in total risk-weighted assets may have induced them to adopt a much lower international profile than in earlier years, particularly in the international interbank markets.
The increased lending by banks to the real estate sector, either directly or through nonbanks, and the large fluctuations in asset prices occurred against the background of financial liberalization and innovation in Japan. During the 1970s and 1980s, developments on both domestic and international fronts created arbitrage opportunities that made many regulations on the domestic financial market difficult to sustain.24 One important influence was the development of a domestic securities market as a result of the increase in government borrowing following the oil price shock of the mid-1970s; another was the removal of capital controls during the 1980s. The surge in supply of government bonds facilitated the development of instruments that enabled depositors to circumvent deposit rate ceilings: government bond repurchase (gensaki) agreements provided a higher-yielding alternative to wholesale time deposits, while medium-term government bond (chukoku) funds provided an alternative to retail time deposits. This created pressure for liberalization of deposit rates, which took place during the 1980s. As a result, at end-September 1992, interest rates on about 70 percent of deposits in city banks were market determined (Chart 8).
Chart 8.Japan: Dependence on Deposits with Deregulated Interest Rates by Type of Bank
Source: Bank of Japan, Economic Statistics Monthly, various issues.
Note: Deposits with deregulated interest rates are time deposits with unregulated interest rates, certificates of deposit, foreign currency deposits, and nonresident yen deposits.
Financial innovation and liberalization set the stage for a significant change in the volume and pattern of bank lending in Japan. Bank lending, after remaining fairly stable relative to nominal GDP in the 1970s, rose from 61 percent in 1980 to a peak of 94 percent in 1990 (Chart 9).25 Expansionary monetary policy in 1987–88, designed mainly to stabilize the exchange rate in the period following the Louvre accord, accommodated both the surge in bank lending and the increase in asset prices in the later part of the 1980s. The pattern of lending also shifted substantially away from manufacturing and distribution industries and toward households and the construction and services industries (Chart 10). As a result of this shift, small and medium-sized enterprises and households became the primary customers for bank loans, replacing the larger enterprises that had previously had preferential access to bank credit.
Chart 9.Japan: Ratio of Bank Credit to Nominal GDP
Source: Nikkei Telecom.
Chart 10.Japan: Composition of Bank Lending
Source: Nikkei Telecom.
The shift in lending away from the large enterprises toward small and medium-sized firms and households enabled banks in part to offset the impact of liberalization of deposit rates on net interest margins. However, these lending rates may not have been high enough to compensate for the greater risks involved in this lending, as can be seen from Table A2.26 During the 1980s and early 1990s, the net interest margins of Japanese banks narrowed significantly and net operating expenses were scaled back, but these cuts were not sufficient to boost banks’ profits before provisions and taxes, which remained broadly stable through 1990. This situation changed somewhat in 1991, as net interest margins and net operating expenses rose substantially without greatly affecting net operating profits. At the same time, loan loss provisions increased sharply. For all ordinary banks, provisions more than doubled as a ratio of total assets, from 0.03 percent to 0.07 percent, whereas for city banks, they tripled from 0.03 percent to 0.09 percent.
One reason for this increased risk taking, which was not adequately offset by an increase in lending margins, was the impact of financial innovation and liberalization on competition in banking and related services.27 The erosion of profits from traditional banking activities that had been sheltered from competition in effect reduced the amount of private equity at stake in the banking system. This meant that the potential losses to banks’ shareholders from increased risk taking had diminished, while the potential gains remained unchanged.28 Typically, the reduction in private equity would be reflected in a decline in banks’ stock market value relative to that of the market as a whole. However, in Japan, bank shares were subject to certain rigidities until 1984 and then to a subsequent adjustment to freer market conditions. Despite the potential distortions in bank share prices, there is some evidence that these prices underperformed relative to a broader stock market index in 1987–92 (Chart 11).
Chart 11.Japan: Ratio of Stock Price Index of Large Banks to Nikkei 225 Index
Sources: Nikkei Telecom and Reuters.
The Japanese authorities expect banks to be able to earn and to cost-cut their way out of current debt problems over the next few years. This strategy has been underpinned by two supplemental fiscal packages that indirectly helped to stabilize equity and real estate prices by stimulating general economic activity. The Government’s supplemental budget in fiscal year 1992 also included ¥ 1.6 trillion (0.3 percent of GDP) for government purchases of land and ¥ 1.1 trillion (0.2 percent of GDP) for equity investments by the Postal Savings Corporation, the Postal Life Insurance Welfare Corporation, and public pension funds.29 In addition, a gradual reduction in the discount rate since July 1991, from 6 percent to 2.5 percent, has allowed banks to increase net interest income.30
The Government also announced in August 1992 several other measures that helped stabilize equity prices and that will assist banks in managing their bad loans. First, the Ministry of Finance reaffirmed that banks can omit reporting in their interim accounts (end-September accounts) the valuation losses on equities whose market values had fallen below book values. Typically, banks would have tried to cover such losses by selling other equities that still carried latent profits, adding to the downward pressure on equity prices. Second, the Ministry of Finance relaxed its limit on the dividend-payout ratios of banks. Some banks have in the past avoided cutting dividends by realizing capital gains on some of their equity holdings to raise net profits. Moreover, the historical volatility of Japanese equity prices suggests that revaluation reserves remain a volatile component of bank capital. Accordingly, the Minister of Finance stated in February 1993 that it is desirable for banks to secure their own capital and attain capital adequacy standards without relying on the revaluation of reserves on the stocks held.
With respect to bad loans, the Ministry of Finance has called upon the tax authorities to demonstrate greater flexibility in allowing banks to make tax-deductible loan loss provisions. For a provision against a specific domestic loan, the requirements for a tax deduction are that either the borrower has initiated bankruptcy procedures or has been in excess liabilities for some time. In September, the National Tax Administration issued a ruling that shortened the definition of this period from more than two years to more than one year.
In February 1993, 162 financial institutions (129 banks, 32 insurance companies, and Norinchukin) launched the Cooperative Credit Purchasing Company (CCPC). The new company’s total paid-in capital is ¥ 7.9 billion, of which the 21 major banks provided ¥ 4.8 billion. The authorities are strongly supportive of this effort, but are inclined at this time to leave the funding of the agency to the banks themselves. CCPC will buy from its members and their nonbank affiliates problem loans collateralized by real estate.31 The loan purchases will be at discount to the face value of the assets, with the precise level of the discount to be determined on the basis of a disclosed third-party assessment of the market value of the collateral.32 The institutions that sell their loans to the CCPC will, in turn, provide the company with the funding for the loan acquisitions at the discounted values. The selling institutions will then record the discounts as losses on their own books, which will be tax deductible; if banks were to make the provisions in their own books against problem loans, full tax deductibility would generally take longer to achieve. Thus, the main benefit created by this new company is to enable those institutions with problem loans to make faster write-offs of these loans.
The success of the above strategy will thus rest, in part, on sustained bank operating profits, on which a number of cyclical and structural developments will have a bearing over the medium term. The net interest income of Japanese banks has increased substantially with the cyclical decline in interest rates.33 The major banks’ net interest income rose to ¥ 5.9 trillion in fiscal year 1992, up from ¥ 5.2 trillion in fiscal year 1991. At the same time, net operating profits (Gyoumu-juneki) of these banks increased to ¥ 3.2 trillion from ¥ 2.4 trillion, as net noninterest expenses were held largely in check. The positive effect of low interest rates has been more than offset, however, by a doubling of specific loan loss provisions and charge-offs to ¥ 1.3 trillion from ¥ 0.6 trillion and by the writing-down of equities to the lower of market or book value. As a result, the pretax profits (Keijo-rieki) of the major banks declined to ¥ 1.4 trillion in fiscal year 1992, from ¥ 2.2 trillion in fiscal year 1991. Thus, in a favorable interest rate environment, the major banks were able to limit the decline in their pretax profits to 36 percent, while provisioning at a rate that would cover within the next two fiscal years the maximum potential losses on reported nonperforming loans.34
The ongoing process of financial innovation and liberalization may affect the net operating income of Japanese banks in the period ahead. In December 1992, the Ministry of Finance announced details of the measures under the Financial System Reform Law (passed by the Diet in June 1992) to ease the segmentation of the financial system. This plan went into effect on April 1, 1993. Its major effect is to allow banks to engage in securities and trust activities through majority-owned subsidiaries, while securities firms will be permitted to establish such subsidiaries engaging in banking and trust business. Regional banks will be able to carry out certain trust activities in-house.
The new subsidiaries will not be allowed, however, to engage in some of the most important activities of the existing institutions, such as equity-related activities of securities firms and loan trusts, tokkin trusts, and pension trusts for trust banks. The creation of new securities subsidiaries will also be restricted at first to the long-term credit banks, trust banks, and centralized cooperative credit institutions (such as Norinchukin), while securities firms, long-term credit banks, and the Bank of Tokyo may establish trust banking subsidiaries.35 The city banks will be able to establish subsidiaries after an interval of about one year. The Ministry of Finance will review the scope of activities permitted of the new subsidiaries in two to three years. Thus, any increased competition resulting from these measures will materialize only gradually.
To prevent certain conflicts of interest, the Financial System Reform Law created a number of “fire walls” between banks and their new subsidiaries. These include a restriction on the dual employment of officers in a bank and its new subsidiary, a requirement that transactions between a bank and its new subsidiary be on an arm’s-length basis, and a prohibition on transactions by a new subsidiary with a bank customer that are tied to an extension of credit by the bank. In addition, the Ministry of Finance announced further fire wall restrictions in December 1992. First, a securities subsidiary will not be allowed to act as lead manager of a securities offering by any issuer that has a “main bank” relationship with its parent bank, unless the issuer has net assets of at least ¥ 500 billion.36 Second, joint visits to a customer by representatives of a bank and its securities affiliate will be prohibited, unless requested by the same customer. Third, the exchange of nonpublic information between a bank and its securities subsidiary will be prohibited unless consented to in writing. Fourth, a securities subsidiary will be expected to raise to 50 percent within five years the ratio of directly recruited staff to total employees, as opposed to transferees from its parent bank. Finally, the headquarters office of a bank and its securities affiliate may not be in the same building.
In addition to the desegmentation of the financial services industry, the liberalization of deposit rates is scheduled to be completed in 1994. In June 1993, the liberalization of interest rates on time deposits was completed with the abolition of the minimum balance requirement on super time deposits, which had been set at ¥ 3 million. Liberalization of interest on demand deposits is scheduled to be completed in 1994.
An alternative source of operating profits for banks is the net gain from the sale of securities holdings, and the banks’ revaluation reserves represent the potential amount of such gains. For the major Japanese banks, these latent profits totaled ¥ 17.8 trillion at end-March 1993.37 Although these reserves are ample in relation to the loans already identified as nonperforming, they are distributed unevenly among the major banks, as are the loan losses. In addition, the realization of capital gains by banks to cover the unfolding loan losses would require them to raise alternative forms of tier 2 capital, such as subordinated debt or perpetual subordinated debt, to maintain their compliance with the Basle ratio for total capital.
For this “earn-your-way-out-of-trouble” approach to make headway, either real estate and equity prices need to approach the levels before the problems arose, or spreads between lending and deposit rates must remain wider than they would be under a fully liberalized financial system (or indeed wider than they were in the late 1970s and early 1980s before deregulation). On the first count, equity prices have recently risen nearly 40 percent above their low levels of 1992, but they still remain about half those achieved during the asset price bubble years of 1986–89. For real estate, despite recent declines, the price index remains at twice the level it reached during the years preceding the boom.38 Prospects for recovery of real estate assets remain uncertain. On the second count, the ongoing efforts to liberalize the Japanese financial system—set in motion in more propitious times—and the progress already made in moving toward a more competitive, market-based system, may well make it harder for the banking system to maintain above-market interest rate spreads. In particular, the factors that are threatening the franchise value of large banks in other major financial markets—a more competitive environment with more liquid money markets and direct debt markets—are putting pressure on the earnings of Japanese banks as well. Corporations have already gained access to competitive international markets, and they are unlikely to allow banks to charge above market spreads. In addition, the growing presence of technically sophisticated foreign firms in Tokyo, unhampered by balance sheet problems, is eating away at the profits from technically upscale securities market activities. The increased competition between the securities houses and the banks, the desegmentation of the banking system, the already substantial liberalization of interest rate controls, and the loosening of ties between banks and large corporations can all be expected to put additional pressure on the earnings of these institutions.
At the same time, several features of the Japanese financial system—as well as some features of the broader economic situation—give the authorities more room for maneuver in dealing with the problem of bad loans than is available in many other countries. First, the accounting system does not yet force the timely recognition of all financial gains and losses carried on balance sheets.39 In addition, market participants explained that some of the common OTC contracts are not subject to margin calls and can be rolled over at original prices. Margin requirements are a method of marking financial contracts to markets to limit credit risk. This lack of transparency can have adverse longer-term consequences, however. By clouding the true extent of the loan loss problem and by pushing it into the future, it may only increase the adjustment that ultimately needs to be made. Second, and reflecting the economy’s high saving history, there are a number of large institutional investors, such as pension funds, life insurance companies, and the postal savings systems, whose portfolio decisions can be directed in part to sectors of the economy that are experiencing temporary difficulties. Finally, the pace of progress toward achieving a more market-based financial system remains to some degree a public policy decision; for example, those reforms that would increase the degree of competition facing banks could be slowed temporarily during the adjustment period.40 However, the Japanese authorities have reaffirmed their commitment to implement financial liberalization as scheduled—and not to allow the present difficulties to interrupt progress toward the creation of a more competitive, transparent, and efficient financial system.
Recent Difficulties of U.S. Commercial Banks
After several years of serious difficulties, the banking industry in the United States has recently been enjoying at least a temporary respite: a strong upturn in profits flowing from the cyclical drop in interest rates.41 In fact, it is this rebound in bank profits that has thwarted earlier forecasts (including that of presidential candidate Perot) of a “December surprise” in bank failures. In the 1980s and continuing into the beginning of the 1990s, U.S. commercial banks had experienced a series of significant losses arising from concentrated credit exposures to regional economies. Sizable losses on loans to developing countries and to oil-related industries had led to a sharp increase in the number of bank failures beginning in the mid-1980s (Chart 12). In 1989, one-half of all bank failures were attributable to oil-patch loans in Texas and the Southwest. More recently, real estate loans (particularly in New England and the West) have been the industry’s greatest weakness.42 As a result of the surge in bank failures, the fund of the Federal Deposit Insurance Corporation (FDIC) that guarantees deposits in commercial banks was almost exhausted in 1991. To alleviate its problems, the fund was augmented with an increased authority to borrow from the U.S. Treasury under the FDIC Improvement Act.43
Chart 12.United States: Number of FDIC-Insured Banks Closed
Source: Federal Deposit Insurance Corporation, Annual Report (1991), Table A, p. 127.
The origin of the current problems in the U.S. banking sector follows the stylized, declining-franchise-value story outlined earlier in this section. As inflation in the United States accelerated in the 1960s and 1970s, rising interest rates hampered banks’ ability to attract deposits because interest rate ceilings became binding. The advent of money market mutual funds in 1972 provided a close substitute to bank deposits at a time when the spreads between market rates and the deposit rate ceilings were large. These funds, by demanding high-quality, short-term corporate paper, also contributed to the deepening of the U.S. commercial paper market. As a result of these developments, banks—especially the money center institutions that operated in the wholesale market—began to experience pressures on their net interest margins.
To enable banks to compete more effectively for funds and to stem the outflow of deposits from banks and thrift institutions as interest rates rose again in the late 1970s, the Government gradually liberalized deposit rates and, in the early 1980s, expanded the investment powers of savings and loans.44 As a result, banks could compete more effectively for deposits.
Chart 13.United States: Office Vacancy Rates
Source: CB Commercial.
Even in an environment in which interest rate ceilings were removed or once again nonbinding, commercial banks found that they had lost much of their traditional corporate business. Banks faced increased competition in their short-term lending activities from the money market. The commercial paper market provided a low-cost alternative to bank loans for blue-chip commercial firms and, increasingly, for less well-known firms as well. By the end of the 1980s, the amount of nonfinancial commercial paper outstanding was about 21 percent of outstanding commercial and industrial loans at banks, compared with about 11 percent at the start of the decade (Chart 14).
Chart 14.United States: Ratio of Commercial Paper to Commercial and Industrial Loans
Source: Board of Governors of the Federal Reserve System, Federal Reserve Bulletin, various issues.
Confronted with increased competition in both deposit taking and lending activities, banks in the 1980s significantly altered the pattern of their lending. First, after remaining below 30 percent for most of 1975–80, the ratio of commercial bank lending to nominal GDP increased significantly from 1981 to 1986 when it reached 34 percent (Chart 15). The ratio remained stable during 1987–90 before declining in 1991. Second, since late 1982, banks have changed the composition of their loans (Chart 16). The share of real estate loans increased to 32 percent of banks’ total loans and securities in mid-1992, from 23 percent in 1980.45 At the same time, commercial and industrial loans, which traditionally formed the core of banks’ lending operations, dropped to 18 percent of banks’ total loans and securities in 1992 from 24 percent in 1980. The shift into real estate lending was prompted in part by the accelerated depreciation allowances for such investments that were in effect from 1981 to 1986. Moreover, as leveraged buyouts became popular in the late 1980s, banks’ commercial lending became increasingly associated with these transactions. By 1990, over 35 percent of the domestic commercial loans made by money center banks were in support of leveraged buyouts.
Chart 15.United States: Ratio of Domestic Banks’ Loans and Leases to Nominal GDP
Source: Data Resources Incorporated.
Chart 16.United States: Composition of Domestic Bank Lending
Source: Data Resources Incorporated.
The significant rise in bank lending relative to nominal GDP and the growing concentration of loans for real estate and leveraged buyouts both point to possible increases in risk taking by banks. The marginal loans by banks are likely to be riskier, based on the premise that safer opportunities are usually taken first, whereas the plunge into real estate and leveraged buyouts cuts against the precepts of broad portfolio diversification. As the banks were taking on what appeared to be greater risks in their lending activities, however, the net interest margins and net operating expenses, both scaled by their total assets, remained virtually unchanged (Table A3). Thus, no additional earnings were being generated to cover the potential increase in losses from the riskier loans.
Banks’ increased tolerance for risk stemmed from the erosion of profits from traditional banking activities as a result of greater competition. Because of the consequent reduction of private equity at stake in the banking system, the potential losses to banks’ shareholders from greater risks had diminished, while the potential gains remained unchanged. One indication of this loss in private capital is the relatively poor stock market performance of the money center banks since the late 1970s, and especially from 1986 onward (Chart 17). These institutions have borne the brunt of the competitive pressures from nonbank financial institutions. In contrast, the share prices of the smaller, retail banks that operate primarily in local markets have outperformed those of both the money center banks and the market as a whole.
Chart 17.United States: Standard & Poor’s (S&P) and NASDAQ Stock Price Indices for Banks in Relation to S&P 500 Composite Index
Source: Data Resources Incorporated.
An increase in risk taking was also a feature of the activities of savings and loans during the 1980s. Large losses experienced by these institutions and the near exhaustion of the industry’s capital in the early 1980s led the troubled segment of savings and loans to take greater risks in an effort to regain solvency. The result was a bifurcated industry by 1989—one segment was profitable and adequately capitalized, whereas the second was deeply insolvent.
Like the savings and loan industry in 1989, the banking industry is now made up of two different segments. The smaller institutions tend to be well capitalized and profitable. In fact, according to an annual international survey of bank profitability by IBCA Limited (1992b), a credit rating agency specializing in banks, three U.S. regional banks (and one U.S. money center bank) were among the top ten institutions in terms of inflation-adjusted profitability. A number of several larger institutions, however, are not well capitalized and are burdened with asset quality problems. Using the definition established under the FDIC Improvement Act for the purpose of implementing the new prompt corrective action regulatory framework (see below), a well-capitalized institution must have a ratio of tier 1 capital to total assets of at least 5 percent. Adequately capitalized institutions are those with a leverage ratio of at least 4 percent, while undercapitalized institutions are those with a leverage ratio of less than 3 percent, the statutory minimum.46 At end-September 1992, 93 percent of all banks with 68 percent of the industry’s total assets were well capitalized, while 5 percent of all banks with 31 percent of the industry’s total assets were only adequately capitalized. A handful of small banks (2 percent of all banks with 1 percent of their total assets) were undercapitalized at that time.
The larger banks, moreover, tend to be more heavily burdened with nonperforming loans. As Table 4 shows, during 1989–91, the ratio of net loan charge offs and loan delinquencies to total loans was significantly larger for money center banks than for all banks. The data also show that, during 1990–91, problem loans in the real estate sector constituted more than half of the total problem loans of all banks. Preliminary data for 1992 indicate that nonperforming loans may well have peaked, as the industry recorded a drop in loan charge offs and delinquencies in the first three quarters of the year.
|All Loans||Real Estate Loans|
|Year and Bank Size||Net|
|Money center banks||2.37||7.90||…||…|
|Other large banks||1.54||5.69||…||…|
|Money center banks||1.93||6.95||…||…|
|Other large banks||1.65||5.36||…||…|
|Money center banks||1.42||7.07||…||…|
|Other large banks||1.34||4.39||…||…|
The recent wave of bank failures pushed the fund of the FDIC that guarantees deposits (up to $100,000) in commercial banks to the brink of insolvency in 1991. Failed banks insured by the FDIC in the previous few years were mostly located in the Southwest and New England. Banks in these regions are still suffering from locally depressed real estate markets. The savings bank sector, with a business similar to that of the savings and loans, constituted one-third of the disbursements made by the FDIC. In 1991, FDIC funds were augmented with increased authority to borrow $25 billion from the U.S. Treasury under the FDIC Improvement Act. Table 5 presents alternative estimates of the net present value of losses to the Bank Insurance Fund (BIF) arising from the failure of troubled institutions over the next several years. These estimates range from $15 billion to $88 billion and center around $30 billion (0.5 percent of GDP). This amounts to roughly one-sixth of the losses sustained by troubled savings and loans during the later part of the 1980s. Although the wide range of estimates points to the difficulty of predicting these losses, it nevertheless is expected that the FDIC could at least meet the losses in the lower end of this range with the resources currently at its disposal, including the $25 billion line of credit with the Treasury to cover insurance losses and the revenue from deposit insurance premiums.47 The 25 basis point levy per dollar of deposits is expected to yield the FDIC about $4 billion in annual revenue from insured commercial banks in fiscal year 1993.
|Office of Management and Budget||15–27||January 1993|
|Congressional Budget Office||29||January 1993|
|Vaughan and Hill||35–88||December 1991|
|Barth, Brumbaugh, and Litan||31–43||December 1990|
There are two key differences between the banking industry today and the thrift industry in 1989. First, the banking industry earned substantial profits in the first three quarters of 1992 ($46 billion before taxes at an annual rate, up from $25 billion in 1991), and the ratio of its capital to total assets was 7.4 percent at end-September 1992. The savings and loan industry, in contrast, sustained heavy losses in 1988, and the ratio of its tangible capital to total assets was less than 2 percent. Second, the approach of the U.S. authorities to current banking problems reflects some lessons that were painfully learned during the saving and loan crisis. In that crisis, regulators initially eased capital adequacy standards by relaxing accounting rules and by expanding the permissible investments of savings and loans. In effect, rather than closing insolvent institutions, the regulators allowed these savings and loans both to benefit from government-insured deposits and to increase the riskiness of their operations in a bid to regain viability. The present value of losses to the Government from the insolvent segment of the thrift industry was about $180 billion.48 This experience led to an important and unique regulatory initiative in the United States, namely, the prompt corrective action regulatory framework implemented in 1992, which greatly reduces the scope for troubled institutions to increase the riskiness of their operations. Rather than expanding the permissible assets and activities of troubled institutions, the new law curbs the operations of such institutions and requires the regulators to initiate the closure of a bank if the ratio of its tangible capital to total assets drops below 2 percent.49
Increased difficulties among the larger U.S. banks remain a possibility, however. The buoyant earnings of U.S. banks in 1992, flowing largely from the cyclical drop in interest rates, may not be sustained as the economic recovery takes hold. Moreover, several large banks fall considerably below the regulatory authorities’ new standard for well-capitalized banks, and the prerequisite to accumulating the necessary capital is a sustained improvement in earnings. In this context, the recent consolidation among some large banks may improve profitability.50
Banking Difficulties in Some Other Industrial Countries
In 1992, concerns emerged about a rapid deterioration in the quality of domestic assets of French banks. The first sign of this deterioration began to appear at about end-1991 as the economy slowed, with signs of strain emerging among many small and medium-sized businesses.51 Since then, the problems have intensified and spread to larger groups. An estimated 20 percent decline in prices in the Paris real estate market from their peak in 1990 has further compounded the deterioration of the banks’ balance sheets. Finally, a rise in short-term interest rates triggered by the defense of the French franc cut interest margins and added to the banks’ problems.
Banks’ exposure to the commercial real estate sector grew very rapidly in 1988–90, and the French Bankers Association estimates that loans to this sector now account for about 8–9 percent of banks’ total loans. These exposures, however, are not distributed evenly throughout the banking system, as a handful of institutions have real estate exposures in the range of 10–30 percent of their total loans. In December 1992, representatives of the banking and insurance industry asked the Government to support the ailing commercial property market. The Government has responded by extending the holiday on the 20 percent transfer tax that is payable on property deals, by reinstating the favorable tax treatment of finance leases to encourage owner occupation, and by easing administrative procedures that were designed to discourage the take-up of office space in Paris. Two features of the Paris real estate market also tend to promote stability. First, close cooperation between lenders and public authorities is helping to prevent fire sales and wholesale liquidations. Second, office vacancy rates are relatively low, at about 6–7 percent, although the trend is upward.
The approach taken toward banking problems in France has traditionally relied on cooperative efforts from within the banking system (including a loss sharing arrangement) to prevent banking problems from spilling over into other sectors. So far, French authorities expect that this approach can again be used successfully to deal with current difficulties. As in some other industrial countries, much depends on the pace of recovery of economic activity and on the path of interest rates. The quicker the recovery and the more pronounced the decline in interest rates, the more likely it will be that the banking industry can solve its current problems on its own.
The performance of banks in the United Kingdom has deteriorated significantly since 1989 owing to the prolonged recession and to a decline in real estate prices of more than 20 percent. Loan losses have come largely from small and medium-sized businesses that were least able to withstand the downturn; however, in 1991–92, there has also been a string of failures of larger corporations and property developers. Significant exposures to names like Bond, Olympia and York, and Maxwell have been particularly painful to the banking sector. More generally, the rapid expansion of credit in the late 1980s has left the private sector overleveraged. U.K. banks have sharply increased loan loss provisions over the past two years. This has in turn been reflected in a marked reduction in pretax profits from their peak in 1988. Indeed, in 1992, although some U.K. clearing banks turned in higher profits than in 1991, one clearing bank reported the first loss in its history.
All this notwithstanding, U.K. financial intermediaries have remained in a relatively stable financial position—at least compared with banking systems elsewhere. U.K. banks have generally maintained higher risk asset ratios than the 8 percent required by the Basle capital agreement. For example, the four clearing banks had a risk-weighted tier 1 capital/asset ratio of nearly 6 percent in mid-1992, or 50 percent more than the required Basle ratio.
Like the banks, U.K. building societies, which specialize in mortgage lending, have been vulnerable to the troubled residential real estate market. A significant positive factor, however, is that whereas 100 percent mortgages have been relatively common in the United Kingdom, most mortgages with a loan/value ratio in excess of 75 percent were insured. Thus, the losses on residential mortgages have been borne in the first instance by the insurance companies.
As elsewhere, the future course of real estate prices and the timing and strength of the recovery are important variables for the health of the U.K. banking system. There are some signs that the U.K. banks may be at a turning point. Bank share prices rose sharply in 1992 and early 1993, outperforming the stock market as a whole by a wide margin. In the period ahead, bank earnings are likely to benefit from the recent drop in U.K. interest rates and from efforts to trim operating expenses.
Three Choices and an Ounce of Prevention
As hinted at above, countries faced with significant banking problems generally have three broad policy options.
One option consists of slowing deregulation and innovation, to create a less competitive environment, to allow institutions time to earn their way out of losses by keeping spreads high. This approach requires regulatory forbearance to push accounting losses into the future, and it places the costs directly on depositors and borrowers. The rub with this buying-more-time approach is that it can undermine the incentives for the banking sector to reduce capacity, lower costs, and reduce risk taking. In the U.S. saving and loan crisis, this approach produced very poor results because many institutions used the extra time not to adjust but to redouble their bets. In the developing country debt crisis, the buying-more-time approach eventually yielded a more favorable outcome but only because the additional time was used by lenders to build up provisions against eventual losses and by borrowers to undertake adjustment and reform programs that improved their debt service capacity.
A second approach consists of a direct government injection of capital into the troubled institutions, conditional on downsizing, to avoid future problems. This approach is often the default option when the troubled institutions are regarded as “too large to fail” and when liberalization and internationalization of the financial system has already proceeded to such a degree that depositors and borrowers cannot be held captive to higher spreads. In any case, when this approach has to be taken, it is essential that the direct capital injections be made conditional on downsizing of the industry and on a strengthening of bank supervision to avoid a recurrence of the same problem.
The third approach is to accept the reality that certain troubled institutions have no remaining value and that liquidating them before they can do more harm is the most sensible long-term solution to the problem. While this approach is, of course, the most direct answer to overcapacity in the industry, it is only likely to be feasible in situations in which the problems are not concentrated in institutions whose closure would have “systemic” effects—either for that economy or for the international financial system at large. In addition, this approach is subject to the judgmental problem of distinguishing a permanent from a transitory, or cyclical, weakness.
This is not the first time during the past ten years that some industrial countries find themselves faced with crises or strains in their national banking systems, and with the significant costs and constraints that resolving those problems can entail for taxpayers and for the stance and effectiveness of macroeconomic (and in some countries also their exchange rate) policies. The current bout of banking problems—which has a large real estate component—certainly owes a good deal to “cyclical” factors (the shift from relatively loose to relatively tight monetary policy, the unwinding of earlier asset price excesses, and the slowdown in economic activity). Moreover, it has to be expected that even the most ably managed banks will occasionally make incorrect credit decisions. Still, the frequency and widespread country distribution of banking problems suggest that two important lessons have not yet been heeded. One is that it is precisely when banks and other financial intermediaries find both that their competitive position is being eroded and that they suddenly have expanded investment opportunities that they are most susceptible to abandoning the principles of sound banking and to taking excessive risks—particularly when much of the risk is effectively being underwritten by implicit and explicit government guarantees. The second, related lesson is that—with less segmentation of permissible activities across types of financial institutions—what was a profitable activity for early entrants can become a significant source of losses if later arrivals expand the size of that activity beyond reasonable risk/return trade-offs and their own expertise.
So much for current banking strains and crises. What can be done in the future to minimize the probability of recurrences? Financial authorities in the industrial countries have been making major efforts—mostly through the Group of Ten Basle Committee on Banking Supervision—to revamp the regulatory and supervisory framework so that it can better cope with the new more competitive international environment. Considerable progress has been made in putting into place more stringent risk-based capital requirements—for credit and market risk—to induce more prudent behavior by banks and other financial institutions and to serve as a financial buffer against losses through official safety nets. By early 1993, international banks in all Group of Ten countries met the January 1993 Basle capital requirements. Similarly, improvements in the consolidated supervision of international banking should reduce the practice of regulatory arbitrage. In addition, several countries are undertaking reforms of their financial disclosure requirements to make banking and finance more transparent. Finally, it is generally recognized that more efficient pricing of various features of the financial safety net, such as deposit insurance, can reduce the moral hazard that has at times led banks to assume greater lending risks.
All these reforms are welcome. Given the cost of resolving banking and financial crises, this is indeed an area where an ounce of prevention is worth many pounds of cure. Still, it would be unreasonable to expect these improvements in the regulatory and supervisory framework to eliminate strains in national banking systems. The reality is that major banking systems are in a structural transition to a changing competitive environment, and current banking problems are to some extent a reflection of that transition. That transition will not be completed overnight—especially in countries where financial innovation is still accelerating.
Bank lending as a percentage of GDP expanded in Japan from 61 to 94 percent between 1980 and 1990; in Norway from 64 to 85 percent between 1983 and 1986; in Sweden from 43 to 68 percent from 1986 to 1990; in Finland from 55 to 76 percent between 1986 and 1990; and, more modestly in the United States, from 30 to 34 percent between 1984 and 1986. This expansion in credit was generally accommodated during the 1980s by the prevailing monetary policy stance.
Monetary policy tightening in the late 1980s and early 1990s slowed, in most cases, credit expansion and produced a sharp correction of key asset prices—a development that impaired the value of bank credit. See International Monetary Fund (1993a).
The consequences of asset price deflation on financial markets has also been examined in Annex I of International Monetary Fund (1993b).
Although it is difficult to generalize, there are often international spillover effects from domestic banking crises. In particular, banks under stress are frequently downgraded by international rating agencies and hence their costs of funding in international markets increase. In some cases, this has led to a fall in international lending exposure.
While the scope of this paper is limited to examining recent banking problems in several major industrial countries, further investigation is clearly needed to determine the central factors explaining the lack of banking difficulties in some other industrial countries.
The Japanese authorities do not agree with this classification. The severity of banking problems in various countries can be measured, inter alia, in terms of the stock of nonperforming loans, the volume of restructured loans, the extent of direct and indirect government support for the banking system, the decline in the equity value of banking firms, and the extent of the downsizing of bank balance sheets.
A full understanding of the effects of financial liberalization on the banking industry would require estimates of the “equilibrium” level at which the ratio of bank lending to nominal GDP would stabilize after a switch to a competitive environment. Further analysis on this issue, however, is limited because, for most industrial countries, deregulation is still an ongoing process.
In Norway, the price of commercial real estate peaked in 1987, having more than tripled since 1981 (Chart 1). By 1991, the price of commercial real estate had returned to its 1981 level, after adjusting for inflation. Real estate values in Finland (residential) and Sweden (commercial) reached their peaks in 1989, 332 percent and 565 percent above their level in 1981, respectively, and have fallen significantly since then. By 1992 the price of real estate in Finland had returned to its 1981 level, after adjusting for inflation. Similar cyclical patterns are evident in the real stock market prices in these countries as well (Chart 2).
However, the average marginal tax rate on capital income remained at about 40 percent until the tax reform in 1992, when a marginal rate of 28 percent was introduced.
Strong conditions are attached to the capital infusions from the GBIF, as regards both the volume of bank lending and cost control. A major part of the capacity adjustment that has taken place in the Norwegian banking industry has been the result of balance sheet reductions in the three largest commercial banks. Manpower requirements have been cut substantially.
At the time, no other authority existed with a mandate to carry out the restructuring of a troubled bank.
The preference shares are convertible into common equity if interest is unpaid for three years or if the bank’s capital ratio falls below the legally required minimum. The interest rate on them is also scheduled to increase over time, creating an incentive to repurchase the shares.
Three additional institutions announced in early 1993 that they may need government support. However, no action on these cases has as yet been taken.
For a further discussion of this proposition in the context of the Nordic countries, see Llewellyn (1992).
One factor behind the sharp increase in bank credit was the tax treatment of interest payments on debt at the time of financial liberalization. The combination of high marginal income tax rates and generous allowances for the deduction of interest expenses led to negative after-tax real interest rates for households in these countries throughout much of the 1980s.
IBCA Limited, Real Banking Profitability (London), various issues.
It is possible that the decline in the value of real estate has in fact been larger, but lack of transactions data makes it difficult to pinpoint the estimate. The decline in real estate values in London and New York in recent years has been estimated at about 50 percent.
By including estimates of nonperforming loans in those segments of the banking industry not covered by the public disclosures, i.e., the regional banks and unbacked trust accounts, as well as restructured loans, the staff’s estimate of total nonperforming loans yields a figure significantly higher than the reported figure.
The problem loans to nonbanks do not always fall into the category of nonperforming loans because some of them are restructured if payments are missed.
Japan, Ministry of Finance (1992). According to IBCA Limited, the total number of nonbank financial institutions in Japan exceeds 20,000, with the top 300 accounting for about two-thirds of their total lending (about ¥ 100 trillion at end-September 1992). This amount represents roughly a fourfold increase in total outstanding loans from end-1986. See IBCA Limited (1992a).
The jusen, in particular, were formed by groups of city banks, long-term credit banks, trust banks, regional banks, insurance companies, securities companies, and agricultural cooperatives.
See IBCA Limited (1992a). Empirical estimates of the value of deposit guarantees to Japanese banks also support the conclusion that the problem of bad loans is somewhat greater for the long-term credit banks and trust banks than for the city banks. See Fries, Mason, and Perraudin (forthcoming).
Under the Basle capital accord, banks are permitted to count 45 percent of unrealized capital gains on securities (but not real estate) as tier 2 capital.
The weakness in equity prices has precluded the issuance of tier 1 capital. Also, as Japanese banks have traditionally paid low dividends to equity holders, there is little scope to increase retained earnings through dividend reductions.
No major bank fell below the target ratios set by the Basle accord for total and tier 1 capital of 8 percent and 4 percent, respectively.
The figures include both the banking accounts and trust accounts of all banks.
For a more extensive discussion of this proposition in the context of Japan, see Tsutsui (1990).
Japanese banks in the second half of the 1980s and early 1990s raised substantial amounts of equity from new share issues and retained earnings, albeit from a relatively low level compared with their assets. Over time, this effort could reverse the incentive for increased risk taking caused by greater competition.
The ceilings on the proportion of the assets of these funds that could be invested in equities were also removed. The ceilings were 30 percent, 80 percent, and 30 percent, respectively.
In Japan, as elsewhere, deposit rates tend to fall faster than the rates of return on banks’ assets.
In principle, credits of housing loan companies and non-banks collateralized by real estate will be eligible only if the financial institutions involved have reached agreement on the allocation of related losses.
The first batch of nonperforming loans of ¥ 681 billion face value was bought by the CCPC at 66 percent of face value on average in March 1993.
The net interest income of banks in Japan and elsewhere tends to mirror interest rate developments, since deposits typically have a shorter maturity than that of bank assets. Thus, as interest rates fall, the average cost of funds to banks declines more rapidly than does their average return on assets, leading to an upturn in operating profits.
The Ministry of Finance reported that of the ¥ 12.3 trillion in nonperforming loans at end-September 1992, ¥ 4.0 trillion was not backed by adequate collateral. At the end of fiscal year 1992, specific reserves for domestic loan losses of the 21 major banks totaled ¥ 1.9 trillion.
The creation of new subsidiaries by banks and securities firms will be subject to licensing requirements that may impact on the pace at which these new affiliates are created as well.
The ¥ 500 billion net asset test leaves unrestricted underwriting for approximately 34 large issuers of securities, the outstanding debt securities of which account for about 40 percent of the domestic corporate bond market.
Japanese banks, particularly those with large branch networks, also have substantial unrealized profits on real estate. These latent profits, however, do not contribute to tier 2 capital and could not be easily realized under current market conditions.
The index of urban land prices in the six major metropolitan areas stood at 150 in early 1986, the beginning of the property boom. It reached a high of 420 in the first half of 1990, and it declined to 300 toward the end of 1992.
In this regard, it is notable that lending by the banking sector to borrowers that are not real estate related declined by 0.3 percent during 1992, whereas lending to the real estate sector grew by 5.4 percent during the same period. The recent growth in outstanding bank loans to the real estate sector has been attributed to the increased lending to (1) help local governments’ land developing entities buy parcels of land; (2) provide operating funds for cash-starved private real estate developers; (3) assist companies in selling their real estate assets as part of their restructuring efforts; and (4) finance redemptions of equity-linked debts issued by private companies.
The authorities retain significant discretion in setting the pace of the introduction of further liberalizing measures, such as the abolition of fixed commissions in securities trading, the removal of the remaining interest rate controls, the increased disclosure of financial information, and the licensing of the new subsidiaries permitted under the Financial System Reform Law.
Banks’ liabilities typically are of shorter maturity than their assets, hence a decline in interest rates temporarily increases the spread between borrowing and lending rates.
As an indication of the difficulties in the commercial real estate market, the vacancy rate for commercial real estate in central business districts increased from 4 percent in 1980 to 17 percent in 1986 (Chart 13). The vacancy rate peaked at 19 percent in early 1992.
The Act authorized the FDIC to borrow up to $70 billion (1.2 percent of GDP) from the Treasury; $25 billion (0.4 percent of GDP) to absorb losses, and the remainder in working capital secured by assets seized from failed banks.
The Depository Institutions Deregulation and Monetary Control Act of 1980 initiated the phased elimination of deposit rate ceilings by 1986 and, in recognition of their weak competitive position, allowed thrift institutions to offer checking accounts. In 1982 the Garn-St. Germain Depository Institutions Act permitted the payment of interest on checking accounts and authorized the introduction of a new money market deposit account. At the same time, the investment powers of thrifts were expanded by federal and state banking agencies to include commercial lending and, in some areas, real estate development.
Such sectoral concentration was aggravated by the inability of the banking sector to branch nationwide (McFadden Act) and thereby achieve geographical diversification. Reform of these restrictions as well as those on product lines (Glass-Steagall Act) are currently under consideration.
The assignment of banks to different capital zones is based primarily on their capital ratios but may in some instances be adjusted on the basis of the numerical ratings assigned to the bank by their supervisors or the existence of supervisory orders concerning the bank.
The prompt corrective action rules prohibit any insured institution that is classified as undercapitalized from paying dividends. Such institutions are also subject to additional mandatory supervisory actions, including increased monitoring by the appropriate federal banking agency, formulation of a capital restoration plan, restrictions on growth in total assets, and limits on new acquisitions, branches, and business activities. Any significantly undercapitalized institution, in addition to the above measures, is subject to mandatory limits on managerial compensation, while any critically undercapitalized institution must be placed in conservatorship or receivership within 90 days. The banking agencies may also use a range of discretionary actions to protect the FDIC against losses.
Such mergers include the acquisition of Security Pacific by BankAmerica and the merger of Manufacturers Hanover with Chemical Bank. BankAmerica estimated at the time of the acquisition of Security Pacific in April 1992 that it would reduce the combined annual operating expenses of the two institutions by $1.2 billion within three years, from a combined base of $7.3 billion, largely through the reduction in personnel costs. By the third quarter of 1992, IBCA Limited reported that BankAmerica had realized approximately one-half of these savings. With respect to the merger of Manufacturers Hanover and Chemical Bank in December 1991, the projected savings in operating expenses were $0.8 billion within three to four years. IBCA Limited reports that the combined unit is ahead of schedule in realizing these cost savings.
Some of the major French banks have been heavily exposed to a string of international corporate failures such the real estate development firm Olympia and York and the MGM studio.