- International Monetary Fund
- Published Date:
- October 1998
The Asian crisis illustrated that countries with weak and underregulated banking systems are less able to manage the negative consequences of volatile capital flows and exchange rate pressures.1 Banking systems in several emerging markets experienced, to different degrees, problems of poor asset quality, outflows of deposits from the banking system or from smaller to larger banks, overexpansion of balance sheets and overexposures to liquidity, market, and credit risks. The policy responses to these difficulties varied across countries and depended, among other things, on the consolidation of the respective banking industry and the extent of regulatory deficiencies.
In this annex, the dynamics of the main Asian banking crises are described, together with an assessment of the performance of the major Latin American and Eastern European banking systems.
Asian Banking Systems
In China, the Asian crisis has increased concern among policymakers about the health of the financial system and has given considerable impetus to accelerating the reform process. The People’s Bank of China (PBC) is itself being restructured along the lines of the U.S. system of federal reserve banks and the number of provincial branches of the PBC is to be reduced from 31 to 12, with greater centralization of control to prevent political interference at the regional level. Attempts to put the banking system on a sound commercial footing have begun with the introduction of a new system for bank loan classification, recapitalization of the four state commercial banks, and restrictions aimed at getting banks out of stock market speculation.
The authorities have decided to adopt the standard five-category classification of bank loans—pass, special mention, substandard, doubtful, and loss—but will not mandate increases in provisions at this stage. Introduction of the new loan classification system is expected to be completed by end-1998, and guidelines for making the transition are to be issued by the People’s Bank. Currently, Chinese banks are only subject to a general provisioning requirement of 1 percent of total loans calculated at the beginning of the year. As yet, there is no specific provisioning requirement and its introduction is to be left to a later stage when the new asset classification is in place. Until end-1996, Chinese banks accrued interest on non performing loans for three years. This was shortened to two years in January 1997 and further reduced to one year in 1998.
There are no official assessments of nonperforming loans under the new asset classification system. In March 1998, the Governor of the People’s Bank reported that the authorities’ estimate of nonperforming loans in the banking system (defined as all loans that were over one day overdue) was 20 percent, with only 6 percent considered as unrecoverable. Some foreign analysts, however, express skepticism in evaluating nonperforming loans of the Chinese banking system and have made estimates higher than the official figure.2 In addition, many believe that the banks’ exposure to the property sector will add to the burden of nonperforming loans as the economy slows down.
To enhance confidence in the banking system, the Ministry of Finance is in the process of issuing Rmb 270 billion of “special” bonds to recapitalize the large state commercial banks, in an attempt to raise their capital adequacy ratios to BIS standards. The recent cut in reserve requirements from 18–20 percent to 8 percent of deposits has freed funds that the banks will use to purchase these special bonds. Although a part of these funds is also to be used by the banks to repay loans from the central bank, full sterilization of the liquidity released is not expected by market participants in view of the growth slowdown. Concerns remain, however, about whether the increase in bank capital ratios will be sustained. Controls on interest rates and the credit allocation process which still prevent any meaningful “commercialization” of banks and competition among them, are likely to again erode capital adequacy ratios. More fundamentally, the authorities need to confront the legacy of policy lending, first by cleaning up banks’ balance sheets and second, by restructuring state-owned enterprises (SOEs) to wean them away from assured funding.
In June 1997, triggered by evidence that banks had been using substantial amounts of their own funds to speculate on the stock markets, the People’s Bank banned state banks from trading in stocks and asked them to unwind their positions. In addition, state banks were prohibited from financing the stock market activities of SOEs, securities houses, and trust firms, from making overdrafts for stock purchases, and from engaging in treasury debt repos.
The current reform of the banking system is being pitched within a three-year horizon, but the process is likely to be more prolonged since the reform of financial intermediaries will have to proceed in tandem with the restructuring of SOEs and fundamental changes in the legal and accounting frameworks. Also, while some small commercial banks have been allowed to operate nationwide, their expansion has not been rapid enough to make inroads into the deposit base of the big state commercial banks. Competition from the nine foreign banks authorized to do renminbi business is minimal and still in the “experimental” phase—they are restricted geographically to Shanghai, permitted to do renminbi business only through their main branch, and their interest rates are subject to the same ceilings as those of domestic banks.
Although the external payments crisis in Indonesia has its more immediate roots in the corporate sector, the weaknesses in the overstretched and underregulated banking system contributed to the initial worsening in market sentiment that precipitated the fall in the rupiah and the ensuing rush to hedge corporate foreign exchange exposures. The share of nonperforming loans for the system as a whole was at around 9 percent in early 1997, but this average masked the existence of seven large and balance-sheet-impaired state banks together with a large number of very small undercapitalized banks.3 Moreover, the group of more dynamic private sector banks began to experience asset quality deteriorations after a rapid expansion of lending fueled by large capital inflows in recent years. Despite an improved regulatory framework, compliance with prudential regulations and credit guidelines was poor, and there were serious doubts on the accuracy and transparency of banks’ financial statements—compounded by a tradition of related-party lending, cross holdings of equities and loans, and liberal options for loan restructuring (“evergreening”). Bank Indonesia’s reluctance to shut down the insolvent banks also raised doubts about the sustainability of the authorities’ strategy for a gradual consolidation of the industry.
Following the widening of the band in July 1997 and the floatation of the rupiah in August, Bank Indonesia tightened liquidity initially but was later on forced to ease its stance as domestic and foreign liquidity conditions sharply deteriorated. The initial tightening intensified the segmentation of the interbank market. Some of the smaller banks were subject to runs as depositors shifted their funds toward state, large private, and foreign banks, or even out of the system. After the closure of 16 small banks in early November, the public accelerated the withdrawal of deposits and Bank Indonesia provided substantial liquidity support, including to some large private banks. Foreign lenders began to selectively reduce and finally cut credit lines, and U.S. dollar deposits fell drastically, leaving some local banks unable to meet their dollar commitments in the international interbank market.
As the rupiah continued to depreciate, fueled by corporates’ attempts to hedge foreign exchange debt and the uncertainties about the extent of the overall external obligations, it became clear that many corporate borrowers would default on their U.S. dollar-denominated debts, and the risks of systemic banking problems rose concomitantly. Moreover, although the banks were believed to run matched currency books, there was uncertainty as to whether the debts of the banks’ offshore affiliates were actually consolidated in the banks’ accounts. Foreign-currency-denominated loans increased by about 200 percent from end-1992 to June 1997, to reach about 20 percent of bank loans, and the exposure to the property sector was even higher—with the big property developers having borrowed heavily in U.S. dollars.
The continuation of liquidity shortages and the escalation of solvency problems in the banking system, which affected also the largest private banks, led the authorities to announce on January 27, 1998, a government guarantee for depositors and creditors (excluding subordinated debt) of the locally incorporated banks, the creation of the Indonesia Bank Restructuring Agency (IBRA), the elimination of restrictions on foreign ownership of domestic banks, and a temporary and voluntary suspension of corporate foreign debt repayments.4 All outstanding liquidity support from Bank Indonesia to the banks—equivalent to 10 percent of GDP by end-January—was transferred to IBRA and the agency took over the responsibility of managing the weak banks as well as the disposition of bad assets. The restructuring agency plans to convert Bank Indonesia’s loans into equity and sell stakes to foreign institutions. By mid-February, IBRA intervened in 54 banks that had used heavily emergency support from Bank Indonesia—after the definition of uniform and transparent criteria for transferring weak banks to IBRA—and 250 examiners were placed in the banks to observe their compliance with prudential regulations. In addition, in an attempt to speed up the consolidation of the banking system, the government announced a sharp increase in minimum capital requirement for banks, redefined nonperforming loans, and issued new guidelines on provisions.5
Significant progress in the restructuring process was made in April 1998, but uncertainties remained about the design and commitment to a genuine restructuring of the banking system as well as to the impact of the corporate sector debt problem on the banks’ balance sheet.6 Over the weekend of April 4–5, seven small banks were closed and seven others (including the second, fourth, and eighth largest private banks) were placed under IBRA’s control. The seven banks closed had been allowed to borrow from Bank Indonesia in excess of five times their equity and 75 percent of their assets, and their depositors’ accounts were automatically transferred to the largest state bank. The seven banks placed under IBRA were supposed to continue to operate normally, but shareholders’ rights were suspended and a governance contract would be drawn with a state-owned bank to provide management and control of the bank. There were at least 40 other banks under IBRA management by early May 1998, and analysts expected many others to follow the same fate as their capital adequacy ratios would be under 5 percent—the ratio at which IBRA supervision is required.
The political and economic environment severely deteriorated by mid-May and concerns about Indonesia’s financial sector grew considerably. A large part of the Indonesian corporate sector is insolvent, owing to translation losses on their foreign currency borrowings as well as the burden of large off-balance-sheet liabilities related to derivative contracts. Other companies are struggling to cope with high interest rates and the inability to import raw materials, as letters of credit from Indonesian banks are no longer being accepted. Standard and Poor’s has stated that banks’ non performing loans could reach 55 percent of total loans over the next 12–18 months, while Moody’s has classified the country’s banking system as insolvent, estimating nonperforming loans at 30–75 percent of total loans. At end-May, IBRA announced that it had taken control of Bank Central Asia, the largest private bank in the country. A week later, the steering committee representing foreign creditors and government negotiators reached an agreement on restructuring Indonesia’s $59 billion of corporate external debt and on the $8.9 billion owed by Indonesia’s banks, allowing the authorities to focus on the domestic side of bank and enterprise restructuring.
The structural weaknesses of the Korean banks—reflected in an average Bank Financial Strength Rating of D, see Table 2.6 in Chapter II—became increasingly apparent during 1997, as GDP growth slowed and the operating environment deteriorated. These structural weaknesses were the result of years of bad lending practices and a weak supervisory and regulatory framework. Historically, the banking system was the vehicle used to support the government’s industrial policies. Although policy loans were scaled back during the 1990s, they left a legacy of poor managerial and credit analysis skills, as well as large exposures to the highly leveraged chaebols (conglomerates). Also, in response to the competition from underregulated nonbank financial intermediaries, commercial banks expanded the use of trust accounts—subject to no interest rate or exposure restrictions—that were maintained separate from the rest of the balance sheet but nevertheless heightened the risk profile of the banks.7 Finally, the system’s lack of transparency and regulatory forbearance significantly underestimated asset quality problems. The authorities had traditionally published as bad loans only the two most seriously delinquent categories of loans, amounting to only about 1 percent of total loans at end-1996, while other reported categories including a larger amount of impaired loans were not disclosed. Also, provisioning requirements for nonperforming loans and for losses on the banks’ securities portfolios were relaxed over 1995–96.
The banking system exhibited increasing signs of stress during the first half of 1997 as a number of major conglomerates went bankrupt, and the authorities finally announced in August 1997 a set of measures to address their rapidly deteriorating liquidity and solvency condition. In the first quarter of 1997, two rating agencies downgraded the three major creditor banks of two bankrupt conglomerates and subsequently placed them under review for further downgrades during the summer, arguing also that the banks’ low profitability and large losses on their securities holdings—among other factors—had left them badly positioned to handle a substantial rise in problem loans. By the first week of September, six highly leveraged chaebols had failed or been placed under bankruptcy protection, raising serious concerns about the banks’ asset quality. Concerns were also raised about the condition of the merchant banks, which were heavily exposed to the conglomerates and suffered severe funding problems as domestic and foreign lenders reduced their exposures to them.8 On August 25, 1997, the authorities announced a set of measures aimed at increasing confidence in domestic and international financial markets. First, official support was provided by the Bank of Korea, in the form of a special loan to Korea First Bank—the fifth largest commercial bank—and the government acquired stock in the bank in exchange for government bonds. In addition, a special funding facility was created to assist merchant banks whose exposure to bankrupt companies exceeded 50 percent of their equity—in the event, 21 out of 30 merchant banks. Second, a special fund was set up within the Korea Asset Management Corporation (KAMC), to which the banks would be allowed to sell their nonperforming loans. Third, guarantees were announced by the government on the foreign liabilities of Korean financial institutions, including both commercial and merchant banks.
These measures were perceived by market participants as insufficient, and as the quality of their balance sheets continued to deteriorate, the Korean banks faced increasing liquidity problems in the international interbank market. Short-term foreign borrowing had soared in the years immediately preceding the crisis, fueled by the ample availability of international liquidity, the perceived stability of the won and the regulatory ceiling on commercial banks’ medium- and long-term borrowing in international financial markets. The situation changed dramatically when international investors focused on Korea’s financial problems, and many banks saw their credit lines either withdrawn or reduced by their correspondent banks. In December, the Bank of Korea allowed the won to float freely and investors and lenders panicked when they learned that the country’s short-term external debt was approximately $104 billion—rather than the $66 billion originally reported—and that usable reserves were lower than expected. As a result, the Korean banks’ short-term external liabilities fell from $62 billion in September 1997 to $29 billion at end-December, and despite substantial foreign exchange liquidity support from the Bank of Korea, a default was averted only when the central banks of a group of industrial countries arranged a three-month extension of the maturing bank debts.9
A new series of measures to restore confidence and restructure the banking system has been undertaken since December 1997. First, following the results of an evaluation of the merchant banks and their rehabilitation programs, 14 merchant banks were closed between January and April 1998 and the other 16 have to comply with a timetable to achieve capital adequacy ratios of at least 6 percent by end-June 1998 and 8 percent by end-June 1999. A Bridge Merchant Bank was established at end-December 1997, to pay out depositors of suspended merchant banks and to take over, manage, collect, or liquidate their assets. Second, liquidity support in won was provided to help banks cope with the suspension of merchant bank operations. Third, the 12 commercial banks whose capital adequacy ratios at end-1997 (under full provisioning)10 fell below 8 percent have submitted recapitalization plans to the supervisory authority that will be evaluated by end-June 1998. On January 30, 1998, the government injected W 1.5 trillion in both Korea First Bank and Seoul Bank—following a capital reduction whereby shareholders’ equity was substantially reduced—and the two banks are being prepared for privatization. Fourth, the administration of deposit insurance funds for commercial and merchant banks and other institutions was consolidated in the Korean Deposit Insurance Corporation (KDIC), and the coverage was extended to almost all the domestic liabilities of these institutions. Finally, a Financial Supervisory Commission was established and has assumed responsibilities for the supervision and restructuring of all bank and nonbank financial institutions.
As the problems in the corporate sector continued to escalate in the first quarter of 1998, the Korean authorities unveiled further plans to restructure the financial system. The Korean banking system is closely interlinked with the conglomerates and the merchant banks. Substantial cross-payment guarantees allow the financial difficulties of a single chaebol company to threaten the solvency of the entire chaebol group, thereby increasing the risks of spillovers from the corporate sector to the banking system. In the first quarter of 1998, more than 10,000 companies went bankrupt—compared with 14,000 for the whole year in 1997 and 11,570 in 1996—with small firms hardest hit as banks continued to lend mostly to the conglomerates. Bankruptcies also continued to grow among the chaebols, and the average debt-to-equity ratio of the top 30 biggest ones jumped to 518 percent at end-December 1997 from 386 percent in the previous year—in part owing to the unhedged foreign exchange component. The Financial Supervisory Commission has asked the banks to appoint outside directors to special teams set up to determine the viability of companies in a transparent way by the end of May, and those declared viable should submit restructuring plans to the banks by July 1998. On May 20, 1998, the authorities announced that they will issue W 50 trillion of bonds as part of a plan to buy nonperforming loans and recapitalize the banks. The KAMC will pay W 25 trillion to buy W 50 trillion of bad loans, while the KDIC will invest about W 16 trillion in banks to boost their capital. The banks are expected to raise a further W 20 trillion by selling equity and subordinated debt to private investors, aided by the removal of foreign ownership restrictions in the sector. Finally, W 9 trillion will be available to pay off depositors of failed institutions. The announced support package brings the government resources committed to the financial system—including support to depositors—to W 84 trillion (14 percent of GDP), while impaired loans reached W 118 trillion at the end of March 1998.
Korea’s economic recovery depends critically on the speed of corporate and financial restructuring, but despite the reform measures announced many market participants have doubts about whether they can be fully implemented and sustained. The Korea Development Institute has warned that the recession could last three to five years if financial and corporate restructuring proceeds slowly. On the one hand, despite a series of announced corporate bankruptcies, insolvent companies have been allowed to continue operating and undercut their competitors.11 On the other hand, the top five chaebols have submitted restructuring plans to reduce debt-equity ratios to below 200 percent by end-1999, cut business lines to no more than five, and eliminate cross-payment guarantees. Also, the government’s requirement that firms be classified according to their chances of survival was announced with no clear guidelines, raising concerns about the banks’ ability and willingness to distinguish solvent firms from insolvent ones in a short period of time.
The Malaysian banking system strengthened considerably following the crisis of 1985–88, owing to very rapid economic growth, buoyant share and property prices, and the enactment of strict prudential regulations. Asset quality improved substantially—the ratio of nonperforming loans to total lending fell from a peak of 35 percent in 1987 to 3.6 percent by mid-1997—and capital adequacy was appropriate going into 1997. Moreover, restrictions on foreign borrowing left the corporate and banking sectors with relatively low exposure to foreign exchange risks, ameliorating the impact of the external credit squeeze affecting the region and providing the authorities more latitude to respond with the appropriate macroeconomic and structural policies.
Despite the relative strength exhibited by the Malaysian banking system going into the second half of last year, the severe pressures in the stock and foreign exchange markets that followed the depreciation of the Thai baht raised concerns about the vulnerability of the financial system. The main source of vulnerability is the high degree of leverage of the economy—with one of the highest loan to GDP ratios in the world (see Figure 2.13 in Chapter II), together with large exposures to property and stock markets. In order to address these vulnerabilities, Bank Negara Malaysia issued guidelines in March 1997 to limit banks’ exposure to property to 20 percent of the loan portfolio and that of stocks and shares to 15 percent of the portfolio, and in October 1997 it required banks to submit individual credit growth plans in an attempt to restrict overall credit growth to 25 percent in 1997 and 15 percent in 1998. In the event, loan growth continued to be high, reaching 30 percent annually during 1996–97.12 As a result, commercial banks became heavily exposed to the property sector (30 percent of loan portfolios), the stock market (15 percent), and consumer lending (13 percent) as of end-1997.13
With the onset of the regional crisis, banks and finance companies experienced a significant decline in profitability and asset quality deteriorated sharply. The level of nonperforming loans increased from 3.6 percent in June 1997 to 5.7 percent at end-1997. Under the new classification system where loans are deemed nonperforming when past due for three months, the level of nonperforming loans was 6.7 percent at end-1997. Moreover, the level of nonperforming loans under the new classification system increased sharply to 9.3 percent in March 1998. The banking system as a whole remained adequately capitalized as indicated by the 10.6 percent risk-weighted capital ratio at end-December 1997—the same level as the previous year—despite heavy losses also in share-related lending. Two banks and three finance companies fell below the 8 percent capital adequacy requirement for end-1997, and more are expected to do so as the level of nonperforming loans continues to increase led by higher defaults in the property sector and further losses at the banks’ stockbroking subsidiaries. The looming over-supply of real estate has not yet been fully felt in real estate prices (see Figure 2.15 in Chapter II) and will put further pressure in asset quality toward the end of the year. In order to ameliorate the expected property price deflation, the government announced on May 20, 1998 that foreign investors are now allowed to buy all types of residential and commercial property without restriction if the cost of the property is a minimum $65,000 and financing is obtained overseas.
Until the first quarter of this year, the policy response of the Malaysian authorities had been to inject liquidity into the interbank market in order to keep interest rates low regardless of the negative impact on the currency. Growing concerns about the vulnerability of the financial system led to a shift in deposits from domestic to foreign and large domestic banks, as well as to increased segmentation in the interbank market. This prompted the central bank to extend significant liquidity support to the affected institutions—primarily Tier 2 banks and finance companies—with Bank Negara’s deposits with financial institutions rising to RM 35 billion (12.5 percent of GDP) at end-January.14 Also, in that month, the authorities announced that all depositors in Malaysian financial institutions would be guaranteed, and the liquidity pressures somewhat eased by the end of the first quarter.
Facing the prospects of a large number of finance companies becoming undercapitalized during 1998, the authorities have taken steps to consolidate the sector. The merger program, directed at consolidating 39 existing finance companies into 6 or 7 anchor companies, is being supported by a one-year guarantee extended by the government to the acquiring institution in the event of any further reduction in the value of the acquired assets. The final approval of the mergers would be contingent upon completion of due diligence, with final payments to shareholders based upon the valuations established through this process. Many domestic and foreign-owned finance companies are expected to be absorbed by commercial banks within the group. The institutions that chose to stay out of the merger program will have to demonstrate that they are able to comply with the capital adequacy requirements, that will rise gradually from 8 percent to 9 percent by end-1998 and to 10 percent by end-1999, and those that are unable to do so will be subject to alternative resolution strategies—including conservator-ship or liquidation.
Despite Bank Negara’s encouragement, only two bank mergers were completed in 1997 and two others were announced in March 1998. As the economy slows down in 1998 some Tier 2 banks are expected to experience difficulties and this will accelerate the pace of consolidation. The separation of banks into two tiers in 1994 encouraged banks to raise capital to reach Tier 1 status—rather than the intended effect of leading to mergers and acquisitions. As a result, Tier 2 banks had an average equity to assets ratio of 8.7 percent in December 1977, higher than the 7.5 percent average of the Tier 1 banks. This is likely to change in the near future, as Tier 2 banks have increased their portfolios much faster than Tier 1 banks and face the prospects of increased provisions. In order to accelerate the process of consolidation, the authorities announced on May 21, 1998, the creation of an Asset Management Company, designed to manage and liquidate bad assets of the financial system. The pace of consolidation will be slower, however, if the relaxation of foreign ownership rules—currently restricted to 30 percent of equity—does not occur in 1998.
The supervisory and regulatory framework is regarded by market participants to be good by regional standards, and it has recently been improved with a series of measures designed to bring it to international best practices. Between December 1997 and March 1998, the authorities approved a shortening of the period for classifying loans as nonperforming from six to three months, an increase in general and specific provisions, the requirement to publish and comply with capital adequacy standards on a quarterly basis, and to subject all banks to monthly stress tests and ensure that they take corrective action before they actually need additional capital—among others.
The Philippines has not been insulated from the financial turmoil in the region, but the country’s banking system shows a more favorable risk profile than many other countries in the region. The fact that rapid loan growth is a fairly recent phenomenon—with the resulting low leverage (see Figure 2.13 in Chapter II)—combined with a fairly adequate regulatory framework and a decisive response from the authorities, has contributed to a relatively lower financial vulnerability. The currency depreciation and higher interest rates that accompanied the reversal in capital flows during the second half of 1997 led to a deterioration in asset quality, and prompted the central bank to induce lower domestic interest rates by (1) reducing banks’ intermediation costs (via lower statutory reserve requirements and a higher portion of banks’ reserve balances with the central bank earning interest); (2) increasing liquidity (through opening of credit windows and outright purchase of government securities); and (3) reducing central bank policy interest rates.
Although the exposure to the property market and foreign currency lending have grown rapidly in recent years, and the risks to the banks of such exposures were heightened by the depreciation of the peso, these risks appear to be of a smaller magnitude than in other countries in the region.15 First, although U.S. dollar lending is estimated to account for some 30 percent of total loans, the extent to which U.S. dollar loans have been made to unhedged borrowers appears to be relatively small.16 Moreover, during the currency crisis, the central bank introduced a currency risk protection program—also known as the nondeliverable forward (NDF) facility—which was offered by the central bank through commercial banks to help eligible borrowers hedge existing foreign exchange liabilities. Second, most real estate developers tend to have a low degree of leverage and the initial concerns of a property glut appear to have been exaggerated, as initial supply forecasts were later on scaled back and residential and office prices have shown more resilience than in other cities (see Figure 2.16 in Chapter II). Nevertheless, prospects for the real estate sector remain weak as uncertainties about the region are likely to affect the property market.
The deterioration in asset quality since mid-1997 has led to difficulties in some small banks, but most of the large commercial banks held capital to risk-weighted asset ratios of 15–20 percent by end-1997 and are likely to withstand the increase in bankruptcies and debt restructurings. Only one small commercial bank and three thrift banks (accounting for 0.33 percent of banking system assets) have experienced difficulties as of May 1998.17 Rural and thrift banks are regarded by market analysts as the weakest segments of the banking system and more failures are expected but with unlikely systemic consequences. The ratio of nonperforming loans to total loans of the commercial banks has increased from 3.4 percent in June 1997 to 4.7 percent in December of 1997 and 6.7 percent in February 1998. Nonperforming loans are expected to increase some 3 percentage points following the reclassification of loans as nonperforming after three months—compared with six months currently—and analysts expect them to reach 10–15 percent by year-end. More important, loan loss provisions covered only 40 percent of nonperforming loans, but provisioning expenses are increasing and the large banks are still reporting healthy profit levels. Overall, the commercial banks were regarded as adequately capitalized as of December 1997, with an average capital adequacy ratio of 16 percent. Even with a 50 percent write-off of the projected nonperforming loans, most of the large banks would remain within the regulatory requirements.
A series of measures have been implemented to strengthen the banking system. Before July 1997, the central bank imposed a limit on real estate loans of 20 percent of the total loan portfolio—with full compliance by May 1998—and reduced the maximum loan-to-value ratio to 60 percent from 70 percent. Also, a 30 percent liquid cover on all foreign exchange liabilities in FCDUs was approved in June 1997. In October, loan classification was tightened (with full compliance by April 1998), and a 2 percent general loan loss provision was imposed (with compliance of 1 percent by October 1998, another half percent by April 1999, and the balance by October 1999). In early 1998, the banks’ minimum capital requirements were raised,18 and increased disclosure of nonperforming loans, provisions, and capital was required for December 1998.
The increasing weaknesses in Thailand’s financial system were already apparent in the first half of 1997, as demonstrated by the solvency problems of several finance companies.19 Although some of the major Thai commercial banks exhibited sound balance sheets, the smaller commercial banks and most finance companies were suffering a serious deterioration in asset quality even before the depreciation of the baht on July 2, 1998, as a result of the economic slowdown and their overexposure to a distressed property sector. Following the suspension of 16 finance companies just before the baht was allowed to float, another 42 were suspended on August 5.
The sharp depreciation of the baht raised further concerns about the quality of the banks’ portfolios, as many corporate borrowers had large unhedged U.S. dollar-denominated liabilities, and this led to growing liquidity—both domestic and external—problems. Notwithstanding the measures announced by the Bank of Thailand on August 5, 1997, including a guarantee to depositors and ordinary creditors (subordinated creditors were explicitly excluded) of the 15 banks and the 33 finance companies still in operation, there were runs by depositors on some of the smaller banks.20 The Bank of Thailand provided liquidity support through the Financial Institutions Development Fund (FIDF), that arranged a “recycling facility” to channel some of the deposits back to the smaller banks. The FIDF liquidity support to finance companies and commercial banks reached about 15 percent of GDP during 1997. Meanwhile, the banks were suffering increasing difficulties rolling over their large short-term offshore liabilities—a large share of which was owed to Japanese banks through the Bangkok International Banking Facilities (BIBFs).21
The continuing financial instability led the authorities to announce a financial sector restructuring package by mid-October 1997, but the markets remained unsettled until after the new government adopted a series of specific measures to begin to address the financial system’s problems. The October package included the creation of a Financial Sector Restructuring Authority (FRA)—to assess the rehabilitation plans of the finance companies and dispose of their assets, an Asset Management Corporation (AMC)—to manage and sell the bad assets, as well as tighter loan classification and provisioning rules and a relaxation of the limit on foreign ownership of financial institutions—from 25 percent to 100 percent, for up to 10 years. However, the October package failed to reassure investors as it continued to delay the final fate of the finance companies and did not address the severe bad loan problem of the other financial institutions. The new government that took over in November subsequently closed down 56 of the 58 suspended finance companies and, in the first quarter of 1998, the Bank of Thailand intervened four commercial banks, and had their management replaced and the capital of existing shareholders written down through debt-equity swaps by the FIDF.22
The strategy to restructure the financial sector became increasingly focused on strengthening the core banking system during the first half of 1998. The authorities took concrete steps to initiate the sale of assets of finance companies and to restructure and recapitalize the banks, but substantial hurdles remain in both tasks. The FRA auctions of noncore assets began in February 1998. The first assets being sold were automobiles, which make up around 6 percent of the total assets to be sold, while actual hire purchase loans (core assets) were auctioned much later in June 1998. The bulk of the assets are commercial real estate loans and these auctions are likely to be more difficult unless stronger property and contractual rights can be enforced. Current laws in Thailand make it difficult to foreclose and liquidate an insolvent debtor, and in order to facilitate this process the authorities have amended the bankruptcy law and are in the process of proposing amendments to the Code of Civil Procedure.23 In addition to the AMC, which will be a bidder of last resort in the sale of finance companies’ assets, the authorities created a Radhanasin (“Good”) bank (RAB), to bid for a limited portfolio of superior assets.
The sheer size of the banks’ recapitalization needs, heightened by the expected deterioration of the loan portfolio, makes further government support quite likely. Officials at the Bank of Thailand estimate the level of bad loans to have at least doubled by the end of December 1997 from around 9 percent earlier in the year. Nonperforming loans on a 90-day-past-due basis were estimated at above 25 percent of total loans at end-1997, and only one-fifth of them were provisioned against. Moreover, as asset prices and GDP continue to fall, nonperforming loans are expected to reach at least 30 percent during 1998, leading to recapitalization needs that could approach 30 percent of GDP (see Table 2.7 in Chapter II). The strategy of recapitalization centers on financial institutions’ raising new capital on their own—except for the four intervened commercial banks—including from potential foreign partners. The banks have to sign memorandums of understanding with the Bank of Thailand by August 15 on their proposed recapitalization plans, but market participants expect that a sizable portion of the recapitalization requirements will be met by the government, especially for small banks. As of early May 1998, about B 230 billion (4 percent of GDP) of new equity had been raised, with B 107 billion contributed by the FIDF.24 Another indication of the likely official support emerged in late May, as the government took over seven other finance companies and Bank of Thailand officials suggested that only 5 to 10 of the remaining 28 finance companies would eventually succeed raising capital on their own. Also, the authorities announced that no more bank closures would be forthcoming.
The Thai authorities have been gradually improving the banks’ regulatory framework as well as accounting and disclosure rules, but the framework remains less stringent than other emerging markets. Recent improvements include the introduction of capital adequacy standards and the tightening of nonperforming loans recognition and provisioning requirements. Thai commercial banks and foreign branches are currently required to maintain a minimum capital ratio of 8.5 percent and 7.5 percent, respectively, while foreign BIBFs are exempt from capital requirements. At present, loans become substandard only after 6 months in arrears. After 12 months the uncollateralized part becomes doubtful, while the collateralized portion remains substandard. This has been tightened and, beginning July 1, 1998, loans with payments overdue more than 3 months will be classified as nonperforming or substandard. The provisioning rate for substandard loans was raised to 20 percent of the value of the loan from 15 percent in 1997, but banks are allowed to subtract 90 percent of the collateral value to calculate the provision.25 This implies that the full impact of the new provisioning requirements cannot be determined until the authorities release the new valuation standards. As the extent to which property values have fallen in Thailand is highly theoretical owing to the illiquidity of the real estate market, rating agencies estimate that the new requirements are only slightly more onerous than those currently in place. Moreover, the provisioning requirements will be phased in gradually, with 20 percent of the new provisions to be made by the second half of 1998, 60 percent by the second half of 1999, and 100 percent by the second half of 2000.
Latin American Banking Systems
The recovery and strengthening of the Argentine banking system was tested by the contagion effects of the Asian crisis and showed a remarkable resilience. Deposits continued to grow during the last quarter of 1997, in sharp contrast with the Tequila shock—when deposits fell by more than 15 percent in four months, and despite the fact that they are partially covered by a limited deposit insurance scheme. The stock of deposits grew by 29 percent in 1997, outpacing the increase in loans of 18 percent over the same period. The most important effect of the sharp increase in interest rates in the fourth quarter of 1997 was the losses in the banks’ fixed income portfolio that contributed to the generalized decline in profitability for the year as a whole.26 Some banks were able to minimize the effects of the higher interest rates on their profit and loss accounts by shifting securities from their trading book to the investment account (where securities are valued at cost but have to be held for one year) or to the available-for-sale account (where securities are marked-to-market but unrealized capital gains/losses are accounted in shareholder’s equity).
Asset quality improved steadily throughout 1997, with the ratio of past-due loans to total loans falling to 10.2 percent in December 1997 from 12.9 percent in December 1996. Although the stock of problem loans as percentage of total loans remained substantially higher among public banks than among private banks, both groups experienced a similar improvement in asset quality. Loan loss reserves remained relatively low for the system as a whole—at 60 percent of past-due loans in December 1997, but they show a positive trend and, more important, net past-due loans were just 22 percent of total equity. This reflects the relatively high level of capitalization of the Argentine banks that ended the year with a ratio of equity to assets of 12.4 percent. The risk-weighted capital adequacy ratio for the system as a whole fell from 19.6 percent in December 1996 to 18.2 percent in December 1997; for private banks the fall was from 15.5 percent to 14.6 percent, but the capitalization levels still remain well above the 11.5 percent requirement of Argentine regulations.
The process of consolidation of the highly fragmented Argentinean banking system continued during 1997, albeit at a slower pace than after the liquidity crisis of 1995. The total number of banks fell from 205 in December 1994 to 153 in March 1996—when several loss-making provincial banks were privatized and a number of private banks were merged or taken over by stronger institutions—and it stood at 138 institutions at end-1997.27 More important, the three largest private banks have continued to increase market share, and they are expected to further accelerate the consolidation process with their recent introduction of new products and the new requirement that all companies pay their salaries electronically through the banking system. This will certainly force medium and small banks to rethink their strategies or sell out to stronger/larger banks.
The monetary authorities have strengthened considerably the regulatory framework after the Tequila crisis, developing a “systemic liquidity policy” and a so-called BASIC system of banking oversight, which contributed to reducing the liquidity and solvency risks of the Argentinean banking system. The systemic liquidity policy takes into account the fact that emerging markets are likely to face sharp and unexpected reversals in capital flows and that this could severely endanger the objectives of macroeconomic and financial system stability. The two main instruments of the liquidity policy are a stand-by repo facility with international banks and legal liquidity requirements for the banks, that together cover some 30 percent of the deposit base. The repo facility is a line of credit arranged with a group of foreign financial institutions, whereby the central bank and the participating local banks have the option to engage in repurchase agreements with Argentine government securities for up to $6.7 billion. The legal liquidity requirements apply to deposits and other liabilities (such as lines of credit from abroad, interbank repo operations, and commercial paper) and have been increased from 15 percent in February 1996 to 20 percent by February 1998.28
The BASIC29 approach to banking oversight combines aspects of the conventional approach to bank regulation and supervision (with on-site inspections and off-site analysis) with the monitoring and discipline imposed by the market (including the use of credit rating agencies). The approach includes the obligation for banks to issue subordinated debt for an amount equivalent to 2 percent of total deposits. In order to ensure an adequate production and dissemination of information about the soundness of the system, the central bank has established a credit bureau and makes available banks’ balance sheet information on the internet on a monthly basis. The accuracy of the information is checked through quarterly external audits, and since January 1998 banks have to obtain two ratings from well-established private credit rating agencies. The regulatory framework has also been enhanced with a tightening of capital requirements through the introduction of a more stringent criterion for calculating risk-weighted assets, requirements for market risks, and strict rules for loan classification and provisioning (see Box 2.8 in Chapter II).
Among the Latin American emerging markets, Brazil’s financial markets were probably the most affected by contagion from Asia. The monetary authorities’ rapid and forceful defense of the real, engineered through massive intervention in foreign exchange markets and a doubling of the basic reference interest rate to 43 percent in late October 1997, led to liquidity pressures and a deterioration in asset quality in the banking system that is still to be felt in the banks’ balance sheets. The deterioration in asset quality is unlikely to create systemic risk as the large domestic banks are very well capitalized and, together with a large number of foreign entrants, have shown that the trend toward consolidation started after the Real Plan in 1994 continues unabated.
The events of October 1997 had a major impact on the liquidity and funding—both domestic and external—of the Brazilian banking system. Banks with large leveraged positions—the major investment banks and some of the smaller banks—were exposed to margin calls and a number of medium-sized commercial banks that suffered from a flight-to-quality of their deposits were the most severely affected by the liquidity pressures. There were apparently no cases of banks obtaining emergency liquidity from the central bank, but the monetary authority injected large amounts of liquidity through rediscount lines and through the purchase of government bonds. Brazilian banks were very active issuers in the Eurobond market during 1996 and 1997, and in order to ameliorate the effects of reduced access to that market, the central bank eased restrictions on bond maturities and allowed short-term foreign borrowings to be invested in high-yielding dollar-linked government bonds.
The impact of the increased interest rates on asset quality will be felt mostly in the second and third quarters of 1998, but the rapid decline in rates to almost precrisis levels suggest that the deterioration may be less than originally expected. The ratio of nonperforming loans as reported by the central bank (loans in arrears and liquidation less accrued interest charges) rose from 7.1 percent in December 1996 to 7.8 percent in December 1997 and 8.3 percent in March 1998. Because of accounting lags, nonperforming loans are expected to peak in the second or third quarter of 1998. However, loan loss provisions also grew considerably and in March 1998 they covered 138 percent of nonperforming loans. At the beginning of the crisis, analysts estimated that nonperforming loans would double from their 1997 level, but the relatively more favorable interest rate environment and increased write-offs in the first quarter of 1998 have led to downward revisions in these estimates.
The public sector continues to have a strong presence in the banking system, controlling 44 percent of total banking system assets by mid-1997, and public-sector banks have weaker financial fundamentals than the private sector ones. Banco do Brasil, the largest commercial bank in the country and one of the main providers of funding in the interbank market, suffered record losses of over $7 billion in 1996 and was recapitalized by the government. Under new management, the bank has taken steps to reduce political interference in its lending decisions, cut costs, and invest in automation technology. In 1997, the bank showed positive profits for the first time since 1994 but nonperforming loans remain at 13.3 percent of total loans. Rapid asset growth in 1997, driven by a 37.8 percent expansion in the loan portfolio, contributed to a fall in the equity to assets ratio from 6.8 percent at end-1996 to 5.5 percent at end-1997. In December 1997, the federal government acquired ownership of the largest state bank, Banespa, following a swap of R$53 billion in state debt for long-term government securities and a clean up of its operations. The bank still has a ratio of nonperforming loans of 23.7 percent but its private sector loans have been fully provisioned, its payroll cut from 34,000 to 22,000, and several branches have been shut down.30
The top-tier private banks have successfully integrated the provision of credit with fee-based services—such as underwriting, capital markets advisory, asset management, and insurance—while a number of medium-sized and small retail banks—with smaller regional bases and branch networks—are having difficulties adapting to the new low-inflation environment. The largest banks expanded quite rapidly during 1997, with asset and loan growth of 45 percent and 33 percent, respectively.31 They have continued to enjoy growing noninterest income—mostly fees and commissions, but many of them suffered large losses in their trading books during the last quarter of 1997. For the past three years these banks have cleaned up their portfolios, writing off substantial amounts of bad loans—that stabilized at 3.2 percent of total loans at end-1997. The compression in interest margins has led to a rapid expansion in consumer lending—particularly in the areas of automobile loans and credit cards—where loan losses are relatively higher. The top three banks (Bradesco, Itaú, and Unibanco) are regarded as very well managed and, with large capital adequacy ratios (in the 12–18 percent range at end-1997), they are well positioned to withstand the deterioration in asset quality anticipated by several analysts, as well as to proceed with further acquisitions. In contrast, the medium-sized and small domestic banks had lower asset growth than the larger ones (at 7.7 percent during 1997), lower noninterest income, and a higher level of nonperforming loans (4.4 percent of total loans by end-1997). While these institutions are still well capitalized, with an equity to assets ratio of 11.4 percent, they have had declining earnings and face increasing competition from the larger and more diversified banks.
The consolidation process initiated since the introduction of the Real Plan in mid-1994 has accelerated in 1997 and analysts foresee that the number of banks will shrink over time from 233 institutions at present—including commercial, investment, multiple, and savings banks—to about 100.32 The successful process of guided consolidation executed by the central bank has led to an increasing participation of foreign investors (currently estimated at around 18 percent of the banking system’s equity) and to a series of mergers and acquisitions among domestic banks. Recent foreign investment in Brazil’s financial system began in March 1997 with Hongkong and Shanghai Banking Corporation’s purchase of Banco Bamerindus, once the country’s third largest private sector bank. In April 1997, Spain’s Banco Santander acquired a small bank that was later on used as a vehicle to purchase a medium-sized bank, and in April 1998 Spain’s BBV agreed to buy a controlling stake in Banco Excel Econômico—the ninth largest private bank. In September 1997, Banco InterAtlantico took over another mid-sized bank and a Portuguese bank announced its intention to buy another one. In addition, several Brazilian investment banks have been acquired by foreign institutions, while others are diversifying into the retail business.
In response to the increased foreign competition, the three largest private banks have acquired medium-sized banks, and they are preparing to bid in the sale of the remaining state banks. In June 1997 Banco Itaú acquired Banerj, in December 1997 Banco Bradesco acquired BCN, and in March 1998 Unibanco acquired a controlling stake in Banco Dibens, in an attempt to consolidate their leadership in the industry and expand their client bases. Two large state banks have already been privatized, two have been liquidated, and the authorities hope to complete the privatization of seven state-owned banks in 1998. Several options are being considered for the remaining ones under PROES (Program for the Reduction of the State Public Sector in Banking), including converting them into development agencies (which would not take deposits from the public), restructuring them, or liquidation.
The central bank has made important progress in strengthening regulation and improving supervision of the banking system, though much remains to be accomplished. An 8 percent risk-weighted capital adequacy requirement was introduced in August 1994, and it was increased to 10 percent in May 1997 (effective December 1997) and to 11 percent in November 1997 (effective January 1999). In addition, in May 1997 Brazil introduced capital requirements for the coverage of counterparty credit risks arising from derivatives positions, but there are no requirements to cover for market risks. Following the events in October 1997, the authorities moved quickly to establish “fire walls” in the banks’ asset management operations, that is to separate the treasury and proprietary operations of the banks from those of the funds they manage for third parties. From a regulatory point of view, rating agencies have expressed concerns with respect to the quality and frequency of disclosure, the exclusion of renegotiated loans from measures of non-performing loans, the booking of operations through offshore subsidiaries, complicated ownership structures designed to avoid taxes and regulations, and the significant holdings of some banks in nonfinancial enterprises. Supervision has shifted from emphasizing compliance with financial ratios to the appointment of a team of supervisors to each bank in order to understand and assess the bank’s risk exposures and controls, with a view to achieve an effective consolidated supervision. Finally, the central bank is establishing a centralized credit risk database which will collect and disseminate information on individual borrowers with outstanding debts of R$50,000 or more.
The Chilean banking system continues to be the strongest in the region, and this strength has allowed the system to withstand the fallout from the recent Asian crisis. As a result of more than a decade of economic growth and stability, combined with one of the best regulatory environments in emerging markets, Chile has the highest loan leverage in the region, with system loans measuring 57 percent of GDP as of 1996 and cumulative loan growth exceeding GDP growth by 33 percent since 1990. Loan growth reached 19.7 percent in 1997–13.7 percent in real terms, almost double GDP growth—and asset quality remained at levels comparable to those of industrial countries. Past-due loans were just 1.06 percent of total loans at end-December 1997, roughly the same level as in 1996, with loan loss reserves covering 167 percent of such loans.33 With roughly one-third of total exports destined for Asian markets, Chile is one of the most vulnerable countries in the region to a recession in Asia. Moreover, the central bank’s response to pressures in the foreign exchange market with a 200 basis points increase in interest rates early this year, is also going to contribute to a slow down in economic activity and a deterioration in asset quality. However, the prudent lending practices of Chilean bankers, strict limitations to currency and maturity mismatches, and the regular upgrading of provisioning requirements by the supervisory institutions have kept the system’s vulnerability to such deterioration in asset quality or to currency depreciation relatively low.
Chilean banks’ earnings have come under pressure in recent years, as the industry has experienced a series of mergers that have intensified competition and reduced interest margins. Profitability continued to be depressed as the net interest margin deteriorated from 3.05 percent in 1996 to 2.64 percent in 1997. Net income for the system grew only 4.4 percent in 1997, while return on average equity declined from 15.7 percent in 1996 to 13.8 percent in 1997. Noninterest income grew at a faster rate of 14.7 percent, bringing this source of revenues to almost 20 percent of total revenues—still a low percent compared with other countries in the region. The contribution of net revenues from affiliates also grew strongly in 1997 and is expected to become a more important source of revenue as the banks take advantage of the opportunities created by the new banking law. The weak profitability was also influenced by large merger-related costs incurred by the system’s two largest banks.
After four years of legislative discussion, a new General Banking Law was approved in October 1997. The new law will allow banks to extend their scope both domestically and internationally, and restore the industry’s profitability levels. Banks will now be allowed to open affiliates and subsidiaries abroad, buy minority or majority stakes in foreign financial institutions, and lend directly to foreign corporations and domestic firms operating abroad. However, banks’ participation in foreign credit operations will be limited to 70 percent of their capital and reserves, and to countries with which Chile has established reciprocal supervisory arrangements. Previous fire walls will be dismantled to permit banks to directly own and operate brokerage, securities underwriting, fund management, and insurance subsidiaries. Banks will also be permitted to offer a second group of domestic businesses such as factoring, credit collection, custodial services, and financial consulting. Also, new, objective criteria for the granting of banking licenses have been published.
Despite their excellent asset quality, Chilean banks show less impressive capitalization levels, and the new law attempts to remedy this with the imposition of stricter rules for capital adequacy. The ratio of equity to total assets fell from 5.3 percent at end-1996 to 5.0 percent at end-1997. The new banking law has eliminated the previous maximum leverage limit and has established a modified version of the Basle Committee risk-weighted capital adequacy requirements beginning in 1998. The financial system’s capital adequacy ratio was 11.5 percent by end-1997, with the domestic banks showing a lower 10.5 percent ratio. Chilean banks and finance companies that were not in compliance with the new requirements as of November 1997 had to present a recapitalization plan to the Superintendency of Banks and Financial Intermediaries and will have two years to comply with the 8 percent ratio. In anticipation of the new law, many banks proceeded to increase their capital ratio during 1996–97, and as of end-January 1998, all financial institutions met the new capital adequacy requirement. Under the new law, bank regulators will also establish a new system of bank ratings according to levels of solvency, and management quality. Banks with higher capital adequacy, solvency, and management ratings will receive fast-track authorization for their new domestic and foreign business expansions.
A consolidation trend is resulting in fewer but stronger financial institutions that will expand into nontraditional financial service areas. In just two years and a half, the 10 largest banks have gained 8 percentage points of market share and account for 85 percent of the system’s total loans.34 Two mega-mergers in 1996–97 gave birth to the two largest banks in Chile, Banco Santiago and Banco Santander Chile, that jointly account for 30 percent of the system loans and are refocusing their businesses to provide universal banking services as other institutions in the region. Competition will remain intense and some of the banks that currently have single-digit market shares are likely to merge.
The Mexican banking system remains entangled in severe asset quality problems, but the deterioration in asset quality appeared to have stabilized toward the end of 1997 after two years of robust GDP growth and significant government assistance. With the change to accounting practices patterned on U.S. GAAP at the beginning of 1997, past-due loans roughly doubled to 13 percent of total loans, but fell to 11.2 percent at the end of the year. However, the measure of past-due loans excludes the downside risk of loans sold to FOB APROA and allows for some degree of transitory regulatory forbearance, especially with respect to mortgage loans. Including the 25 percent downside risk on loans sold to FOBAPROA, the share of problem loans rises to 18 percent on average for the largest banks assuming no recovery value on these loans. Moreover, the level of loan loss provisions remains at 61.2 percent, a relatively low level given the large amount of past due loans that will have to be charged-off in the near future. Banks have recently been required to start provisioning for the expected losses in the FOBAPROA loans.
Although the Mexican banks appear to be well-capitalized, with the capital adequacy ratio for the system at 13.7 percent at end-1997,35 the proportion of equity in their capital bases is relatively low. Moreover, the high capital adequacy ratio also reflects the low risk-weight assigned to government securities (including FOBAPROA loans) and the large revaluation gains of shares in subsidiaries and affiliates. As part of their recapitalization programs, some banks issued mandatorily convertible subordinated bonds, which are counted as Tier 1 capital by Mexican regulations but would be considered as Tier 2 by the Basle Committee guidelines.
The largest Mexican banks recorded increasing profitability in 1997, albeit from the very low levels of 1995 and 1996. Many banks showed some real loan growth, especially in the last quarter of 1997, but this was barely sufficient to cover the decline in spreads. The latter reflects the heavy past-due loan portfolio and the mortgage discount programs launched by several banks in 1997, as well as competition from lower cost capital market funding. For most large banks, nonrecurring trading gains account for over one-third of total revenues, and they were quite substantial in the second and third quarters of 1997 but fell substantially in the final quarter. Most banks reported some growth in fees and commissions, in particular from credit card and fund transfer fees. Many of the medium-sized and small banks suffered losses during 1997, as they focused on loan provisioning and recoveries as well as in a rapid increase of their branch networks—to reduce their dependence on money market funding.
The consolidation of the Mexican banking system accelerated in 1996 and 1997, spurred by the entry of foreign participants and the merger of some domestic medium-sized banks. By the time recently announced mergers and acquisitions have been completed, about 75 percent of the banking system’s assets will be concentrated in just five banking groups, with foreigners owning minority stakes in three of them. Foreign participation in terms of total assets reached 19.9 percent by end-1997, including foreign banks’ affiliates and controlling stakes in domestic banks, and it is expected to grow following the approval of legislation allowing unrestricted foreign ownership in the largest banks. Moreover, many of the financial groups own insurance companies and private pension fund administrators (Afores), in most cases jointly with foreign entities. The distribution of insurance and pension products through banks branches is expected to increase the banks’ noninterest income and improve their operating efficiency indicators.
The quality of the regulation at the National Banking Commission (CNBV) has continued to improve, with new measures on disclosure and self-regulation, but rating agencies note some remaining deficiencies. The adoption of U.S. GAAP standards and measures related to the mark-to-market of securities are major improvements in the transparency of information about banks’ positions. Also, the CNBV has adopted measures to provide that information to the market and to strengthen the role of agents that validate that information (external auditors, rating agencies, and credit bureaus). The merger of the National Banking and Securities Commissions to create the CNBV in 1995 was a first step in the process of consolidated supervision, but the absence of consolidated accounting still makes it difficult to establish effective limits on group lending. However, accounting criteria for financial groups were issued by end-1997 and the controlling groups are required to publish consolidated financial statements since March 1998. The CNBV is preparing regulations aimed at creating a risk management unit inside the banks and improving corporate governance.
The Mexican authorities have submitted to Congress a set of measures to further strengthen the banking system and modernize the regulatory framework. The federal government would assume FOBAPROA’s liabilities, a move that would increase the transparency of the restructuring process and could eventually, through refinancing, improve the liquidity of the banking system.36 Under the new legislation, FOBAPROA would be divided into two agencies, one to accelerate the disposition of the bad portfolio taken over from the banks and another one to administer a new limited deposit insurance scheme. FOBAPROA has a loan portfolio of $45 billion and the difficulties of marshaling the approximately 440,000 loans in the absence of a secondary market for bank loans has led the government to explore the possibility of creating coinvestment partnerships with private asset managers who would manage, collect, and sell the loans.37 The new deposit insurance scheme would cover only small depositors, with a view to induce more market discipline and reduce moral hazard. Finally, the CNBV would get new powers and autonomy under the new legislation.
The recovery of the Venezuelan banking system continued in 1997, underpinned by strong GDP growth and high oil prices. The crisis of 1994–95 had caused a remarkable contraction of banking activities, with the ratio of loans to GDP falling to 8 percent by end-1996, one of the lowest levels among emerging markets (see Figure 2.13 in Chapter II). Higher economic growth and disposable income, combined with negative real interest rates, fueled an increase in the demand for credit that resulted in real loan growth of 61 percent during 1997. The loan to GDP ratio rose to 12.4 percent at the end of 1997, as the economy continued to remonetize with the fall in inflation and the loan portfolio reached 50 percent of total banks’ assets—from a low of 34 percent in 1996.
The banking system has emerged from the crisis with relatively good asset quality and adequate capitalization.38 Nonperforming loans, which had reached 23.4 percent of total loans at the peak of the crisis in March 1995, fell to 7.4 percent in December 1996 and to 3.9 percent in December 1997. Moreover, loan loss reserves remain adequate at 91.7 percent of nonperforming loans. The banks are well-capitalized on average, but the ratio of net worth to risk-weighted assets fell from 17.9 percent at end-1996 to 15.6 percent at end-1997. This was largely a result of the change in the structure of the banks’ balance sheets that showed an increase in loans (carrying a 100 percent weight) relative to government securities (carrying a zero weight), as demonstrated by the fact that equity to total (unweighted) assets rose from 12.9 percent at end-1996 to 13.1 percent at end-1997.
The drastic fall in oil prices since the end of 1997 and the sharp increase in interest rates required to defend the exchange rate of the bolivar, caused liquidity problems at some institutions at the beginning of 1998 and is likely to lead to a considerable deterioration in asset quality toward the second half of the year. The changes in the monetary policy instruments used by the Central Bank of Venezuela in 1998 demonstrated that the interbank market becomes segmented as liquidity dries up.39 The rapid expansion in loan portfolios during 1997 was concentrated in the consumer and mortgage sectors, with a large share of loans directed to the purchase of assets like real estate, equities, and durable goods. These sectors are quite sensitive to an asset price deflation and could be a source of deterioration in asset quality. Although the large banks appear to have developed adequate credit scoring systems and to have maintained relatively strict lending policies, most of the small and medium-sized banks have also been expanding their loan portfolios aggressively and lack the risk controls and information systems to support safe loan growth. Analysts estimate that 12 out of the 40 universal and commercial banks have capital adequacy ratios that might be too low to compensate for the increased provisions necessary to maintain an adequate coverage following the expected asset quality deterioration.
The two most important structural factors that have influenced developments since the banking crisis have been the entrance of foreign banks and the adoption of the universal bank paradigm. Starting in December 1996, foreign participation in the Venezuelan banking system changed from being almost marginal to representing 41 percent of the total system assets in December 1997 (see Figure 2.16 in Chapter II). During this period, three of the five leading banks came under the control of foreign institutions, and the trend continues in 1998. The increased foreign presence has strengthened the system and is imposing competitive pressures that will lead to further consolidation in the system. Consolidation will also be aided by the transformation of many institutions into universal banks, and 11 banks have already been approved to operate as such. Universal banks are allowed to perform all the activities of the different specialized institutions that operate in the country (commercial and investment banking, mortgages, leasing, and asset management), subject to certain operational restrictions.
Banking regulation and supervision has improved after the crisis, but market participants note that the institutional capacity and enforcement powers of the Superintendency of Banks and Financial Institutions (SBIF) need to be strengthened further. The SBIF has introduced a new chart of accounts, general and specific loan loss reserves, and minimum standards for the approval and management of consumer credit and has strengthened on-site inspections. More recently, the SBIF has moved to improve the reach of consolidated supervision through the signature of agreements with several of its regional counterparts—but it still needs to implement procedures for the consolidated supervision of financial groups and offshore operations of Venezuelan banks, and for monitoring connected lending and credit concentration.
European Banking Systems
In the Czech Republic, the four largest banks still account for over 60 percent of banking system assets, although their market share has been falling, mainly owing to competition from the 23 foreign-owned banks and branches that have increased their share to over 20 percent.40 Until recently, the large banks all remained under state control, even though three of them had been substantially privatized via voucher privatization in 1992–93. However, the government has announced its intention to sell its shareholdings to strategic investors so as to improve the financial strength and operating performance of the banks. The first such sale occurred in March 1998 when the state’s 36 percent stake in the third largest bank, Investicni a Postovni Banka (IPB), was sold to Nomura Europe. Preparations for the sale of controlling shareholdings in the other three large banks are under way, and foreign observers have noted that the nationwide presence of these banks would be attractive to foreign banks, but that uncertainty over the quality of their loan books will depress sales prices.
Preliminary data for 1997 for the banking system (excluding the Consolidation Bank and banks under Czech National Bank conservatorship) suggest that banks were relatively successful at increasing gross profits from banking activity faster than assets, and at holding growth in operating costs somewhat below asset growth. However, owing to a large increase in provisions, net profits grew by only 6 percent. For the sector as a whole, net profits fell to only 0.36 percent of assets, with the profitability of foreign-owned banks increasing and that of both the large and small domestic-owned banks weakening. The weak balance sheets have also been reflected in slow asset growth: lending fell in real terms in 1997 by around 2 percent. The risk-weighted capital ratio for the banking system as a whole was around 10½ percent at end-1997. Within this aggregate, capital ratios of the largest banks were only modestly above the National Bank’s 8 percent minimum requirement, while the capital ratio for the group of small domestic banks appears to have fallen to below 8 percent.
While the asset quality of many of the banks remains poor, this largely reflects lending decisions taken several years ago, and there are indications that credit risk management has recently improved substantially. The May 1997 currency crisis and increase in interest rates do not appear to have had substantial direct effects on bank profitability, reflecting reasonably good market risk management by Czech banks. While classified loans (all loans for which interest and principal are overdue more than 30 days) represented 27.0 percent of all loans at end-1997, this represented a fall from 29.3 percent of all loans at end-1996 and 33.0 percent at end-1995. A high proportion of these classified loans are long-standing loss loans, which tend not to be written off because of accounting and tax regulations, and because of legal hindrances to bankruptcy proceedings and sales of collateral. After the deduction of the value of collateral or guarantees on classified loans and the weighting of the uncovered component by the National Bank’s provisioning requirements (which reflect probabilities of losses), risk-weighted classified exposures were equivalent to 9.5 percent of all loans, and were fully covered by provisions. Some observers have, however, questioned the valuation of collateral by some large banks, especially with regard to real estate, and have suggested that there is substantial underprovisioning by some banks. Indeed, all of the four major banks substantially increased provisions in 1997, with some indicating the intention to also allocate the majority of 1998 operating profit to increasing provisions.
Recent changes to the Czech regulatory and supervisory framework for banks and financial markets have focused on eliminating some of the factors that made possible the various types of fraudulent behavior and insider dealing that flourished in the early years of reform. Some of the important aspects of an amendment to the Banking Act that came into force in February 1998 are (1) the introduction of limits on bank participation in nonfinancial institutions; (2) the introduction of “Chinese walls” between commercial and investment banking activities within banks; and (3) strengthened requirements for securities transactions performed by banks either on their own account or on clients’ account. A second amendment to the act is under discussion and would include, inter alia, stronger requirements on bank management with respect to their relationship with shareholders. Finally, an amendment to the Bankruptcy Law approved in January 1998 should strengthen the position of banks as creditors by (1) providing for a wider and more precise definition of insolvency; (2) requiring faster bankruptcy filing; and (3) allowing faster decisions by judges and faster liquidation of companies.
While the banking sector in Hungary—where bank lending to the nongovernment sector is about 25 percent of GDP—is substantially smaller than in the Czech Republic, the process of restructuring the banking system is further advanced as the result of greater privatization and previous government policies to clean up the balance sheets of banks and enterprises.41 This process has not, however, been costless: it is estimated that the cost of restructuring the banking sector over 1992–96 was around 10 percent of GDP, and that less than half this amount was regained in subsequent privatizations. The process of sales of state shareholdings in large banks continued in 1997, and none of the seven largest banks—all of which had their origins in the state sector—now have majority state-ownership. For the most part, privatization has been via strategic investment by foreign financial institutions, and institutions from Belgium, Germany, Ireland, Italy, the Netherlands, and the United States now hold strategic—typically majority—shareholdings in five of the seven largest banks. The two exceptions in this regard are the largest bank, the National Savings Bank (OTP), which was sold via public offerings, and Postabank, the least healthy of the major banks, which has a dispersed ownership but is now looking for a strategic investor. As a result of privatizations, capital injections by the new owners, and growth of new foreign-owned banks, it is estimated that by early 1998, over 70 percent of bank capital was foreign owned, and that foreign controlled banks accounted for well over 60 percent of the assets of the banking system, up from 46 percent at end-1996.
While the Hungarian banking system is clearly among the healthiest in Central and Eastern Europe, profitability has been only moderate and capital adequacy ratios are declining—albeit from relatively high levels. Pretax profits were at around 2 percent of average assets in 1996 and the first half of 1997, lending spreads have fallen sharply, fee income is still relatively low, and overhead costs remain fairly high. Classified credit exposures totaled 10.9 percent of total banking system credit in mid-1997, with a majority of these being in the “watch” category. Specific provisions for the banking system as a whole appear adequate, although analysts have reported that some banks appear underprovisioned. As of mid-1997, banking system capital stood at 16.9 percent of risk-weighted assets, down around 2 percentage points from end-1996 and by more from its peak at end-1995. Looking ahead, the challenge for the large Hungarian banks will be to ensure that efficiency is improved and the rapid expansion in balance sheets that is under way—loans grew around 14 percent in real terms in 1997—does not translate into a worsening in asset quality.
The supervisory and regulatory framework in Hungary is also stronger than in many transition economies. New laws on financial institutions and supervision became effective in January 1997, along with the establishment of a single supervisor for banks and other financial institutions, as part of a shift to universal banking and in preparation for eventual EU membership. Since January 1998, branches—rather than only subsidiaries—of foreign banks have been permitted, and in 1999 banks will be permitted to provide a range of brokerage and investment banking services currently provided through fully-owned subsidiaries. Finally, Hungary also has a relatively effective bankruptcy framework: indeed, the bankruptcy law put in place in 1992 was found to be too strong and was replaced by one slightly less stringent, but still substantially more effective than most other countries in Central and Eastern Europe.
In Poland, the expected liberalization of foreign bank entry from 1999—as part of Poland’s OECD membership and EU association agreement—appears to have contributed to an acceleration of the privatization and restructuring of the banking sector. In the early years of transformation, Poland was characterized by a relatively fragmented banking system, privatization had not proceeded especially far, and foreign participation in banking was low. This has changed recently, however, with the privatization of five of the regional banks over 1993–97, and of one of the four specialized banks—Bank Handlowy, the third largest bank—in 1997, via its sale to institutional and individual investors, employees, and three foreign financial institutions, which together acquired 28 percent of the bank. As a result of privatization, the share of assets of state-controlled banks had fallen to 52.4 percent at end-1997, down from 86.1 percent at end-1993. Sales of state-owned shareholdings have continued in 1998, and it is expected that the privatization of the largest bank—the Pekao group, that accounts for around 20 percent of sector assets—will begin in the second half of 1998. However, two other state-owned specialized banks—including the second largest bank, the State Savings Bank (PKO), which holds about 18 percent of total sector assets—are unlikely to be sold before 2000.
Until recently, competition in the banking sector has not been especially strong, and profitability in 1996 was reasonably good, with healthy interest margins and a return on average assets of 2–2½ percent. Preliminary data for 1997 from the National Bank of Poland suggest, however, that profitability declined in 1997, with after-tax earnings growing by only 3 percent, a substantial fall both in real terms and in relation to the strong growth in assets (28 percent in nominal terms). The major factors behind the pressures on profits appear to have been a decline in interest spreads and increases in reserve requirements, overhead costs, and specific provisions, which are likely to continue with the increased foreign bank entry.
The asset quality of Polish banks is relatively good, owing to the bank and enterprise restructuring of 1993–94 and the effective prudential framework. Based on the National Bank’s quite stringent loan classification system—loans that would be classified as “watch” loans in other countries are included in the substandard category—“irregular assets” fell during 1997 from 13.2 percent to 10.4percent of gross claims of commercial banks, with the proportion of loss loans falling from 7.4 percent to 5.3 percent.42 While part of this decline may be due to the recent strong growth in new loans, the longer-term improvement from end-1993—when irregular and loss loans were around 31 and 17 percent, respectively—is impressive. For the commercial banks as a group, irregular loans were essentially fully provisioned at end-1997. Capital adequacy for most commercial banks was also healthy, with a median risk-weighted capital ratio of around 17½ percent, and around three-fourths of all banks reporting ratios above 12 percent.
The prudential supervision framework in Poland is relatively strong and was boosted by the implementation of a new Banking Law that became effective in January 1998. Prudential regulations are generally consistent with, and sometimes exceed, the EU and Basle Committee standards. The main challenges for the General Inspectorate of Banking Supervision will be to move toward consolidated supervision, to implement capital requirements for market risk, and to increase staff resources to be able to conduct more frequent on-site inspections.
In Russia, the phenomenon of lax licensing in the early years of transition was more marked than in any other country in the region, with the number of commercial banks growing to around 2,500 at end-1994. Since 1995, however, the Central Bank of Russia (CBR) has worked aggressively to close down many of the weaker banks, and by end-1997 the number of banks had fallen by around one-third. Despite the large number of banks, the banking sector is relatively underdeveloped, with the ratio of M2 to GDP at only around 16 percent, reflecting the high inflation of earlier years and lack of trust in the ruble and the banking system. There has been some consolidation among the larger banks, and the share of the top 200 banks in total assets has grown from 81.9 percent to 88.4 percent at end-1997. The Sberbank (State Savings Bank) is by far the largest bank in Russia and accounts for around one-third of banking system assets and around 78 percent of household deposits. Other important banks include the large banks that form part of the “financial industrial groups” (FIGs) that have large holdings of some of Russia’s leading companies, obtained in some cases during the “loans for shares” transactions of 1995. As of end-1997, there were 165 banks with foreign equity, including 16 fully owned subsidiaries: the share of nonresidents in total equity was less than 6 percent, substantially below the current 12 percent ceiling on foreign participation.
Commercial banks in Russia are characterized by the extremely short-term nature of both sides of their balance sheet, widespread use of foreign currency denominated assets and liabilities, and a relatively high proportion of securities in their assets. Loans to the nongovernment sector appear, however, to have begun to grow, with a real increase in 1997 of around 14 percent. This, along with an increase in holdings of government securities, was funded through a substantial real increase in deposits and an increase in foreign borrowing which contributed to a swing of nearly $10 billion in the net foreign asset position of banks.
After strong profitability in 1996 because of high yields on government securities and large interest rate spreads, banking sector profitability fell sharply in 1997. Preliminary data from the CBR$indicates a fall of 47 percent in 1997 ruble pretax profits. The return on equity of the largest 100 banks was reported to have fallen from around 80 percent in 1996 to around 20 percent in 1997, with the return on assets falling from around 3 percent to less than 1 percent. Contributing factors include lower interest rate spreads, lower earnings on government securities, and losses on securities holdings in the market turmoil of the fourth quarter of 1997. Some losses from the latter turmoil may, however, have been carried forward into 1998, following a CBR decision to allow banks to carry forward revaluation losses on holdings of government securities. It is likely that the continuing market turmoil in 1998 has put further pressure on the profitability of some banks, and one major bank—Tokobank, about the seventeenth largest bank—was placed under CBR administration after liquidity problems in May 1998. Many of the major banks underwent ratings downgrades in early 1998, with ratings agencies citing the sovereign downgrade, and the likely effect of the market turbulence on already weak financial positions. This weakness has shown up in reduced access by banks to international markets in early 1998 after substantial growth in access in 1997.
The asset quality of Russian banks is mixed, with substantial holdings of assets with little credit risk and smaller holdings of loans with high credit risk. There is little data on aggregate nonperforming loans, but one official estimate suggested that around 22 percent of loans were nonperforming at end-1997, a modest improvement from earlier higher levels. It has long been suggested by observers, however, that loan provisioning may not be adequate. Systemwide capital adequacy data are not available, but many banks reported risk-weighted ratios of over 20 percent at end-1996. The high capital ratios relative to the CBR minimum requirement of 7 percent reflect the relatively small holdings of assets with high risk-weights, and suggest that many banks may be well protected against credit risk. The capital adequacy regulations are not, however, designed to protect against market risk, and it remains to be seen how banks have been affected by the recent large fluctuations in the value of their portfolios of government securities.
The prudential supervision framework in Russia was substantially weaker in the early years of transition than in many of the other countries in the region. However, there was a substantial improvement in January 1996 with the introduction of Instruction No. 1 of the CBR. This instruction sets requirements, which have been gradually improved and tightened, for capital adequacy, bank liquidity, large exposures, connected lending, shareholdings in other companies, and issuance of bank bills. During 1997, the CBR tightened rules on the calculation of bank capital, and on loan classification and provisioning requirements, and issued instructions to improve the internal control procedures of banks. It has also established limits on open foreign currency positions which must be met and reported on a daily basis, and has introduced limits on overall foreign borrowing. From January 1998, the CBR has required banks to use a new chart of accounts and to follow accounting standards that are closer to international standards. Prudential supervision is still mainly done by off-site supervision, but the CBR is now conducting on-site supervision of the largest banks. It also monitors the health of 14 large banks—which account for about two-thirds of all assets—on an ongoing basis in a special department. A law on bank bankruptcy is under discussion and would allow the CBR to act more swiftly in closing banks and removing management. Market observers have recognized the major steps taken over the last two years in improving the prudential framework, but have highlighted important further priorities which include the full implementation of international accounting standards—including consolidated accounting, the clarification of links with affiliated companies (especially in the case of the financial industrial groups), and the incorporation of derivatives and off-balance sheet activities into the prudential framework.
1997, “Currency Board Arrangements: Issues and Experiences,” IMF Occasional Paper No. 151 (Washington: International Monetary Fund).
International Monetary Fund, 1997, International Capital Markets: Developments, Prospects, and Key Policy Issues, World Economic and Financial Surveys (Washington).
International Monetary Fund, 1998, Czech Republic: Selected Issues, IMF Staff Country Report 98/36 (Washington).
1996, “Monetary and Exchange System Reforms in China: An Experiment in Gradualism,” IMF Occasional Paper No. 141 (Washington: International Monetary Fund).
National Bank of Poland, 1998, Summary Evaluation of the Financial Situation of Polish Banks (Warsaw).
The surges of capital into emerging markets, the sharp declines in equity prices in times of crises, and—in some cases1—the subsequent robust recoveries of these markets raise questions about the rationality of asset market dynamics. Insights into these dynamics can perhaps best be gleaned from the extensive body of research into the U.S. equity market, where high-quality data for thousands of stocks are available over a 70-year period. Using these data, financial economists have studied the dynamics of equity prices by examining the behavior of (notional) portfolios of stocks that are formed based on their prior return performance. In many studies, stocks are sorted into deciles, with those stocks previously showing the largest gains referred to as the “winner” portfolio and those showing the largest falls referred to as the “loser” portfolio. The performance of these portfolios is then simulated in the subsequent “test period,” to see if there are any consistent differences in the subsequent returns on the different portfolios.
A standard finding (for example, Jegadeesh and Titman, 1993) is that U.S. stocks that have been winners and losers over prior periods of one month to one year show “momentum” (or positive autocorrelation) in their relative performance over the corresponding subsequent test period: that is, winners continue to yield above-average returns and losers continue to yield below-average returns at these horizons.2 Other studies have shown similar momentum effects in the performance of stocks in European countries (Rouwenhorst, 1998) and of national markets within the global market (Asness, Liew, and Stevens, 1997). The magnitude of these effects is often quite large and persistent, with return differentials between winners and losers often around 1 percent a month over periods of 6 to 12 months.
At longer horizons, however, U.S. stocks appear to demonstrate reversals (or negative autocorrelation) in their relative performance. A number of researchers, beginning with DeBondt and Thaler (1985), have demonstrated that stocks that have been losers over a period of two to five years go on to subsequently yield higher rates of return than the corresponding prior winner stocks, with return differentials of up to around 9 percent a year. Other researchers (for example, Lakonishok, Shleifer, and Vishny, 1994) have found similar return differentials between “glamor” and “value” stocks, that is, stocks with respectively high or low prices relative to their fundamentals (earnings, cash flows, and so on) and that are likely to have been winners or losers in the recent past. Furthermore, winner-loser reversals of up to about 6 percent a year have also been found in the performance of the national market indices of various mature stock markets (Richards, 1997), and there is also some evidence for reversals in emerging markets (Richards, 1996).
Since predictable patterns in asset prices are suggestive of market inefficiency, the phenomena of short-term momentum and longer-term reversals have generated substantial debate. Researchers investigating the momentum phenomenon (for example, Jegadeesh and Titman, 1993) have, however, found it difficult to explain using conventional asset pricing models. Accordingly, some financial economists (such as Fama, 1998) consider momentum to be evidence of temporary underreaction to news, just as the phenomenon of price drift after earnings announcements is generally considered evidence of slow reaction in stock prices. However, other economists (for example, Lakonishok, Shleifer, and Vishny, 1994) argue that momentum is the result of investors’ overreaction to current trends.
The voluminous literature on price reversals has noted that reversals could be due to risk factors, since changes in required rates of return have immediate effects on asset prices in one direction and an offsetting influence in subsequent periods: for example, losers could be stocks that have fallen sharply in price because they have become riskier, with their subsequent higher returns simply reflecting their now-higher risk. Indeed, some researchers have suggested that reversals can be fully explained by risk differentials and by the disproportionate effect of small or low-priced stocks (for example, Ball, Kothari, and Shanken, 1995). However, other researchers of the related glamor-versus-value effect have demonstrated that value stocks, which go on to outperform glamor stocks, are not riskier based on conventional notions of risk (Lakonishok, Shleifer, and Vishny, 1994). Instead, they suggest that reversals are related to momentum effects, with the pattern of autocorrelations due to irrational fads or investor misperceptions that systematically take prices away from fundamental values, requiring an eventual correction that is reflected in negative autocorrelation at longer horizons. As evidence for this, Lakonishok, Shleifer, and Vishny (1994) have shown that ex post differences in the growth rates of fundamentals of glamor and value stocks turn out to be far smaller than the differences that must have been (irrationally) expected based on the initial difference in valuations.
While proponents of market efficiency may disagree, many financial economists would now argue that the phenomenon of short-term momentum and long-term reversals is both pervasive3 and the result of behavior by market participants that is not fully compatible with full market efficiency. There is now a long tradition of arguments (for instance, Graham (1959), Shiller (1981), Arrow (1982), DeBondt and Thaler (1985), and Lakonishok, Shleifer, and Vishny (1994)) that stock prices do not merely reflect rationally discounted expected cash flows but often also reflect irrational investor sentiment or systematic errors in expectations formation. In addition, there are a number of possible explanations or rationalizations for such behavior by investors. First, the literature on individual decision making (Kahneman and Tversky, 1982) suggests that individuals may systematically give weight to recent information in forming judgments, which could lead investors to amplify price movements resulting from recent news. Second, the behavioral literature also suggests the possibility of judgment errors of the type that investors might equate good companies—or those that have recently performed well—with good investments, regardless of price (Shefrin and Statman, 1995). Third, there may exist a class of traders—“noise traders”—who are able to move prices away from fundamental values without necessarily inviting arbitrage activity that would cause them to lose money (DeLong and others, 1990). Indeed, theoreticians are currently working to build some of these ideas into formal models that can yield the observed pattern of autocorrelations in returns with as few deviations as possible from the standard assumptions about rational agents in fully efficient markets.4
1997, “Parallels Between the Cross-Sectional Predictability of Stock Returns and Country Returns,”Journal of Portfolio Management, Vol. 23 (Spring), pp. 79–87.
1995, “Problems in Measuring Portfolio Performance: An Application to Contrarian Investment Strategies,”Journal of Financial Economics, Vol. 38 (May), pp. 79–107.
“Investor Psychology and Security Market Under- and Over-Reactions,” Journal of Finance (forthcoming).
“Market Efficiency, Long-Term Returns, and Behavioral Finance,” Journal of Financial Economics (forthcoming).
1959, The Intelligent Investor: A Book of Practical Counsel (New York: Harper and Brothers, 3rd ed.).
1993, “Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency,”Journal of Finance, Vol. 48 (March), pp. 65–91.
1982, “Intuitive Prediction: Biases and Corrective Procedures,” in Judgment Under Uncertainty: Heuristics and Biases (Cambridge, England; New York: Cambridge University Press).
1994, “Contrarian Investment, Extrapolation, and Risk,”Journal of Finance, Vol. 49 (December), pp. 1541–78.
1996, “Volatility and Predictability in National Stock Markets: How Do Emerging Markets Differ?”Staff Papers, International Monetary Fund, Vol. 43 (September), pp. 461–501.
1997, “Winner-Loser Reversals in National Stock Market Indices: Can They Be Explained?”Journal of Finance, Vol. 52, pp. 2129–44.
1995, “Making Sense of Beta, Size and Book-to-Market,”Journal of Portfolio Management, Vol. 21 (Winter), pp. 26–34.
Given the global integration of financial markets over the last decade, large capital flow reversals can occur quite quickly, considerably shortening the time in which appropriate policy responses have to be made. As a consequence, and in no small part stimulated by the recent crises in Europe, Mexico, and Asia, researchers are taking a fresh look at the determinants of currency and banking crises and attempting to develop early warning signals of brewing trouble in currency markets and banking systems.1 The lack of transparency in the operation of financial systems, especially in emerging markets, considerably complicates such a task.
To start, researchers have to identify situations that can be termed full-fledged currency and/or banking crises. Defining currency crises as instances when a “large” currency depreciation takes place excludes situations where a currency was under substantial pressure but the authorities managed a successful defense by, among other measures, raising interest rates and/or intervening in the foreign exchange market. As a result, most researchers define currency crises by using indices that weight changes in the exchange rate, foreign exchange reserves and (if available) short-term interest rates—the construction of these indices as well as the thresholds used for identifying crises differ across researchers. Even this definition may not completely capture crisis situations because in several instances, the authorities have responded to exchange market pressures by introducing capital controls.
Stresses in the banking system are even more difficult to quantify. The data necessary for making an assessment are generally not available and, as a result, dating of banking crises must rely on events such as the closure of banks and official support for (and/or government takeover of) financial institutions. Generally, banking sector weaknesses emerge because of deterioration in asset quality. Reliable and timely data on nonperforming assets is not always available and even indirect evaluations of asset quality require information on bankruptcies, exposures of financial intermediaries to different sectors, and movements in real estate and other asset prices—information that is generally not available in many developing and transition economies.
After dating crisis periods, two types of empirical methodologies have been used in the search for leading indicators of currency crises. Many researchers have identified leading indicators by comparing the behavior of a variable prior to crises with its behavior in tranquil periods.2 A variable is a useful leading indicator if it displays anomalous behavior prior to crises while not providing false signals of an impending crisis in normal or tranquil times. What is construed as anomalous behavior for a particular variable is defined by choosing a selection rule that achieves a balance between decreasing the probability of not predicting crises and decreasing the probability of giving false signals of stress. The advantage of such “univariate” event analyses is that they are easy to implement and do not impose much a priori structure on the data. However, when multiple indicators are available one has to address the question of combining them for predicting the possibility of a crisis. Efforts to do this are still at a preliminary stage (see Kaminsky, 1998).
A second approach has been to directly estimate the probability of a currency or banking crisis (using limited-dependent variable models) and identify the variables that statistically aid in predicting crises (see Frankel and Rose, 1996). This approach has the advantage that indicators are evaluated simultaneously and the statistically significant ones can then be used to calculate the probability of a crisis occurring in a specific period. It should be noted, however, that this methodology has been used with annual data and further refinement of leading indicators would require a large number of observations on the “rare” events categorized as crises. Mere use of, say, quarterly or monthly data is not enough. While such disaggregation potentially allows for greater refinement of the dynamics leading up to crises, the complexity of estimation requires more information on a larger number of the key informative events, the crises. For currency and banking crises, such large data sets are, typically, not available.
Differing methodologies, different time periods and selection of countries, and the diversity in defining what constitutes exchange market pressure make it difficult to compare results across the various studies and come up with a clear-cut answer to the question: What set of leading indicators of currency and banking crises are likely to prove most useful? That said, some tentative conclusions about indicators of vulnerability can be drawn. Currency crises tend to be preceded by an overvaluation of the real exchange rate, rapid domestic credit growth, expansion of credit to the public sector, a rise in the ratio of broad money to foreign exchange reserves, an increase in the domestic inflation rate, a decline of FDI flows, and an increase in industrial country interest rates. Other factors that receive some, though less, support as leading indicators of currency crises are a widening of the trade deficit, an increase in the fiscal deficit, a deterioration in export performance, and a slowdown in real GDP growth. It is noteworthy that current account and fiscal deficits do not seem to garner a lot of support as important indicators.3 With regard to banking crises, these are often preceded by large inflows of short-term capital, rapid expansion of domestic credit (frequently a consequence of financial liberalization coupled with inadequate supervision by bank managements as well as regulators), slackening of real activity, and declines in the stock market and prices of other assets. Case studies suggest that, in many instances, liberalization without adequate strengthening of the regulatory regime not only sets the stage for a banking crisis but also makes it more difficult to cope with a crisis if it erupts.
How well do current models perform in out-of-sample crisis prediction? Recent events raise the following question: Using data until end-1996, would these models have alerted policymakers to the possibility of the kind of turmoil that has been witnessed in Asia? Berg and Patillo (1998) and International Monetary Fund (1998) attempt an answer to this question by comparing the out-of-sample performance of different approaches in predicting the Asian currency crises. They conclude that, while the forecasts are informative, these models do not as yet provide much improvement over informed guesses. Using aggregate (publicly available) data to predict banking crises, Demirgüç-Kunt and Detragiache (forthcoming) meet with similar success. In this context, two points should be noted: First, leading indicator models are still in their infancy and more rigorous data reporting requirements for financial and nonfinancial institutions that are just beginning to be introduced may enhance the usefulness of such models. Second, the entire sovereign credit rating industry did not foresee the vulnerable situation of many Asian economies and was taken by surprise when the crisis broke.
The timing of events in the economic arena are notoriously difficult to analyze. Economic theory, while relatively good at characterizing equilibrium situations, tends to be less informative about the dynamics that could lead from one equilibrium to another. To predict the timing of rare events such as financial crises, which may critically depend on factors that are hard to capture such as structural features of the economy, institutional developments, changes in the political landscape, and expectations of domestic and foreign players in various markets, is likely to be even more demanding. More important, the process of policymaking and the policy responses themselves have a crucial bearing on whether situations of stress degenerate into crises. And, typically, these cannot be taken into account in modeling exercises. Hence, it is not surprising that models based on quantifiable factors that do not endogenize policy responses have not met with much success.
The paucity of data on crisis episodes is a major hurdle in the further refinement of current models that examine such events and attempt to identify leading indicators. For example, researchers are forced to assume that the parameters characterizing the behavior of certain variables in the buildup to crisis situations (and their aftermath) are similar across time and across countries. Given the diversity in institutional arrangements, the dramatic changes that have taken place in industrial and developing country financial systems in the last decade, and the increased integration of global markets, such assumptions may well be untenable. Thresholds defining what are acceptable levels for certain variables are likely to differ across countries and could well change over time for the same country. Lack of adequate data makes it difficult, if not impossible, to test such assumptions.
In the end, the holy grail of crisis prediction may be intrinsically unattainable. Indeed, the very success of such models in predicting crises would eliminate the phenomenon they were trying to predict if policymakers took appropriate action in response to early warning indicators. Further, since foreknowledge of crises would typically allow trading profits to be made, the existence of a successful prediction model is unlikely in efficient markets.
Crises that result because weak fundamentals make a country vulnerable to adverse shocks may be predictable. Crises that arise because of a unique concatenation of events, or from pure contagion effects, or because technology, new instruments, and new ways of doing business transform the financial system in unforeseen ways, or because some widely held belief is falsified by events are less likely to be foreseen by economic models. The Latin American debt crisis of the 1980s shattered the then prevailing myth that sovereign states “could not default.” The 1992 ERM crisis showed that countries, even industrial ones, with high unemployment may find it preferable to exit a fixed exchange rate system than live with the consequences of higher interest rates for a brief period. The 1994 Mexican crisis taught us about vulnerabilities associated with short-term sovereign foreign currency debt and a weak banking system. The Asian crisis, though inextricably linked to domestic macroeconomic and financial developments, has put the spotlight on structural features of financial systems more broadly and has revealed that debt exposures and currency imbalances of private corporations and financial institutions can be as lethal as those of the public sector. Hence, what is needed is not only a better understanding of the run-up to crises past but also a better grasp of what factors could precipitate crises in the faster paced and evolving new international financial environment.
“Currency Crises: In Search of Common Features,”IMF Working Paper (Washington: International Monetary Fund, forthcoming).
“Are Currency Crises Predictable? A Test,”IMF Working Paper (Washington: International Monetary Fund, forthcoming).
1997, “Banking Insolvency: Bad Luck, Bad Policy or Bad Banking,” Annual World Bank Conference on Development Economics (Washington: World Bank).
1998, “The Determinants of Banking Crises in Developing and Developed Countries,”Staff Papers, International Monetary Fund, Vol. 45 (January), No. 1, pp. 81–109.
“Monitoring Banking Sector Fragility: A Multivariate Logit Approach with an Application to the 1996–97 Banking Crises,”IMF Working Paper (Washington: International Monetary Fund, forthcoming).
1995, “Exchange Market Mayhem: The Antecedents and Aftermath of Speculative Attacks,”Economic Policy (October), pp. 249–312.
1998, “Staying Afloat When the Wind Shifts: External Factors and Emerging-Market Banking Crises,”NBER Working Paper No. 6370 (Cambridge, Massachusetts: National Bureau of Economic Research).
1996, “Currency Crashes in Emerging Markets: An Empirical Treatment,”Journal of International Economics, Vol. 41 (November), pp. 351–66.
1996, “The Roots of Banking Crises: The Macroeconomic Context,” in Banking Crises in Latin America, ed. by R.Hausmann and L.Rojas-Suárez (ed) (Washington: Inter-American Development Bank).
1996, “Presumptive Indicators/Early Warning Signals of Vulnerability” (unpublished; Washington: Institute for International Economics).
1996, “Banking Crises in Emerging Economies: Origins and Policy Options,” BIS Economic Papers No. 46 (Basle: Bank for International Settlements).
1997, “Determinants of Banking System Fragility: A Case Study of Mexico,”Staff Papers, International Monetary Fund, Vol. 44 (September), pp. 295–314.
International Monetary Fund, 1997, International Capital Markets: Developments, Prospects, and Key Policy Issues, World Economic and Financial Surveys (Washington).
International Monetary Fund, 1998, World Economic Outlook, May 1998: A Survey by the Staff of the International Monetary Fund, World Economic and Financial Surveys (Washington).
1998, “Currency and Banking Crises: A Composite Leading Indicator,” (unpublished; Washington: Board of Governors of the Federal Reserve System).
1996, “The Twin Crises: The Causes of Banking and Balance-of-Payments Problems,” Discussion Paper No. 544 (Washington: Federal Reserve Board).
1998, “Leading Indicators of Currency Crises,”Staff Papers, International Monetary Fund, Vol. 45 (March), pp. 1–48.
“Current Account Reversals and Currency Crises,” IMF Working Paper (Washington: International Monetary Fund, forthcoming).
The recent developments in Asia have renewed the debate on the usefulness and effectiveness of capital controls. A number of economists have advocated “throwing sand in the wheels” of international markets to inhibit volatile, short-term capital flows. This view stems from the belief that massive inflows of foreign capital, mainly in the form of short-term portfolio investment and bank loans, have proved to be too unstable and unpredictable to be a major source of external financing for emerging markets. Thus, in contrast to earlier periods where the focus was on using capital controls to limit capital flight, the current proposals have examined the use of controls to alter the volume and composition of capital inflows. In this context, Chile’s capital controls have been viewed by some observers as the type of instrument that can be used to manage short-term inflows.
In the early 1990s, Chile experienced a surge in capital inflows that created a conflict between the authorities’ internal and external objectives: the problem was how to maintain a tight monetary policy without hindering Chilean export competitiveness. In 1991, the central bank attempted to resolve this dilemma by imposing a one-year un remunerated reserve requirement (URR) on foreign loans, which was primarily designed to discourage short-term borrowing without affecting long-term foreign investments. The fixed holding period of the reserve requirement implied that the financial burden diminished with the maturity of the investment. Between 1991 and 1997 the rate of the URR was increased and its coverage extended in several steps to cover most forms of foreign financing except foreign direct investment. Currently, there is a one-year minimum holding period on capital inflows (applying to all inflows above US$10,000 except for short-term borrowing and holdings of American Depository Receipts (ADRs)). Bonds issued abroad by Chilean companies must have an average maturity of at least four years. In addition, there is a 10 percent unremunerated reserve requirement, also with a one-year holding period, for all external liabilities that do not result in an increase in the stock of capital.1 In practice, this means that loans, fixed-income securities, and most equity investments are subject to the URR, and only FDI and primary issuances of ADRs are exempted from the reserve requirement. However, primary issues of ADRs are also subject to two minimum rating requirements (BB), granted by internationally recognized credit rating agencies.
The Chilean experience has been viewed by many as a means of controlling the composition of foreign borrowing without hindering the volume of capital inflows to the country. However, the empirical evidence regarding the effectiveness of the Chilean controls in reducing short-term external debt is somewhat ambiguous. For example, while it is difficult to be conclusive in the absence of a counterfactual, national data for Chile’s external debt suggest that the introduction of capital controls affected the maturity composition of net capital inflows only after 1995 when the controls were strengthened (Figure A4.1). However, data from the BIS describe a somewhat different picture. Table A4.1 reports the claims of all banks with head offices in the BIS reporting area for 1997 for 24 emerging markets.2 The BIS figures for short-term external borrowing substantially exceed those reported in Chilean sources and suggest the existence of a large amount of foreign currency loans issued by Chilean affiliates of foreign banks, outstanding import credits (both types of loans are not included in official short-term external debt data), or significant misreporting of external liabilities. Moreover, the maturity structure of foreign bank borrowing appears quite different from what is implied by the national data. At the end of 1997, loans with maturity up to one year represented 49 percent of total foreign currency loans, whereas the Chilean debt data for the equivalent component of the total external debt is 11 percent.3
Figure A4.1.Chile’s Short-Term External Debt
Source: International Monetary Fund, World Economic Outlook.
|Hong Kong SAR||211,968||167,954||79.2|
|Taiwan Province of China||26,173||21,402||81.8|
A Brief Survey of the Literature on the Effectiveness of the Chilean Controls
There is a large empirical literature addressing the issue of the effectiveness of capital controls. An extensive review, Dooley (1996), observed that the general conclusion is that capital controls allow countries to temporarily maintain a wedge between domestic and international yields. However, it appears that this effect is eroded over time as the private sector develops new techniques to avoid the restrictions.4
Empirical studies typically test for the effectiveness of capital controls by examining the evolution of offshore/onshore interest rate differentials and whether covered interest rate parity is violated. Since Chile did not have either a well-developed forward exchange market, or an offshore deposit market for Chilean pesos, empirical work on Chile has relied on alternative procedures to test effectiveness. As will be discussed, the empirical evidence regarding the ability of the Chilean controls to drive a sustained “wedge” between domestic and external monetary conditions is mixed. However, there is another dimension to the effectiveness of Chilean-type capital controls, namely their ability to limit the accumulation of short-term external debt by financial and nonfinancial entities. Unfortunately, there is little empirical evidence regarding whether Chilean controls have been effective in this dimension.
Quirk and Evans (1995) observe that net short-term private capital inflows recorded in the balance of payments decreased in 1991 with the introduction of capital controls. However, they also observe that “net errors and omissions” and the estimated trade misinvoicing also increased sharply in the same year. One possible interpretation of that evidence is that the change in “errors and omissions” represents an increase in unrecorded short-term flows reflecting an attempt by the private sector to circumvent the capital restrictions.
Using the tax revenues generated by the URR as a proxy for the effectiveness of the restrictions (and including errors and omissions from the balance of payments in their definition of short-term capital inflows) Valdés-Prieto and Soto (1997) find that the URR did not have a significant effect on short-term borrowing before 1995, when the implicit tax increased from 3.6 percent to 6.7 percent, after the Central Bank changed the regulations and required investors to hold their reserves in U.S. dollars.5 However, the paper does provide evidence that the URR was effective in the period 1995–96.6 Nevertheless, the authors present data (but not formal tests) suggesting that other forms of short-term borrowing increased over that period as the private sector substituted exempt short-term flows—not always classified as short-term credit in the Chilean statistics—for taxed short-term flows, as the authorities gradually changed the tax design over time in an attempt to counteract new methods of evasion.7
Soto (1997) runs a VAR analysis on capital flows, interest rates, and the level and volatility of the real exchange rate for Chile. The VAR approach allows controls for the endogeneity problem that may invalidate the estimates in studies like Valdés-Prieto and Soto (1997). He finds that capital controls had the desired effect of reducing capital inflows, maintaining higher interest rates and lower real exchange rate, and reducing the share of short-term capital inflows. However, the size of these effects is estimated to be very small.
Edwards (1998a) tests Chile’s capital controls effectiveness indirectly. To make up for the lack of offshore interest rate and forward exchange rate data, he focuses on how the evolution of the real exchange rate and interest rate differential was influenced by capital controls. This hypothesis was that under the assumption of “effectiveness,” the introduction of capital controls will significantly affect the relationship between domestic and foreign interest rates and the time-series characteristics of the real exchange rate. His empirical results suggest that the impact of capital restrictions on the behavior of the real exchange rate has been very limited and short-lived. The paper also provides some evidence suggesting that the degree of persistence of interest rate differentials increased somewhat after the introduction of capital controls.
Cardoso and Laurens (1998) find that the introduction of capital controls had only temporary effects on the composition of external financing, which is consistent with the view that the private sector will attempt to circumvent any restriction to capital movements, and that over time it will succeed. They regress a direct measure of net private capital inflows on an index of capital account restrictions and a vector of control variables, including real interest rate differentials, domestic GDP, and seasonal dummies. Their results suggest that capital controls were effective in the six months following their introduction, but ceased to be effective afterward. However, their analysis does not control for the possibility of simultaneous causal ity in the regression. In particular, it seems difficult to establish the direction of causality between interest rate differential and capital inflows.
Selective Capital Controls and Prudential Regulation
To the extent that most international flows in emerging markets are intermediated by the banking system, prudential measures that prevent some particular cross-border activities and regulate banks’ open foreign positions can have effects similar to restrictions to capital movements (International Monetary Fund, 1995).
In Chile, strong and well-designed prudential regulations complement capital account restrictions in protecting the financial system from capital flow swings (Table A4.2). Banks cannot lend domestically in foreign exchange, with the exception of foreign-trade-related credits. Moreover, there is a limit on the open foreign exchange position set at 20 percent of banks’ capital and reserves, and there are other limitations on maturity mismatches (Eyzaguirre and Lefort, 1998). Some observers have argued that the combination of these prudential measures and capital account restrictions has accomplished two objectives. First, it has limited the foreign exchange exposure of both bank and nonbank entities. Second, in the event of sudden capital outflows, the limitations on maturity mismatches would allow the central bank to defend the exchange rate parity by raising the interest rate, without damaging banks’ profitability.
|Capital Transactions||Provisions Specific to Credit Institutions|
American Depository Receipts (ADRs) and other securities investments
|One year minimum holding period. 10% unremunerated reserve requirement (URR). The issuance of primary ADRs is exempt from the URR, but issuers are subject to a minimum international credit rating.||Open foreign exchange position||Limit of 20% of capital and reserves. Foreign financial investment ceiling of 25% of capital and reserves.|
|Foreign direct in vestment||One year minimum holding period.|
Minimum amount of US$10,000.
Exempted from the URR.
|Lending domestically in foreign exchange||Only foreign-trade-related loans are allowed.|
|Import-export credit||Exempted from the URR.||Other restrictions on lending||Limits on maturity mismatches.|
Banks are not allowed to keep shares in their portfolios.
|All other inflows||10% URR.|
The importance of the Chilean prudential framework is stressed in Zahler (1993), who argues that financial liberalization and capital account opening should be accompanied and preceded by a comprehensive reform of supervisory regulations. He suggests that the lack of adequate banking legislation is the major cause of the failure of most Latin American liberalization processes. Implicit insurance of banks’ liabilities and the absence of supervision led the banking system to excessive risk taking that later resulted in a general crisis. Along the same lines, Edwards (1998b) suggests that Chile owes its stability not to capital controls but to banking regulations. He argues that during the 1970s and early 1980s Chile also imposed an URR on capital inflows that did not prevent a major banking and currency crisis in 1982. He stresses that the major difference with the current situation is the existence of sound banking regulations that were substantially improved with a major reform in 1986.
Bank for International Settlements, 1998, The Maturity, Sectoral and Nationality Distribution of International Bank Lending (Basle, May).
1998, “The Effectiveness of Capital Controls on Inflows: Lesson from the Experience of Chile” (unpublished; Washington: International Monetary Fund).
1996, “A Survey of the Literature on Controls over International Capital Transactions,”Staff Papers, International Monetary Fund, Vol. 43 (December), pp. 639–87.
1998a, “Capital Flows, Real Exchange Rates, and Capital Controls: Some Latin American Experiences,” paper presented at the National Bureau of Economic Research Conference on Capital Flows to Emerging Markets, Cambridge, Massachusetts (February).
1998b, “The Americas: Capital Controls Are Not the Reason for Chile’s Success,”Wall Street Journal (April3), p. A19.
1992, “Structural Determinants and Economic Effects of Capital Controls in OECD Countries,” in Financial Openness and National Autonomy, ed. by T.Banuri and J.Schor (ed) (Oxford: Clarendon Press).
1998, “Capital Markets in Chile, 1985–1997: A Case of Successful International Financial Integration” (unpublished; Washington: International Monetary Fund).
International Monetary Fund, 1995, International Capital Markets: Developments, Prospects, and Key Policy Issues (Washington).
International Monetary Fund, 1997, Annual Report on Exchange Arrangements and Exchange Restrictions (Washington).
1995, “International Capital Mobility in the 1990s,” in Understanding Interdependence: The Macroeconomics of the Open Economy, ed. by P.Kenen (ed) (Princeton: Princeton University Press).
1995, “Capital Account Convertibility: Review of Experience and Implications for IMF Policies,” IMF Occasional Paper No. 131 (Washington: International Monetary Fund).
1997, “Controles a los Movimientos de Capital: Evaluación Empiríca del Caso Chileno” (unpublished; Santiago: Banco Central de Chile).
1997, “The Effectiveness of Capital Controls: Theory and Evidence from Chile” (unpublished; Universidad Católica de Chile).
The structural changes that have occurred in national and international finance during the past two decades can be seen as part of a complex process best described as the globalization of finance and financial risk. The key elements of this ongoing transformation have been (1) an increase in the technical capabilities for engaging in precision finance, that is, for unbundling, repackaging, pricing, and redistributing financial risks; (2) the integration of national financial markets, investor bases, and borrowers into a global financial market place; (3) the blurring of distinctions between financial institutions and the activities and markets they engage in; and (4) the emergence of the global bank and the international financial conglomerate, each providing a mix of financial products and services in a broad range of markets and countries. These changes have altered investor and borrower perceptions of financial risks and rewards around the world, and their behavior across national and international financial markets.
This annex documents the broad areas of structural change that have occurred in the past decade or more. The first subsection examines the consolidation and restructuring that has occurred in the international financial services industry comprised of banks, investment banks, institutional investors, and insurance companies. The second subsection describes the increased integration of capital markets, including the greater linkages between trading exchanges and national markets. The final subsection describes the impact of information technology and mathematical models on finance, and their ability to unbundle, repackage, price, and trade precisely defined elements of financial risk, and some of its implications for risk management.
Consolidation and Restructuring of the Global Financial Services Industry
The global financial services industry has been transformed during the past two decades, and aspects of this transformation appear to have accelerated in the 1990s. Two basic characteristics have defined this transformation. First, traditional banking institutions have been transformed into new financial services firms taking on new business lines and new risks—including those of institutional securities firms, insurance companies, and asset managers. Second, non-bank financial institutions—such as mutual funds, investment banks, pension funds, and insurance companies—now actively compete with banks both on the asset and liability sides of banks’ balance sheets. In effect, the financial services industry has become desegmented, which is increasingly blurring the distinction between banks and nonbanks.
The Changing Business of Banking
The motives for expanding beyond traditional banking have been twofold and have operated both domestically and internationally. First, the lowering and removal of regulatory barriers has meant that banks could enter businesses that had been off limits, and this has allowed them to diversify their revenue sources by taking on related activities in different markets. Second, bank disintermediation and the further development and deepening of capital markets worldwide has allowed corporations to raise funds directly through bond and equity issues. As a result, the traditional source of bank profits—lending to small and large firms financed by low-cost deposits—has suffered due to competition from securities markets and institutional asset managers. These competitive pressures on traditional bank franchise values have forced banks to seek more profitable sources of revenues, including new ways of intermediating funds. The U.S. bank data provide a good illustration of the impact of these pressures: between 1980 and 1995, U.S. banks’ share of personal financial assets fell by 50 percent to 18 percent, and nonbank financial institutions’ (pension funds, insurance companies, and mutual funds) share rose by almost the equivalent amount to 42 percent. Of course, these trends are less evident in some countries where capital market deepening has not yet occurred to the degree that it has in the United States, but even in these countries banks are aware that their traditional franchises are becoming more difficult to maintain.
The degree of disintermediation has not been shared equally by all banks and in all countries; banks that are active in smaller markets have experienced less competitive pressure, and these pressures have been slower to take hold in countries that have historically relied more heavily on banks than on securities markets. Nonetheless, aggregate national data show four main trends. First, deposits as a share of total bank liabilities have declined since 1980 in all of the Group of Seven countries except Japan and the United Kingdom, and there is an indication that this trend has accelerated in the 1990s (Table A5.1). This trend is particularly pronounced in the United States, where deposits as a share of total bank liabilities declined by 10 percentage points during the first half of the 1990s. Second, tradable liabilities of banks as a percentage of total liabilities have increased (Table A5.2). In other words, banks are increasingly funding their activities by issuing securities. Third, loans as a percentage of bank assets have generally declined since 1980 (Table A5.3). Fourth, bank assets have shifted toward investments in securities (Table A5.4).
Data refer to commercial banks; private checkable, time, and savings deposits divided by total liabilities.
Deposits from the domestic sector as a percentage of total liabilities.
Total Canadian dollar deposits as a percentage of total Canadian dollar liabilities.
A11 insured commercial banks; subordinated debt divided by total liabilities.
Issued bonds and negotiable debt securities.
Negotiable liabilities are defined as certificates of deposit and bonds outstanding.
U.S. commercial banks’ total loans divided by total financial assets.
Banks’ loans to the domestic sector as a proportion of total assets, exclusive of bad loans.
Nonmortgage and mortgage loans as a percentage of Canadian dollar liabilities.
All insured commercial banks; total securities, trading securities, and repurchase agreements divided by total assets.
For 1990, securities operations, treasury bonds, and negotiable debt securities; for 1995, trading securities, securities held for sale, and investment securities.
The changing business of banking is most evident in banks with an international focus. As a proxy, consider the largest 50 banks in the world (see Table A5.5).1 During the 1990s, three changes in the composition of these banks’ balance sheets are noteworthy. First, there has been a clear displacement of lending by other activities: the proportion of “other earning assets” relative to total assets has increased noticeably in recent years from 33 percent to 37 percent. Moreover, excluding Japanese banks, where the trend is reversed, this ratio has risen even further (to 39 percent).
|Other earning assets/total assets||33.35||31.85||33.78||36.40||37.24||37.14|
|Off-balance-sheet items/total assets||14.58||17.02||20.81||19.50||19.67||20.33|
|Other operating income/net interest revenue||49.18||54.10||62.00||61.77||56.16||67.06|
|Commission and fees/other operating income||86.56||80.37||61.09||61.28||60.87||57.54|
|Number of reporting banks||25||33||34||30||30||36|
|Trading income/other operating income||24.80||18.50||22.06||12.92||14.69||18.15|
|Number of reporting banks||5||10||10||11||11||11|
Average across all banks.
Second, off-balance-sheet items have grown relative to total assets: between 1991 and 1996, the average ratio of off-balance-sheet items to total assets has increased by almost 6 percentage points, and in 1996 stood at more than 20 percent. These data mask variation in the importance of off-balance-sheet activities across banks: one bank registered as much as 112 percent of off-balance-sheet items relative to total assets and another as little as 0.03 percent. Often little indication of the types of instruments comprising off-balance-sheet items is provided in financial reports, but an important component of off-balance-sheet activities among the major international banks has been derivatives instruments.
Third, as banks have shifted from lending to other activities, the income of banks has tilted away from traditional deposit-loan spread income and toward other types of income. For example, the proportion of “other” operating income to net interest revenue grew from 49 percent to 67 percent during 1991–96. In some of these banks, other operating revenue is two to three times net interest revenue. An increasingly more important source of revenue for internationally active banks has been their activities in derivatives and fee-based income from investment services: derivatives-based earnings for the larger banks is estimated as roughly 15–20 percent of their noninterest income.
The restructuring that is under way in banking systems is also reflected by banks expanding into other segments of the financial industry and by consolidation within banking industries. First, banks in many countries have stepped-up their securities market activities. This is evident by the well-publicized acquisitions of securities firms by some of the major global banks, by the relaxation of restrictions separating commercial and investment banking in several countries (for example, Canada, the United States, and Japan), and by domestic and foreign banks establishing securities market subsidiaries.2
Second, banks have entered the insurance business. Most of the insurance business now conducted by banks has been domestic, and much of it aims to distribute insurance products—annuities and variable life policies that mirror other long-term investment products—to retail customers. In Europe, by using their low-cost distribution channels, banks have gained market share versus insurance companies in virtually every major European market for relatively simple, standardized savings-type policies, referred to as “bancassurance.” Although some European banks have attempted to enter the insurance business by growing it internally, most have acquired insurance companies. In the United States, rigid regulatory constraints have historically meant the banks have had little latitude to penetrate the insurance industry, but the recent relaxation of some of these restrictions has made banks one of the fastest growing distributors of annuities and basic life insurance policies. For instance, 84 percent of banks with assets over $10 billion and over 60 percent of banks with assets between $100 million and $10 billion were selling insurance in 1995.3 Further, it is expected that the life insurance industry will lose its exclusive underwriting right to annuities in the near future, leaving greater latitude for U.S. banks to enter this part of the insurance business. The recent announcement of a proposed merger between a large U.S. global bank and a large U.S. insurance company might be interpreted as indicating that banks’ involvement in the insurance business is set to expand significantly into all areas of the insurance industry.
Third, banks have entered the asset management business, both by establishing their own asset management units and by acquiring independent asset management firms. Banks seemingly see two potential benefits from expanding their asset management business: (1) fee income from providing investment management services; and (2) providing a wider range of financial services to their traditional customers in order to counter the disintermediation of their deposits. Even in Europe, where universal banks in some countries have long been in the asset management business, it is likely that competition from independent asset managers may cause these banks to become more aggressive in increasing their asset management operations.
Finally, the competitive challenges faced by banks have fostered consolidation in banking industries in North America, Japan, and Europe. This consolidation has typically occurred among domestic banks, but the objectives of the larger bank mergers have often been rooted in the view that by becoming larger they stand a better chance of competing both domestically and internationally. International competition is expected to continue to be a motivating factor underlying mergers, because in many countries restrictions on the entrance of foreign financial institutions are being removed. A recent merger proposal between two Swiss banks, for example, would create a financial institution with close to $1 trillion in assets—a magnitude that is considerably larger than the GDP of the smallest Group of Seven country (Canada). In the rest of Europe, many factors have motivated merger activity and such activity is likely to continue and perhaps accelerate with the introduction of the euro in 1999.4 Overcapacity, deregulation, loss of national protection, disintermediation in wholesale banking, weak earnings growth in many banking business sectors, the need for scale to spread growing information technology and processing costs, and the rising demands of shareholders for a competitive return on their investment are some reasons that have been cited.5
The ability of banks to increase scale and broaden their scope has resulted in two trends developing simultaneously. The first is consolidation among already large banks driven by the goal to be global players in a financial market characterized by financial institutions providing a large number of services worldwide. For instance, an annual report for one of the top 50 banks states that the institution enjoys worldwide relationships with 500 multinational industrial and service corporations, with 4,000 institutional investors and with 1,300 financial institutions. These firms will maintain extensive distribution channels, be at the forefront of product development, and transfer risks around the globe. The second trend is disaggregation at the national and regional levels where banks and other financial institutions will become more specialized, niche players. These institutions will take advantage of the increasing “commoditization” of some types of products and will specialize in only a few areas that meet particular customer demands.
Desegmentation of Financial Services and the Institutionalization of Asset Management
Traditionally, intermediation between borrowers and savers occurred through banks and securities firms, with banks lending depositors’ funds directly to firms, and securities firms providing the distribution of new issues of debt and equity to individual investors, pension funds, and insurance companies. Two notable trends have eroded this traditional view of financial intermediation. First, from the supply side, nonbank financial institutions have been slowly competing away banks’ traditional assets, by facilitating the securitization of finance and also by offering financial services that have historically been provided almost exclusively by banks. Investment banks, securities firms, asset managers, mutual funds, insurance companies, specialty and trade finance companies, hedge funds, and even telecommunications, software, and food companies are starting to provide services not unlike those traditionally provided by banks. Second, on the demand side, households have bypassed bank deposits and securities firms in order to hold their funds with institutions better able to diversify risks, reduce tax burdens, and take advantage of economies of scale. The result has been dramatic growth in the size and sophistication of institutions that specialize in investing money, increasingly on a global basis, on behalf of households.
Nonbank financial sectors in the major advanced economies are very large. In the Group of Seven countries, insurance companies, pension funds, investment companies, and other institutional investors managed assets totaling more than $20 trillion in 1995 (Table A5.6). To put this in perspective, this amounts to about 110 percent of GDP of the Group of Seven countries, it is more than half the value of all bonds and equities outstanding in these countries, and it represents 90 percent of all assets in the banking systems in these countries. In comparison, total assets of institutional investors in 1980 in the major advanced economies was only about one-tenth of what it was in 1995, and as a share of GDP in no country did institutional assets exceed GDP in 1980.6 The United States has progressed the furthest in the process of the institutionalization of savings: U.S. institutional assets under management totaled $11.5 trillion (159 percent of GDP) in 1995, compared with total assets in the U.S. banking system of $5 trillion in the same year.
|Other forms of institutional saving|
|Total assets of all investors (in percent of GDP)|
Pension fund assets of Japan are combined with those of insurance companies and trust accounts of trust banks.
Other nondepository financial companies (bank personal trusts, finance companies, and real estate investment trusts).
Trust accounts of trust banks excluding investment trusts.
As institutional investors have grown in size, they have diversified their portfolios internationally. In 1980, institutional investors in most countries had fewer than 5 percent of their assets invested in foreign securities (Table A5.7). By the mid-1990s, the share of foreign assets in their portfolios had increased to roughly 20 percent on average. For illustrative purposes, if all institutional investors in the Group of Seven countries had 20 percent of their assets invested abroad in 1995, this translates into about $4 trillion of funds invested in foreign markets.
|Life insurance companies|
Only bills and bonds.
The growth of institutional investors has been especially marked in the U.S. mutual fund industry. U.S. mutual fund assets have risen at double-digit growth rates since 1970 when they amounted to just $48 billion.7 The magnitude of wealth that has accumulated in mutual funds since the mid-1980s is striking: by April 1998, U.S. mutual funds managed assets of more than $5 trillion, which is more than the assets of all U.S. banks combined. In addition, from April 1970 to April 1998, the number of U.S. mutual funds increased from 361 to almost 7,000, and the number of individual accounts with mutual funds increased from about 11 million to more than 170 million. U.S. mutual funds are currently estimated to own 20 percent of all U.S. equities.
Large-scale shifts in households’ saving behavior and deregulation of financial industries in many industrial countries have made the fund management industry one of the most dynamic segments of the financial industry in recent years. This dynamism is particularly visible in the hedge fund industry. Although hedge funds have been an acknowledged industry since about the mid-1960s, their growth has accelerated in the 1990s with assets under management increasing 12 times between 1990 and 1997. Since hedge funds are typically offered only to institutional investors, companies, or high-net-worth individuals their investment strategies are only limited by their prospectuses, giving them a large range of investment opportunities, including the ability to go short and use leverage.8 Given their use of leverage many view hedge funds as a high risk/high return investment. However, risk-adjusted returns calculated from a large hedge fund data vendor show that, on average, across a variety of types of hedge funds, they have higher returns with lower risk than the S&P 500 index, at least partly demonstrating the advantages of their investment styles.
Demographic changes and the increased sophistication of small investors around the world, in tandem with the deregulation of financial markets, have intensified competition for savings among banks, mutual funds, insurance companies, and pension funds. The response of the industry to intensified competition for funds has been merger and acquisition activity, mostly for strategic reasons, such as the capacity to build and strengthen their business abroad, the ability to add more assets to existing products in order to create significant operating leverage, and a desire to add to the product mix.
The merger and acquisition activity has been apparent in two recent developments. First, gains in information technology have virtually eliminated the importance of geographic location. Fund management companies have begun consolidating their operations geographically, often in locations that are not usually thought of as major financial centers—for example, San Francisco and Boston. Second, there is evidence that the growth of large asset management firms has exceeded the growth of small ones. In 1985, the top 10 institutional investors in the United States managed assets worth $969 billion expressed in 1995 dollars. A decade later, the top 10 institutional investors managed assets of $2.4 trillion.9 The largest institutional investor in the United States currently manages more than $900 billion in assets, or roughly five times the assets (in constant dollars) of the largest institutional investor in 1985.10 In comparison, the 300th largest asset manager at the end of 1995 controlled $2.7 billion in assets, just slightly more than the $2.4 billion (in constant dollars) managed by the 300th largest asset manager in 1985. This is consistent with a consolidation of assets, with the largest asset managers growing much more rapidly than the smaller asset managers. In Europe, too, the growth of large fund managers has taken place in recent years. The announcement of a merger in 1997 of two Swiss banks aimed at creating the world’s largest asset manager, with close to $1 trillion under management. The desegmentation of the financial services industry is reflected in the fact that banks and securities firms have been particularly active participants in recent mergers and acquisitions in the asset management industry—four of the top six deals of 1997 involved banks and securities firms.
Accompanying the move of banks and securities firms into the asset management industry is the penetration of nonbank financial institutions into traditional bank activities in credit markets. For example, nonbank financial institutions have become involved in loan syndications and bridge loans. Insurance companies, pension funds, asset managers, and mutual funds have entered the credit market via bridge loans, syndicated loans, new structured vehicles such as CLOs and credit derivatives.11 And some European insurers have sold home and automobile loans as well as products that compete directly with bank deposits.
Closer Integration of Financial Markets
Liberalization of domestic capital markets and of international capital flows since the early 1970s, coupled with rapid gains in information technology, has been the catalyst for financial innovation and the growth in cross-border capital movements. In part, the globalization of financial intermediation has occurred in response to the demand to intermediate these growing cross-border capital movements. Firms in most countries currently enjoy access to financial services from a more diverse and more competitive array of providers, and at lower cost, while investors have better information and access to an expanded menu of investment opportunities.
There are many ways of assessing the extent of globalization of financial markets, because markets become integrated in a number of ways: through the increasing web of connections among financial institutions, through exchange linkages, and through less formal trading and information linkages. Before exploring some of the mechanisms by which markets are connected globally, evidence is presented that indicates the growing extent to which financial market integration is taking place.
Cross-Border Finance in a Global Securities Market
The integration of national financial systems into a single global financial system is indicated by more diversified investment portfolios, the larger number of firms tapping foreign sources of funds, and the growth of highly sophisticated asset managers, an important subset of which focus exclusively on identifying and exploiting arbitrage opportunities around the globe. Gross flows and net flows of capital have increased markedly since 1970 (see Table A5.8). The 32 times increase in gross direct investment in the industrial countries is impressive, but it pales in comparison to the growth in gross portfolio flows, which has increased by almost 200 times.
|Foreign direct investment||14.45||34.25||82.82||75.94||283.24||369.01||357.53||448.32|
|Foreign direct investment||-4.05||-9.93||-8.14||-12.66||-59.58||-83.18||-87.41||-92.60|
Group of Seven countries.
Another measure of capital market integration is cross-border securities transactions.12 Cross-border transactions in bonds and equities in the major advanced economies amounted to less than 5 percent of GDP in 1975, but in 1997 they amounted to between one and seven times GDP (Table A5.9). Securities transactions between U.S. and foreign investors, for example, totaled $17 trillion in 1997. Foreign participation in securities markets in Europe is even higher than in the United States and Japan. This accords with the stylized fact that about half of all equity transactions for firms located in the European Union (EU) take place outside the home country.13
Gross purchases and sales of securities between residents and nonresidents.
Mirroring the expansion in cross-border trading in financial assets, firms are increasingly turning to international securities markets to raise funds. International issues of equity have risen almost sixfold during the 1990s for firms located in the industrial countries (Table A5.10). The nominal increase in outstanding issues of international debt securities has been even more impressive (Table A5.11): in early 1998, the outstanding amount of international bonds was $3.7 trillion, or more than six times larger than in 1985. Nonresident holdings of public debt have also increased substantially. Such holdings in Belgium, Canada, Germany, and the United States have more than doubled since 1983, and in Italy there has been a threefold increase since 1990 (Table A5.12).
|Net he rlands||432.0||536.0||65.0||1,267.0||3,330.0||4,071.0||5,817.0||3,693.0||754.9|
|Developing countries and regions|
|Hong Kong SAR||0.0||271.0||230.0||837.0||320.0||1,206.0||3,278.0||3,568.0||…|
|Developing countries and regions|
|Hong Kong SAR||7.5||13.7||15.4||26.1||33.2||34.2|
|Lat in America|
End-of-year data; definitions vary across countries.
Market integration is reflected also in the trading of the same securities in multiple geographic areas. The New York Stock Exchange (NYSE), for instance, lists 343 foreign firms, and American Depository Receipts (ADRs)14 traded on the NYSE cover 315 non-U.S. companies headquartered in 42 different countries. Similarly, at the end of 1997, the London Stock Exchange listed 526 foreign firms. Likewise in markets for derivatives contracts, in the major international financial centers one can trade in derivative securities on a variety of foreign assets. For example, both LIFFE in London and Deutsche Terminbörse (DTB) in Germany trade a German bund contract. In Singapore, the Singapore International Mercantile Exchange (SIMEX) trades a Japanese Nikkei 225 futures contracts, as does the Osaka Securities Exchange in Japan. On U.S. derivatives exchanges, one can trade contracts on Brady bonds and a wide variety of foreign exchange contracts, including contracts on the Brazilian real, the Mexican peso, the South African rand, the Russian ruble, the Malaysian ringgit, the Thai baht, and the Indonesian rupiah.
Financial globalization has been a counterpart to international trade in goods and services, the growing financing needs of countries, and the globalization of national economies.15 This is reflected by the observation that trading in the global foreign exchange market has far outpaced growth in international trade in goods and services. Since 1986, daily nominal foreign exchange turnover has risen sixfold (Table A5.13). World exports of goods and services in 1995 totaled about $6.1 trillion, compared with almost $1.2 trillion in daily foreign exchange market turnover. Put on the same basis, daily turnover in foreign exchange markets was on the order of 50 times exports of goods and services, almost three times what it was a decade earlier. Foreign exchange trading growth rates of these magnitudes, net of the growth rate in trade in goods and services, is clear indicator of the globalization of financial markets.
|Global estimated turnover1|
|(In billions of U.S. dollars)||188||590||820||1,190|
|As a percent of|
|World exports of goods and services||7.4||15.8||17.4||19.1|
|Total reserves minus gold (all countries)||36.7||75.9||86.0||84.3|
Daily average turnover, on spot, outright forward, and foreign exchange swap transactions, adjusted for local and cross-border double counting and for estimated gaps in reporting.
Finally, the integration and globalization of capital markets has been reinforced by the yield-seeking behavior of investors across national borders, most apparent by the cross-border arbitraging of differences in yields on investments with similar risks. Onshore/offshore interest differentials have declined markedly since the 1970s, and are now negligible for most advanced economies. Similarly, covered interest parity holds more tightly across most advanced economies than in the early 1980s.16 Indeed, a sophisticated and significant segment of the financial industry in the major international financial centers is singly concerned with arbitraging often minute mispricing of financial assets around the globe.
In summary, by many measures national financial markets have become increasingly integrated into a single global financial system. The magnitudes of cross-border transactions in securities, foreign exchange turnover, and financing volumes make international trade in goods and services appear small in comparison. The integration process has advanced considerably over the past two decades, and especially so in the 1990s, but there is further room. Banks and other financial institutions have only recently begun to adjust to the new reality of a developing global financial market, and investment portfolios are not anywhere near most benchmarks of optimal international diversification. For instance, a well-known rule of thumb from modern portfolio theory is that an optimally diversified portfolio should have country weights corresponding to the ratio of a country’s market capitalization to world market capitalization.17 The U.S. stock market represents about 42 percent of the world stock market, Japan 15 percent, the United Kingdom 9 percent, other industrial countries 23 percent, and emerging markets 11 percent.18 As the average for institutional investors in the industrial countries currently is somewhere around 20 percent of assets invested abroad, it is apparent that there could be a good deal more cross-border capital flows in the years ahead. The elimination of national currencies in the EU when the euro is introduced in 1999 is an event that might significantly accelerate this process.
Exchange Trading Links
One way in which global markets are becoming more integrated is that exchanges are linking up across borders. The motivation is economic: cost cutting and the introduction of incentives such as lower trading fees and longer trading hours. With the increasing use of technology, trades can be executed more cheaply, and the accompanying lower fees have spurred competition among the exchanges. Estimates suggest that a doubling of volume reduces the trading cost for each contract by about 25 percent: economies of scale make getting bigger, better.
The first overseas joint venture linked SIMEX and the Chicago Mercantile Exchange (CME) in September 1984, in which the popular Eurodollar contract was traded in two major time zones with cross margining, allowing the opening and closing of positions in either location. Now such linkages are common. The grandest scheme, announced in the run-up to the EMU, is the planned development of a European-wide exchange, to be called Eurex. So far, the DTB (the German derivatives exchange) and Soffex (the Swiss options and futures exchange) have formed the axis of the new exchange. A memorandum of understanding with the Marché à Terme International de France (MATIF, the French futures exchange), SBF-Paris Bourse (the French stock exchange), and Monep (the French options exchange) will serve as a basis for a contractual agreement to be signed in the coming months in the formation of the EURO Alliance. A trading alliance between Eurex and the Chicago Board of Trade (CBOT, a U.S. futures exchange) has already been agreed upon.
Exchanges are also attempting to expand participation in their markets by relaxing their membership criteria to include offsite members. A switch from floor trading to screen-based trading opens the door to remote membership and broader participation, since floor trading essentially requires onsite membership. Broader membership means access to more capital and less risk for the clearinghouse, and, usually, increased volume. Some exchanges are attempting to marry floor trading with electronic trading by allowing some of each. MATIF, for example, plans to introduce parallel trading—screen and open outcry—which will aim to offer French, German, and Swiss products on a single electronic trading platform. The DTB will be helped by its new electronic software that will enable traders to put on more complicated trades simultaneously, such as butterfly options and other multiposition trades, which may be difficult to execute in a pit environment.
Similarities Between OTC and Exchange-Traded Markets Increase Market Integration
A consensus is emerging that the open outcry method used on exchange floors will disappear eventually. An unresolved question is whether exchange trading can remain competitive with the OTC markets. OTC trading has grown at a phenomenal pace, far outstripping the exchange-traded markets. Since OTC markets more easily accommodate global trading by the use of telephone, fax, telex, and other communications technologies that remain untied to a specific geographical location, globalization of financial markets has occurred predominantly through OTC markets. The largest markets—foreign exchange and government securities—are predominately traded OTC. Still, the process of globalization can be facilitated by some of the features of exchange trading.
OTC market participants as well as exchange members are attempting to alter their respective markets to take advantage of the attractive features of both types of trading. Exchange-traded markets provide liquidity, price transparency, and credit risk mutualization through the auspices of their clearinghouse framework. The OTC markets have adopted ways of mitigating the credit risk that clearinghouses so efficiently manage: use of netting and sophisticated collateral arrangements are now the norm in the OTC market. Additionally, several attempts are being made to create clearing and settlement facilities for OTC contracts, taking into account the idiosyncracies of the contract negotiated. The London Clearing House is planning to introduce a swaps clearinghouse. Exchanges are also introducing more tailored products to capture the advantages of OTC markets and those moving to electronic means of trading are able to distribute their screens geographically, mirroring the disbursed location of participants in the OTC market. And OTC markets now provide more price transparency on plain-vanilla-type instruments and other instruments whose attributes can be easily summarized.
Over time, the differences that separate OTC trading from exchange trading may slowly disappear as each market migrates closer to the other by choosing those aspects that add value to its participants. To some degree, this outcome is being driven by the joint interests that many of the core, global institutions have in these two trading mechanisms. Many clearing members of exchanges, for example, are also active OTC market participants. Moreover, this is a world-wide phenomenon since many global institutions, with heavy emphasis on OTC trading, are members of exchanges in multiple jurisdictions.
The financial information business, which is dominated by four firms—Reuters Holdings, Bloomberg, Dow Jones Markets, and Bridge Information Services—is also facilitating globalization. The line between information provision and trading is becoming blurred in the race to provide globally accessible financial services. Reuters, for instance, is not only providing financial information but also has successfully devised entire trading systems. Both Instinet, an electronic trading system for primarily retail customers, and the R2000-2 system for the most sophisticated foreign exchange dealers, are examples of Reuters’ success in providing real-time trading systems.
At the same time, the Internet is breeding a host of niche players with connections to financial institutions and investors. While it is unclear whether such players can realistically compete with the global information providers, companies specializing in certain types of information or in combining information with the ability to trade via an electronic brokerage unit are a growing industry in their own right. Even large financial institutions are using the Internet to conduct financial business connecting investors and borrowers without regard to geographic location. For example, Santander Investment and Dresdner Luxembourg launched the first Latin American syndicated loan via the Internet in January 1998.
New Markets and Products for Unbundling, Pricing, Trading, and Managing Risk
Financial instruments are bundles of risks. For example, a floating-rate loan in yen from a U.S. bank to a Japanese bank contains three major risks from the perspective of the U.S. bank: foreign exchange risk (one type of market risk); interest rate risk (another type of market risk); and the risk that the Japanese bank may default on its obligations in the loan contract (credit risk).19
A party to a financial arrangement may not want all the risks associated with that arrangement, or perhaps it may want to leverage certain risks. For example, the Japanese bank might want the loan denominated in yen, but the U.S. bank might want to avoid exposure to the yen/dollar exchange rate—perhaps to avoid a mismatch with currency exposures elsewhere on its books, or because it is discouraged from adding to its currency exposure by managers or regulators. Similarly, the U.S. bank might not want the credit or interest rate risk on its books.
Markets currently exist that enable either party to the arrangement to reconfigure the risks of the arrangement, independently of the other party. To unload the currency exposure, the U.S. bank could enter a swap that exchanges the yen payments for dollars at agreed exchange rates. To unload the interest rate exposure, the U.S. bank could enter an interest rate swap that exchanges payments at LIBOR for payments at a fixed rate of interest. And to unload the credit exposure, the U.S. bank could enter a credit-derivative transaction that transfers the risk of default to the counterparty of the transaction.20 In fact, undertaking all three transactions could, in principle, turn the floating-rate yen-denominated loan into a riskless fixed-rate dollar-denominated security. Although the transfer of risks is most easily seen in swap-type arrangements, other derivatives (such as futures, forwards, and options) and even other securities with derivative-like components (such as convertible debt, warrants,21 and structured notes) are also used to unbundle and distribute risks.
Recent data that cover the major markets bear out the striking size and growth of the markets for these products. For example, in 1997, the notional principal of major exchange-traded futures and options came to $12 trillion, more than two and a half times its level in 1992, and almost 17 times its level in 1987 (see Chapter IV, Table 4.6). In 1996, notional principal in major OTC markets (interest rate and currency swaps and interest rate options) came to $25 trillion, almost 5 times its level in 1992, and almost 30 times its level in 1987 (see Chapter IV, Table 4.7). By comparison, in 1997, estimated world GDP was about $30 trillion. More comprehensive surveys of global derivatives markets paint an even more striking picture of their size. According to a survey conducted by the BIS, at end-March 1995 financial institutions participating in the survey were involved in (after adjusting for double counting and including estimated gaps in reporting) about $64 trillion, by notional principal, of OTC and exchange-traded derivatives contracts. To put this in perspective, the aggregate market value of all bonds, equity, and bank assets in Japan, North America, and the 15 EU countries totaled $68.4 trillion at end-1995, which is about 7 percent larger than the size of derivatives markets as measured by the above survey.
Recent Developments in Markets for Unbundling Risks
The globalization, rapid growth, and increasing sophistication of capital markets has increased the scope for products that can perform this unbundling. The degree of innovation in financial product development has been large. The objective in this section is not to catalogue all of the recent innovations, but rather to provide some perspective on the types and range of products that have recently entered the marketplace. Such risk-products cover a range of risks, from simple market risks, such as interest rates and currencies, to more unusual risks, such as weather-related catastrophes. They also run the gamut from simple to quite complex. Simple, plain-vanilla arrangements, such as interest-rate swaps, have a relatively long history, are well understood, and are fairly straightforward to price. By contrast, more exotic instruments, such as structured notes, are relatively new, less well understood, and can be technically challenging to price and hedge. The increasing complexity of financial products has resulted in increased emphasis on technical model-building for pricing and managing the risks of these products.
A number of products now enable insurance companies to augment the types of risks that they carry on their balance sheets. For instance, futures contracts based on indices covering property and casualty insurance began trading in 1995, allowing insurance companies and others to trade geographical concentrations of underwritten policies. Similarly, in the real estate market, new specialized instruments enable investors to trade different types of real estate risk, and the CBOT and CME plan to list real estate futures contracts in the near future. Finally, there have been recent considerations to base futures contracts on macroeconomic variables: several global banks and securities firms are currently experimenting with using macroeconomic variables, such as GDP or inflation, as the basis for derivative contracts or, even, as the basis for payments of bonds.
Among the new products devised to split risks and recombine them into new ones, credit derivatives are one of the fastest growing markets, and based on the size of the underlying credit markets, they may be set to become one of the largest segments of derivatives markets. U.S. Federal Reserve bank call reports show that the notional amount of credit derivatives held by insured commercial banks and foreign branches in the United States increased from $41 billion in the first quarter of 1997 to $97 billion by the final quarter.22 Although there does not exist data on the size of the global credit derivatives market, estimates suggest it is in the neighborhood of $150–200 billion. While the credit derivatives market may appear small relative to the interest rate swaps market, the notional principle of a credit derivatives contract may be a better measure of exposures than is the case for interest rate swaps.23
The credit derivatives market is currently dominated by four principle types of products: credit default swaps, total rate of return (TROR) swaps, credit-linked notes, and credit spread options. All credit derivatives transfer (for a price) credit risk between two parties. Of the four principle types, TROR swaps and credit spread options are most common in the market for emerging market debt, constituting about half of the credit derivatives market. In contrast, credit default swaps and credit-linked notes are most commonly associated with the trading of bank loans. Not surprisingly, this segment of the market is the fastest growing. In addition to these four principle types of credit derivatives, structured products that contain credit derivatives as a component are becoming more common. The usefulness of these structured products is related to the trading of very specific types of risks; for example, a portion of the credit risk associated with default on a bridge loan. Newer structured deals are being applied to the lease and insurance markets, even in some equity deals, mixing and matching the types of risk that customers desire to trade.24
Developments in Risk Management
A contributing factor to the increased interest in credit derivative products stems from advances in risk management. The ability to simulate the outcomes of credit risk, market risk, liquidity risk, and other types of exposures, and develop estimates of the covariances between the various types of positions a financial institution has, are essential in understanding the overall degree of risk to the institution. There has been considerable progress during the past three years or so in modeling market risk, and recent advances have applied that approach to credit risk, liquidity risk, and operational risk.
The ability to precisely measure and manage market risk (the risk of movements in prices) was the first area of risk management to develop and is now at a stage where further increases in sophistication are unlikely to lead to large changes in the way this risk is managed. The value-at-risk (VAR) methodology, an outgrowth of portfolio theory, was a natural first step to using improved data processing techniques to better measure statistically the probability of losses. Most of the current research is devoted to refinements of existing techniques and better stress testing, while the adoption of standard VAR models is now being undertaken by second- and third-tier banks and other financial and corporate institutions. The Asian crises served as a wake up call to many risk managers who found that their VAR models, since they were backward looking, were unable to predict the true extent of possible losses on their portfolios. Even those models that were built to be more sensitive to recent events failed to account for the correlation between market risk and credit risk. This has led to even greater emphasis on better stress testing and scenario analysis whereby several unlikely configurations of events is considered to evaluate the risks associated with the institution’s exposures. In addition, the most advanced risk managers are considering ways of integrating market risk models and credit risk models in ways that can better identify the overall risk of the institution, a movement that was in train prior to the Asian crises, but that is now considered vital.
Credit risk management is now the focal point for many of the large financial institutions. Credit risk refers to the potential nonpayment of a counterparty to another (counterparty default risk), often associated with inability of the counterparty to make an owed payment. For a bank, credit risk is typically the largest business risk. Ironically, it is also the risk that has, until recently, received the least analytical attention. This situation has improved with the advent of credit derivatives and the realization that better credit risk management can lead directly to an improved bottom line. There are now a number of systems and data bases available to help piece together credit risk profiles and provide aids to managing large credit portfolios.
The two credit-risk systems receiving the most attention are J.P. Morgan’s CreditMetrics and Credit Suisse Financial Products (CSFP) CreditRisk+. Most observers have noted that while the goals of the two products are identical—to evaluate the loss distribution of a portfolio of credit exposures and the capital necessary to support the exposures—they use different methodologies. CreditMetrics follows the method used for RiskMetrics, in which the probabilistic behavior of individual assets is analyzed and then the correlations among the individual assets is used to generate a loss distribution for the portfolio as a whole. CreditRisk+ examines the average default rate associated with each rating or score in a credit rating scheme and the volatilities of the default rates. When added to the exposures these elements produce a loss distribution and an estimated capital allowance.
A problem faced by both systems, and by all those looking to analyze credit risk, has been the lack of data. Actual defaults on securities, in general, are rare and frequently defaults on loans are unreported. Thus, bond defaults, rather than loans, along with data on ratings changes underlie the popular models. More recently, plans to collect proprietary data on loans have been implemented. For instance, the Loan Pricing Corporation (LPC) in New York has accumulated ratings changes and default histories for 20,000 performing loans and 1,400 defaulted loans, drawing from some 30 banks and publicly available information. KPMG, a consulting firm, has entered the credit risk area by providing a product to help with loan valuation called the Loan Analysis System (LAS). This is an attempt to follow the loan from “cradle to grave” to determine how the structure and embedded options (for example, prepayment options) of the loan influenced its price. Others, such as KMV, a San Francisco-based consulting firm, examine the probability of default as related to the firm’s equity value, which can be modeled as a call option on the firm’s value.25 These techniques demonstrate not only that credit risk can be dissected, but that there are multiple ways of doing so.
The end-game for those purchasing or developing in-house models to analyze credit risk is twofold: (1) better, and more standardized, methods for analyzing and managing credit risk; and (2) lower credit risk capital requirements with the use of an internal model, or at least capital requirements more closely tied to demonstrable credit risk. Most believe that there need to be a number of models being used in the market before one can be chosen as a market standard: the methodologies are not yet well enough developed to assess their accuracy. Moreover, as with VAR models for market risk, each firm is likely to customize their own model to fit their business needs and portfolio characteristics. The key will be to have enough standardization that certain principles or qualitative features of the models will be deemed essential to these types of models, allowing regulators to attain a degree of conformity of regulatory capital across similar institutions.
Recent turmoil in emerging markets has illustrated that VAR models do not adequately account for liquidity risk. Liquidity risk has been defined as the risk that the holder of a financial instrument may not be able to sell or transfer that instrument quickly and at a reasonable price. Liquidity risk includes the risk that a firm will not be able to unwind or hedge a position.26
Several initiatives are currently under way to examine and more rigorously capture liquidity risk. One approach27 incorporates three potential losses due to liquidation: (1) a “liquidity discount,” that is, the amount by which the price of a security is decreased when large sales are required; (2) the volatility of the liquidity discount; and (3) the volatility of the time horizon to liquidation. These elements can be built into a VAR model to result in a “liquidity-adjusted VAR” model. Another approach28 begins with the observation that hedging risks may require firms to pay out margin on one side of a hedge when no cash flows are being received on the other. Firms typically do not have unlimited funds from which to make these payments and thus limitations on the hedgeable quantities should take account of the firm’s ability to meet margin payments. Thus, the “liquidity-at-risk” concept is developed as the maximum of the cumulative margin calls requiring cash payments during a relevant time horizon.
Many of the recent losses experienced by some financial institutions have been the result of operational malfunctions. “Operational risk is the risk that improper operation of trade processing or management systems will result in financial loss. Operational risk encompasses the risk of loss due to the breakdown in controls within the firm including, but not limited to, unidentified limit excesses, unauthorized trading, fraud in trading or in back office functions including inadequate books and records and a lack of basic internal accounting controls, inexperienced personnel, and unstable and easily accessed computer systems.”29 A firm with high operational risk may be viewed as a high credit risk, since the probability of a default may rise when operational systems are inadequate, linking operational risk with credit risk.
Recent reports by the Basle Committee on Banking Supervision (1998), and the International Organization of Securities Commissions (IOSCO) (1998), have stressed the importance of operational controls. For instance, in the IOSCO discussion paper it is noted that “the lack of an adequate control environment and ‘control consciousness’ on the part of a firm’s governing body and senior management has been at the root of such recent losses at Barings, Daiwa, Kidder Peabody, and NatWest.” Some of the problems experienced in these firms were the result of improper separation between the front- and back-office functions and inadequate record-keeping as well a general lack of separation of trading and support functions. IOSCO recommends the control of operational risk “through proper management procedures including adequate books and records and basic internal accounting controls, a strong internal audit function which is independent of the trading and revenue side of the business, clear limits on personnel, and risk management and control policies.”
Some financial institutions are now evaluating how to better account for operational risk in their internal allocations of capital and many are expecting to expend considerable resources in managing operational risk in the future. A Coopers and Lybrand/Louis Harris study of top management at 80 of the world’s largest capital markets participants found that almost all (98 percent) of sell-side firms surveyed expect to make significant investments in risk management and other enterprise control systems over the next five years.
Discussions of risk management seldom isolate settlement risks, but globalization puts increasing strain on settlement systems, particularly foreign exchange settlement systems. Settlement risk is the risk of nonpayment through a settlement system and, depending on the source of the risk, is related to both credit risk and operational risk. For instance, nonpayment may arise because of the counterparties’ inability to pay (a credit problem) or due to technical difficulties (an operational problem). As the largest market requiring ongoing use of national payments systems, foreign exchange settlements represent the area comprising the greatest systemic risk from settlement difficulties. A number of recent studies, including the 1996 report issued by the Committee on Payment and Settlement Systems (1996), have pressed the private sector to reduce foreign exchange settlement risks.
Aside from beginning to measure and monitor the settlement exposures involved, some private sector institutions are developing new methods of reducing the foreign exchange settlement risks. Chase Manhattan has offered for consideration a new product, entitled “contracts for differences,” that would avoid the need to deliver spot foreign currency by creating a spot deal that pays the difference between the original spot deal and a valuation index (consisting of a combination of a spot rate adjusted for overnight interest rates, a TOM outright rate30). The product is in a trial phase and if there is continued interest in the product it will be introduced in the fall. Given that delivery in this contract is the difference between two rates, the “contract for differences” is suited for foreign exchange counterparties who do not have the need to receive or pay actual currency.
Another avenue is being considered for transactions in which at least one of the parties must obtain foreign currency. In 1996, the Group of Twenty, originally consisting of 17 large banks active in foreign exchange markets, formulated a plan to set up a clearing bank that would work on a principle dubbed “continuously linked clearing” that is expected to substantially reduce settlement risks by simultaneously settling the two legs of the transaction. The new firm, CLS Services, Ltd., has obtained an agreement from the two largest multilateral netting facilities, ECHO and Multinet, that transactions flowing through their systems would be settled through the CLS bank. The bank still needs the approval of U.S. regulatory bodies before it can begin functioning.
Another area prone to settlement difficulties is securities settlement. The increased use of delivery-versus-payment systems are helping to reduce securities settlement risks. In addition, the risk of a large participant defaulting can be partially mitigated by the establishment of a clearinghouse. For instance, the Emerging Market Clearing Corp. (EMCC), is being set up within the United States to mitigate these risks. Prior to the development of the EMCC, one bank dominated the clearing and settlement of the Brady market, possibly creating a situation in which its failure, or a failure of one of the interdealer brokers who dominate trading, could spread a problem throughout the system. Initially it will handle clearing and settlement of Brady bonds, but intends to expand to include sovereign Eurobonds and other emerging market debt instruments. The EMCC will process the trades, guarantee them, provide risk management services, and send settlement instructions to Euroclear and Cedel Bank.
In sum, accompanying the unbundling process has been an increase in the sophistication of private risk management systems covering a number of areas. Overall, the developments in risk management systems examining market risk have improved greatly and are being distributed beyond the institutions located in the advanced countries. These systems can be used to diversify and control consolidated market risks globally. However, private risk management systems, even in the most sophisticated of global players, still lack a robust methodology of connecting market risk with credit risk and still have difficulty quantifying and managing operational risks. Globalization has made these elements of risk management ever more important. When credit extension accompanies new, nontraditional instruments (counterparty risks of derivatives, particularly credit derivatives, collateralized loan obligations, and so on), the connections between credit and market risk are important for managing a global portfolio. Further, when the complexity of global institutions increases dramatically, monitoring and insuring against operational risks and settlement risks become critical for institutions to maintain their reputational capital and their functioning as an ongoing concern. Thus, while globalization has permitted a steady increase in the degree of diversification through better risk management systems, potentially lowering private risks and systemic risks, it has also added new dimensions, and new connections, to old risks that need careful attention by private risk managers.
Basle Committee on Banking Supervision, 1998, “Framework for the Evaluation of Internal Control Systems” (Basle, January).
1994, “The Internalization of Financial Markets: Measurement, Benefits and Unexpected Interdependence,”Cashiers Economiques et Monetaires, Bank of France, Vol. 43.
Committee on Payment and Settlement Systems of the Central Banks of the Group of Ten Countries, 1996, Settlement Risk in Foreign Exchange Transactions: Report (Basle: Bank for International Settlements).
1998, Hedge Funds and Financial Market Dynamics, IMF Occasional Paper No. 166 (Washington: International Monetary Fund).
1993, “The Integration of World Capital Markets,”IMF Working Paper No. 93/95 (Washington: International Monetary Fund).
1998, remarks before the Annual Financial Markets Conference of the Federal Reserve Bank of Atlanta, Miami Beach, Florida, February27. Available via the Internet: http://www.bog.frb.fed.us/boarddocs/speeches/19980227.htm
International Monetary Fund, 1995, International Capital Markets: Developments, Prospects, and Key Policy Issues, World Economic and Financial Surveys (Washington, August).
International Monetary Fund, 1997, World Economic Outlook, May 1997: A Survey by the Staff of the International Monetary Fund, World Economic and Financial Surveys (Washington).
International Organization of Securities Commissions, 1998, “Risk Management and Control Guidance for Securities Firms and Their Supervisors” (Montreal, March).
World Economic and Financial Surveys
This series (ISSN 0258-7440) contains biannual, annual, and periodic studies covering monetary and financial issues of importance to the global economy. The core elements of the series are the World Economic Outlook report, usually published in May and October, and the annual report on International Capital Markets. Other studies assess international trade policy, private market and official financing for developing countries, exchange and payments systems, export credit policies, and issues discussed in the World Economic Outlook. Please consult the IMF Publications Catalog for a complete listing of currently available World Economic and Financial Surveys.
World Economic Outlook: A Survey by the Staff of the International Monetary Fund
The World Economic Outlook, published twice a year in English, French, Spanish, and Arabic, presents IMF staff economists’ analyses of global economic developments during the near and medium term. Chapters give an overview of the world economy; consider issues affecting industrial countries, developing countries, and economies in transition to the market; and address topics of pressing current interest.
$36.00 (academic rate: $25.00); paper.
1998 (May). ISBN 1-55775-740-2. Stock #WEO-198.
1997 (Dec). ISBN 1-55775-714-3 (English only). Stock #WEO-1797.
1997 (Oct.). ISBN 1-55775-681-3. Stock #WEO-297.
Official Financing for Developing Countries
by a staff team in the IMF’s Policy Development and Review Department led by Anthony R. Boote and Doris C. Ross
This study provides information on official financing for developing countries, with the focus on low-income countries. It updates the 1995 edition and reviews developments in direct financing by official and multilateral sources.
$25.00 (academic rate: $20.00); paper.
1998. ISBN 1-55775-702-X. Stock #WEO-1397.
1995. ISBN 1-55775-527-2. Stock #WEO-1395.
Issues in International Exchange and Payments Systems
by a staff team from the IMF’s Monetary and Exchange Affairs Department
The global trend toward liberalization in countries’ international exchange and payments systems has been widespread in both industrial and developing countries and most dramatic in Central and Eastern Europe. Countries in general have brought their exchange systems more in line with market principles and moved toward more flexible exchange rate arrangements in recent years.
$20.00 (academic rate: $12.00); paper.
1995. ISBN 1-55775-480-2. Stock #WEO-895.
Staff Studies for the World Economic Outlook
by the IMF’s Research Department
These studies, supporting analyses and scenarios of the World Economic Outlook, provide a detailed examination of theory and evidence on major issues currently affecting the global economy.
$25.00 (academic rate: $20.00); paper.
1997. ISBN 1-55775-701-1. Stock #WEO-397.
International Capital Markets: Developments, Prospects, and Key Policy Issues
by a staff team led by Charles Adams, Donald J. Mathieson, Garry Schinasi, and Bankim Chadha
The 1998 report provides a comprehensive survey of recent developments and trends in the advanced and emerging capital markets, focusing on financial market behavior during the Asian crisis, policy lessons for dealing with volatility in capital flows, banking sector developments in the advanced and emerging markets, initiatives in banking system supervision and regulation, and the financial infrastructure for managing systemic risk in EMU.
$25.00 (academic rate: $20.00); paper.
1998. ISBN 1-55775-770-4. Stock #WEO-698
1997. ISBN 1-55775-686-4. Stock #WEO-697
Private Market Financing for Developing Countries
by a staff team from the IMF’s Policy Development and Review Department led by Steven Dunaway
The latest study surveys recent trends in flows to developing countries through banking and securities markets. It also analyzes the institutional and regulatory framework for developing country finance; institutional investor behavior and pricing of developing country stocks; and progress in commercial bank debt restructuring in low-income countries.
$20.00 (academic rate: $12.00); paper.
1995. ISBN 1-55775-526-4. Stock #WEO-1595.
Toward a Framework for Financial Stability
by a staff team led by David Folkerts-Landau and Carl-Johan Lindgren
This study outlines the broad principles and characteristics of stable and sound financial systems, to facilitate IMF surveillance over banking sector issues of macroeconomic significance and to contribute to the general international effort to reduce the likelihood and diminish the intensity of future financial sector crises.
$25.00 (academic rate: $20.00); paper.
1998. ISBN 1-55775-706-2. Stock #WEO-016.
Trade Liberalization in IMF-Supported Programs
by a staff team led by Robert Sharer
This study assesses trade liberalization in programs supported by the IMF by reviewing multiyear arrangements in the 1990s and six detailed case studies. It also discusses the main economic factors affecting trade policy targets.
$25.00 (academic rate: $20.00); paper.
1998. ISBN 1-55775-707-0. Stock #WEO-1897.
Available by series subscription or single title (including back issues); academic rate available only to full-time university faculty and students.
For earlier editions please inquire about prices.
The IMF Catalog of Publications is available on-line at the Internet address listed below.
Please send orders and inquiries to:
International Monetary Fund, Publication Services, 700 19th Street, N.W.
Washington, D.C. 20431, U.S.A.
Tel: (202) 623-7430 Telefax: (202) 623-7201
This annex discusses recent developments in the systemically important emerging markets’ banking systems to assess the vulnerabilities in these systems and their potential consequences for macroeconomic developments and policies. The analysis draws on publicly available material published by national authorities, banks, and rating agencies. Because of space and resource constraints, this section does not describe developments in banking systems in all emerging markets or even in all countries where there are significant banking problems. Instead, it discusses developments in a few of the more important emerging markets in Asia (China, Indonesia, Korea, Malaysia, the Philippines, and Thailand), Latin America (Argentina, Brazil, Chile, Mexico, and Venezuela), and Europe (the Czech Republic, Hungary, Poland, and Russia).
See, for example, Lardy (1997).
By June 1997 Indonesia had 239 banks, of which 27 were regional banks, 11 were foreign banks, and the remainder were domestic private banks. The seven public banks held over 40 percent of the banking system assets, and the large number of private banks was the result of the deregulation wave that began in 1988 (there were only 64 private commercial banks in 1987).
While the first two measures are already in place, the government is expected to pass a law in July 1998 removing all existing restrictions on foreign ownership of banks.
A new special mention category of loans was created, other categories’ definitions were tightened and simplified, and provisions for substandard loans (as well as general provisions) were increased substantially. Since provisions in 1998 will be extraordinarily high, provisions for the existing loan portfolio can be amortized over four years at a minimum of 25 percent of the total amount a year.
These uncertainties, among other factors, led Moody’s to lower the Bank Financial Strength Rating of all banks not already rated E to that level on March 20.
The Korean financial system comprises 26 commercial banks (53.4 percent of end-1997 total assets), specialized and developments banks (16.5 percent of total assets), and several nonbank financial institutions (30.1 percent of total assets). The commercial banks are divided into 16 nationwide banks and 10 regional banks, and the sector also includes 52 branches of foreign banks. The trust accounts of the commercial banks accounted for 20 percent of total assets of the financial system at end-1997.
Most merchant banks were first established as finance companies and converted to its new status in the 1990s, to bring funds from the curb market into the formal market.
According to commercial banks’ end-December balance sheets, the ratio of short-term foreign currency-denominated assets to liabilities was 58 percent, below the 70 percent ratio required by the regulations.
The recapitalization needs may be underestimated because of the lax Korean loan classification standards. Using that classification and including nonperforming loans of W 7 trillion purchased by the KAMC, the ratio of nonperforming loans to total loans increased from 6.9 percent at the end of the third quarter of 1997 to 8.5 percent by end-1997.
The insolvency laws were reformed in the first quarter of 1998, establishing economic criteria for the courts to assess a company’s viability, creditor committees, as well as setting time limits on court decisions and rehabilitation processes.
Malaysia’s banking system comprises 35 commercial banks (69 percent of total banking system assets at end-1997), 39 finance companies (22 percent), 12 merchant banks (6 percent), and 7 discount houses (3 percent). Many finance companies and merchant banks are subsidiaries of commercial banks, and government involvement in the banking sector is relatively high.
A few of the major banks also have exposures to other countries in the region, through their subsidiaries in the Labuan Offshore Financial Center. However, these exposures are relatively small and are expected to add 2–4 percentage points to the banks’ nonperforming loans ratios. Off-balance-sheet commitments and contingencies amount to 15 percent of assets on a credit-equivalent basis and although they are taken into account in capital adequacy ratios banks are not required to make provisions against them. Except for a few banks, these exposures are estimated to be relatively small.
In February, the central bank cut its statutory reserve requirements by 3.5 percent to 10 percent, but offset this by cutting back on its interbank lending. Progressive hikes in the reserve requirements, from 8.5 percent in end-1993 to 13.5 percent by mid-1996, had been the main instrument used to manage the surge in capital inflows during the early 1990s.
Off-balance-sheet exposures of Phillippine banks, mostly trust accounts and derivatives, are not included in capital adequacy ratios but they are estimated to be small relative to the credit risk exposures in banks’ loan books.
Philippine banks’ Foreign Currency Deposit Unit (FCDU) loans expanded substantially during 1995–96, but prudential regulations require banks to keep balanced FCDU books.
The Philippines has 54 commercial banks (89.6 percent of total system assets), 112 thrift banks (8.4 percent of total system assets), and 900 rural banks (2 percent of system assets). The level of foreign involvement in the banking system has increased recently and the 54 commercial banks include 14 branches and 3 subsidiaries of foreign banks.
Additional loan provisions for “loans especially mentioned” (5 percent, regardless of collateral) and for secured loans classified as substandard (25 percent) with full compliance by April 1999 were imposed.
By mid-1997, Thailand had 15 domestic commercial banks, 91 finance companies, and other smaller nonbank financial intermediaries. The finance companies account for around 25 percent of financial sector assets and many of them are controlled by the commercial banks. The three largest commercial banks account for 50 percent of total banking system assets and the largest six banks account for 75 percent of total bank assets. There are also 19 foreign banks that held only about 9 percent of assets, but this underestimates the importance of foreign banks that execute much of their business offshore.
At the same time, the authorities sought to minimize moral hazard by imposing a deposit cap of 3 percentage points over the average rate offered by the five largest banks.
Of the total external debt of $92 billion at end-1997, about $39 billion was in the banking system. Foreign banks account for a large share of the debt maturing in 1998, but most of the funds borrowed from their own headquarters are being rolled over. The lower rollover rates have been experienced by the domestic Thai banks and finance companies.
The four banks are First Bangkok City Bank PCL, Siam City Bank PCL, Bangkok Metropolitan Bank PCL, and The Bangkok Bank of Commerce PCL, respectively, the seventh, eighth, ninth, and eleventh largest in terms of total assets by January 1998.
Although important amendments to the 1940 Bankruptcy Act were passed in March 1998, the limits to the special protection normally granted to new creditors of distressed firms was only modified—rather than repealed, and further procedural clarifications are slated to be approved by parliament toward the end of October.
See Chambers and Karacadag (1998). Two of the three largest banks in the country led the recapitalization move with good international response, but unfavorable market conditions forced several finance companies to postpone equity issues in April–May 1998.
Doubtful loans will be those that have not been serviced for 6 months, compared with 12 months before, while those not serviced for over one year will be classified as bad loans. Provisioning levels will be 50 percent and 100 percent, respectively. There will be a 1 percent general loan provisioning requirement and a 2 percent requirement for special-mention loans.
Overnight interest rates shot up to over 12 percent in late October 1997, but returned to precrisis levels of around 6 percent by end-November. The return on average equity of the 18 largest Argentinean banks fell to 5.5 percent in December 1997 from 6.8 percent in December 1996, and the net interest margin also fell to 3.3 percent, from 5 percent by end-1996 (see Salomon, Smith Barney, 1998a).
Five other provincial banks are in the process of privatization and the largest state-owned commercial and mortgage banks are slated for future privatization.
Although this may appear to be a relatively high liquidity ratio, it is lower than that required in other countries with currency board arrangements. For instance, liquidity ratios are 25 percent in Hong Kong SAR and 30 percent in Estonia and Lithuania (see Baliño and Enoch, 1997).
The acronym BASIC stands for bonds, auditing, supervision, information, and credit rating (see Banco Central de la República Argentina, 1997).
The other public banks are more specialized in their operations. Caixa Econômica Federal is the main institution providing housing finance and BNDES is the primary development bank.
This discussion is based on 74 private banks for which the financial statements for end-1997 were available from Fitch IBCA Ltd. The large banks are those with assets above $4 billion and the group of medium-sized and small banks comprises 52 institutions with assets below that level.
At the outset of the Real Plan, there were 271 banks operating in Brazil. During the first three years of the plan, 40 banks were intervened by the central bank, of which 29 were liquidated, 4 failed, 6 were placed under temporary administration, and 1 continued to operate. A further 32 banks went through restructurings that resulted in the transfer of control, some of them with government support through the restructuring program PROER (see International Monetary Fund, 1997). The cost of restructuring the commercial banks, including the recapitalization of Banco do Brasil, is estimated at around 4 percent of GDP, but this does not include the costs associated with the clean up of the state banks.
This discussion of Chilean banking developments is based on data of the Superintendencia de Bancos e Instituciones Financieras collected and published by Salomon, Smith Barney (1998b).
By end-1997, the Chilean financial system comprised 32 institutions: 29 commercial banks (of which 17 are foreign-owned and one is government-owned) and 3 finance companies.
This ratio incorporates the impact of market risks; if only credit risks are considered, the capital adequacy ratio increases to 16.9 percent.
The fiscal cost of the bank rescue programs is currently estimated at 14.4 percent of GDP, up from an estimate of 11.9 percent of GDP in October 1997, owing to the higher costs of restructuring several regional banks and a lower loan recovery rate than projected earlier.
In July 1997, FOBAPROA started the sale of assets, through its subsidiary Valuacion y Venta de Activos. The first pool of loans was valued at 49 cents on the dollar. The subsidiary was absorbed by FOBAPROA in August and the sale of assets has not resumed yet.
The official assistance to the banking sector amounted to $11 billion—18 percent of 1994 GDP—and the authorities expect to recover $2.6 billion through the sale of the assets (as of May 1998, FOGADE had recovered $1.6 billion from asset sales).
Reflecting concerns about the costs of an increasing stock of central bank bills, the central bank shifted to treasury bonds and nonnegotiable CDs as the main instruments of monetary policy. The lower liquidity of the latter instruments led to liquidity problems in many of the medium-sized and small banks.
Further details on the Czech banking system can be found in International Monetary Fund (1998).
For further details on the Hungarian financial sector, see van Elkan (1998).
For further details of the performance of the Polish banking system in 1997, see National Bank of Poland (1998).
In the Mexican crisis, for example, stock prices in the five hardest-hit emerging markets (Argentina, Brazil, Mexico, the Czech Republic, and Poland) fell by an average (capitalization-weighted) of around 51 percent between October 1994 and March 1995, but then recovered by around 79 percent over the following two years. By contrast, the average initial fall in 25 other emerging markets was around 9 percent, with a subsequent two-year increase of 14 percent.
At shorter horizons, however, the evidence for momentum is less clear, with some studies showing negative autocorrelation around large daily or weekly price movements.
Cutler, Poterba, and Summers (1991) document this common pattern of “speculative dynamics” in a range of markets including equities, currencies, land, and collectibles.
See, for example, Daniel, Hirshleifer, and Subrahmanyam (forthcoming).
For analyses of leading indicators, see Eichengreen, Rose, and Wyplosz (1995); Frankel and Rose (1996); Goldstein (1996); Aziz, Caramazza, and Salgado (forthcoming); Berg and Patillo (forthcoming); Demirgüç-Kunt and Detragiache (forthcoming); International Monetary Fund (1998); and Kaminsky, Lizondo, and Reinhart (1998). For recent analyses of the determinants of currency and banking crises, see Gavin and Hausmann (1996); Kaminsky and Reinhart (1996); Sachs, Tornell, and Velasco (1996); Goldstein and Turner (1996); Caprio and Kliengebiel (1997); González-Hermosillo, Pazarbaşioğlu, and Billings (1997); Demirgüç-Kunt and Detragiache (1998); and Eichengreen and Rose (1998). A brief survey of the determinants of currency and banking crises is provided in International Monetary Fund (1997), pp. 245–49.
See, for example, Eichengreen, Rose, and Wyplosz (1995); Frankel and Rose (1996); Aziz, Caramazza, and Salgado (forthcoming); Kaminsky, Lizondo, and Reinhart (1998); and International Monetary Fund (1998).
See, Milesi-Ferretti and Razin (forthcoming) for further discussion of current account deficits as predictors of currency crises.
The rate of the URR, which had been 30 percent since May 1992, was reduced to 10 percent in June 1998.
The BIS reporting area includes the Group of Ten countries plus Austria, Denmark, Finland, Ireland, Luxembourg, Norway, and Spain. The data include all cross-border bank claims plus claims in nonlocal currency of local affiliates of banks in the BIS area. The BIS data are disaggregated only by maturity or by sector so it is not possible to analyze the maturity structure of only the private sector debt.
Another explanation for this discrepancy maybe in that BIS data classify loans by their actual maturity, while Chilean data consider the maturity at the date of issuance.
Previously, reserve requirements could also be constituted in other currencies, including yen. Since interest rates in yen were lower than in U.S. dollars, investors preferred to constitute their reserve requirements in yen as they carried a lower implicit tax.
A shortcoming of the methodology applied in this paper is the possibility of a simultaneity bias of the estimated parameters, as some of the variables used as regressors may in fact be endogenous.
This argument, albeit weak, seems consistent with the behavior of the Central Bank, which at the end of 1995 strengthened the regulation extending the reserve requirement to fixed income securities and to equities.
The top 50 banks by total assets were identified by The Banker (July 1997), using end-1996 data. Data regarding these banks was obtained from Fitch IBCA Ltd. Using the assets from the largest 1,000 banks as reported by The Banker (July 1997), the top 50 banks account for almost 50 percent of the assets showing that the top 5 percent of banks hold 50 percent of the assets.
For example, during 1996–97, 394 banks filed applications to establish an office in the United States, compared with just 29 in 1995 (International Banking Regulator, various issues). A number of institutions applied to U.S. bank regulators for permission to engage in a variety of nonbanking activities, with brokerage and investing receiving the highest number of applications, followed by other nonbanking, which was then followed by requests to use forward contracts, futures, options, and swaps. A bank can submit an application for multiple requests so it is difficult to judge the number of banks entering these various nonbanking activities.
From Glossman and Plodwick (1998), p. 17.
See Chapter V for a more detailed discussion of financial sector consolidation in Europe.
See Lee (1998).
See “The Increasing Importance of Institutional Investors” in International Monetary Fund (1995) for a discussion of the trends in institutional asset management between 1980 and 1992.
Data on U.S. mutual funds are from the Investment Company Institute.
For a discussion of hedge funds, see Eichengreen and others (1998).
Institutional Investor, 1996 (London). July.
The figures reported here and below on institutional investors are calculated from figures reported in Institutional Investor (various issues) and from Fidelity Investments.
The nature of these instruments are described below.
Even gross portfolio flows are just a net measure of international securities market activity—purchases and sales of foreign securities are reported as a single net entry. This measure approximates the change in stocks of foreign securities held, but says nothing about the level of cross-border financial market activity underlying the change in stocks.
ADRs and Global Depository Receipts (GDRs) represent shares listed on local exchanges.
See Greenspan (1998).
In theory, these country weights should be based on all assets (stocks, bonds, real estate, and so on). A common simplification is to use stock market capitalization.
Data for 1996 from International Finance Corporation (1997).
There are other risks in financial instruments, such as liquidity, operational, legal, and settlement risks, particularly when the instrument is traded in a secondary market. These risks are less obvious than credit and market risks, but no less important.
In addition to the credit risk traded in the derivative, OTC derivatives contracts themselves contain counterparty credit risk owing to the possibility that the counterparties to the contract fail to make agreed upon payments.
A convertible is a corporate bond or preferred equity issued by the company which allows the holder to exchange the bond for equity in a fixed ratio anytime prior to maturity of the bond. Sometimes the numbers of shares to be exchanged for each bond is lowered over time to accommodate a generally rising stock price. A warrant is an option to buy the equity security at a fixed price prior to a specific expiration date. Warrants differ from regular options in that they are issued by the company and thus increase the number of shares outstanding when they are used. Often the warrants are attached to an issuance of equity and are not “separable” meaning that only current holders of equity can exercise them.
Two motivations for banks’ use of credit derivatives are (1) freeing up capital for further loan and bond origination and maintain client relationships, and (2) capital arbitrage. Banks can reduce their regulatory capital reserves by cash-collateralizing existing exposures that have low returns on regulatory capital. The arbitrage opportunity occurs because investors in these trades are not required to hold 8 percent capital against the securitized notes or derivatives they buy. Insurance companies, one of the largest users of credit derivatives, have risk-based capital guidelines, but they are far less than 8 percent against investment grade credit.
The true exposure of an interest rate swap is quite small because amounts being transferred depend on the difference between a fixed and floating interest rate, using the notional amount to calculate payments, and the notional amounts themselves are not swapped. However, payments being transferred between buyers and sellers of credit derivatives are likely to be much closer to the notional principle, since the value of the security in the case of a credit event will be much closer to its initial value than the difference in two interest rates. Thus, the seller’s possible exposure can be fairly close to the notional principle designated in the derivative, implying the exposures being taken using credit derivatives may not be comparable to those in most other derivatives markets.
An interesting new credit derivative is inconvertibility options. These options insure investors against a currency becoming inconvertible: their payoff occurs when a central bank prevents a specified currency from being converted into another currency.
When the firm’s value falls below its obligations (debt) the firm defaults. Thus, the strike price for the call is the value of the debt and the volatility of the firm’s business risks can be used in an options pricing framework to calculate the probability of default.
See Jarrow and Subramanian (1997), pp. 170–73.
See Singer (1997), pp. 86–87.
TOM refers to tomorrow.