Back Matter

Back Matter

International Monetary Fund
Published Date:
December 1997
    • ShareShare
    Show Summary Details
    Background Material—Part I Recent Developments and Trends in Capital Markets and Banking Systems

    Annex I Recent Developments in Emerging Capital Markets

    As the effects of the Mexican peso crisis on investor sentiment continued to wane, a number of factors helped propel private capital flows to the emerging markets from $192.8 billion in 1995 to a new peak of $235.2 billion during 1996.1 These factors included, first, the low level of interest rates in Japan and Germany and the compression of corporate bond spreads in the United States, which prompted fixed-income investors in the mature markets to move down the credit spectrum and search for higher yields on emerging market debt. Second, improved economic performance in many emerging markets reduced perceived credit risks. Third, institutional investors in the mature markets continued to seek the benefits of portfolio diversification in the emerging markets. Fourth, innovations in financial markets improved the ability of investors to manage exposures and risks to emerging markets, increasing the attractiveness of such investments. Fifth, continued financial and capital account liberalization in many emerging markets encouraged inflows. Finally, improvements in the availability and quality of information on emerging markets facilitated improved asset selection and assessment.

    Underlying the surge of total private flows in 1996 were both strong foreign direct investment (FDI) and portfolio flows. FDI continued to grow rapidly, representing the largest component of flows, while portfolio flows almost doubled. As portfolio flows rebounded vigorously, bank lending flows fell off, though they continued to grow strongly to particular regions, such as Asia. Across the emerging markets, during 1996 and into 1997, investor sentiment shifted away from Asia in view of the regional slowdown, concerns about the current account deficits of some countries, and uneasiness about the stale of the property and financial sectors, in favor of Latin America, where growth picked up, inflation slowed, and there was visible progress in strengthening and restructuring banking systems. The growth of total flows to Asia moderated, while flows to Latin America more than doubled, rising above the previous highs of 1993. While flows to the Middle East and Europe grew strongly, flows to Africa and the transition economies declined. As through the first half of the decade, the aggregate reserves of the emerging market countries continued to grow during 1996, and almost half of the net inflows were accumulated as reserves. Compared with the turbulence during late 1994 and 1995, emerging foreign exchange markets were relatively calmer in 1996 and early 1997. Though certain systemically important emerging markets remained susceptible to speculative attack, these pressures remained localized. In mid-May 1997, however, as the Thai baht came under severe speculative attack, pressures spilled over to a number of other countries, both within and outside the region, where international investors saw parallels in economic circumstance and structure.

    The surge in portfolio flows during 1996 was associated with a spectacular boom in emerging debt markets, while emerging equity markets continued to recover from the trough following the Mexican crisis. There were dramatic improvements in the liquidity of emerging debt markets and steep reductions in the volatility of returns on both debt and equity markets. The bond market rally sparked a sharp shift in the structure of emerging market primary external financing toward increased bond issuance and a reduced reliance on syndicated bank lending. Spreads on new bond issues fell across the board, while maturities lengthened. The favorable environment encouraged a number of new entrants into the market and led several borrowers to restructure existing liabilities at improved terms. By early 1997 spreads on emerging market debt had declined to their previous historic lows—of late 1993 and early 1994—leading to concerns that yields may have reached their lower limits in adequately compensating for risk. Although trading activity continued to increase, returns fell off sharply during the first quarter of 1997. Expected returns on emerging market equity—earnings-price ratios adjusted for growth—rose steadily during 1996 and into 1997, buoyed by upward revisions to forecasts of growth, while volatility declined. Adjusted for volatility, returns, particularly in Latin America, looked increasingly favorable relative to those in the mature markets. The increase in emerging market equity prices during 1996 accelerated in the first quarter of 1997, again particularly in Latin America. In the international syndicated loan market, a reduced demand for bank financing by emerging market borrowers coincided with rising supply, and strong competition among banks created considerable pressures on pricing and weakened loan structures, also raising concerns as to whether risks were being sufficiently priced. Refinancings accounted for almost a fifth of new syndications of medium- and long-term loans in 1996, and over a third in Latin America.

    This annex discusses emerging market financing, with a focus on recent developments during 1996–97. The first section discusses net capital flows in the balance of payments, the behavior of international reserves, and developments in foreign exchange markets. The following sections discuss developments in emerging debt markets, equity markets, mutual funds dedicated to emerging markets, and international bank lending.

    Capital Flows, Reserves, and Foreign Exchange Markets

    Capital Flows in the Balance of Payments

    In spite of several unfavorable developments, total private capital flows to emerging markets during the 1990s have proven remarkably resilient (Table 13 and Figure 21). Increases in interest rates in the mature markets during the course of 1994, the Mexican peso crisis and “Tequila” (contagion) effects that followed, and occasional high volatility in the mature assets markets all had only temporary and localized effects on these flows. Similarly, during 1996 the strong performance of many of the mature equity markets, uncertainties relating to the course of interest rates in the mature markets, and perceived vulnerabilities in some of the systemically important emerging market countries failed to deter the overall volume of private flows to emerging markets, which grew by 22 percent to a new record of $235.2 billion. For the first time in the 1990s, private capital flows to the emerging markets exceeded total (private plus official) capital flows in 1996, and $13.2 billion in net repayments of official flows meant that total capital flows actually declined from $232.0 billion in 1995 to $222.0 billion in 1996. Net official flows were negative not only to Latin America, reflecting the substantial repayments by Mexico of the official assistance extended in the aftermath of the crisis, but also to the Middle Eastern, European, and transition economies.

    Figure 21.Net Private Capital Flows to Emerging Markets

    (In billions of U.S. dollars)

    Sources: International Monetary Fund, World Economic Outlook; and IMF staff estimates.

    1Total net private capital inflows equal net foreign direct investment plus net portfolio investment plus net other investment.

    A key characteristic of the surge in private capital inflows to the emerging markets during the 1990s, and one that has been critical in underpinning the resilience of total private flows during the period, has been the steady growth of FDI flows. Encouraged by continued capital account liberalization and the easing of restrictions on FDI in emerging market countries, multinational corporations swiftly relocated and purchased existing production facilities in the relatively lower-labor-cost emerging market countries. FDI flows to emerging market countries expanded between 1991 and 1995 at an average annual rate of 37 percent and continued to grow robustly during 1996, increasing by 21 percent. Since 1995, net FDI flows have accounted for the largest proportion of flows, and in 1996, at $105.9 billion, accounted for some 45 percent of total private capital flows.

    Unlike FDI flows, portfolio flows to the emerging markets have been volatile. After falling off sharply during 1994 and 1995 in the wake of the Mexican peso crisis, total portfolio flows to emerging markets recovered robustly, increasing by 86 percent, from $31.6 billion in 1995 to $58.7 billion in 1996, accounting for 25 percent of total private flows. Despite the rebound, however, portfolio flows remained well below—at just over half—the peak levels reached in 1993 when they accounted for 66 percent of private flows. “Other” flows, which largely reflect bank lending, after having risen sharply during 1995 as the increased costs of borrowing on international capital markets in the wake of the Mexican crisis caused emerging market borrowers to turn to bank financing, declined modestly during 1996 to $70.6 billion. As a proportion of total private flows, however, they declined from 38 percent in 1995 to 30 percent in 1996.

    During the last few years, investors have displayed an increasing tendency to discriminate between regions and countries in response to changes in economic fundamentals, and this has been reflected relatively quickly in the behavior of capital flows. The Mexican peso crisis resulted in a reallocation of flows away from Latin America toward Asia and the transition economies. As total flows to Latin America fell off during 1994–95, flows to Asia continued to increase and flows to the transition countries rose steeply. During 1996, as investor sentiment turned away from Asia, the growth of flows to that region slowed and there was a sharp rebound in flows to Latin America. Portfolio flows have been more responsive than FDI flows, and in 1996 Latin American countries were the largest recipients of portfolio flows among the emerging markets.

    Total private capital flows to Latin America more than doubled from their depressed levels in 1995 of $35.7 billion to $77.7 billion in 1996. The sharp rebound raised total flows to the region above the previous peak of 1993, completing the recovery of total flows from the effects of the Mexican crisis. FDI flows, which had declined only modestly during 1995, grew by 50 percent, to $29.9 billion in 1996. After net outflows of $7.5 billion during 1995, net portfolio inflows resumed, totaling $27.1 billion. Portfolio flows remained, however, well below half their peak of $61.1 billion in 1993. Reflecting the declining reliance on bank lending, “other” flows contracted by 11 percent to $20.7 billion in 1996. As a share of total net inflows, they contracted more sharply, falling from 65 percent in 1995 to 27 percent in 1996. While total flows to the region recovered during 1996, there was a drastic change in the composition of these flows relative to 1993. Net portfolio inflows, which equaled total net private inflows during 1993, represented only 35 percent of flows during 1996.2 On the other hand, the share of FDI in total flows to the region rose from 22 percent in 1993 to 38 percent in 1996, and compared with net “other” capital outflows during 1993, there were net “other” inflows during 1996 representing 27 percent of total net inflows.

    While total private flows to Asian emerging markets continued to grow during 1996, rising to $106.8 billion, the rate of growth decelerated sharply, from 32 percent in 1995 to 8 percent. The increase was due primarily to increased FDI flows, which grew by 14 percent, to $58.0 billion. Portfolio flows to Asia, after having declined modestly from their peak of $23.8 billion in 1993, remained steady at $20.1 billion during 1995 and 1996, while “other” net inflows increased modestly. In comparison with Latin America, FDI flows account for a substantially larger proportion of flows to Asia—54 percent in 1996—while portfolio flows account for substantially less—19 percent during 1996. Japan represents an important source of FDI flows to the Asian emerging markets, and the relocation of Japanese manufacturing to the region since the mid-1980s has been a major driving force behind the growth of FDI to the region.

    Following a substantial increase in capital flows to the transition economies in 1995, inflows declined sharply in 1996. There were sharp declines to both the Czech Republic and Hungary, where all categories of flows declined, and a somewhat more modest decline in flows to Poland, where FDI continued to grow. Net flows of capital to the Middle East and Europe rose from $15.3 billion in 1995 to $22.2 billion in 1996, reflecting an increase in “other” flows, while FDI flows remained modest and steady, and portfolio flows declined. Private capital flows to Africa, which rose modestly during 1995, fell in 1996 to $9.0 billion. Africa is the only region that has not shared significantly in the resurgence of private capital flows to the emerging markets during the 1990s, not receiving any significant portfolio flows over the period, and during 1996, official flows accounted for over 40 percent of flows to the region.

    The rapid and unfaltering growth of FDI flows to emerging markets during the 1990s and the steady increase in the share of FDI flows in total private flows have led many observers to conclude both that the risks of a reversal of sentiment against the emerging markets have concomitantly diminished and that, were such a reversal to occur, the consequences would not be severe. Underlying this belief are several notions. First, that FDI flows, by their nature, tend to be “long-term,” in that they are driven by positive longer-term sentiment in favor of emerging markets and, therefore, less likely to be reversed than relatively “short-term” portfolio flows. Second, since FDI entails physical investment in plant and equipment, it would, in fact, be difficult to reverse.

    The events surrounding the Mexican crisis certainly help support this view. Even as portfolio flows to Latin America switched from a net inflow of $60.8 billion during 1994 to a net outflow of $7.5 billion in 1995, substantial net inflows of FDI continued, declining only modestly, from $21.5 billion to $19.9 billion. However, there are a number of features of both the data on FDI flows, and the historical behavior of FDI flows, that suggest caution in interpreting the growth in importance of such flows as imparting an enduring resilience to capital flows to emerging markets.

    Several factors suggest that the proportion of FDI in total flows as measured by balance of payments data may overstate the importance of these flows. First, the balance of payments differentiation between FDI flows and portfolio flows is arbitrary. Foreign investment in the equity of a company above a critical proportion of outstanding equity is classified as FDI, whereas that below the critical threshold is classified as portfolio equity investment. In reality, small differences above the critical level are unlikely to represent any substantially longer-term intentions of the investor, as compared with those below. Second, if the foreign company undertaking the FDI borrows locally to finance the investment, say from a local bank, depending on the form of incorporation of the company locally, the setup of the plant may count as FDI while the bank lending could show up as a capital outflow, reducing the proportion of net bank lending in overall flows and raising the proportion of FDI flows. Finally, there are sometimes tax or regulatory advantages to rerouting domestic investment through offshore vehicles and these factors have likely overstated the growth of FDI in recent years. The most commonly cited example of such rerouting of domestic investment is that by Chinese enterprises through Hong Kong, because of the tax advantages of doing so. With regard to the reversibility of FDI flows, while it may, in principle, be more difficult and expensive to sell physical rather than portfolio assets, physical assets, nevertheless, can still be sold, albeit typically at a discount, and in the end the sentiment for reversal will be weighed against the discount. There is little reason to expect the discount to always be prohibitive. With regard to the predictability of FDI flows, the experience of the Mexican crisis discussed above notwithstanding, research indicates that, historically, for both industrial and developing countries, FDI and other flows labeled “long-term” according to the traditional balance of payments definition have generally been as volatile as, and no more predictable than, flows labeled “short-term.”3

    Reserve Accumulation

    As has consistently been the case throughout the 1990s, the aggregate reserves of the emerging markets continued to grow during 1996 (Table 13 and Figure 22). Of the $222.0 billion total capital flows to the emerging markets during 1996, $104.8 billion—47 percent—was accumulated as central bank foreign exchange reserve assets, while the remainder was used to finance current account deficits. The increase was larger in Asia ($61.8 billion—59 percent of total) than in Latin America ($26.2 billion—25 percent of total), though as a proportion of flows (54 percent in Asia and 40 percent in Latin America) it was substantial for both regions. Central bank reserve assets of the major Asian emerging market countries rose across the board, with the exception of the Taiwan Province of China, where reserves declined sharply early in the year but recovered most of these losses by year-end. The most rapid growth was in China, where reserves increased by $31.7 billion during the year. In Latin America, as capital inflows rebounded, reserves accumulated rapidly in Argentina, Brazil, and Venezuela, not only recovering from their losses during 1995, but rising well above previous levels. In Mexico, after recovering from their losses during the crisis by late 1995, reserves rose during the latter half of 1996 but remained below the levels of early 1994. In Eastern Europe, reserves declined modestly in both the Czech Republic and Hungary, though they remained at relatively high levels. In South Africa, which has persistently had perhaps the lowest level of reserves among the major emerging markets, reserves continued to fluctuate at low levels.

    How large has the 1990s buildup of reserves been? Of the $ 1.2 trillion in total net flows to emerging markets during 1990–96, some $575 billion—49 percent—was accumulated as reserves. This raised emerging market central banks’ reserve assets to $823 billion by end-1996, a more than threefold increase since end-1989 and representing about half of the stock of reserve assets of the world’s central banks. Five of the world’s 10 largest holders of reserves are now emerging market economies. These holdings are concentrated in Asia—China, Taiwan Province of China, Singapore, and Hong Kong, China. Growth in China’s reserves has been the most dramatic—rising by $89 billion during the 1990s—making it the largest holder of reserves among the emerging market countries. Other notable increases over the period include Singapore ($56 billion). Brazil ($51 billion), and Thailand ($27 billion).

    Figure 22.Total Reserves Minus Gold of Selected Emerging Markets, January 1990—May 1997

    (In billions of U.S. dollars)

    Source: International Monetary Fund, International Financial Statistics.

    The large buildup in emerging market central bank reserve assets during the 1990s reflects in part direct central bank intervention to prevent nominal exchange rate appreciation in the face of the substantial capital inflows. It also reflects concerns about the risks of a sudden reversal of capital flows. Recent history, in particular the sharp loss of reserves during the reversal of capital flows to Mexico during 1994 when, within a few days in December, the central bank lost $5 billion in reserves, and portfolio management considerations in a world of increased capital mobility suggest that traditional import-cover measures are no longer appropriate for judging the adequacy of the level of reserves. Reserve coverage needs to be measured instead in relation to a broad range of monetary aggregates and banking system and government short-term liabilities. Relative to these aggregates, the buildup in reserves has been more modest. At end-1996, for example, while Thailand’s reserves were sufficient to cover over six months’ of imports, they represented only about a quarter of broad money.

    The substantial accumulation of reserves by emerging market country central banks raises several issues about the efficiency of allocation of capital. First, reserve assets represent a component of national wealth but are typically held in low-yield (albeit liquid and credit-risk-free) assets such as government securities in the mature markets—particularly U.S. treasury securities; thus, excess holdings of reserves would imply an inefficient allocation of national wealth. (Asset-liability management at a national level is discussed in Annex V of the Background Material.) Second, it is ironic that some 49 percent of the $1.2 trillion of net capital flows into emerging markets in search of higher returns has ended up accumulated as reserves, a substantial proportion of which has then been reinvested back in low-yield instruments in the mature markets. This implies that the differential between the higher yield demanded and earned by investors from the mature markets in emerging markets, and that earned on reserves reinvested back into the mature markets, represent a cost that will ultimately be borne by residents of emerging markets. This flow cost could be substantial, and present yield spreads on emerging market debt suggest these costs could be of the order of $10 billion annually. These costs, of course, need to be weighed against the benefit of the liquidity provided by the reserves and the objective of alleviating downward pressures on the exchange rate in the event of a reversal of capital flows from emerging markets.

    The large buildup of reserves also implies that emerging markets now have a bigger presence in world securities markets. As a measure of their importance, consider that the $823 billion of emerging market reserves represented over 20 percent of the stock of marketable U.S. government securities at the end of 1996.4 The buildup of reserves also creates channels for the interaction and feedback of disturbances in financial markets among the emerging and mature markets. First, the recycling of capital inflows into emerging markets back into the mature markets means that disturbances in cither could have multiplier effects. A disturbance that leads to a decline in interest rates in the mature markets, for example, and stimulates flows into the emerging markets, but results in almost half of it being reinvested back in the mature markets, is likely to place further downward pressure on interest rates in the mature markets, cause further outflows to emerging markets, and so on. Second, very similarly, a reversal of flows from the emerging markets, to the extent that it prompts a sell-off of reserve securities by emerging market central banks and puts upward pressure on interest rates in the mature markets, is likely to exacerbate outflows from the emerging markets.

    Foreign Exchange Markets

    A cornerstone of macroeconomic management in most emerging markets in response to the surge in capital inflows during the 1990s has been sustained central bank intervention to prevent nominal exchange rate appreciation, and emerging market currencies have, with few exceptions, either been pegged or depreciated in nominal terms over the period (Figure 23). In response to episodes of reversals in flows, authorities have relied on their reserve holdings to resist downward pressures on nominal exchange rates. Since the adjustment of real exchange rates can take place through the adjustment of either nominal exchange rates or domestic prices, preventing nominal exchange rate adjustment shifts the pressure to domestic prices. Forcing adjustment through goods prices can—if goods prices are slow to adjust—reduce the volatility of real exchange rates in the event that the sources of pressure for change—capital flows—themselves tend to be reversed frequently. An important consideration, therefore, is the nature of the capital flows—whether they are temporary and likely to be reversed or they are of a more permanent nature. The substantial buildup of reserves during the 1990s, and the limiting of nominal exchange rate movements, suggests that capital inflows into the emerging markets have tended to be treated as short term. The strategy of intervention and of limiting nominal exchange rate movements has increasingly given rise to uncertainty on the part of market participants as to the sustainability and future course of exchange rate management.

    Figure 23.Exchange Rates of Selected Emerging Markets, January 1990–May 1997

    (Local currency/U.S. dollar)
    (Local currency/U.S. dollar)

    Source: The WEFA Group.

    Compared with the turbulence in foreign exchange markets during late 1994 and 1995 when several emerging market currencies came under attack in the aftermath of the Mexican peso crisis, exchange markets were relatively calmer in 1996 through April 1997. In Asia, the Indian rupee, after falling early in 1996, came under strong upward pressure, then stabilized during the latter half of the year. The depreciation of the Indonesian rupiah against the U.S. dollar slowed, while the volatility of the Philippine peso continued to decline, and it remained relatively stable against the U.S. dollar. The substantial appreciation of the yen against the U.S. dollar through mid-1995 and its subsequent reversal significantly affected some of the Asian emerging market currencies. In particular, the Korean won first appreciated through late 1995 and then depreciated substantially through 1996 and into 1997. The Thai baht followed a similar pattern, though within the much smaller bands set by the Bank of Thailand (BOT). The Malaysian ringgit and the Singapore dollar are the only major Asian emerging market currencies to have appreciated against the U.S. dollar over the 1990–96 period. In Latin America, there was a marked reduction in the volatility of the Mexican new peso and the Brazilian real during 1996 through May 1997. The Mexican new peso depreciated modestly during the period, while the real depreciated steadily against the U.S. dollar. In April 1996, after the Venezuelan bolivar was floated, it depreciated by some 62 percent during the month, then remained relatively stable until a system of crawling bands was implemented in July. Elsewhere, in the transition economies, the Czech koruna was relatively stable until it came under attack in mid-May 1997 (discussed below), while the Hungarian forint and the Russian ruble continued to depreciate relatively steadily.

    During 1996 through April 1997, among the larger emerging markets, the currencies that were subject to substantial pressures were the Thai baht and the South African rand. The Thai baht was subject to periodic bouts of speculative pressure amid a host of concerns: a slowdown in exports and growth, a current account deficit at 8 percent of GDP in 1996, a buildup in short-term debt, a glut in the property sector, and weaknesses in the domestic financial system. Such bouts of speculation were often driven by the possibility that the BOT would alter the basket of currencies against which it traditionally determines the value of the baht because of changes in trading patterns. As the BOT maintained interest rates at relatively high levels to relieve pressures on the currency, this further depressed economic activity and increased pressures on the domestic financial system. This policy conundrum caused speculative pressures to erupt periodically in the belief that eventually interest rates would have to be lowered out of concern for the state of the economy, and the baht devalued.

    Market participants report that these speculative pressures were driven primarily by foreign investors. Several conditions facilitated the ability of foreign investors to speculate against the currency, relative to other emerging market currencies: specifically, Thailand maintains an open foreign exchange system, there are well-developed spot and forward foreign exchange markets, and foreign residents can obtain baht credit from domestic banks. The ability of speculators to obtain domestic currency credit, either implicitly or explicitly, is a key element in currency attacks. Speculation against the baht included directly taking positions on the forward market—selling baht forward—creating pressure on the forward rate to depreciate. When the speculator enters into a forward contract, typically with a domestic bank, the bank bears the investor’s credit risk, and the forward contract represents an implicit extension of credit. If the domestic bank enters into an offsetting transaction with the central bank to hedge its position—say, by the central bank buying baht forward—this implicit extension of credit ultimately reverts to the central bank (see Chapter IV Appendix 2). Settlement of forward sales of baht by a foreign speculator also typically involves the extension of credit. Speculation against the baht also included the use of explicit baht credits, which, when converted into foreign currency, created a short position on the baht. The conversion of baht credit into foreign currency represented a capital outflow, placing downward pressure on the spot exchange rate and, to the extent that these pressures were offset by central bank intervention, they resulted in a loss of reserves.

    In South Africa, the predominant source of pressure in the foreign exchange market between late February and early May 1996, when the rand plunged by 18 percent, was political uncertainty. While the pressure originally began with rumors about the health of President Mandela, as these rumors proved unfounded, attention focused on other political concerns. Unlike Thailand, South Africa, with a long history of capital controls on domestic residents, has much less developed spot and forward foreign exchange markets. In addition, foreign residents are not permitted to obtain rand credit from domestic banks without an underlying transaction. These features make it, in principle, more difficult to short the rand. Despite the fact that Thailand had a substantial stock of reserves, and South Africa did not, a common feature of the central banks’ defense of their currencies was intervention in the forward foreign exchange market.

    Amid widespread concerns that Japanese interest rates were likely to be raised with negative consequences for capital flows to emerging markets, and following adverse economic news, starting May 7, 1997, the Thai baht once again came under severe speculative pressure. For the first time since the contagion in the form of Tequila effects following the Mexican crisis, these pressures quickly spilled over to a number of other emerging market currencies. In Asia, the Indonesian rupiah, the Malaysian ringgit, and the Philippine peso all came under pressure. In Eastern Europe, the Czech and Slovak currencies came under attack. There were no notable immediate spillover effects to Latin America. Central bank defenses in support of the currencies included a combination of exchange market intervention, interest rate hikes, and measures aimed specifically at reducing foreign investors’ access to domestic currency credit. Interbank overnight interest rates rose to varying degrees and over differing time spans across countries: the rupiah rate rose from 14 percent on Friday, May 9, to 16 percent by Friday, May 16: the ringgit rate rose from 7 percent to 19 percent by Tuesday. May 20: the peso rate rose from 11 percent to 20 percent on Monday, May 19; koruna rates reached 200 percent on Thursday, May 22.5 On baht, the sharpest increase in interest rates was not onshore but in the offshore market, where rates shot up to 1,300 percent. The Bank of Thailand directed banks, usually the primary providers of baht, both onshore and offshore, to segment the two markets. The limitation of baht credit offshore drove up interest rates substantially more than onshore, causing speculators to settle their forward positions through the spot market, which put upward pressure on the exchange rate. Domestic banks also segmented the customer base by restricting baht lending to foreign clients, or charged them prohibitive swap rates, and stopped buying back baht-denominated commercial paper from offshore. Similar pressures were reported, though to a lesser extent, on both ringgit and rupiah offshore rates, and Malaysian and Philippine banks restricted the lending of local currency to foreign customers. The Czech National Bank limited access by nonresidents to the domestic money market.

    In Asia, market participants widely reported coordinated exchange market intervention among the Asian central banks, particularly in support of the baht, and though it was unclear as to whether the recently established network of regional bilateral repurchase agreements had been utilized, the perception that they could be appeared to deter speculation. Since there are substantial offshore markets for these currencies—in Singapore and Hong Kong, China—some of the apparently coordinated intervention by the Monetary Authority of Singapore and the Hong Kong Monetary Authority was simply on behalf of other central banks. Some market participants reported pressures on U.S. bond markets during the period as Asian and Eastern European central banks sold treasuries. While the baht withstood the pressures in May, and the pressures on the other Asian currencies abated, on May 26 the Czech National Bank abandoned its policy of maintaining the currency inside a trading band against a hard currency basket.

    The contagion of speculative pressures on emerging market currencies in May was selective. The countries to which the run on the baht spread had, in the view of investors, a number of features in common with Thailand. Within Asia, Malaysia, Indonesia, and the Philippines had all been affected by the slowdown in the region, though to varying degrees. All had current account deficits, though of a smaller magnitude than that of Thailand, and most had accumulated debt rapidly during the 1990s, though again to a lesser extent. All had undergone booms in the property sector, and all had varying degrees of financial sector fragilities. The Czech Republic shared many of these features and had perhaps even more similarities with Thailand than the affected Asian countries did. Among currencies not affected by the contagion was the Korean won, even though there were many parallels in economic circumstance with Thailand. Several observers have noted that this was perhaps because Korea’s debt levels were lower, because the substantial depreciation of the won during the last year and a half had left it at a more appropriate level, or because the recent appreciation of the yen would have greater benefits for Korea than its neighbors. While these factors may have played a role, it should be noted that, unlike the Czech Republic and the Asian economies that were attacked. Korea restricts won credit to foreign residents, and the foreign exchange markets, particularly the forward market, are undeveloped. Simply put, this makes it difficult for foreign investors to speculate against the won.

    In the wake of the volatility in emerging market currencies following the Mexican peso crisis in late 1994, a strong demand developed for products with which foreign investors could hedge exchange rate risk on emerging market investments. Such hedging has often been hindered by underdeveloped local forward and futures foreign exchange markets in these countries or by capital controls prohibiting or limiting such transactions, and this situation has led to the development of a number of products offshore. Since May 1995, futures exchanges in New York and Chicago have offered a variety of products including options on the Mexican new peso and Brazilian real futures. A particularly notable development has been the use of OTC nondeliverable forward (NDF) contracts in emerging market currencies. While markets exist for a variety of currencies in London and New York, the Asian segment of the NDF market has been particularly active. The market, which operates between banks and brokers in Singapore and Hong Kong, China, is estimated to have daily volumes of between $500 million and $800 million, and market participants expect it to continue to grow over the coming year. NDF contracts allow agents to take notional forward positions in currencies for which restrictions exist in the forward market or for which an established forward market is absent. NDFs in New Taiwan dollars, Korean won, Philippine pesos, Indian rupees, Chinese yuan, and Vietnamese dong trade actively in Singapore and Hong Kong, China. A typical contract works as follows. Counterparties establish a price for the contract at the start date. The contract is then settled at maturity based upon a rate indexed to the underlying currency. Settlement is made in U.S. dollars and no local currency is paid or received. Agents can, therefore, manage foreign exchange exposures without violating local exchange control restrictions.

    Bond Markets

    Several factors acted in concert to create a spectacular rally in emerging debt markets during 1996.6 These included, first and perhaps foremost, the low-yield environment in the mature markets. While interest rates remained at low levels in Japan and Germany, there was a compression of spreads on the U.S. corporate bond market as improved business prospects lowered perceived corporate credit risk. This spurred fixed-income investors from the mature markets to search for higher yields on emerging debt markets during 1996 and into 1997. Second, improvements in underlying fundamentals in many emerging markets resulted in both formal upgrades of sovereign credit ratings and in perceptions of reduced credit risks. Third, the continued diversification of the portfolios of institutional investors from the mature markets into the emerging markets boosted the ongoing process of securitization in international capital markets. Fourth, Japanese and European retail interest in emerging market debt continued to be sustained at high levels.

    The coincidence of these factors interacted to reinforce interest in emerging debt markets. First, as lower perceived credit risks narrowed spreads, one of the ways investors sought to pick up yield was to seek out longer-maturity issues. This favorable environment prompted several sovereign borrowers to launch new issues to restructure existing liabilities at improved terms and reduce refinancing risk by extending the maturity profile of their external debt. This further lowered perceived credit risks, reinforcing demand and narrowing spreads. By creating more comprehensive yield curves for emerging market debt, the new, longer-term sovereign issues improved the ability of international investors to manage, diversify, and hedge their exposures, enhancing the desirability of these instruments. These issues also set benchmarks for domestic corporate bonds, thereby increasing the access of these entities to international bond markets. Second, the decline in spreads also led investors to move down the credit spectrum in search of higher yields, facilitating the entrance of several new—that is, first-time—borrowers, both sovereign and corporate, hence increasing the size and breadth of the market. Finally, increased investor interest was associated with dramatic improvements in the liquidity of emerging debt markets, enhancing the attractiveness of these instruments.

    While the liquidity of emerging debt markets improved substantially during 1996, two characteristics suggest lingering market imperfections. First, yield spread differentials between the Brady and Eurohond sectors endured, suggesting continued market segmentation. Second, the dramatic decline in emerging market spreads in an environment of low interest rates in the mature markets raised questions about whether the compression of spreads had been excessive.

    Secondary Markets

    Spreads and Returns

    Spreads on emerging market debt, which have been declining since the peak reached in the spring of 1995 in the aftermath of the Mexican crisis, continued to decline during 1996 (Figure 24).7 Sovereign yield spreads, for example, in the J.P. Morgan Emerging Market Bond Index (EMBI), fell from their peak of 1752 basis points in March 1995 to 1044 basis points at the end of 1995, then to 537 basis points by the end of December 1996.8 Total returns on the EMBI rose from a robust 27 percent in 1995, to 34 percent during 1996. These returns were in contrast to sharp declines in returns in the mature markets, with returns on the J.P. Morgan Government Bond index for the United States (GBI) dropping to 3.4 percent in 1996 from 17 percent in 1995, and returns on the Merrill Lynch High-Yield (MLHY) index of U.S. corporate bonds dropping to 11 percent from 20 percent.9

    Figure 24.Bond Markets: Selected Returns, Yields, and Spreads

    Source: Bloomberg Financial Markets L.P.

    In early 1997, emerging market spreads continued to decline, falling by the third week of February to about their previous historic lows of around 400 basis points, last reached in late 1993 and early 1994. As spreads had last reached their historic lows in the period preceding the run-up in U.S. interest rates during 1994 that ushered in the Mexican crisis, these levels gave rise to concerns that yields had reached their lower limits in adequately compensating for risk. Spreads then fluctuated, widening at first, then narrowing again, and returns on the EMBI dropped off to 2.6 percent during the first quarter of 1997, though they continued to exceed those of the GBI, with losses of 1.1 percent, and the MLHY, with returns of 1.4 percent. Starting in the last week of February 1997 there was a sharp correction, and by the time the U.S. federal funds rate was raised in the third week of March, emerging market spreads had risen by around 60 basis points. Following the 25 basis point hike in the U.S. federal funds rate, emerging market spreads widened by an additional 60 basis points through mid-April, having risen a total of 120 basis points over a two-month period. Spreads then fell by about 75 basis points through May 1997, to around 450 basis points, some 50 basis points above their historical lows. (Factors driving the compression of emerging market spreads over the recent period are discussed at the end of this section.)

    Figure 25 shows that the decline in spreads on emerging market debt during 1996 and early 1997 and the subsequent correction were, with few exceptions, across the board. In the Brady market, the decline in stripped yield spreads for each of the major Latin countries brought them below precrisis levels by mid-1996. While Mexico had enjoyed a spread substantially below the other major Latin countries prior to the crisis, it has not done so since. Bulgaria was a notable exception to the broad-based decline in spreads during 1996, with the stripped yield spread on its Bradys widening early in the year, and then declining sharply in early 1997 with the announcement of plans to proceed with a currency board. On the secondary market for Eurobonds, Hungarian spreads fell by more than 100 basis points over the period, to 70 basis points, while those for China fell by 50 basis points.

    Figure 25.Yield Spreads for Selected Brady Bonds and U.S. Dollar-Denominated Eurobonds1

    (In basis points)

    Sources: Bloomberg Financial Markets L.P.; Salomon Brothers; and IMF staff estimates.

    1Yield spreads on Brady bonds are “stripped” yields.

    2Latin America: Republic of Argentina bond due 12/03 and United Mexican States bond due 9/02.

    3Other: National Bank of Hungary bond due 6/98 and People’s Republic of China bond due 11/03.

    Reflecting both changes in perceptions of credit risk and the relatively lower liquidity of emerging debt markets, returns on emerging market debt have been considerably more volatile than those on mature market debt (Figure 26, top panel). The volatility of returns on the EMBI rose steadily, from about 1 percent in early 1993 through the Mexican crisis, peaking in mid-1995 at 3 percent. Volatility has since declined steadily, falling by May 1997 to 1.5 percent. The close correspondence between the level and volatility of spreads—both rising and falling together—indicates that while the rise in yields during 1994–95 and the subsequent period of turnaround tended to be erratic, suggesting increased uncertainty of credit risk, the subsequent decline in spreads was accompanied by diminishing uncertainty.10 Despite the reversals in yields during the early part of 1997, volatility continued to diminish. Throughout the period, the volatility of returns on the GBI and the MLHY have remained relatively stable around 0.5 percent and 0.4 percent, respectively.

    Figure 26.Emerging Market Debt: Volatility and Correlation of Returns with Mature Markets

    Sources: Bloomberg Financial Markets L.P.; and IMF staff estimates.

    1Computed as the standard deviation of weekly changes in (the logarithm of) the total return index over the preceding year.

    An important consideration for investors from the mature markets in emerging market debt is the gains from diversification of their portfolios. These gains depend on the correlation of returns between the emerging and mature markets. The bottom panel of Figure 26 presents the correlation of returns on the EMBI and the mature markets.11 It shows that after a low in early 1995 following the Mexican crisis, returns on emerging market debt and both U.S. treasuries and high-yield U.S. corporate bonds have tended to be highly positively correlated, with the correlation of returns recently reaching almost 0.8. This suggests that the benefits of diversification among the emerging and mature debt markets have been diminishing.


    The surge of investor interest combined with the growing volume of new issuance resulted in a tremendous growth of trading in all types of emerging market debt instruments and derivatives. After remaining unchanged in 1995, transactions in emerging market debt instruments increased by 93 percent, to $5,296 billion in 1996 (Table 15).12 Brady bonds remained the most traded instrument, with transactions increasing by 70 percent to $2,686 billion. Despite the sharp increases in turnover of Brady bonds, their share in total trading has continued to decline steadily, falling from 61 percent in 1994 to 58 percent in 1995, and 51 percent in 1996. Similarly, the share of loans traded in the market has continued to decline as the stock of loans traded has fallen, and in 1996 accounted for 4 percent of activity, compared with a third of all activity in 1992. These declines have been offset by robust increases in the trading of corporate and non-Brady sovereign bonds, of local market instruments, and of derivative instruments on emerging market debt. With the rapid increase in primary issuance, turnover of corporate and non-Brady sovereign bonds rose almost threefold to $658.1 billion, and their market share expanded from 8.5 percent in 1995 to 12 percent in 1996. Trading of local emerging market instruments doubled during 1996 to $1,187.9 billion.13 The steady growth of this segment of the market, which in 1996 accounted for 22 percent, is a particularly significant development since it suggests the increasing acceptance of emerging market debt into the mainstream. Reflecting the growing maturity of the secondary market for emerging market debt, trading in options and warrants also continued to grow rapidly, contributing 9 percent to total turnover in 1996.

    Table 15.Secondary Market Transactions in Debt Instruments of Emerging Markets(In billions of U.S. dollars)
    Total turnover1,978.92,766.22,738.85,296.91,620.6
    By region
    Eastern Europe104.5172.3314.2612.7161.5
    Middle Rust2.82.65.321.26.6
    Western Hemisphere1,621.62,259.32,284.24,265.91,366.8
    By instrument
    Brady bonds1,021.31,684.01,580.12,686.0671.4
    Corporate and non-Brady sovereign bonds176.6164.9233.3658.1361.4
    Local market instruments1361.9518.9571.11,187.9427.4
    Options and warrants on debt57.4142.4179.2471.090.8
    Un specified88.111.645.30.7
    Source: Emerging Markets Traders Association.

    Trading in Latin American instruments continued to dominate the market, accounting for 80 percent of total trading in 1996, while the debt of four countries—Brazil. Argentina. Mexico, and Venezuela—accounted for 77 percent of all trading activity, Brazilian instruments were the most commonly traded ($1,441 billion—27 percent), followed closely by Argentine ($1,292 billion—25 percent) and Mexican ($946 billion—18 percent) instruments. There was a significant increase in the volume of trading in several Asian instruments. In particular, trading in Indonesian. Thai, and Malaysian debt totaled $75 billion, compared with negligible amounts in 1995. In other segments, there was also tremendous growth in volumes of South African instruments, which rose fourfold to $170 billion, and in Russian instruments, which grew by 160 percent to $380 billion.

    Trading volumes surged again in the first quarter of 1997, reaching $1,620 billion, with all of the major trends evident in 1996 continuing. As increases in turnover have exceeded new issuance, there have been substantial improvements in the liquidity of emerging market debt instruments. The turnover on Brady bonds, for example, increased from $1,580 billion in 1995 to $2,686 billion in 1996, while the outstanding stock of Brady bonds increased only modestly over the period, from around $148 billion to $156 billion (these figures should be viewed as only broadly indicative as there are wide disparities in turnover across Brady bonds). This implies that, on average, each dollar face value of Brady bonds turned over 17 times in 1996 compared with 11 times in 1993. While such improvements in liquidity helped further stimulate demand for the instruments, these markets remain small relative to the mature debt markets. The U.S. government bond market, for example, with a marketable stock of around $3.5 trillion, is estimated to have a daily turnover of $200 billion—more than the entire stock of outstanding Brady bonds. As discussed below, the small size of emerging debt markets, and the potential ability of large trades to move the market, have contributed to inefficiencies and arbitrage opportunities between segments of the markets. In particular, persistent spread differentials between Brady bonds and Eurobonds with equivalent sovereign risk have raised questions as to whether these securities are priced appropriately and why these differentials have not been arbitraged. Since several countries have recently retired their Brady bonds, and others are expected to do so in the near future, this has raised further concerns about the size and liquidity of emerging debt markets as the stock of Brady bonds diminishes (see Box 2).

    Box 2.The Brady Bond Market Comes of Age

    Since the first restructuring of Mexico’s defaulted sovereign loans into Brady bonds in 1990, the Brady market has grown to become the largest and most liquid emerging debt market. The investor base, composed originally of commercial and investment banks, gradually widened to include mutual funds, insurance companies, and other institutional investors. The number of distinct issuers and diverse characteristics of the different classes of Brady bonds—fixed- and floating-rate, collateralized and uncollateralized—and more recently the availability of derivatives, facilitated a rich set of sovereign and interest rate investment strategies. However, seven years after Mexico turned its defaulted sovereign loans into the first Bradys, some market participants are forecasting a rapid demise of the market. With the conclusion of a debt restructuring deal in March 1997 for Peru, the stock of outstanding dollar-denominated Brady bonds reached a peak of around $156 billion and has been declining following a series of buybacks and exchanges for uncollateralized global and Eurobonds. Côte d’voire and Vietnam are expected to be the last significant entrants to the market, but their additions to the stock of Bradys is unlikely to offset the amounts recently retired by Brazil, Ecuador, Panama, and Poland.

    As in previous Brady deals. Peru’s debt restructuring operation offered a menu of options to creditors, with the government repurchasing $2.6 billion of principal and past-due interest and issuing $4.8 billion of Brady bonds. Creditor preferences determined the issuance of $2.4 billion in past-due interest bonds (PDIs), $1.7 billion in front-loaded interest reduction bonds (FLIRBs), $560 million in discount bonds, and $182 million in par bonds. The PDI bonds and FLIRBs carry below-market interest rates for the first 10 years, paying LIBOR plus ½ percent thereafter, and have a graduated amortization schedule to maturity in 2017. The discount and the par bonds are collateralized and mature in 2027.

    Improved conditions in emerging debt markets following the sustained rally since The Mexican crisis have led several countries to buy back and/or exchange their outstanding Brady bonds, mainly the collateralized instruments, at significantly lower spreads. Following the high-profile exchange in April 1996, Mexico used the proceeds of a 20-year global bond to retire $1.2 billion of discount bonds in September and called the remaining $1.1 billion of Aztec bonds in early 1997. In a deal that mimicked the Mexican swap, the Philippines exchanged one-third of its par bonds for a $690 million 20-year uncollateralized Eurobond in September 1996. The exchange freed up $183 million of collateral in U.S. treasury bonds. Ecuador, Panama, and Poland also followed this strategy and bought back some $250 million, $600 million, and $1.7 billion of Brady bonds, respectively. More recently, Brazil—the largest Brady country, with almost $50 billion in bonds outstanding—exchanged $2.7 billion of Brady bonds for a 30-year uncollateralized global bond.

    Market Segmentation: The Brady-Eurobond Differential

    Figure 25 suggests that yields on Eurobonds have typically been lower than on Brady bonds. (Note the differences in scale for the Latin Brady bonds and Eurobonds.) The reported differential, however, reflects in part the fact that these bonds have different maturities, and the yields are not directly comparable. Moreover, many Brady bonds are partially collateralized by U.S. treasury discount bonds, while the more recently issued Eurobonds are not. The yield spreads reported on those Bradys that are collateralized are “stripped” yields, that is, yields after the value of the collateral has been subtracted from the value of the bond.14 Since the Bradys and Eurobonds have very different cash-flow patterns, rather than comparing yields and maturities. Figure 27 compares yields relative to duration.15 It is apparent that spreads on Brady bonds exceed those on Eurobonds, and the differentials can be substantial. On the date shown (April 11, 1997), for example, the Argentine floating-rate bond was some 175 basis points above the comparable Eurobond, while the par bond was some 285 basis points above. The yield differential between the Mexican par and discount Bradys and the Eurobond yield spreads was some 210 basis points.

    Figure 27.Brady and Eurobond Spreads1

    (In basis points)

    Source: J.P. Morgan.

    1As on April 11, 1997.

    2Restructured bonds; Brady Par = Brady par bonds due 2023; Brady FRB = floating rate bond due 2005; Pre 2 = U.S. Pensioner I due 2001; and Bole 10 = 10-year U.S. dollar-denominated domestic government debt due 2000. Eurobonds; Rep 99, etc. = Republic of Argentina bond due 1999: and Rep FRN = floating-rate note due 1999.

    3Brady bonds: Par = par bond and Disc = discount bond due 2019. Eurobonds: UMS 01, etc, = United Mexican States bond due 2001; and BNCE 04, etc. = Banco Nacional de Comercio Exterior bond due 2004.

    Market participants have offered various explanations for the persistence of these yield differentials. Many of the explanations put forward are unconvincing. Some suggest a lack of investor sophistication, and some a lack of liquidity. First, it has been argued that since Brady bonds represent restructured loans, they carry the stigma of prior defaults, whereas Eurobonds are original-issue debt. For such aspects of debt to affect yield requires that investors perceive that there is a greater risk of default on the Brady bonds than on Eurobonds, despite the fact that rating agencies assign identical ratings to sovereigns’ Brady and other foreign currency debt,16 Second, the actual “stripping” of the Brady bonds of their collateral to earn the stripped yield—which requires shorting the collateral, U.S. treasury discount bonds, in a portfolio—entails costs. Market participants place these costs at 40–80 basis points, and they cannot, therefore, provide a complete explanation of the 175–300 basis point differentials. Third, the unusual cash-flow patterns—such as be low-market coupons—of Brady bonds may have prompted investors to demand higher yields on Brady bonds. Durations are employed in Figure 27 precisely to make different cash flows comparable. Fourth, since many of the Eurobonds are bearer securities, some investors may be willing to pay a premium—give up yield—for anonymity that allows them to forgo registering the securities. Fifth, Eurobonds have lower volatilities than Brady bonds, and so investors may require a lower yield. While this may be true, it may reflect the fact that Eurobonds trade less frequently than Bradys, The relative magnitudes of turnover discussed above suggest that the Bradys are more liquid than Eurobonds. Sixth, all Brady bonds are callable at par while most of the more recently issued Eurobonds are not. As the prices of emerging market debt have risen rapidly over the past two years with, for example, the Mexican discount bond trading in the low 90s recently, the value of the call feature on Brady bonds has become a consideration. (The value of the call option is also a consideration across different Bradys—with, for example, the Mexican par bond trading recently in the high 70s compared with the discount bond trading recently in the 90s.) For most of the period since their inception, however, the call option has been so far out-of-the-money that its value has been insignificant.17 Finally, carrying out an arbitrage trade of buying Bradys and selling Eurobonds, which requires carrying out a repo, is expensive. As the size of particular Eurobond issues has been relatively small, and some are often traded infrequently, many of the Eurobonds trade “special” in the repo market, rendering arbitrage prohibitively expensive. Any sizable transaction would, therefore, he likely to move the market.

    Primary Issues

    The rally in emerging debt markets provoked a sharp shift in the structure of external financing for emerging markets toward bond issuance. International bond placements by emerging market entities soared in 1996 to $102 billion, far exceeding the previous record of $63 billion in 1993 (Table 16 and Figure 12). Issuance from all regions—with the exception of Africa—rose sharply. The increase was particularly marked for Latin America, where issuance more than doubled, from $23.1 billion in 1995 to $47.2 billion in 1996, while its share in total issuance rose from 40 percent to 47 percent. The three largest borrowers-Mexico, Argentina, and Brazil—raised $18 billion, $14 billion, and $11 billion, respectively, accounting for over 90 percent of issuance from the region. Asian issuance rose from $25.3 billion in 1995 to $43.1 billion in 1996, while the share in total issuance remained steady at 43 percent. Following the liberalization of external borrowing restrictions, Korean entities were the most active issuers, raising $16 billion, accounting for 38 percent of issues from the region, while Hong Kong, China, Indonesia, and Thailand each raised about $4 billion. While issuance from the European emerging markets rose, their share in total issuance fell to 7 percent, as increases from Russia and Romania were offset by declines from Hungary. Issuance from the Middle East rose to account for some 3 percent of the total, while that from Africa continued to decline, accounting for 1½ percent.

    Table 16.Emerging Market Bond Issues, Equity Issues, and Syndicated Loan Commitments(In millions of U.S. dollars)
    Bond issues1
    Emerging markets7,78913,94524,39462,67256,54057,619101,92627,723
    Middle East040002,0522,9937102,570275
    Western Hemisphere2,8507,08512,93328,79417,99023,07147,15711,876
    Equity issues
    Emerging markets1,1665,5747,24711,91518,03811,19316,4143,203
    Middle East705062813368925689493
    Western Hemisphere983,8913,8766,0224,6049623,66180
    Syndicated loan commitments
    Emerging markers28,37750,66942,48843,01555,15674,93379,73721,367
    Middle East1,08912,1193,6022,4566,9957,3235,8361,436
    Western Hemisphere3,4806,2556,8517,6183,7299,76810,2973,888
    Short-term commitments2
    Emerging markets4,4235,2478,24511,86814,31221,56530,4587,352
    Middle East3504436925521400
    Western Hemisphere1001,7704,3554,7784,2953,4873,248390
    Emerging markets41,75575,43582,374129,470144,046165,310228,53559,645
    Middle East1,50913,0694,2524,86910,1298,3039,3001,804
    Western Hemisphere6,52819,00128,01547,21230,61837,28864,36316,234
    Sources: Capital Data Bondware and Loanware.

    Sovereign issuance, which had risen sharply in 1995 in the aftermath of the Mexican crisis, as private sector issuance fell off, continued to rise, accounting for 37 percent of issuance in 1996 and 40 percent in the first quarter of 1997. The growth of sovereign issuance was driven by, as noted earlier, the restructuring of previous liabilities at improved terms, the entrance of several new sovereign credits, and the establishment of benchmarks for domestic corporate bonds. In a number of headline deals, several sovereigns bought back or swapped existing Brady bonds for uncollateralized Eurobonds with lower sovereign risk spreads and relatively long maturities, and freed up the collateral on the Brady bonds. With strong demand for the new Eurobond issues, many of the placements were heavily oversubscribed. In September 1996, Mexico followed up its April exchange of Brady bonds for $1.8 billion uncollateralized dollar-denominated 30-year global bonds, by using the proceeds of a $1 billion issue to retire a further $1.2 billion of discount bonds. In the same month, the Philippines launched a $690 million 20-year bond in exchange for Brady bonds. The Mexican and Philippine transactions were widely viewed as forerunners for other such exchanges and as heralding the eventual demise of the Brady market. More recently, in June 1997. Brazil swapped a 30-year $3 billion global bond for $2.7 billion (face value) of Brady bonds and $750 million of new money. Among the newly rated sovereign credits, in November 1996, Russia placed a $1 billion five-year issue, its first since 1917 and the largest-ever debut issue by an emerging market sovereign.

    There was a broad-based improvement in the terms of issuance for borrowers (Figure 28 and Table 17). Spreads declined, maturities lengthened, the proportion of fixed-rate issues increased, and the proportion of callable bonds rose. While the average spread for unenhanced dollar issues actually increased from 218 basis points in 1995 to 244 basis points in 1996, this reflected higher average spreads on private sector issues as the spectrum of borrowing entities expanded and maturities lengthened. For sovereign issues, spreads declined from an average of 383 basis points in 1995 to 307 basis points in 1996. There was an impressive lengthening of yield curves as several issuers placed 10-, 20-, 30-, and even some 100-year bonds. During 1996 and the first quarter of 1997, some 175 issues, or almost a quarter of all issues by emerging market entities, had maturities of 10 or more years, while six entities issued 100-year bonds.18 The average maturity of new issues in the dollar sector rose from 6.6 years in 1995 to7.7 years in 1996, while sovereign maturities, after having shortened to 4.5 years in 1995, rose dramatically to 9.5 years during 1996. The search for higher yields also appeared to shift issuance in favor of fixed-rate issues, as investors traded yield for interest rate risk. The proportion of fixed-rate issues rose from 67 percent in 1995 to 70 percent in 1996, and to 72 percent in the first quarter of 1997. Similarly, bonds with call options, offering a pickup in yield for the risk of the call being exercised, rose from 13 percent in 1995 to 21 percent in 1996.

    Figure 28.Spreads and Maturities for Sovereign Borrowers1

    Source: Capital Data Bondware.

    1Unenhanced U.S. dollar-denominated bonds.

    Table 17.Emerging Market Bond Issues: Fixed-Rate, Floating-Rate, and Call Options1(In percent)
    Fixed-rate issues us a percentage of total issues
    Number of issues61.348.969.567.743.452.357.454.8
    Issue amount69.464.575.074.248.367.169.673.6
    Floating-rate issues as a percentage of total issues
    Number of issues13.313.510.810.829.732.727.631.5
    Issue amount8.
    Callable as a percentage of total issues
    Number of issues20.030.811.218.530.425.127.821.0
    Issue amount11.318.48.515.525.112.821.113.7
    Source: Capital Data Bondware.

    The U.S. dollar has traditionally been the primary currency in which international issues of emerging market debt have been denominated, accounting for some 70 percent during 1990–94 (Table 18). A remarkable change in the last two years has been the growing diversity of currencies of issuance. The currency sectors targeted by issuers have not always corresponded to the pattern of their export earnings, as issuers have sought to take advantage of pockets of strong local investor—often retail—interest, sometimes tailoring issues to investor preferences, and suggesting that primary markets for emerging market debt remain segmented. The spate of issues in a host of currencies has continued to raise concerns about pricing in nontraditional sectors—that investors may be underpricing credit risk and issuers underestimating exchange rate risk. Little information is available on the extent to which the proceeds from emerging market borrowing have been swapped into currencies matching their export earnings patterns and, therefore, on their exposure to exchange rate movements among the major currencies.

    Table 18.Emerging Market International Bond Issues by Currency of Denomination(In percent)
    Share in total issues by emerging markets
    U.S. dollar6068707475576968
    Deutsche mark2612973101116
    U.S. dollar6562708172707983
    Deutsche mark125545181720
    Western Heinisphere
    U.S. dollar8588928889566463
    Deutsche mark125545181720
    U.S. dollar2411221231184421
    Deutsche mark6049213412142958
    Share in total issues in global bond markets
    U.S. dollar3531373941404343
    Deutsche mark771011714119
    Sources: Capital Data Bondware; and IMF staff estimates.

    Following the surge in yen issuance after the liberalization of rating requirements on the Samurai market in 1995, the value of yen issuance (in yen) remained steady in 1996, though the share in total issuance (measured in dollars) fell from 26 percent to 14 percent. Retail investor demand in Japan remained strong and is estimated to have accounted for about half of the purchases. Emerging market issues also continued to receive positive receptions in the deutsche mark sector, and the share of deutsche mark issuance remained steady at around 10 percent in 1996. Strong retail investor interest was evidenced by the successful repackaging by investment banks of outstanding Latin American Brady bonds and other bonds into deutsche mark—denominated bonds that were predominantly sold to German retail investors (see Box 3). Although the proportion of “other” currencies of issuance remained unchanged, there was a shift to new and largely untapped currency sectors as borrowers sought new niches. The lira sector was particularly active, with an unprecedented $3.7 billion of issuance since the beginning of 1996, with large sovereign issues by Argentina and Mexico. With the decline of Italian interest rates prompting a strong interest in these issues, the authorities limited emerging market issues in the sector to two a month. Moreover, there were a number of issues in French francs, Dutch guilders, Australian dollars, and pounds sterling.

    Box 3.Repackaged Brady Bonds

    “Repackaged” or “synthetic” Brady bonds are structured asset-backed securities in which the underlying asset is a portfolio of Brady bonds and the structure is provided by a credit derivative providing for a reduction, or suspension, of payment if a credit event involving the issuer of the Brady bond occurs (see Appendix 1, “Credit Derivatives,” to Annex III). These credit-linked notes are issued by an offshore trust or special purpose vehicle that holds the underlying Brady bonds, usually with a significant degree of over collateralization. Most repackaged Brady bonds are sold to retail investors in Germany and are denominated in deutsche mark at fixed interest rates, so the issuer will, if necessary, swap the income from the Brady bonds into fixed-rate deutsche mark. Hence, the investor acquires a hedged exposure to emerging market credit that earns a significant premium over German government bonds.

    The first public repackagings of Brady bonds in 1992 involved Venezuelan Debt Conversion Bonds, but the market for Brady repackagings really only developed in 1996—in 1993–94 there had been a large number of repackagings of Mexican tesobonos and some other non-Brady debt. Since 1992 there have been at least 76 public repackagings of emerging market debt with a total value of $6.6 billion. Repackaged Brady bonds have accounted for $2.1 billion—most of the remainder was composed of Brazil Multi-Year Deposit Facility Agreement bonds ($1.3 billion) and repackaged Mexican tesobonos ($1.1 billion). The most common sovereign risks identified in the repackagings were Brazil ($2 billion in repackaged bonds), Mexico ($1.2 billion). Venezuela ($737 million), and Argentina ($513 million). Other countries whose bonds have been repackaged include Ecuador, Mexico, Russia, and Turkey.

    These bonds provide a means of arbitraging yield differentials between different investor bases—a comparatively high demand in Germany for deutsche mark-denominated emerging market credit—and between different classes of bonds (Eurobonds versus Bradys). However, if such transactions increase in popularity, credit-linked bonds may have a detrimental effect on liquidity in the markets for emerging market debt—since the Brady bonds are stored in trusts and replaced by relatively illiquid Eurobonds. Also, credit-linked bonds are issued by private firms but provide exposure to sovereign credit risk, and therefore compete for investor interest against new sovereign debt, possibly increasing the borrowing costs for emerging market issuers.

    The share of dollar issuance by all emerging market borrowers rose from 57 percent in 1995 to 69 percent in 1996—about its level during 1990–94. While the share of total dollar issuance returned to its historical level, there remained some notable shifts in currency composition across regions. The most notable shift has been the increase of the non-dollar segment for Latin America, where dollar issues, which had accounted for almost 90 percent of issuance during 1990–94, represented only 64 percent during 1996. The proportion of dollar issuance by Asian entities on the other hand was some 10 percentage points higher in 1996 than it has been historically.

    International issuance of emerging market debt has traditionally been in the major convertible currencies, requiring issuers to bear or manage the inherent exchange rate risk. While investment in domestic currency debt has represented a viable alternative to foreign investors in many emerging debt markets, this channel has been relatively limited.19 By avoiding domestic currency instruments and thus avoiding exchange rate risk, however, given the typically higher interest rates in emerging markets, they sacrificed yield. With investors searching for higher yields, and in another sign of the coming of age of emerging market debt as an international asset class, the last two years have seen the international issuance of debt—by entities from both the emerging and mature markets—in previously untapped emerging market currency sectors. Many of these markets were created by inaugural issues by supranational (including the European Bank for Reconstruction and Development, the European Investment Bank, the Inter-American Development Bank, the International Bank for Reconstruction and Development, and the International Finance Corporation), and some have grown rapidly to include sovereign, bank, and corporate borrowers. Since the development of these sectors, many supranational have continued to find them attractive sources of funding, tapping them repeatedly as investors have traded higher yields for emerging market currency risk.

    Among the Asian currencies, the New Taiwan dollar, the Philippine peso, and the Korean won sectors have been quite small, with issuance of under $1 billion each. In Eastern Europe, the Polish zloty and the Slovak koruna have also featured among the smaller markets. The Czech koruna market, on the other hand, has been very popular, growing quickly to reach $3.4 billion. The market has largely been tapped by supranational through short-dated bonds at high yields with the supranationals accounting for about a third of issuance, and Austrian, German, and Dutch banks accounting for the remainder. The Argentine peso ($650 million) market generated considerable interest following the issue in January 1997 of a 10-year Arg$500 million bond that focused attention on relatively cheap peso debt and extended the local currency yield curve from 2 to 10 years. The most active emerging market currency sector has been the South African rand, growing from its inception in 1995 to $17.9 billion of issuance by end-May 1997 (see Box 4).

    Box 4.Emerging Market Currency Eurobonds: The Eurorand Market

    Recently, offshore issuance and trading in South African rand-denominated debt have grown rapidly. Following its inception in September 1995, issuance activity in the Eurorand market remained relatively modest, with issuance of around $1 billion annually in 1995–96. During the first half of 1997, issuance surged to over $15 billion, while the yield curve was extended out to first 10 and then 30 years. The sector has been particularly popular with supranational issuers, who have accounted for about half of the issues. While a wide range of other entities have been active in the sector, including international banks and corporations, these have been almost exclusively from the mature markets. By May 1997, only two South African entities had tapped the sector. It is estimated that less than 10 percent of the funds raised in the sector have been for use in South Africa.

    For investors, the attraction of Eurorand debt has been the combination of high yields and highly rated issuers. Investors have, therefore, been able to earn rand interest rates, but at lower perceived credit risks than if they invested directly in South Africa, where entities are bound by the sovereign ceiling. By permitting the separation of exchange rate and (sovereign) credit risk, Eurorand debt has been extremely popular with retail investors—particularly in Europe—willing to accept rand exchange rate risk but preferring the lower credit risk of an investment grade issuer, and with institutional investors bound by fund management rules to investment grade issues. The attraction for issuers to the Eurorand market has been the low cost of funding. By offering rand exposure without sovereign risk, triple-A-rated issuers have been able to price primary deals typically some 75 basis points below the South African gilts yield curve. The fact that investors have been willing to accept lower yields from the more highly rated issuers than is available on South African gilts has created a yield gap that has allowed issuers to swap the proceeds with, for example, a highly rated international investment bank or a South African counterparty, to obtain dollar funding rates of 35–40 basis points below LIBOR.

    The fact that a majority of the funds raised have not been intended for use in South Africa raises the question of what effect the Eurorand market has on capital flows to South Africa and on the value of the rand. In the first instance, when the rand required for purchase of a Eurorand issue is obtained on the domestic spot market by surrendering dollars, there is a capital inflow into South Africa. There are then a variety of possibilities, and the net effect could be neutral or positive for capital flows and foreign exchange markets. First, the issuer could exchange the rand raised for dollars on the domestic spot market, implying a capital outflow dial offsets the original inflow, and the net effect is zero. Second, the issuer could invest the proceeds in South Africa. In this case there is a net capital inflow equal to the value of the issue. Third, after exchanging the proceeds into dollars, the issuer could enter into a swap to buy rand forward from a domestic South African counterparty, and this would reduce pressure on the forward rand exchange rate. If the issuer enters into a swap with an international investment bank, the investment bank in turn could hedge its risk by, for example, making a leveraged purchase of gills. In this case, there would be some net inflow, but less than the value of issue.

    The favorable environment facing potential issuers in emerging debt markets caused a frenzy of first-time ratings by the major international credit-rating agencies. During 1996, 16 sovereigns, primarily from Eastern Europe and the Middle East (7 from each), were assigned ratings by at least one of the two major international credit-rating agencies. By comparison, only five new sovereigns were assigned ratings in 1995. By May 1997, an additional five countries had received ratings. The number of countries with sovereign ratings from at least one of the major ratings agencies has grown rapidly from 11 in 1989 to 58 in 1996. In addition to the increase in the number of sovereigns rated, there has been a rapid increase in the number of nonsovereign entities in emerging markets that have been rated. This latest round of ratings has more or less completed the waves of regional ratings through the major emerging market regions, again with the exception of Africa.20 The fact that a majority of the new ratings have been in the investment grade category, coupled with upgrades of existing ratings over the last few years, has resulted in a steadily increasing proportion of investment grade emerging market sovereigns. Compared with 44 percent in 1993, the proportion had risen to 55 percent by May 1997 (based on Moody’s ratings). Systematic regional differences among the emerging markets persist, how-ever, as a majority of the Asian emerging markets are rated investment grade, while in Latin America, though ratings have been improving, the majority are still rated below investment grade. The European emerging markets are somewhat evenly split.

    Spread Compression in Emerging Debt Markets

    Numerous market participants and observers have argued that the low level of interest rates in the mature markets caused investors to substitute risk for increased returns, and that the resulting increase in demand pushed down spreads on riskier securities. This is evidenced, it is argued, not only by the decline in spreads on emerging market debt instruments but also on other risky assets, such as U.S. corporate bonds. Several market participants have also argued that spreads on sovereign emerging market debt instruments can be expected to decline further, to levels comparable with similarly rated U.S. corporate bonds.

    Looking back at Figure 24, it is apparent that sovereign spreads moved closely in line with U.S. interest rates from early 1993 through early 1996, albeit with a lag of three to four months, suggesting that the level of interest rates in the mature markets has played a role in affecting spreads. The historic low in emerging market spreads reached in January 1994 closely followed the bottoming out of U.S. interest rates in October 1993, while the peak in spreads in March 1995 followed the peak in the U.S. interest rate cycle in November 1994. Subsequently, as U.S. interest rates continued to decline through 1995, sovereign spreads also fell, but then continued doing so even as U.S. rates began to rise in February 1996 and then remained at a relatively higher level. The behavior of high-yield U.S. corporate bond spreads, on the other hand, appears to have been relatively independent of the behavior of the level of U.S. interest rates. What is evident is a persistent and substantial decline in MLHY yields over the period, from around 6 percent in early 1992 to 2.8 percent in May 1997, while U.S. interest rates fluctuated considerably over the period.

    The differential between sovereign spreads and high-yield U.S. corporate spreads has fluctuated considerably over the period. At the start of 1992, for instance, average yield spreads on the EMBI and the MLHY were equal. Then, after moving apart in early 1992, they converged briefly again in December 1993, before diverging substantially as sovereign spreads widened in the period leading up to and following the Mexican crisis. As both yields fell during 1996, with sovereign spreads declining by more, the differential has once again narrowed, though sovereign yields remained 140 basis points above high-yield U.S. corporate spreads in May 1997.

    In a liquid and efficient bond market, such as the U.S. corporate bond market, yields on (straight) corporate bonds relative to a (credit) risk-free benchmark of the same duration should depend on premiums for default risk, market risk, and the correlation of returns with other assets (the market portfolio). Therefore, unless movements in the level of the (credit) risk-free interest rate directly have an impact on the creditworthiness of corporate borrowers, on volatility, or on their correlation with returns on other assets, there is no reason to expect changes in corporate bond spreads to be related to the general level of interest rates. There are a number of features of emerging debt markets, however, that suggest that in contrast to the U.S. corporate bond market, investors would demand additional premiums for holding these instruments. These include the liquidity problems discussed above which, for example, resulted in persistent yield differentials between Bradys and Eurobonds, incomplete yield curves in emerging market debt instruments, and the lack of a well-developed or liquid derivatives market. Spreads on emerging market debt can in general, therefore, be thought of as representing a combination of premiums for default risk; market risk from volatility due in part to changes in perceptions of default risk: the correlation of returns with other assets; liquidity problems; and the availability—or lack thereof—of related instruments that could be used to manage or hedge market risks from the particular instrument.

    Several factors suggest that there are reasons for each of these components to have declined recently. First, improved economic performance in many emerging market countries, including improved prospects for growth, particularly in Latin America, and declines in inflation, lowered perceptions of default risk. The tremendous increase in emerging market holdings of reserves has probably also contributed to declining investor perceptions of default risk on external borrowings. Second, the decline in interest rates in the mature markets directly lowered the required servicing of existing external floating-rate debt for emerging market borrowers, thus lowering overall debt-servicing requirements and improving creditworthiness. In addition, as noted above, the favorable external financing environment allowed many emerging market borrowers to restructure existing liabilities at improved terms, lengthen the maturity profile of external debt, and further improve creditworthiness. Third, since the abatement of the Mexican crisis, the volatility of emerging market debt instruments has declined considerably, suggesting that the premium for volatility should have declined. Fourth, liquidity in emerging market debt instruments has been growing steadily with increases in both the size and turnover in these markets as discussed above. Fifth, the ability of investors to manage risks from holdings of emerging market debt instruments in their portfolios has steadily improved with the lengthening of yield curves on these instruments, permitting increased diversification across maturities, and a filling out of the spectrum of debt-issuing entities across geographic regions, sectors of economic activity, and different risk classes within countries. Risk management has also improved with the growth of derivative products in emerging market instruments.

    Has the compression of emerging market spreads been excessive? As discussed above, the behavior of EMBI and MLHY yield spreads suggests that, relative to historical differentials, there remains some room for emerging market spreads to decline to U.S. corporate levels. However, history provides a limited guide to determining whether yield spreads are sufficient to compensate for the risk of future default. The question of whether yields are sufficient to compensate for default risk could be answered directly if the probability of default implicit in market spreads could be identified and compared with independent estimates of the probability of default. Market spreads would clearly be judged to have declined too far if they were insufficient to cover the probability of default—that is, if the spread were decomposed into (1) a premium estimated to be due to the probability of default and (2) a liquidity premium to encompass the variety of other factors potentially affecting demand, and the liquidity premium were found to be negative (since there is no obvious reason to expect that investors would be willing to pay a premium—give up yield—to hold emerging market debt). Estimating default risk for sovereigns requires simulating a country’s balance sheet to do a financial risk exercise as is done for a bank or other enterprise. It would then be possible to calculate directly the probability distribution that a country would not be able to meet its payments and hence the probability of default. Few countries, however, publish such balance sheets. This is an important difference between sovereigns and corporates: debt-issuing corporations publish balance sheets and are subject to stricter disclosure requirements. Moreover, the legal framework in the event of corporate default is relatively clear. The volatility of perceived credit risk for emerging market sovereigns is, therefore, likely to be greater. Inherently higher volatility of perceptions of credit risk for sovereigns suggests that yields on sovereigns should exceed those for corporates.

    Equity Markets

    As several of the mature equity markets reached new highs in 1996, emerging equity markets continued to recover from the trough in early 1995, though cumulative returns since the peak in 1994 remained negative. The effects of the Mexican crisis continued to fade, and the volatility of equity prices—in both Latin America and Asia—subsided during 1996 and into 1997 to levels prior to the crisis, while the recovery of economic prospects in Latin America boosted forecasts of earnings growth. These factors combined to make emerging market equity look increasingly attractive relative to the mature markets, and price increases in Latin America accelerated in early 1997. The recovery in emerging equity markets in 1996 was accompanied by increased liquidity for most markets as turnover rose but new issuance remained subdued. While overall flotations of new equity by the emerging markets continued to decline, there were marked differences across regions, as placements by Latin American entities rebounded while those by Asian entities fell. There was an increased reliance on international issuance across regions, however, and the volume of international issuance increased.

    Secondary Markets

    During 1996 emerging equity markets posted their first collective positive return since the boom of 1993, with total dollar returns measured by the International Finance Corporation’s Investable (IFCI) Composite Index rising by 9.4 percent (Figure 29).21 Relative to the mature markets, however, emerging equity markets performed modestly. Returns were substantially higher on the S&P 500 index (23 percent) in the United States, for example. The relatively modest overall performance of emerging equity markets during 1996 masked divergent performance across regions. While Latin American equity markets rose by 17 percent and Asian markets by 10 percent, European. Middle Eastern, and African markets fell by 2.3 percent. During the first half of 1996, emerging equity markets rose along with those in the mature markets, with the major regional component indices of the IFCI rising along with the S&P 500 index. In July, as U.S. share prices gyrated downward amid concerns about possible increases in U.S. and international interest rates, the S&P 500 fell by 4.4 percent. Concerns of a spillover into emerging markets, combined with domestic developments, led to a simultaneous decline in emerging stock markets, and the IFCI Composite Index fell by 6.6 percent, as both Asian and Latin equity prices declined sharply. While U.S. share prices recovered quickly and rose rapidly (17 percent) during the remainder of the year, the IFCI Composite Index recovered only somewhat, and was virtually unchanged by the end of the year.

    Figure 29.Emerging Equity Markets: Selected Returns, Price-Earnings Ratios, and Expected Returns

    Sources: Bloomberg Financial Markets L.P.; International Finance Corporation (IFC), and Emerging Markets Data Base.

    1All return indices are expressed in U.S. dollars.

    2Price index.

    The rise in emerging market equity prices during 1996 accelerated in early 1997, with the IFCI Composite Index rising by 9.5 percent during the first quarter. Latin American markets rose particularly sharply, increasing by 15 percent, while Asian markets rose by 1.2 percent. Again, as during the first seven months of 1996, the emerging and mature stock markets moved in tandem. The S&P 500, and the IFCI Asian and Latin American indices all rose during January and February, and fell in March. In contrast to 1996, however, the collective increase in emerging market equity prices of 9.5 percent during the quarter was well in excess of the increase in the S&P 500 of 2.7 percent. While Asian markets then recovered modestly through May 1997, returns in Latin America once again accelerated, with returns of 13 percent during April and May. Despite the recovery in emerging market equity prices during 1996 and 1997, at their recent peaks they remained below the previous highs of September 1994. While Asian markets in the aggregate regained their previous peak levels in February 1997, before falling again, Latin markets in May remained 5.5 percent below their peak in September 1994.

    Price-earnings (P/E) ratios are the most commonly employed summary measures used to discern relative value in equity prices.22Figure 29 (middle panel) compares P/E ratios in emerging equity markets with those in the United States and the United Kingdom from January 1993 through May 1997. It is apparent that P/E ratios in emerging markets have not generally been lower than those in the mature markets. P/E ratios for the Asian emerging markets, in fact, consistently exceeded those in all the other markets over the period, and sometimes substantially so. Latin American P/E ratios, on the other hand, starting from well below those in the mature markets in early 1993, rose steadily to the levels of mature markets as share prices in the region increased through early 1994 without a commensurate increase in earnings. Despite the sharp declines in share prices in early 1995, and their relatively lower level since, P/E ratios for the region have been in the same range as those in the mature markets.

    Earnings-price ratios—the inverse of P/E ratios—provide a measure of the yield or expected return on equity in the event earnings are expected to remain constant. When earnings are expected to grow, however, current earnings per share underestimate expected returns. Since growth rates in the emerging markets exceed those in the mature markets—and those in Asia have been well above those in Latin America—a better comparison of expected returns is provided by actual earnings-price ratios plus expected earnings growth. Figure 29 (bottom panel) compares earnings-price ratios plus a proxy for expected earnings growth constructed using forecast GDP growth from the IMF’s World Economic Outlook.23 This comparison flows that expected returns on equity in the emerging markets, once adjusted for expected earnings growth, consistently exceeded those in the mature markets during the period. The differential in expected rates of return between the emerging and mature markets has fluctuated within 5 percentage points. From January 1996 through May 1997, the expected return on the composite of emerging market equity rose by about 1½ of a percentage point to 11½ percent, while that on the S&P 500 fell by ¾ of a percentage point to 7 percent.

    Among the emerging markets, expected returns in Asia have more or less consistently exceeded those in Latin America. Returns in the two regions have, however, tended to remain relatively close together, with changes in the differential among the two stemming largely from the volatility of Latin American returns. This volatility has implied that for brief periods the difference has been as high as 3 percentage points and at other times it has been negligible. There have also been extended periods, such as during much of 1993, when the differential was only ¼ of a percentage point to ½ of a percentage point. From January 1996 through May 1997, expected returns on Latin American equity edged up, albeit somewhat erratically, by 1½ percentage points to 10½ percent, of which some ½ of a percentage point is attributable to upward revisions to growth, while those on Asian equity rose modestly by ¼ of a percentage point to 12½ percent.24 Consequently, the differential of 3 percentage points at the start of 1996 narrowed to 1¾ percentage points by May 1997.

    Higher expected returns on equity in emerging markets relative to the mature markets have been associated with generally higher price and return volatility (Figure 30, top panel), with that in Latin American markets exceeding that in the Asian emerging markets.25 Volatility in emerging equity markets rose steadily during 1994 in the run-up to the Mexico crisis, plateauing at a considerably higher level during 1995. The rise was substantial in both Latin America and Asia. Volatility in Latin American markets rose from 2 percent in late 1993 to 5 percent by mid-1995, and in Asian markets from 1½ percent to 3 percent. Volatility then declined dramatically during the course of 1996 and continued doing so through May 1997, with that in Latin America falling below its previous low, to 1¾ percent, and that in Asia returning to its previous low of 1½ percent. The behavior of volatility in emerging equity markets was in contrast to that in the United States, where the volatility of the S&P 500 rose steadily during the course of 1996 and into 1997. In fact, by July of 1996, the volatility of the S&P 500 exceeded that of the composite of emerging markets, by October it exceeded that of the Asian emerging markets, and by early May 1997 it exceeded that of the Latin American markets. The middle panel of Figure 30 compares the ratio of expected returns to volatility. It flows that during 1996 and into 1997, as expected returns on emerging market equity rose and volatility declined, risk-adjusted rates of return rose dramatically for both Asian and Latin American markets. In the United States, on the other hand, the run-up in share prices and increase in volatility caused the S&P 500 to look less and less attractive, while in the United Kingdom, after declining during the first half of 1996, the ratio fell during the later part of the year and then stabilized in 1997.26

    Figure 30.Emerging Equity Markets: Selected Volatilities, Return-Volatility Comparisons, and Correlations

    Sources: Bloomberg Financial Markets L.P.; and IMF staff estimates.

    Market participants have pointed to the ongoing process of increased portfolio diversification by institutional investors in the mature markets as playing an important role in driving portfolio flows into emerging markets during the 1990s. The benefits of diversification into the emerging markets depend on the (lack of) correlation among returns between the emerging and mature markets. The correlation of price changes between the S&P 500 and the emerging markets, over the period 1992 through May 1997, were as follows: composite emerging 0.26, Latin American 0.27, and Asian 0.09. The correlations between the FT 100 and the emerging markets were as follows: composite emerging 0.32, Latin American 0.24, and Asian 0.21. The correlation between the Latin American and Asian emerging markets was 0.14, and that between the S&P 500 and the FT 100 was 0.29. All of these correlations are relatively small, suggesting considerable benefits from diversification. The weakest correlations, suggesting the greatest benefits from diversification, are between the United Stales and the Asian emerging markets, and between the Asian and Latin American emerging markets, It is notable that the correlations among the mature markets do not appear to be substantially greater than between the mature and emerging markets. Figure 30 (bottom panel) presents rolling correlations, calculated over the preceding year, and provides an indication of the evolution of co-movements between markets. The correlations of both the Asian and Latin American emerging markets with the S&P 500 increased during the latter half of 1996 through the first quarter of 1997, and were at their highest levels over the period by the spring of 1997.

    The aggregate regional indices mask substantial diversity in individual country returns within regions.27 In Asia, there was a wide range of returns during 1996–97. At one end of the spectrum, stock markets in China, in Taiwan Province of China, and in Hong Kong, China, performed very strongly. The dissipation of tensions between China and Taiwan Province of China helped raise cumulative returns in Taiwan Province of China to 51 percent during 1996 through May 1997, while China’s stock market posted returns of 57 percent, driven by a sharp increase in retail interest with the abolition of inflation subsidies on bank deposits. As concern over the return of Hong Kong to China diminished, returns in Hong Kong increased to 50 percent. Returns in Indonesia (21 percent), India (19 percent), and Malaysia (12 percent) were more moderate, while there were modest losses in the Philippines (-2.3 percent). At the other end of the spectrum, stock markets in Korea and Thailand were some of the worst performers in the world. Losses on the Korean stock market reached 32 percent, as the effects of economic slowdown were exacerbated by financial scandals, bankruptcies of large conglomerates, and labor unrest. Despite the government’s repeated easing of restrictions on foreign ownership limits, the Korean stock market failed to revive through much of the period, doing so only modestly in mid-1997. The Thai stock market registered a cumulative loss of 58 percent over the period, reflecting the host of concerns noted earlier.

    In Latin America, performance was almost uniformly positive. All the major countries in the region showed substantial gains. The most spectacular increase was in Venezuela, where dollar returns rose to 144 percent during 1996 through May 1997, as confidence surged following the adoption of a macroeconomic adjustment program in early 1996, a return of flight capital, strong oil prices, and some large privatizations. Brazil (94 percent) was not far behind, while Argentina (46 percent) and Mexico (39 percent) recorded robust returns. The Mexican total return index remains—in dollars—well below (60 percent at the end of May 1997) its high in January 1994. Eastern European equity markets registered tremendous gains from January 1996 through May 1997. Notable among them was Hungary, where total returns reached 166 percent, and Poland with 67 percent, both experiencing strong foreign investor interest. Elsewhere, in South Africa, dollar returns were -8.1 percent, reflecting the depreciation of the rand.

    As emerging equity markets recovered, liquidity, measured by the turnover ratio of shares traded to market capitalization, rose from 54 percent in 1995 to 76 percent in 1996, but remained well below the high of 94 percent in 1994 (Table 19). The increase in liquidity in emerging markets was similar to that in the mature markets, where the turnover ratio rose to 71 percent from 63 percent. Liquidity generally improved in Asia, with the exception of the Thai market, where turnover declined to 37 percent from 42 percent, continuing the decline begun in 1993. By contrast, turnover in China rose threefold to above 300 percent, leading the world’s equity markets. The dramatic increase has been ascribed to increased retail participation in the stock market—estimated at 25–30 million individuals in 1996 and expected to continue growing. Latin markets showed mixed results, as turnover increased in Brazil (to 61 percent from 47 percent) and Mexico (to 44 percent from 31 percent), while it declined in Argentina and Chile. In Europe, stellar returns were associated with improved liquidity in Hungary (to 42 percent from 17 percent) and Poland (to 86 percent from 73 percent).

    Table 19.Stock Market Turnover Ratio and Value of New Equity Issues in Selected Countries and Regions
    (In percent)
    Annual stock market turnover ratios1
    Developed markets47.
    All emerging markets2132.
    South Africa13.96.710.4
    Taiwan Province of China425.4322.5213.4234.0321.8176.6204.1
    Czech Republic46.749.9
    Western Hemisphere
    (In millions of U.S. dollars)
    Value of new equity issues3
    Emerging markets219,364.116,712.116,856.235,377.952,903.553,077.841,631.1
    South Africa851.71,903.32,178.4
    Taiwan Province of China7,444.61,688.01,044.62,526.93,676.93,878.03,983.1
    Czech Republic4,453.82,327.2
    Middle East723.0
    Western Hemisphere507.33,534.33,479.513,976.59,667.04,595.712,633.7
    Source: International Finance Corporation (IFC), Emerging Markets Data Base.

    Primary Markets

    As emerging market equity prices remained below their previous highs, total equity flotations by companies, including both domestic and international placements, declined by 21 percent in 1996 to reach $42 billion for the year (Table 19). Again, however, this figure masks diverse regional trends, as Asian issuance plummeted some 42 percent while Latin issuance rebounded 173 percent. Nevertheless, Asian equity placements continued to account for the major proportion of overall issuance, though to a considerably lesser extent, and the region’s share fell from 75 percent in 1995 to 55 percent in 1996, The decline in overall equity issuance by Asian companies was largely due to lower amounts of equity capital being raised in India. Indonesia, and Korea as other countries in the region continued to issue al the previous year’s pace. In Latin America, the level of equity placements surged largely owing to massive privatization issues in Brazil, which accounted for some $6 billion of placements. In Europe, issuance dropped off sharply, by $3.2 billion, as companies failed to take advantage of soaring equity prices to raise capital. Companies in the emerging markets relied more heavily on international placements. As a share of total issuance, international placements accounted for 38 percent in 1996, double the share in 1995. Both Asian and Latin American companies exhibited this increased reliance on international markets, albeit to different degrees. The share of international placements by Asian companies rose to 41 percent from 22 percent, while that for Latin American companies rose to 29 percent from 21 percent.

    International equity placements rose from $11 billion in 1995 to $16 billion in 1996, compared with a record of $18 billion in 1994(Table 16 and Figure 11). Equity flotations by the telecommunications sector accounted for a quarter of the issuance during 1996, representing one of the largest concentrations ever in a single sector, while the proportion of equity placements by the financial sector—mostly bank issues—continued to rise steadily, reaching 14 percent in 1996. Issuance by Asian entities rose only slightly to $9.8 billion, though they continued to account for the major proportion of international placements by the emerging markets with companies in Hong Kong. China, placing more than a third of the issues. Latin American placements rebounded in 1996 with large telecommunications privatizations in Venezuela (CANTV) and Peru (Telefónica del Peru), but remained modest compared with previous years, reaching some $3.7 billion. After staying out of the market in 1995, entities from Argentina and Mexico placed modest amounts during 1996. Placements by entities in the transition economies have continued to grow, reaching $1.3 billion, double the level in 1995. Issuance by Russian entities accounted for some $800 million, most notably through a $429 million ADR placement by Gazprom, a gas and oil company. In the first quarter of 1997, international placements continued at about their pace in the first quarter of 1996. One of the more notable issues was that of VSNL, a telecommunications firm from India, which raised $526 million through a GDR placement in the country’s largest share offering to date.

    Mutual Funds

    The increased delegation by individual investors of their portfolios to professional fund managers and the institutionalization of savings have dramatically increased the importance of institutional investors in international capital markets.28 The importance of institutional investors in intermediating portfolio flows to emerging markets has been no exception. Some estimates place the proportion of total portfolio flows to emerging markets intermediated—either directly or indirectly—by fund managers as high as 90 percent.29 The sheer size of institutional investor assets in the mature markets has meant that small changes in their portfolio allocations to emerging markets could have enormous effects on flows to these markets. The OECD reports the value of institutional investor assets in the G-7 countries in 1995 at $20.6 trillion. This compares with a total market capitalization of emerging equity markets at the end of 1996 of just $2.1 trillion, and cumulative net portfolio flows to the emerging markets during all of the 1990s of $410 billion, which amounts to just 2 percent of the value of institutional investor assets in the mature markets. Estimates of the portfolio shares of institutional investors in the mature markets dedicated to emerging markets vary considerably across types of institutions and countries. These estimates unanimously suggest, however, that in spite of the increased allocations during the 1990s, the share of institutional investor portfolios dedicated to emerging markets remains well below—by a factor of 3 to 5—that suggested by portfolio theory.30 The pace at which this gap is closed will be an important determinant of flows to the emerging markets.

    Data on emerging market investments for the spectrum of institutional investors are unavailable, but some indication of the rapid increase in institutional flows during the 1990s, their regional allocation, and portfolio composition is provided by the behavior of emerging market mutual funds, presented in Figure 31 for the aggregate of open- and closed-ended funds. (Note that mutual funds have represented a vehicle for investment into emerging markets not only for retail but also for institutional investors,) The net asset value of emerging market mutual funds rose rapidly during the 1990s from $6.8 billion in 1990 to $112.2 billion by the first quarter of 1997, and was remarkably resilient to the Mexican crisis. Among the region-specific funds, the distribution continues to favor Asian funds, which accounted for almost half of asset values in the first quarter of 1997, while those dedicated solely to Latin America accounted for some 15 percent. The broader “emerging markets” funds, which are not dedicated to a particular region, represent a significant proportion—some 30 percent in terms of asset value—of all funds. Equity funds continue to represent the majority of funds, around 90 percent in 1996, though this share has declined steadily from an estimated 98 percent in 1990, while the share of bond funds, which account for almost all of the remainder, has steadily increased from 2 to 10 percent, and multi-asset funds have remained negligible. Bond funds have been more significant among the dedicated “emerging market” and Latin American funds, where their shares had risen to 19 and 17 percent, respectively. Among Asian funds, on the other hand, bond funds continue to represent a modest share, around 3 percent at end-1996.

    Figure 31.Emerging Market Mutual Funds

    (In billions of U.S. dollars)

    Source: Lipper Analytical Services, Inc.

    1Africa, Europe, and Middle East.

    2Non-region-specific funds dedicated to emerging markets.

    Net flows into emerging market mutual funds rebounded from their low of $963 million in 1995 to reach $7.2 billion during 1996. There were particularly robust inflows of $4.5 billion in the first quarter of 1996, which subsequently declined through the year, culminating in an outflow of some $326 million in the fourth quarter, a pattern that was common across both region-specific and broad emerging market funds. In the first quarter of 1997, there was a strong recovery of flows with purchases reaching 52.7 billion, concentrated in the broad emerging market and Latin American funds, while there were net redemptions from Asian funds of $532 million. Generally, purchases and redemptions of individual country-specific funds also mirrored the performance of the local markets. The pattern of flows—both total volumes and regional allocations—corresponded relatively closely to the behavior of emerging market equity prices discussed above.

    International Bank Lending

    Syndicated Loans

    The international syndicated loan market for emerging market borrowers during 1996 and the first quarter of 1997 was, albeit to differing extents across regional segments, characterized by moderating demand for bank lending that coincided with a rising supply of loanable funds. This mismatch created considerable downward pressure on pricing and caused loan structures to weaken. The favorable pricing of emerging market bonds caused borrowers increasingly to turn away from bank lending in favor of the longer maturities and less burdensome restrictions offered by fixed-income instruments, while favorable conditions in the loan market itself encouraged refinancing, which accounted for almost a fifth of new medium- and long-term syndications. On the other hand, the low level of interest rates in the mature markets and the tightening of interest margins on loans in these markets caused banks to look increasingly to the emerging markets for higher yields. The resulting intense competition among banks for lending to emerging market entities pushed down spreads, cut fees, increased tenor, and resulted in a weakening of loan covenants.

    Following the sharp increase in syndicated bank lending to emerging markets during 1995 of over 36 percent due to the increased costs of borrowing on bond and equity markets in the aftermath of the Mexican crisis, the total volume of syndicated lending rose more modestly during 1996 by 6.4 percent to $79.7 billion (Table 16 and Figure 11). Lending to Asian countries continued to grow robustly, however, increasing by 22 percent and accounting for the largest share of bank lending, 62 percent in 1996, up from 54 percent in 1995. Lending to the European emerging markets also rose strongly, increasing by 9.0 percent, while the volume of loans extended to Latin America grew more modestly, by 5.4 percent, and declined to Africa and the Middle East. In the first quarter of 1997, the total volume of syndicated lending to the emerging markets dropped off by 2.4 percent relative to the average quarterly pace during 1996. Lending to the European emerging markets fell sharply, to half its pace during 1996, while lending to Asia (-2.4 percent) and the Middle East (—1.6 percent) declined moderately. By contrast, loan volumes to Latin America picked up, growing by a strong 38 percent, and lending to Africa recovered. The proportion of refinancings in new syndications for emerging market entities grew from the unusually high level of 17 percent in 1995 to 19 percent in 1996 (Table 20), with particularly strong increases to Latin America, Europe, and the Middle East. Average interest margins on new loans to the emerging markets as a whole, after having risen modestly in 1995, declined from 105 basis points to 88 basis points in 1996 (Table 20). The rise in spreads during 1995 had been localized to loans to Latin American entities, while average spreads on loans to the other emerging market regions had narrowed. The compression of spreads during 1996 was most evident in the Latin American and European emerging markets, where margins sometimes fell to levels close to those of the most highly rated international borrowers, again raising concerns as to whether risks were being mispriced.

    Table 20.Emerging Market Medium- and Long-Term Syndicated Loan Commitments: Interest Margins and Refinancings
    (In basis points)
    Interest margins1
    Emerging markets5680103901001058892
    Middle East98518610991608372
    Western Hemisphere74881131275818110998
    (In percent)
    Emerging markets3.
    Middle East0.813.44.314.7
    Western Hemisphere5.828.42.92.724.022.536.326.5
    Source: Capital Data Loanware.

    During 1996, refinancings accounted for around 36 percent of new medium- and long-term syndications to Latin America, while average margins on loans declined steeply, from 181 basis points in 1995 to 109 basis points. An example of the levels to which competition pushed spreads was provided in September by the five-year $500 million refinancing loan by Codelco, a Chilean copper conglomerate, which priced at LIBOR plus 22.5 basis points. The proportion of refinancings in new syndications to Latin America moderated in the first quarter of 1997 to 27 percent, margins appeared to bottom out, and average margins remained unchanged. The deterioration in loan covenants, such as restrictions on the gearing ratio of the borrower and the double pledging of assets as collateral, was most evident in mid-1996 when three unsecured loans by Argentine companies were put up for syndication without financial covenants. In another unprecedented deal, in December 1996 the Central Bank of Argentina arranged a collateralized $6.1 billion contingent repo facility with international private sector banks to provide liquidity to the domestic banking system.31

    Average margins on loans to the European emerging markets fell from 131 basis points in 1995 to 88 basis points in 1996, and further to 73 basis points in the first quarter of 1997. The National Bank of Hungary pushed pricing to an all-time low for the region by refinancing a $350 million loan, priced at LIBOR plus 50 basis points signed in August 1996, at LIBOR plus 20 basis points in December. This rapid decline, also enjoyed by other Hungarian borrowers, coincided with the country’s membership in the OECD, which reduced capital requirements for lenders against loans to various Hungarian entities. The keen competition among banks for high-yielding loans was evidenced in April 1997 by the $2.5 billion syndication for Gazprom, the Russian gas and oil company, which priced at LIBOR plus 200 basis points and was three times oversubscribed.

    The robust growth of syndicated lending to the Asian emerging markets was underpinned by the continued expansion of lending to the financial and property sectors and for infrastructure financing. Infrastructure financing increasingly took the form of project finance which, by allowing for a separation of risks specific to the project from the overall balance sheet of the parent company, can provide higher yields. Project finance accounted for 32 percent of medium- and long-term syndications to the region during 1996. In terms of the number of project financing arrangements worldwide, the emerging markets of Indonesia, Thailand, China, India, and Hong Kong, China, have been among the lop 10 most active countries in the world. These have included projects in power, telecommunications, water, and transport. Project financing has been particularly popular in Hong Kong, China, with the largest number of deals outside the United States, driven by the surge in building ahead of the territory’s handover to China this year. Compared with the sharp pickups in the share of refinancings in total syndications to Latin America and the European emerging markets during 1996, the share in Asia remained steady at around 16 percent and was essentially unchanged in the first quarter of 1997. Average interest margins on syndicated lending to Asia remained unchanged in 1996 at 85 basis points, then rose to 103 basis points in the first quarter of 1997.

    Interbank Market

    In addition to syndicated lending, interbank loans account for an important share of bank lending to emerging markets. Table 21 documents the recent evolution of interbank credit from BIS-reporting banks to banks in several emerging market countries.32 In Asia, while the stock of net interbank loans to Singapore and Hong Kong, China, has historically been high because of their roles as regional financial centers, net interbank lending flows during 1994–95 to Thailand ($48.9 billion) and Korea ($23.2 billion) were very high, with net credits outstanding by the end of 1995 of $69.9 billion and $43.8 billion, respectively. International banks’ reassessment in the face of unfavorable economic developments and measures by the authorities aimed at reducing reliance on international short-term bank borrowing caused a sharp slowdown in lending to Thailand in 1996. The flow of interbank lending to Thailand declined from $31.7 billion in 1995 to $9.6 billion in 1996, with flows in the last quarter dropping off to below $1 billion. The How of interbank lending to Korea moderated during the early part of 1996 but, as membership in the OECD approached, picked up again, and for the year as a whole reached $14.4 billion, barely below that of $14.9 billion in 1995. Considerable attention has been focused on the buildup of interbank debt by the Asian economies, in particular in Thailand and Korea, and the fact that a substantial proportion of this debt is short term and needs to be rolled over frequently. The progressive liberalization of financial systems, in an environment of relatively undeveloped local equity and bond markets, combined with rapid economic growth and restrictive monetary policies that have kept interest rates high, has created strong incentives for lending to these countries.

    Table 21.Changes in Net Assets of BIS-Reporting Banks vis-à-vis Banks in Selected Countries and Regions1(In millions of U.S. dollars)


    Credit at

    South Africa-3238422671,104-195471-638916,706
    Hong Kong15,62910,84640,24626,0496,50710,2672,0837,192216,866
    Czech Republic-770497818-395-366-1,6331,674-70472
    Middle East
    Saudi Arabia1,9103,256-3,520-1,053-6,2583,5417,036-5,372-27,415
    United Arab Emirates6,3381,430-4,479-5,149-3,086492-1,340-1,215-16,937
    Western Hemisphere
    Source: Bank for International Settlements (BIS).

    The situation is very different among the Latin American countries where Brazilian, Chilean, and Colombian banks, for example, were net lenders to BIS-reporting banks, with outstanding credits of $10 billion, $3.8 billion, and $1 billion, respectively, at the end of 1996. Similarly, Argentine banks have not been significant borrowers, with $4.1 billion in loans outstanding at the end of 1996. Mexican banks made substantial net repayments during 1995 and the first three quarters of 1996, totaling $15.4 billion, and had a modest outstanding debt of $1.7 billion at end-1996. Among the European countries, Russian banks are the largest debtors to BIS-reporting banks, with an outstanding amount of $33.3 billion at the end of 1996, while Polish banks have been net lenders with a net stock of claims of $7.9 billion at the end of 1996. In the Middle East, Egyptian banks were the main recipients of funds in 1996, and they, along with banks in Kuwait and the United Arab Emirates, remain net lenders.

    Annex II Developments and Trends in the Mature Capital Markets

    During the period under review, prices and volumes in the mature international capital markets responded favorably to low or declining inflation rates, efforts toward fiscal consolidation, continued strong growth in the United Stales and the United Kingdom, and the expected strengthening of economic activity in most other countries. An improved macroeconomic environment—along with ample liquidity—also curtailed volatility in the major markets. Equity prices in most advanced countries surged as corporate earnings forecasts were progressively marked up. Large yield spreads among the three largest economies—favoring dollar-denominated securities—attracted unprecedented capital inflows to U.S. securities markets, which were a key factor behind the dollar’s sharp rise against the yen and deutsche mark during the past two years. The first section of this annex discusses developments since early 1996 in foreign exchange markets. The second and third sections summarize recent developments in bond and syndicated loan markets, respectively, and the next section reviews developments in international equity markets. The fifth section reviews developments in global derivative markets, and the final section discusses supervisory and regulatory developments.

    Exchange Rates


    The dollar appreciated sharply against the deutsche mark and yen in 1996 and through May 1997, in large part because of record-high capital inflows into U.S. securities markets and the relatively strong performance of the U.S. economy vis-à-vis Europe and Japan. From its all-time low of ¥80 and DM 1.35 in the spring of 1995, the dollar rose fairly steadily over the subsequent two years, peaking in May 1997 at a level that amounted to a cumulative appreciation of about 60 percent against the Japanese yen and about 30 percent against the deutsche mark (Figure 32). The dollar has since reversed some of its gains against the yen in anticipation of higher Japanese interest rates as well as concerns that the Japanese monetary authorities might sell some of their dollar reserves, though it remains more than 40 percent above its low in early 1995.

    Figure 32.Major Industrial Countries: Exchange Rates, January 1994–May 1997

    (Local currency/U.S. dollar)

    Source: Bloomberg Financial Markets L.P.

    In percentage terms, the movement in the dollar-deutsche mark rate over the past two years has been large but not unprecedented over similar time frames during The past two decades. By comparison, the movement in the dollar-yen rate from the spring of 1995 to its peak in May 1997 was one of the largest moves ever for this exchange rate. The dollar’s depreciation from ¥260 at the end of February 1985 to below ¥ 122 in November 1988 was the same order of magnitude in percentage terms as the recent experience (although opposite in sign), but it occurred over a longer period of time. The 42 percent depreciation of the dollar from October 1978 through February 1980 is perhaps the only instance in the postwar period that rivals the magnitude and speed of the recent change in the dollar-yen rate.

    The dollar’s appreciation in 1995–97 occurred against the backdrop of relatively high yields on dollar assets. With interest rates low in Japan and core Europe, investors’ search for higher yields produced large capital inflows not only into the U.S. markets but into higher-yielding markets globally. As a result, traditionally higher-yielding currencies—in Europe and in the dollar bloc—were supported by capital inflows, as well as improved macroeconomic fundamentals. Within Europe, lower inflation, efforts toward fiscal consolidation, and renewed optimism about European Economic and Monetary Union (EMU) were instrumental in the appreciation of most EU countries’ currencies against the deutsche mark after early 1996 (Figure 33). Despite improved exchange rate fundamentals in Europe and in the dollar-bloc countries, however, record-high capital inflows to dollar markets prevented most currencies—the notable exception being the pound sterling—from strengthening appreciably against the U.S. dollar.

    Figure 33.Major European Countries: Local Currency vs. Deutsche Mark, January 1994–May 1997

    Source: Bloomberg Financial Markets L.P.

    Interest Differentials, Capital Flows, and the Dollar’s Rise

    Market participants in the major financial centers attribute much of the dollar’s momentum to abundant liquidity in international financial markets that has disproportionately been funneled into the dollar markets, especially dollar-denominated fixed-income markets. This concept of liquidity in international financial markets refers to both looser monetary policies in some countries as well as simply large flows of capital internationally. Monetary stimulus in the major industrial countries has on balance shifted toward an easier stance since early 1996, with slightly tighter monetary conditions in the United States and the United Kingdom partially offsetting the monetary stimulus from Europe, Japan, and Canada (Figure 34). In addition, growth in Firms’ demand for bank financing in some of the major countries has been low in relation to monetary growth rates, and thus monetary easing may have significantly affected liquidity (see Figure 34).

    Figure 34.Major Industrial Countries: Real Growth in Broad Money Supply and Claims on Private Sector1

    (In percent from four quarters earlier)

    Sources: International Monetary Fund, International Financial Statistics database; and The WEFA Group.

    1M3 for broad money supply for all countries except the United Kingdom, for which M4 is used. Claims on private sector are taken from International Monetary Fund, International Financial Statistics (line 32d). Data for 1980: Q1 through 1997: Q1.

    The difference in cyclical positions and in the stances of monetary policies—which reflect the difference in cyclical conditions—among the major countries have created large interest differentials between countries. Of particular significance are the higher interest rates in the United States versus the two next-largest economies, Japan and Germany (Figures 3537; see also Figure 2). The interest differential between yen- and dollar-denominated fixed-income securities has been especially large during the period under review. At the short end of the yield curve, the differential between three-month yen and dollar rates has been more than 4.5 percentage points since January 1996, and in May 1997 exceeded 5 percentage points. At the long end of the yield curve, the spread on long-term yen and dollar government bonds rose steadily from early 1996, and in May 1997 stood at more than 4 percentage points. The spread between deutsche mark and dollar yields has increased steadily since early 1996, reaching about 2.5 percentage points in May 1997 for short-term rates and about I percentage point for long-term rates.

    Figure 35.Major Industrial Countries: Short-Term Interest Rates1

    (In percent a year)

    1Three-month certificate of deposit rates for the United States and Japan; three-month treasury bill rate for Italy; rate on three-month prime corporate paper for Canada; and three-month interbank deposit rates for other countries. Weekly averages of daily observations are plotted for all countries other than Italy and Canada. For Italy, results of fortnightly treasury bill auctions are shown. For Canada, weekly observations are plotted.

    21987 GDP weights.

    Figure 36.Major Industrial Countries: Long-Term Interest Rates1

    (In percent a year)

    1Yields on government bonds with residual maturities of 10 years or nearest. Weekly averages of daily observations.

    2 1987 GDP weights.

    Figure 37.Bilateral Exchange Rates and Short-Term and Long-Term Interest Differentials vis-à-vis the U.S. Dollar1

    Source: International Monetary Fund.

    1Interest differentials shown are U.S. interest rates minus domestic interest rates in percent a year. Exchange rates are drawn on logarithmic scales and are defined in terms of national currency units per U.S. dollar, except for the United Kingdom, where it is defined as U.S. dollars per pound sterling. The figures show monthly averages of daily data from January 1985 through May 1997.

    The relatively attractive yields on dollar investments encouraged Japanese and European investors to increase the weight of dollar bonds in their portfolios. This is reflected most clearly by survey data on currency exposures of investors: since early 1995, investors have clearly tilted their portfolios in favor of assets denominated in dollars, pounds sterling, and other high-yielding currencies (Statistical Appendix Table A1). The appreciation of many of the higher-yielding EU countries’ currencies, as well as currencies in the dollar bloc (Canada, Australia, New Zealand), against the deutsche mark and yen, is consistent with yield-seeking international capital flows. Further, the attractiveness of the relatively higher yields in the United States (and elsewhere) has been compounded by uncertainties about investing in core Europe and Japan, which might have caused risk-adjusted interest differentials to be even larger. Specifically, uncertainty about financial system problems in Japan and uncertainty about the future value of the euro—which has been highlighted by renewed optimism about EMU—might have further increased the relative attractiveness of dollar-denominated investments.

    The available data on capital flows strongly suggest that investors have been seeking higher yields by investing abroad. Outward portfolio flows during the past 18 months have been associated with aggressive buying of foreign securities by U.S., Japanese, French, Spanish, and German investors, as well as positions booked in the major international financial centers outside the United States—the United Kingdom, Singapore, and Hong Kong, China (Statistical Appendix Tables A2A3). As for the recipients of these capital flows, the United Slates has clearly been a major target of non-U.S.-based investors. Foreign purchases of U.S. treasury and government agency bonds and notes reached $293.7 billion in 1996, and there was a further $78 billion of foreign purchases of U.S. corporate bonds. Similarly strong capital inflows to U.S. securities markets have been apparent in the first quarter of 1997: foreign purchases of government and corporate bonds during the first quarter of 1997 were slightly above the quarterly average during 1996. In comparison, despite the sharp increase in U.S. equity prices in recent years, there were only $13.2 billion of foreign inflows into the U.S. equity market in 1996.

    As noted in Chapter II of the report, particularly wide interest differentials between the United Stales and Japan, in conjunction with the belief that the Bank of Japan did not want the yen to strengthen in 1996–97, were viewed by some large global hedge funds as a potentially lucrative situation. These so-called yen-carry trades involved borrowing in yen, selling the yen for dollars, and investing the proceeds in relatively high-yielding U.S. fixed-income securities. In hindsight, these trades turned out to be considerably more profitable than simply the interest differential, for the yen depreciated continuously over the two years from May 1995 through May 1997, which reduced the yen liability relative to the dollar investment that it financed.

    With available data, it is difficult to determine the scale of yen-carry trades implemented over the past two years. It is noteworthy, however, that while Japanese banks reduced total cross-border assets by $20 billion in 1996, they increased lending by almost $19 billion to nonbank entities located just in the Cayman Islands (British West Indies)—a home for some of the major hedge funds. Over the same period, entities located in the British West Indies accumulated $20 billion of U.S. long-term bonds. Further, lending by Japanese banks to U.S. nonbank entities expanded by an additional $28.8 billion during 1996.1 Viewed in light of the significant contraction in total cross-border assets of Japanese banks in 1996, the fact that Japanese banks increased their cross-border claims on nonbanks in the Cayman Islands and the United States by almost $48 billion is consistent with parties in these regions instituting significant yen-carry trades.

    The volume of international inflows into U.S. bond markets in 1996 is by far the largest ever—70 percent larger than the previous record set in 1995. Net foreign purchases of U.S. government bonds alone ($294 billion) accounted for 250 percent of the increase in the stock of privately held public debt securities in 1996. This pushed the share of foreigners’ total ownership to one-third of the stock of public debt securities, up from 26 percent in 1995 (Figure 38).

    Figure 38.Foreign and International Holdings of U.S. Public Debt1

    (Percentage of gross marketable public debt)

    Source: Board of Governors of the Federal Reserve System, Federal Reserve Bulletin.

    1Foreign and international holdings are U.S. Treasury estimates and consist of investments of foreign balances and international accounts in the United States.

    Yield-seeking by private investors accounted for much of the large capital inflows to dollar markets during the past eighteen months, but foreign exchange reserve accumulation by central banks was also an important source of inflows to dollar markets. Central banks accounted for 35 percent of total foreign net purchases of U.S. treasury bonds in 1996. Much of this official accumulation of U.S. treasury securities has been a consequence of efforts by developing countries to manage the impact of large inflows of foreign capital on their respective currencies, which required the accumulation of official reserves. Indeed, more than half of reserve accumulation by all central banks in 1996 was by central hanks in developing countries. Among industrial countries, the Japanese monetary authorities have been most aggressive in accumulating reserves during the past two years. During 1995–96, the Bank of Japan accumulated more than $90 billion in reserves, bringing the total in early 1997 to about $218 billion (14 percent of global official reserves). The Bank of Japan’s reserve accumulation in 1996 was almost four times greater than that of the other six G–7 countries combined, and it represented 45 percent of reserve accumulation by central banks in all industrial countries (20 percent of global official reserve accumulation). By contrast, the G–7 countries excluding Japan were responsible for only 5.4 percent of total official reserve accumulation over the year, and about i2 percent of reserves accumulated by all industrial countries during the year.

    The magnitude of capital inflows into the dollar markets has also been large relative to the U.S. current account position (Table 22). The U.S. current account deficit in 1996 was $148 billion (or about 1.9 percent of GDP). Capital inflows amounted to $547 billion, or 370 percent of the current account deficit. Accumulation of U.S. securities by foreign central banks alone was only $26 billion less than what was required to finance the current account deficit. Private security purchases by foreigners, however, contributed inflows of $289 billion. Thus, total accumulation of U.S. securities by foreigners amounted to 5.4 percent of GDP, or 275 percent of what was necessary to finance the current account balance. This record level of portfolio inflows—both in absolute value and as a percentage of GDP—was intermediated in U.S. financial markets and invested abroad through purchases of foreign securities by U.S. investors ($108 billion) and by net lending abroad by U.S. banks ($98 billion).

    Table 22.United States: Selected External Account Variables
    U.S. Assets Abroad, Net1Foreign Assets in the United States2Current Account BalanceU.S. Assets Abroad, Net1Foreign Assets in the United States2Current Account Balance
    (In billions of U.S. dollars)(In percent of GDP)
    Period averages
    Sources: International Monetary Fund, World Economic Outlook database; and U.S. Department of Commerce, Survey of Current Business.

    The aggressive purchase of foreign securities by U.S. investors is consistent with the continued international diversification of U.S. investors’ portfolios. The relatively large net amount of cross-border bank lending by U.S. banks reflects two factors. First, U.S. banks have onlent to their foreign subsidiaries to meet strong demand in the Eurodollar market and to finance the demand for dollar securities by foreigners. Second, as discussed under “International Syndicated Loan Markets” below, the international interbank markets have increasingly used repurchase agreements for interbank funding, and U.S. treasury securities are the predominant form of collateral in these markets.

    The volume of foreign purchases of U.S. bonds has also been large relative to current account positions in many of the other major countries. In Japan, residents accumulated foreign fixed-income securities in 1996 amounting to 142 percent of the current account surplus of Japan, with well above half of these purchases being bonds issued by the U.S. government and U.S. corporations. Similarly, the German current account deficit was ¾ of a percent of GDP, but residents accumulated U.S. government and corporate bonds equivalent to 1 percent of GDP. And in France, purchases of U.S. government and corporate bonds were almost 3 percent of GDP, or more than double the current account surplus. Finally, although the U.K. current account was roughly in balance, net purchases of U.S. government and corporate bonds booked in the United Kingdom were equal to more than 10 percent of GDP.

    In summary, there have been large inflows into dollar markets during the past two years, and these capital flows have been attributed to three factors; liquidity spilling over into international capital markets associated with a loosening of monetary policy in Europe and Japan, which has been facilitated by the weak demand for funds by firms in several European countries and in Japan; wide interest differentials among the three largest economies, which have favored dollar assets: and uncertainties associated with EMU in Europe and with the financial system in Japan. It is difficult to quantify the role of these various factors in recent movements of the exchange rates among the major currencies. Perhaps the least tangible factor is that associated with uncertainties about EMU and banking problems in Japan. It is noteworthy, however, that recent analysis by some market participants attribute about half of the dollar’s rally against the deutsche mark since mid-1996 specifically to EMU optimism.2

    Volatility in Foreign Exchange Markets

    Ample liquidity in the international financial markets and the lack of inflationary pressures helped to maintain low volatility in foreign exchange markets despite large movements in exchange rates among the major currencies (Figures 3940). The absence of volatility has been especially marked for the currencies of the traditionally higher-yielding industrial countries, both because capital inflows to higher-yielding markets added liquidity to foreign exchange markets, and because the same factors that worked to strengthen many of these currencies also reduced risk premiums for holding these currencies. Specifically, a change in the stance of macroeconomic policy, which has emphasized low inflation and fiscal conservatism, contributed to a sharp reduction in uncertainty regarding the key currency market fundamentals—inflation and fiscal policy.

    Figure 39.RiskMetrics Daily Price Volatility for U.S. Dollar Spot Exchange Rates, January 19, 1995–May 30,1997

    (In percent)

    Source: J.P. Morgan.

    Figure 40.Implied Volatility: Yen and Deutsche Mark Three-Month Forwards

    Source: Bloomberg Financial Markets L.P.

    Note: Implied volatility is a measure of the expected future volatility of the currency based on market prices of the call options on forwards on the currency.

    In the European Union, improved fundamentals have been closely tied with increased optimism that EMU would proceed in 1999. As a result, markets have priced in not only observed improvements in fundamentals, but also an improved outlook for the future course of macroeconomic policy as EU countries seek to participate in EMU at an early stage. As a result, the currencies of most EU countries, and particularly those of the higher-yielding—or “non-core”—countries, have appreciated against the deutsche mark. This facilitated the reentry of the Italian lira, and the entry of the Finnish markka, into the ERM in late 1996. The strongest currencies in Europe since early 1996 have been the pound sterling and the Irish pound, both of which have received additional support from strong economic growth and expectations of rising interest rates. The strength of the Irish pound has attracted considerable attention because of its inclusion in the ERM and the fact that it has risen about 10 percent since mid-1996 above its central rate against the deutsche mark. The market’s interest in the Irish pound’s strength within the ERM stems from concerns about what entry rate will be used for the pound when EMU begins.

    Reduced volatility in foreign exchange markets in 1996 strongly affected turnover in spot foreign exchange markets, particularly European foreign exchange markets. Although recent data on turnover in the global foreign exchange market are not available, lower volatility has been widely pointed to as an explanation for the marked scaling back of European foreign exchange trading operations of the major market participants. Activity in currency derivative markets was unaffected by the absence of turbulence in the foreign exchange markets, as turnover continued to expand briskly. In the over-the-counter markets, the notional principal of currency swaps rose 18 percent at an annual rate during the first half of 1996. On the exchanges, currency futures and options volumes rose 19 percent in 1996 over 1995, in part reflecting a rebound after the sharp drop in 1995. Reduced volatility and turnover in spot foreign exchange rates undoubtedly slowed the demand for currency derivatives, but the structural growth of derivative markets associated with the prevalence of risk management continues to expand derivative markets (see the “Derivative Markets” section below).

    This increased use of foreign exchange derivatives for risk management is reflected by reportedly brisk activity in binary range options. These instruments are an effective tool for hedging volatility because their payoff structure is tied to whether or not the exchange rate stays within a specified range. Market participants report that positions of this type were widely used in the French franc-dollar, dollar-yen, deutsche mark-French franc, and many of the other European bilateral exchange rates in 1996. On the long side of these positions, U.S. hedge funds and other high net worth investors are reported to have placed large bets that EMU or official intervention by the Bank of Japan would ensure stable exchange markets in 1996 and early 1997. A notable instance occurred in the fall of 1996, when there were reportedly a large volume of dollar-yen range barriers issued with a range of ¥112–115, On October 29, the dollar reached ¥114.88, just below the knock-out level. Market participants reported that at the time the dollar was temporarily prevented from strengthening further on account of massive selling of dollars by dealers, hedge funds, and others with large long positions in dollar-yen range barriers.

    One method of gaining some insight into the market’s assessment of the direction of the major exchange rates in the future is to study information contained in asset prices. Distributions of the major exchange rates computed from foreign currency options premiums are a way to gauge the market’s expectation about the range of possible future values for exchange rates (see Appendix 1 at the end of this annex). This technique reveals considerable dispersion in the market’s assessment of future values for the yen-dollar rate.

    Bond Markets


    Low interest rates in core Europe and Japan (see Figures 3536), as well as continued international diversification of U.S. investors’ portfolios, led to substantial capital flows from the major industrial countries into the higher-yielding bond markets. In conjunction with low, and in some countries declining, inflation rates, these capital flows to the higher-yielding markets were instrumental in reducing yields relative to the major benchmark yield curves and in curtailing volatility in bond markets. The narrowing of spreads attracted considerable attention within the context of EMU as convergence plays were once again established in those markets, but the compression of spreads in Europe was a reflection of a global phenomenon that included the higher-yielding industrial countries outside of Europe- Canada, Australia, and New Zealand—the emerging markets, and corporate bond markets (Figures 4143). This favorable environment for borrowers caused new issuance of fixed-income securities to reach record levels in the international markets and in most of the higher-yielding domestic bond markets. The yield-seeking behavior of investors was reflected also in international bond markets by the strong demand for dollar-denominated paper: the share of dollar-denominated bonds issued in the international markets more than doubled in 1996 over the previous year, whereas the share of yen bonds fell almost 80 percent and the share of deutsche mark issues fell close to 40 percent.

    Figure 41.Selected European Long-Term Interest Rate Differentials with Germany

    (In basis points)

    Source: International Monetary Fund.

    Figure 42.Yield Differential for the 10-Year Government Bonds of Australia, Canada, and New Zealand

    (In basis points against the yield on 10-year U.S. government bonds)

    Source: Bloomberg Financial Markets L.P.

    Figure 43.United States: Yield Spreads of Corporate Bonds over U.S. Treasuries1

    (In percent)

    Sources: Bloomberg Financial Markets L.P.; Board of Governors of the Federal Reserve System, Federal Reserve Bulletin; and Merrill Lynch.

    1Yields on 10-year U.S. treasury bonds of constant maturities are used for U.S. treasuries. Junk bonds are all high-yield bonds weighted by par value.

    The narrowing of sovereign interest rate spreads globally has in large part been due to low inflation and progress toward fiscal consolidation. Nonetheless, the degree of spread compression may have been amplified by plentiful global liquidity in international financial markets and the related yield-seeking behavior of investors, which has accompanied the sharp decline in interest rates in Japan and core Europe. Additionally, capital inflows into U.S. securities markets were much greater than what was necessary to finance the current account deficit and may, therefore, have contributed to the narrowing in spreads as this capital was effectively intermediated in the United States and reinvested in foreign bond markets (Table 22; see also Table 3).

    European Monetary Union and Convergence Plays

    Spreads in European fixed-income markets (relative to German benchmarks) peaked in early 1995 as a result of the flight to quality associated with the global bond market correction in 1994 as well as the Mexican crisis that developed later in the year. The subsequent narrowing of spreads continued in 1996 because of further easing of monetary policy in Europe and an improved outlook for EMU. The most notable narrowing of spreads occurred in those countries with high spreads at the start of the period. Specifically, the Italian 10-year yield spread narrowed by 350 basis points from early 1996 through end-May 1997, Spanish spreads fell 300 basis points, and Swedish spreads fell 130 basis points. In core EMU countries, where spreads were thin at the beginning of the period, they narrowed to the point that French and Dutch long-term yields traded below deutsche mark yields by late 1996. The United Kingdom is the notable exception to this latest round of convergence plays, a fact that has been attributed to the asynchronous cyclical position of the United Kingdom versus much of continental Europe, and to considerable uncertainty about when the United Kingdom will participate in EMU.

    For the higher-yielding EU countries, from January 1996 to end-May 1997 yield spreads had narrowed by similar magnitudes at the short and long ends of the yield curves. Specifically, from January 1, 1996, through end-May 1997, the difference between rates paid on deutsche mark one-year deposits and Spanish peseta and Italian lira one-year deposits narrowed by 300–350 basis points, compared with the 270–285 basis point reduction in spreads on seven-year deposits for these currencies. Thus, the downward shift in yield curves relative to the deutsche mark curve was only slightly greater at the short end than at the long end of the curves.

    The convergence of interest rates in Europe is often attributed to the improved fundamentals (inflation and fiscal accounts) in peripheral countries. The renewed focus on EMU, and the increased likelihood that it will begin on schedule, have shifted expectations of the future path of fundamentals in the higher-yielding EU countries in close alignment with the criteria of the Maastricht Treaty. As discussed below, there is considerable consensus among market participants that these improvements in current and expected future fundamentals have been the primary driving force behind the convergence process, rather than extraneous factors such as excess liquidity.

    The combination of monetary easing and convergence plays that narrowed interest rate spreads produced a favorable environment for fixed-income investors, especially for investors in the higher-yielding countries (again with the notable exception of the United Kingdom). Total returns on long-term bonds of maturities exceeding 10 years in 1996 amounted to almost 50 percent in Italy, 34 percent in Spain, 25 percent in Sweden, and on the order of 10–15 percent in core Europe. By comparison, with monetary policy leaning in the opposite direction, the total return on U.S. treasury bonds in 1996 was reduced by price depreciation (Table 23).

    Table 23.Major Industrial Countries: Bond and Equity Index Returns(In percent)
    19951996January–May 1997






    Bond indices1
    United States23.7323.731.511.511.051.05
    United Kingdom18.6917.537.5518.964.800.20
    Equity indices
    United States (S&P 500)34.1134.1120.2620.2614.5214.52
    Japan (Nikkei 225)0.74-3.09-2.55-13.203.653.56
    Germany (Commerzbank)5.2613.4322.5113.2623.2512.33
    France (CAC 40)-0.498.1923.7117.0111.580.32
    Italy (Banca Commerciale)-6.78-4.7812.3816.7914.342.85
    United Kingdom (FT-SE 100)20.3519.1811.6423.4812.217.28
    Canada (TSE-300)11.8614.9325.7425.197.687.17
    Source: Bloomberg Financial Markets L.P.

    Statistics on “buy-and-hold” positions understate the (annualized) returns earned on aggressive convergence plays. Market participants report that these convergence plays were very different from those implemented in 1992. In the earlier episode, convergence plays typically exploited yield differentials in spot markets in an environment of managed exchange rates. These spot market convergence positions would involve funding a long position in the higher-yielding bond with a short position in German Bunds, or even more simply—but more capital intensive—establishing a long position in the higher-yielding bond. In either case, if the yield on the long position fell in relation to German Bunds, then the capital gains on the position represented the excess return to the convergence position.

    By contrast, in the recent episode, convergence plays were more sophisticated in that the capital cost associated with establishing positions in cash markets was circumvented to a large extent by taking spread positions in the interbank swaps market. This was reflected in the sharp expansion in swap market activity in 1996: according to the International Swaps and Derivatives Association (ISDA), swaps activity denominated in deutsche mark—the “pay side” associated with convergence positions in the swaps market—soared by 44 percent in the first half of 1996.

    To illustrate how convergence plays are executed in swap markets, consider the following simple example. Suppose an investor believes that the spread between five-year lira and deutsche mark rates will narrow from its “current” level of, say, 300 basis points. The investor enters into a swap contract in which he or she agrees to pay a stream of deutsche mark fixed interest payments (calculated based on a given underlying notional sum) in exchange for a stream of lira fixed interest payments. One year later, suppose the spread on four-year lira and deutsche mark rates is 100 basis points, and to simplify things also assume that the lira-DM exchange rate has not changed—alternatively, the investor could have swapped out the currency risk of the original position. The investor could then unwind his position by entering into an offsetting swap in which he or she pays fixed lira rates for four years and receives fixed deutsche mark rates. In sum, the swap portfolio has no open currency position. However, the investor is receiving net lira income of 200 basis points guaranteed for the next four years.

    With substantial convergence in interest rates having occurred by late 1996, there was a measure of consensus in financial markets that, based on current fundamentals (including the available information on which countries would participate in EMU in 1999), the convergence process had largely run its course. Convergence positions were, therefore, largely unwound by late 1996, This does not imply that it was unlikely that there would be further narrowing of some spreads in the run-up to EMU, but rather that any further narrowing of spreads would hinge on new information about entry into EMU or further improvements in fundamentals.

    The most widely watched indicator of the “maturity” of the convergence process is the difference between current yield spreads and forward yield spreads after 1998 (as implied by current deposit and swap rate curves). For core EMU countries, spreads (over deutsche mark) on seven-year swaps at end-May 1997 were within 20 basis points of implied forward spreads in early 1999, and the levels of these spreads were small (Table 24). In other words, there is little room for further convergence in core Europe in the run-up to the introduction of the single currency in 1999, as the markets have effectively priced in participation in EMU by all core European countries in 1999.

    Table 24.One-Year and Seven-Year Interest Differentials with Germany, February 28, 1995, January 1, 1996, January 30, 1997, and May 31, 19971(In basis points)
    One-Year Interest DifferentialSeven-Year Interest Differential
    February 28, 1995January 1, 1996January 30, 1997May 30, 1997February 28, 1995January 1, 1996January 30, 1997May 31, 1997
    January 1, 199922915-22-23-25-27
    January 1, 2000230-4-29-28-26-29
    January 1, 20012254-26-21-25
    January 1, 1999242439815418992116
    January 1,200024153341051487898
    January 1, 2001238532281116
    January 1, 1999243531086854633
    January 1, 2000241026954493730
    January 1, 20012378254334134
    January 1, 19992206137173100123
    January 1, 2000215513115898111
    January 1, 2001 295112130
    United Kingdom
    January 1, 19992165122293289132103
    January 1, 20002157942112029963
    January 1, 2001214386146132
    European currency unit
    January 1, 199929179622514
    January 1, 2000270121014412
    January 1, 200125618-512
    Source: Bloomberg Financial Markets L.P.

    In Italy and Spain, spreads on seven-year swaps at end-May 1997 were about 40–50 basis points above spreads on implied forward swaps in early 1999. As current long-term interest rates are a function of short-term rates in the run-up to EMU and beyond, one can interpret the difference between current and implied forward long-term interest rate spreads as a gauge of the maturity of convergence only by first separating out the (possibly large) component of these spreads that is attributable to the convergence that has yet to lake place in short-term rates in the run-up to EMU. There is, in general, substantial room for convergence in short-term rates for the peripheral countries because risk premiums for credit and currency risks have become concentrated to a considerable degree at the short end of yield curves. Specifically, spreads over deutsche mark one-year deposits were less than 100 basis points for core EMU countries in May 1997 as well as for implied one-year swaps in 1999 and beyond, but were several hundred basis points for lira and peseta deposits in May 1997, and the implied forward spreads narrow by early 1999 by roughly 50 percent for Spain and Italy. Thus, the fact that the difference between current and forward swap spreads suggests that there will be considerable further convergence in the interest rates of the non-core EMU countries reflects to a considerable degree the fact that short-term rates are expected to converge sharply over the next few years.

    Trading positions in European fixed-income markets have most recently focused either on the prospects of participation in EMU by the various countries and exploiting what is considered to be overly optimistic or pessimistic pricing, or else on buying insurance (for spot positions) against an unraveling of the convergence process.3 A particularly notable instance of the former was the negative spread between French and German rates prevailing in late 1996 and early 1997. Strong domestic demand (especially by insurance firms) for French government bonds—caused by a large shift of funds from money market mutual funds into insurance products and administered savings products—reportedly was instrumental in pushing the French yield curve below the German yield curve.4 Market participants report that this caused a massive deconvergence position taking: these convergence and subsequent deconvergence trades are estimated to be among the largest speculative positions ever taken in the international capital markets—some estimates put U.S. hedge funds’ and proprietary trading desks’ positioning in the French franc-deutsche mark sector at end-1996 in excess of $50 billion. U.S. hedge funds in particular amassed considerable positions in forward swaps designed to profit from a narrowing in the negative spread of French franc forward rates to deutsche mark.

    Spreads in Dollar-Bloc Countries and Corporate Markets

    The terms “spread compression” and “convergence” also well describe developments in the bond markets of industrial countries outside Europe, in the corporate bond markets, and in the emerging markets. As discussed in Chapter II, the narrowing of spreads on emerging market credits in 1996 was no less impressive than that which occurred in the context of EMU. Spreads in the U.S. corporate bond market fell sharply for all credit qualities, and there was compression in spreads across credit ratings—high-yield (“junk”) bond spreads neared the all-time lows of the mid-1980s (see Figure 43). Spreads on Canadian, Australian, and New Zealand credits also narrowed dramatically (Figure 42), as reflected most clearly by the fact that the Canadian government yield curve through 10-year maturities slipped below the U.S. curve for the first time in two decades, and 30-year Canadian bonds traded at par with U.S. 30-year treasury bonds for the first time ever.

    This spread tightening in global bond markets occurred in an environment of low volatility in both foreign exchange markets and bond markets (see Figures 39 and 44, and Appendix 2 at the end of this annex). As discussed above under “Exchange Rates,” this should not be surprising, for the same factors that led to improved exchange rate fundamentals—low inflation and efforts toward fiscal consolidation—had similar consequences in fixed-income markets—as well as on volatility in these markets—as risk premiums decreased. A notable exception to this is Japan, in which volatility in fixed-income markets has been high in recent years (see Appendix 2). In the United States, although anticipation of a tightening of monetary policy caused periodic retrenchment from U.S. fixed-income markets and caused volatility to rise—as with the massive sell-off in early 1996—these concerns subsequently dissipated in the absence of compelling signs of inflationary pressures. When the Federal Reserve finally did raise the federal funds target rate by 25 basis points in late March, the markets had little reaction as the action had already been discounted in asset markets.

    Figure 44.RiskMetrics Daily Price Volatility for 10-Year Government Bonds, January 19, 1995–May 30, 1997

    (In percent)

    Source: J.P. Morgan.

    Fund-Raising in Fixed-Income Markets

    Stability in the major bond markets and convergence in the higher-yielding markets provided favorable conditions for issuance of debt securities by private sector institutions in the domestic and international markets. Issuance was strongest by private sector entities from the dollar-bloc countries and from the higher-yielding EU countries—particularly those that experienced the sharpest narrowing in interest rate spreads. Issuance from Japan and from core Europe was less robust. In addition, efforts toward fiscal consolidation had the effect of dampening overall new issuance activity in most domestic debt securities markets.

    In the international securities markets, net new issues grew by 73 percent in 1996, setting a new record (Tables 5, 25, 26), Dollar issues accounted for almost half of all new issues in 1996—compared with less than a quarter the previous year—and this trend has continued in 1997. In terms of growth rates, dollar-denominated issues rose 353 percent in 1996, compared with a decrease in the amount of new issues denominated in yen and deutsche mark. To some degree, the surge in new issues of dollar-denominated securities reflects the strong demand for funds by U.S. corporations and financial institutions, as issues by private sector entities from the United States rose 164 percent in 1996 and accounted for 25 percent of total new issues in 1996, compared with 19 percent in 1995. This suggests, however, that the strong surge in dollar issues reflects also issuance of dollar-denominated bonds by entities located outside the United States. It seems plausible that this is attributable to the demand by investors for dollar issues, as discussed above.

    Table 25.Outstanding Amounts of International Debt Securities(In billions of U.S. dollars)
    All countries2,037.82,441.22,802.53,225.93,240.7
    Industrial countries1,650.31,977.62,282.12,594.32,600.9
    Of which:
    United States176.9209.6272.8402.6418.1
    United Kingdom186.7212.5227.2274.2275.5
    Developing countries121.8161.9192.1276.3289.7
    Offshore centers111.318.319.436.339.4
    Source: Bank for International Settlements.
    Table 26.Outstanding Amounts and Net Issues of International Debt Securities by Currency of Issue(In billions of U.S. dollars)
    Net Issues
    Amounts Outstanding199619961997
    U.S. dollar836.4910.1983.71,245.931.573.474.2262.146.786.555.5
    Japanese yen272.3412.6496.7517.633.8106.8108.381.224.519.014.1
    Deutsche mark192.8244.0318.8347.131.227.555.154.810.012.613.9
    French franc92.7131.6149.0168.134.527.
    Italian lira37.757.569.799.713.018.410.427.33.711.712.1
    Pound sterling154.8178.2186.7237.331.714.510.130.83.717.222.9
    Canadian dollar81.783.583.777.020.56.7-2.1-6.3-2.6-2.9-0.2
    Spanish peseta10.610.713.217.93.5-
    Netherlands guilder44.965.984.595.37.914.813.518.
    Swedish krona3.
    Swiss franc149.1161.2189.0165.7-2.3-
    Belgian franc2.22.34.313.4-0.4-
    Source: Bank for International Settlements.

    Much of the growth in new issues in the international markets came from entities located in the United States and the United Kingdom. Issuance by other industrial countries generally increased as well, but at much slower rates. The strong demand for funds in the United States derived from both corporations and financial institutions, whereas in most other countries the demand for funds came mostly from financial institutions. This points to a more general development in the international securities markets: financial institutions are accounting for a larger percentage of new issues by industrial countries in these markets. Indeed, currently almost three-quarters of issuance in the international markets by industrial countries is attributable to financial institution fund-raising.

    In domestic debt securities markets, issuance grew at a much more moderate pace, just over 3 percent in 1996. Private sector issuance grew somewhat faster at about 11 percent, whereas public sector issuance fell in all of the major countries except Germany, Italy, Spain, and the United States. Among the major countries, private sector issuance was particularly strong in most dollar-bloc countries—Canada, Australia, the United Kingdom, and the United States. By contrast, issuance actually fell by private sector entities from most of the EU countries, although there was some modest increase in new issuance by entities in the higher-yielding EU countries.

    Risks in Fixed-Income Markets

    Low, and in many cases declining, inflation in the major countries, progress on the fiscal front, and low volatility in foreign exchange markets have created a favorable environment for fixed-income investors as well as lower borrowing costs for issuers. The bond market rally in 1996–97 rests on some assumptions about the future, which could change.

    First, the markets have priced in an improved outlook for macroeconomic fundamentals—most important, low inflation—in all the major markets. The balance of risks points to accelerating economic activity and thus a tightening of monetary policy in at least some of the major countries. In the United States, capacity constraints evidently have little slack, and economic growth in Japan and several European countries appears to be gathering momentum. The main risk to the compression of yield spreads—in Europe, in the U.S. high-yield market, in emerging markets, and in smaller industrial countries—is that cyclical factors will increasingly imply a significant global tightening of monetary policy. If, however, growth does not accelerate in Japan and the major continental European economies, a further tightening of monetary policy in the United Stales will result in a further reallocation of global fixed-income funds toward the dollar markets.

    In early 1994 the tightening of policy by the Federal Reserve Board was associated with one of the greatest corrections in global bond markets in history. Increased interest rates in the major industrial countries caused portfolio rebalancing in fixed-income markets away from the riskier markets, in part because of expectations of slower economic activity and thus a weakened ability of riskier credits to service their debts. In 1994, the correction was exacerbated by the unwinding of highly leveraged positions along the then very steep U.S. yield curve—when the Federal Reserve tightened monetary policy in February 1994, the rush to cover these short positions hastened and steepened the correction in bond markets. By contrast, market participants report that currently there is not a lot of leverage in the fixed-income markets. As noted above, there have been leveraged positions associated with the yen-carry trades, but these positions appear to have been unwound in 1997 because of the yen’s strengthening and concern that interest rates may soon rise in Japan.

    Markets have largely priced in the supposition that EMU will go ahead on time and that there is a high probability that many of the non-core countries will participate, if not in early 1999, then within the subsequent one to two years. It is widely believed that the major risk to EMU going ahead on time is that weak economic growth in Germany and France would prevent both countries from meeting the deficit criteria specified in the Maastricht Treaty. In such an event, proceeding with EMU by adopting a looser interpretation of the Maastricht criteria would open the door to a larger number of countries being eligible to participate in EMU from the start. The implications of continued weak growth and a larger number of initial participants in EMU could lead to increased volatility in European bond markets.

    Spreads in Europe, whether current spot spreads or implied forward spreads, could in general reflect two factors: first, improved fundamentals—lower inflation, fiscal consolidation, and currency stability—and, second, the possibility that the conversion of currencies into a new (blended) currency has the effect of imputing value to fixed-income assets denominated in the weaker currencies. Short-term yield spreads are still quite large for the traditionally high-yield countries, and this, in tandem with the fact that yield curves for these countries are much flatter than for the core EMU countries, explains why implied forward yield spreads are projected to narrow significantly over the next few years. An important uncertainly is whether this expectation of further convergence is attributable to further improvements in economic fundamentals or to expectations that EMU will impute a direct benefit—in terms of lower yields—to the higher-yielding currencies.

    Empirical studies have found that most convergence can be attributed to improved economic fundamentals.5 There is also a consensus among market participants that current spreads reflect improved inflation and fiscal outlooks in the higher-yielding countries, and by implication are not simply reflecting the expectation that some countries will benefit directly from a new currency that is supported by a firm monetary policy of the future European Central Bank (ECB). Nevertheless, concerns have been raised by market participants about two issues. First, it has been suggested that there is some risk that positions currently in place in European fixed-income markets could be sensitive to, say, a change in German monetary policy. In such a scenario, this could be associated with a sell-off in the U.S. markets to cover losses in the European swaps market. Second, market participants report that there exists a risk of the convergence plays on the high-yielding currencies unwinding if concerns arise in financial markets about the EMU process that call into question the future macroeconomic policies of peripheral countries. While an announcement of delayed entry for some of these countries is widely thought to imply some widening of spreads in these markets, such consequences could be managed so long as entry itself was not threatened.6

    International Syndicated Loan Markets

    Volumes and Margins

    Keen competition from securities markets and from within loan markets maintained high volumes and slim margins in the international loan markets (Figure 45, and Table 6). Stimulative monetary policy in Europe boosted liquidity in European banking systems and, in tandem with sluggish growth in Japan and continental Europe, this liquidity spilled over to the international banking markets, and from there into those countries with buoyant economic activity (e.g., the United Kingdom. Australia, and the United States). Syndicated lending volumes have been spurred by refinancing operations, funding associated with mergers and acquisitions (particularly in the United Kingdom), backup facilities associated with commercial paper and asset-backed securities issues, and project financing—including loans arranged as components of financing packages involving securities and bank loans. Overall, announced syndicated credits rose 68 percent in 1996 over 1995, or more than double the level in 1994. Syndicated loans to U.S. borrowers, however, soared 386 percent in 1996, an increase of $218.4 billion, which is more than the $213 billion increase in announced credits to all other countries combined.

    Figure 45.Spreads on Eurocredits1

    (In basis points)

    Source: Organization for Economic Cooperation and Development (OECD).

    1Weighted average of spreads (over LIBOR) applied to Eurocredits signed during the period. Tax-sparing loans as well as facilities classified under “other debt facilities” are excluded.

    Japanese banks continued their retreat from international lending markets in 1996, and this worked to hold back overall activity, especially in the interbank market: international lending by Japanese banks contracted by $20 billion in 1996, while fund-raising by Japanese banks in the international markets fell by $7.5 billion (Table 27 and Statistical Appendix Table A4). In terms of the share of international banking assets by nationality of banks, Japanese banks dropped 3 percentage points in 1996, which pushed their share to a 13-year low of 22 percent.7 Activity by U.S. banks in the international markets was buoyed by increased demand in the Eurodollar market and financing associated with the increased demand by foreigners for U.S. bonds. Much of the activity in international banking markets was associated with the aggressive pursuit of foreign business by European banks, and especially German. Italian, Benelux, Swiss, and U.K. banks. Market participants report that convergence plays associated with EMU have been an important factor in the increased activity of European banks.

    Table 27.Changes in Net Assets of BIS-Reporting Banks vis-à-vis Banks in Selected Countries and Regions1(In billions of U.S. dollars)
    All countries-38.373.795.4267.1-299.0206.8
    Industrial countries-22.787.122.2179.9-146.7172.151.9
    Of which
    United States28.13.874.1124.43.453.2-21.3
    United Kingdom-32.144.427.4116.0-141.988.719.0
    Developing countries-
    Middle East-
    Western Hemisphere43.9-
    Source: Bank for International Settlements (BIS).

    It has been widely reported that margins continued to come under pressure in 1996, but this is not reflected in data on weighted-average spreads for OECD countries (see Figure 45). These aggregate loan margins may mask the fact that lending activity has expanded into new areas geographically as well as to new, lesser-known names with reportedly slim risk-adjusted margins. Indeed, the evidence is clear that margins on loans to non-OECD credits did narrow in 1996, In any case, these considerations prompted once again warnings from regulators that diligence must not be ignored in extending credits at razor-thin margins. U.S. regulators, in particular, expressed concerns also with the lengthening of maturities and relaxation of covenants to higher-risk borrowers.

    Structural Developments

    A notable development in cross-border banking is an increased displacement of traditional interbank credit by repurchase agreements (repos). There are many varieties of transactions that could be classified as repos, but they all consist of a contract that functions as a collateralized loan, and an agreement to repay the loan—repurchase the collateral—by a specified time (typically less than a week). The U.S. repo market is the oldest and also the largest such market, not least because most other countries have only recently introduced repo markets—most often owing to deregulation of money markets. For instance, repo markets opened in 1997 in Germany and in 1996 in the United Kingdom.

    Only the international repo market comes close to the size of the U.S. repo market, a fact that reflects in large part the key role of U.S. institutions in the international markets and, thus, the integration of the U.S. domestic repo market with the international repo market. Recent estimates place outstanding repos at about $1 trillion in the international market—or roughly 10 percent of the stock of gross international bank lending—which is similar in magnitude to the U.S. market (Table 28).8 Assuming an average life of about one week, this suggests annual turnover in the neighborhood of $40–$50 trillion. More important, the growth rates of repo markets have been high: the U.S. repo market has grown at about 20 percent annually in the 1990s, and during their first year of operation, repo markets in the United Kingdom and elsewhere have expanded very quickly.

    Table 28.United States: Outstanding Repurchase Agreements (Repos)(In billions of U.S. dollars)
    Overnight and

    Continuing Repos

    Source: Board of Governors of the Federal Reserve System.

    The proliferation of repo agreements in interbank markets is attributable to several factors. First, there has been a heightened awareness of the credit risk associated with banks’ expansion into less familiar geographic markets and also perhaps increased concern about the credit risk associated with advancing traditional lines to some of the major banks active in the international markets. Second, with banks increasingly active in securities markets, the repo market provides access to cheaper short-term funds than uncollateralized funds. Similarly, the other side of the repo transaction—a “reverse repo” (a purchase with an agreement to resell at a specified price on a future date)—provides greater security because of the collateralization of the loan advanced. Third, (reverse) repo transactions are a more efficient use of a bank’s capital than traditional interbank lines: capital requirements on banks’ lending activities—including those of the Basle Capital Accord of 1988—provide for capital weights on collateralized loans corresponding to the weight attached to the collateral, which is zero for OECD government securities. Fourth, the increased emphasis on repos has fostered the development of standardized collateralization procedures and documentation which has, in turn, further facilitated the use of this method of interbank funding.

    The rapid proliferation of repos in interbank markets has resulted in a closer integration of securities markets and banking markets. This integration has also been promoted by the participation of institutional investors and investment banks in the international loan markets. This is evidenced by the prevalence of securities structures such as note issuance facilities—including commercial paper facilities and medium-term notes programs—that have reduced the distinction between bank loans and securities issues. Further, financing packages that entail a blend of securities financing and bank loans have also become common. The distinction between loans and securities will undoubtedly narrow as new tools are developed for pricing (and thus trading) credit risk and as mergers between commercial and investment banks proceed.

    The increased integration of securities markets and banking has also been facilitated by progress in securitizing loans, which has been fostered by the desire of many banks to economize on capital and thus pare loans from their balance sheets. The securitization of loans has proceeded along two dimensions. One dimension is establishing secondary markets for loans, which has already been implemented in the United States and is the objective of the recently formed Loan Markets Association in London. The second dimension is bundling portfolios of loans with the objective of securitizing the bundle. The issuance of these asset-backed securities (ABS) has grown rapidly in recent years. According to the Federal Reserve Board, in the United States asset-backed securities outstanding reached $738.1 billion in 1996. Although this represents less than 4 percent of all U.S. credit market debt—amounting to $19.9 trillion in 1996—its rate of growth has been two to three times that of overall credit market debt. Although asset-backed securities markets include the packaging of claims on bank loans to consumers and firms, credit card debt, and, in the international markets, the securitization of various types of receivables from developing countries, the market for securities-backed syndicated loans and junk bonds—collateralized loan obligations and collateralized bond obligations, respectively—are two segments of this market that have been particularly robust in the 1990s, after suffering a setback associated with the collapse of the high-yield bond market in the United Slates in the late 1980s.

    Equity Markets

    Recent Developments

    Equity prices in industrial countries rose sharply in 1996 and this trend has continued in 1997, In local currencies, European markets have generally been the star performers, with most markets comfortably outperforming the 20 percent advance in U.S. equity prices in 1996 and the further 14 percent gain in U.S. equity prices during the first five months of 1997 (see Table 23 and Figures 89). Some of the momentum in European equity markets has been attributed to improved prospects for the exporting sector associated with the depreciation of most currencies in continental Europe against the dollar. This depreciation of local currencies against the dollar is clearly reflected by the fact that European equity prices rose substantially less when measured in U.S. dollar terms, although they still generally posted gains on a par with the U.S. market.

    Underlying the strong performance of equity prices has been a downward trend in inflation and interest rates in the industrial countries, both of which are important for the discounted real value of future corporate earnings. In Japan, very low inflation and interest rates in 1996 were not sufficient to boost equity prices, largely because of lingering concerns about the health of the financial sector. On a risk-adjusted basis, the performance of Japanese equities relative to money market rates in recent years has lagged behind the performance of other major industrial country equities (Table 8 and Figure 8), although Japanese equity prices have shown evidence of strengthening in 1997.

    Among the major industrial countries, the U.S. equity market has clearly been the star performer in the 1990s (see Figure 8). Canada has trailed the U.S. market somewhat, and Japan has followed a different course after the collapse of its asset-price bubble at the beginning of the decade. In Europe, the U.K. market has been the strongest performer, reflecting the relative performance of the U.K. economy and thus corporate earnings. Equity prices have picked up recently in France and Germany, while following a more erratic course in Italy. The key questions that these very different performances raise are, first, whether there are indications of overvaluations in the United States given the duration of the market’s rally and, second, why the Japanese equity market has languished.

    Japanese equity prices have been weighed down by at least two factors. For one thing, there has been net selling pressure by important segments of the domestic investor base—life insurance companies, banks, nonfinancial companies, and equity investment trusts. Negative cash flow experienced by some of the major institutional investors in Japan (e.g., life insurance companies) has forced these investors to be net sellers of equities; in addition, weak balance sheets of other financial sector companies have led them to liquidate their equity holdings. Prior to the collapse of the asset-price bubble in 1989, investment trusts had more than half of their assets invested in Japanese equities. Since 1989, this share has fallen steadily, and by the first quarter of 1997 was just over 20 percent. This drop in portfolio allocation has been due in part to lower equity prices, but in recent years there have also been net redemptions from equity investment trusts. The large amount of liquidity in the nonfinancial corporate sector—which, as discussed above, has resulted in stagnant loan growth in the banking system—has not had the effect of supporting equity prices through share buybacks because of the current tax code. Indeed, if not for the net buying of Japanese equities by foreign investors and public institutions, Japanese equity valuations would have been even weaker.

    Second, lower equity prices have raised concerns about their potentially damaging effects on business and consumer confidence as well as on bank balance sheets. The latter concern stems from banks’ direct holdings of equities, their reduced ability to issue convertible bonds to raise capital, and the possible deterioration in both the quality of bank loans that weakness in equity prices might reflect as well as the value of the collateral underlying loans.9 Reflecting these concerns, bank stocks have underperformed the broader market by a substantial margin: from the beginning of 1996 to May 1997, the TOPIX index recorded a drop of about 10 percent, whereas the TOPIX bank index fell about three times as much.10

    A rebound in Japanese equity prices, therefore, rests on the resolution of some important sources of uncertainty. There can be little doubt about the positive prospects for Japan’s exporters—automobile, machinery, and electrical equipment companies—in the current environment, but the overall index is likely to be held back by financial stocks and the lack of local investor support to the market. A resolution of the financial system problems in Japan would undoubtedly be an important positive development for Japanese equity prices.

    In the United States, the rise in equity prices since early 1996 has been matched by stocks in many other industrial countries. However, this rise in the U.S. market came on top of larger and more sustained price gains in the first half of the 1990s. Indeed, since the beginning of the decade, U.S. equity markets have outperformed most other industrial country markets, in some cases by a factor of two or more. To some extent, these differences reflect the different stages of the economic cycle, with the U.S. expansion at a much more mature phase than that in Europe and Canada, It is also worth noting that over a longer historical period—going back to 1970—the performance of the U.S. market adjusted for exchange rate changes is quite similar to the performance of the French, German, and U.K. markets and still falls well short of the Japanese market (Figure 46). While the Japanese market has been particularly weak in the 1990s, it remains at relatively high levels given the very rapid price growth prior to the 1989 crash.

    Figure 46.Major Industrial Countries: Stock Market Indices1

    (Indices, January 1970 = 100)

    Sources: Bloomberg Financial Markets L.P.; International Monetary Fund, International Financial Statistics database; and The WEFA Group.

    1Monthly averages of daily observations from January 1970 through May 1997.

    Although improved profits and lower interest rates may account for a large part of the rise in U.S. equity markets, the size and pace of the upswing, as well as its longevity, have raised questions about possible overshooting. The Dow Jones Industrial Average has taken just 4 years to double in value since 1992, compared with a previous historical average of 17 years. From 1900 through early 1997, the Dow Jones index has risen at an average annual growth rate of just under 5 percent (excluding dividends), but it rose more than 25 percent in 1995 and then again in 1996. Indeed, about two-thirds of the market’s gain since 1970 has occurred in the 1990s, and about half since the beginning of 1995. Measured relative to GDP, U.S. market capitalization has increased from 69 percent in 1970 to about 107 percent in 1996.

    To put the market’s recent performance in context, it is helpful to compare it with the other great bull markets. Two previous periods of rapid increases in U.S. equity prices stand out. The first ran from October 30, 1923, through September 3, 1929, and the second great bull market ran from April 28, 1942, through January 18, 1966.11 During the 1923–29 period the Dow Jones index rose by 343 percent, which translates into a 25 percent annual increase. During the 1942–66 period the index rose 969 percent, which represents an annual growth rate of 10 percent. In comparison, the Dow has risen about 800 percent from August 1982 through early 1997, or about 15 percent annually. Focusing on a narrower window covering just the past two or three years would double this annual rate of increase. In any case, the current bull market is of the same order of magnitude as the two previous great bull markets. It should be borne in mind, however, that two of the greatest bear markets in the United States occurred after these two bull markets.

    Improved fundamentals clearly underlie an important part of the rise in U.S. equity prices since the beginning of the decade. First, reductions in inflation and interest rates are generally correlated with movements in stock values—Japan is the lone, notable exception in recent years. Second, the rise in corporate profits, absolutely and as a share of national income, since early in the decade corresponds to an important part of the rise in market valuation relative to the size of the U.S. economy. The share of corporate profits in national income has reached its highest level in 20 years (Figure 47). But the key question is whether current market valuations also reflect expectations of further increases in the share of profits that may prove unrealistic.

    Figure 47.United States: Corporate Profits and Market Capitalization1

    Source: The WEFA Group.

    1Corporate profits are before deducting the federal, state, and local taxes. Data for 1965:Q1 to 1997:Q1.

    On March 5, 1997, Federal Reserve Board Chairman Alan Greenspan addressed this question when he stated: “If you look at a normal pricing model for stocks, what you get is a not unreasonable level of prices if the earnings forecasts which the analysts are publishing are accurate.”12 These earnings forecasts are indeed high in a historical context, especially longer-term forecasts. In the mid- to late-1980s, five-year real earnings growth forecasts fluctuated between 5 and 6 percent. Since 1990, they have gone straight up and by the second quarter of 1997 they were about 10 percent.13 An important reason for these unusually high earnings forecasts was noted above: actual earnings have been remarkable in the past five years or so, and thus earnings forecasts have simply been brought into line with recent experience. Earnings for the Standard & Poor 500 companies in the first quarter of 1997 grew by about 15 percent over the level a year earlier, and a majority of companies’ actual earnings exceeded analysts’ forecasts.

    Risks to U.S. Equity Prices

    Perhaps the key question in assessing the sustainability of U.S. equity prices is whether companies can continue to grow their earnings at double-digit rates. Average five-year real earnings growth since 1960 has been just over 2 percent; this raises the likelihood that the balance of risks to earnings is below analysts’ forecasts. These risks derive from the observation that profit margins cannot continue to improve in an environment of near-capacity economic activity and tight labor markets. The unemployment rate in the United States reached a quarter-century low of 4.9 percent in April 1997.

    There are other reasons to be cautious about current valuations of U.S. equities (see Figure 10). First, dividend yields at about 2 percent have fallen to historic lows and are about half their long-term average. This fact may be partly explained by changes in tax laws and an increasing share of tax-exposed investors. Moreover, current dividend yields are not unusual in an international perspective: dividend yields in 1996 were under 2.5 percent in Japan, Germany, Italy, Canada, Sweden, Switzerland, Austria, Denmark, Finland, and Norway, and above 3.5 percent only in the United Kingdom. Australia. Belgium, and New Zealand.14 Second, the market-price-to-book ratio, and the closely related Tobin’s q ratio, indicate substantial departures from their historical ranges. Although largely explained by the shift away from capital-intensive industries to services, the rise in the ratio of equity prices to book value is still extraordinary. And third, the average price-earnings ratio, while not yet outside its long-term range, is clearly approaching the upper end. On the other hand, given the favorable interest rate environment, the equity-yield gap, which measures the difference between long-term bond yields and the inverse of the P/E ratio, remains within its historical range and, significantly, is still well below the levels reached prior to the 1987 stock market crash. Also, some analysts have argued that the dividend yield is now a less relevant indicator because many investors hold equities for capital gain rather than current income, and also that price-asset ratios have become less meaningful in view of the growing importance of service-based industries.

    Some market commentators have suggested that it is inappropriate to judge current valuations by comparing standard valuation indicators to historical ranges. The argument put forward is that the U.S. business cycle has fundamentally changed insofar as the length of economic expansions has increased while the length of economic recessions has decreased.15 Possible reasons for this structural change include just-in-time inventory management, trade liberalization, more flexible labor and capital markets, and financial market deregulation. Thus, the argument goes, corporate earnings have become less volatile, so the equity risk premium (over government bonds) has declined. Interpreting this argument in the context of the usual model for equity prices as representing the present discounted value of dividend payments, the associated decrease in the discount rate (from a drop in the risk premium component) can easily produce the conclusion that equity prices could increase sharply over a short period of time just owing to a modest decrease in the risk premium. Note that this hypothesis is equally applicable to explaining the narrowing in corporate bond yield spreads discussed under “Bond Markets” above. It is difficult to test this hypothesis as the risk premium is unobservable and estimates can vary widely. Nonetheless, although equity price volatility has shown some increase recently, earnings volatility has dropped sharply, which is consistent with this hypothesis (Figure 48).

    Figure 48.United States: Corporate Profits Volatility1

    Source: IMF staff calculations using data from The WEFA Group.

    1Calculated as standard deviation for the previous two years ending in the quarter shown of corporate profits (in percent of national income). Data for 1980:Q1 to 1997:Q1.

    On balance, it does not seem that U.S. equity markets are as far out of line as they were in 1987 or as Japanese markets were in 1989, when bond yields were high and rising, corporate earnings were much weaker, monetary policy was stimulative for several years leading up to the markets’ corrections, and general asset-price inflation was apparent, especially in real estate markets (see Appendix 3 at the end of this annex). However, they clearly are at levels that make them vulnerable to negative shocks in the form of higher interest rates, which played an important role in the last two major stock market crashes (the 1987 U.S. and 1989 Japanese corrections), or lower corporate earnings. Also, market volatility has increased significantly in 1996, though this appears to represent a return to more normal levels after a period of unusual stability (see Appendix 2). Finally, the rapid run-up in equity prices in late 1996 and into 1997—which seems more difficult to justify on the basis of improved fundamentals—also may suggest a cause for concern.

    One of the factors that has helped propel the U.S. market upward has been significant net inflows from small investors, routed mostly through the conduit of equity mutual funds (Table 29 and Box 5). From 1989 to 1995, household (direct and indirect) holdings of equities rose from 32 percent of all households to 41 percent, and equity holdings as a share of total financial assets of households rose from 26 percent to 40 percent.16 The net inflow into U.S. equity mutual funds from January 1995 through April 1997 totaled $424 billion, almost $300 billion of which occurred since the start of 1996. To put the role of equity mutual funds into perspective, in April 1997 they managed $1.88 trillion in assets, which is equal to 20 percent of the market capitalization of the New York Stock Exchange, the American Stock Exchange, and NASDAQ (the over-the-counter market) combined.17 The relative attractiveness of expected returns on equities, in comparison with other savings vehicles, has been at work here. But the historically low volatility of stock prices in recent years (see Appendix 2) may also have reassured investors. Recently, volatility, as reflected in options premiums, has been picking up (Figure 49). It remains to be seen how recent equity investors will react to an environment where equities appear more risky and turn in less spectacular gains.

    Table 29.U.S. Mutual Funds: Net New Cash Flow and Total Assets(In billions of U.S. dollars)
    Net New Cash FlowAssets
    To equity fundsTo bond and income fundsTo money market fundsTotalEquity fundsBond and income fundsMoney market funds
    Source: Investment Company Institute.

    Box 5.Trends in Funds Management

    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. In 1985, the 10 largest 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.1 Growth has been especially marked in mutual funds. U.S. mutual fund assets have risen at double-digit growth rates since 1970 when they amounted to just $48 billion.2 By the mid-1980s mutual fund assets had reached $495 billion, and by April 1997 they totaled $3,729 billion. Over the 1970–97 (April) period, the number of U.S. mutual funds increased from 361 to almost 6,500, and the number of individual accounts with mutual funds increased from about 11 million to 151 million. Although the institutionalization of savings, and especially (he shift by households from bank accounts toward mutual funds, has not been as marked in most other industrial countries as it has in the United States, the trend is apparent in other countries also and this process is widely expected to gather momentum in coming years.

    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. In part because the fund management business is a low-overhead business, the response of the industry to intensified competition for funds has been consolidation. This consolidation activity has been evident in two main features of the fund management business.

    First, in an increasingly global financial market, the importance of geographic presence has lessened, and thus fund management companies have responded to competitive pressures by consolidating their operations geographically. Global asset management companies are increasingly consolidating operations in one center, such as San Francisco, Boston, or London. For instance, Dresdner Bank, Germany’s second-largest bank, and Barclays Bank, the largest bank in the United Kingdom, both announced in late 1996 that they were consolidating their global asset management operations in San Francisco. This geographic consolidation has been facilitated by the ability of fund management companies to contract out aspects essential to the business of fund management, but which are distinct from the management of funds per se. In particular, the development of mutual fund “supermarkets” that offer the services of a wide variety of fund management companies at the retail branch level has led to a geographic separation of the fund manager and the investor in those funds. Similarly, fund management companies have increasingly contracted out back-office functions to third parties, which themselves may be geographically far removed from the fund managers.

    Second, there has been a great deal of merger and acquisition activity among fund management companies, particularly in the past few years. Fidelity Investments, the largest institutional investor in the United States, managed $426.7 billion in assets at end-1995, almost two-and-a-half times the assets (in 1995 dollars) Of the largest institutional investor in 1985, Prudential.3 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 1995 dollars) managed by the 300th-largest asset manager in 1985. This points clearly to a consolidation of assets, with the largest asset managers growing much more rapidly than the smaller asset managers. Although Fidelity’s growth in total assets under management slowed in 1996—it received just 10 percent of net equity mutual fund inflows versus 20 percent in 1995—it is a very large player, with some estimates attributing 12–15 percent of turnover in U.S. equities to Fidelity alone. In Europe, too, consolidation in the fund management industry has taken hold in recent years. In late 1996, two French insurance groups, AXA and UAP, announced plans to merge to create one of the world’s largest asset managers, with combined assets of $420 billion (end-1995 figures), rivaling Fidelity of the United States.

    So, while it is clear that consolidation is having profound effects on the size of the larger asset managers, can one conclude that investment assets are concentrating in the hands of just a small number of mammoth asset managers? The evidence does suggest that consolidation is working in this direction, but the pace is not as fast as might be imagined, particularly in the United States. In 1985, the top 10 asset managers accounted for 23 percent of the assets of the largest 300 asset managers, and this share was the same five years later.4 By end-1995, however, this figure had increased modestly, to 27 percent. Moreover, the top 100 asset managers increased their share of the assets managed by the top 300 asset managers by 9 percentage points over 1985–95, accounting for 83 percent of assets at end-1995. Similarly, in Europe, the top 10 asset managers increased their share of assets managed by the top 100 asset managers by 7 percentage points over the period 1991–95, from 31 to 38 percent of the assets of the largest KM). Consolidation activity, therefore, seems to have increased the relative size of the largest asset managers much more in Europe than in the United States. In the United States, consolidation activity has been more broadly based, increasing the relative size of the largest hundred or so asset managers. However, it is noteworthy that classification of asset managers geographically is becoming increasingly meaningless—as mentioned above, recently some large European asset managers have consolidated global asset management activities in the United States. Moreover, consolidation activity has increasingly been across borders, reflecting a tendency toward the evolution of global asset managers.

    In light of the forces affecting the fund management industry in recent years and the response of the industry to those forces, it is widely held that the outlook for the industry contains considerably more consolidation in the industry as well as geographically. An oft-painted scenario for the industry early in the next century is one in which there are a relatively small number of very large global companies each managing assets well in excess of $150 billion and a number of smaller management companies surviving in regional niche markets.

    1Institutional Investor (July 1996).2Data on U.S. mutual funds are from the Investment Company Institute.3The figures reported here and below on institutional investors are calculated from figures reported in Institutional Investor (various issues) and the IMF’s International Financial Statistics.4Institutional Investor (July 1996).

    Figure 49.Implied Volatility: S&P 500, Nikkei 225, and DAX Indices

    Source: Bloomberg Financial Markets L.P.

    Note: Implied volatility is a measure of the expected future volatility of the index based on market prices of the call options on futures on the index. The annualized percent rate of change plotted in the figure is a weighted average of the estimates of the implied volatility of call option futures.

    There are reasons to expect that a sharp correction in equity prices—which would reduce price-earning ratios to near the long-term average—need not have serious consequences for the U.S. economy. Consumer spending has not been driven sharply upward by the rise of the market during 1996, and a market fall back would presumably not induce a sharp fall in final demand. Moreover, the strength of the economy and financial institutions along with an improved securities market infrastructure all suggest the consequences of a significant correction could be managed without having major consequences for the health of the economy. Such expectations are consistent with the experience after the 1987 crash (see Box 6 on Circuit Breakers).

    Box 6.Circuit Breakers

    As equity markets around the world continue their upward momentum, many of them reaching new heights, concerns about possible overshooting and a subsequent sudden drop in equity prices are beginning to emerge. Following the equity market crash of 1987, many countries instituted various forms of circuit breakers. Ten years later, these countries and others are taking a second look—deciding whether to reset the conditions for their use or whether to adopt some form of them for the first time. Despite their growing use, especially among emerging market countries, there continue to be misconceptions about the purpose and effectiveness of circuit breakers. Circuit breakers are only a temporary measure for reducing market volatility or unidirectional price movements.1 If fundamental information is the basis for the price movement, and not features of destabilizing trading strategies or panics, then circuit breakers simply slow the eventual price movement: they do not reverse it.

    The two most common circuit breakers are trading halts and price limits. Trading halts can be initiated in two ways: a specialist, or other exchange official, may have the authority to halt trading; or a trading hall may be imposed after a price change of a given amount or percentage. The length of the halt can be pre-established or can be discretionary. Price limits place bounds on the price change but do not limit the period of nontrading: if a price hits a limit, trading beyond that price limit cannot take place; only when prices fall back within the limits can trading resume. Another type of circuit breaker is a limitation on the types of trades or strategies that can be initiated during a period of high volatility, or even in a normal period. For example, some exchanges routinely disallow short sales unless the price has experienced an “uptick.” that is, the price had to have moved up by the trading increment. Another example is that all index arbitrage trades must be executed through an electronic system that delays their execution by five minutes when price changes exceed a given amount.

    To choose the circuit breaker mechanism that will be most effective, it is important to define the goals of the circuit breaker and to assess the surrounding environment. In most cases, the goal of circuit breakers is either to dampen price movements caused by speculative activity or to slow down the price effects of trading strategies that are thought to have destabilizing or overshooting effects (e.g., portfolio insurance, the dynamic hedging of options). Even when destabilizing speculative activities or trading strategies are absent, there is often a belief that sharp movements in prices, regardless of their cause, are likely to engender a panic mentality, causing investors to act irrationally, further reinforcing existing price movements. Similarly, even when price movements reflect underlying fundamentals, a limit on the maximum amount lost in a given period may allow participants who would not have been able to pay for their losses had the full price decline occurred to pay on a timely basis. Regardless of which type of circuit breaker is chosen, to operate effectively the market needs to be centralized, information needs to be disclosed during the halt, and there needs to be a well-known method for the resumption of trade.

    Trading halts can only truly hall trading when trading is centralized. In the United States, for example, when trading halts were introduced after the 1987 market break, close coordination between the stock exchanges and the futures exchanges, on which associated futures contracts were traded, was required. In October 1996, the Dhaka Stock Exchange in Bangladesh instituted a circuit breaker to limit price movements to a daily 5 percent, only to have its effectiveness undermined by the unofficial curb market where no such impediment to trading could be maintained.

    An integral element to using any circuit breaker mechanism is the disclosure of information. It is imperative that market participants learn something during a trading halt that helps them determine the instrument’s price: both fundamental information and order flow information are important. Depending on the trading mechanism, either indicative quotes or postings of bids and offers and the amounts underlying them should be given at intervals during the halt to provide information about order imbalances. If open outcry is used, market participants should freely announce their willingness to buy or sell at various prices.

    In addition, there needs to be an established and well-known method for resuming trade. A single call auction, whereby a specialist gathers the bids and offers over a set period of time and establishes a market-clearing price at which all the existing orders receive the same execution price, is thought to be one of the most equitable. It helps to relieve an element found in most panics—the desire to get an order executed before the price falls farther.

    In addition to the immediate microstructure issues surrounding the trading environment, there are also infrastructure issues that may require attention. When clearing and settlement procedures are not well established or take extended periods of time to operate, uncertainty regarding the solvency of the participants can arise, limiting liquidity and participation when it is most needed and, in some instances, inhibiting the use of the exchange entirely as a venue for trading.

    As an alternative to circuit breakers, share repurchases by corporate issuers may help stem a dramatic price decline in equity markets. At some point, a firm may deem the price of its stock low enough to buy it back and reissue it at a later date for a profit, thereby obtaining additional equity from the market. Share repurchases signal to the market that the firm, with inside knowledge of its value, believes that the shares are undervalued. Similarly, purchases from a major participant in the market can show confidence and help inhibit further sales. To allow these mechanisms to operate, a country may need to relax or eliminate restrictions regarding corporate repurchases.

    Other institutional features that may reduce the incidence of a crisis include (1) restrictions on bank lending for stock purchases by requiring various amounts of collateral;2 (2) better audit trails to detect market-trading abuses, such as price manipulation that may start a panic; and (3) education of market participants, especially small retail investors, to enable them to understand the practices and procedures surrounding trading during normal limes as well as the different procedures that may occur during stressful periods.

    While certain types of circuit breakers may achieve some goals, they are all impediments to a freely functioning market—preventing buyers and sellers from executing trades at mutually agreed prices—and have some deleterious effects. Circuit breakers may limit trading by participants that are attracted only by large price moves, and hence eliminate a stabilizing factor. In most instances, the existence of circuit breakers is likely to alter participants’ behavior around their imposition. For instance, a “magnet effect” may occur when participants recognize that as the price approaches a price limit they will be unable to execute their desired trades and so they execute early. Alternatively, an opposite “repelling effect” may occur when participants prevent the limit from being hit because they know their ability to trade will be impaired. While these behavioral trading effects are certainly present, the most basic criticism of circuit breakers is that when fundamental information implies a large price movement, circuit breakers merely lengthen the time involved in obtaining the new price level.

    1In fact, some empirical studies find that circuit breakers may increase volatility (Lauterbach and Ben-Tsiyon, 1993; Lee, Ready, and Seguin, 1994).2The collateral should not be the same as the instrument being purchased because this would reinforce price movements. For example, if a bank loan is used to purchase equity and the collateral underlying the loan is also equity, a fall in equity prices means that, to maintain the collateral, the borrower needs to sell equity in an already falling market, adding further pressure on equity prices.

    Derivative Markets

    Recent Developments

    Growth of the global derivative markets during the past decade has been phenomenal both in the exchange-traded sector and in the over-the-counter markets (Statistical Appendix Tables A5A7; see also Tables 910). During the period 1986–96, the annual trading volume of exchange-traded contracts—including interest rate futures and options, currency futures and options, and stock market index futures and options—nearly quadrupled, reaching 1.16 billion contracts at end-1996. The growth of these markets has been even larger when measured by outstanding notional principal: the average annual growth rate of outstanding notional principal of exchange-traded contracts has been 32 percent over the past decade and stood at $9.9 trillion at end-1996. The growth and the size of the over-the-counter markets is even more impressive: the notional principal of outstanding currency and interest rate swaps and interest rate options reported by members of the International Swaps and Derivatives Association (ISDA) rose from $0.9 trillion in 1987 to $24.2 trillion in 1996, representing an annual average growth rate of 45 percent.

    More comprehensive surveys of the current size of global derivative markets paint an even more striking picture.18 According to the survey conducted by the Bank for International Settlements in early 1995, the notional value of outstanding OTC foreign exchange, interest rate, equity, and commodity derivative contracts totaled $47.5 trillion (after adjusting for double counting and including estimated gaps in reporting) at the end of March 1995. About 98 percent of this total is accounted for by interest rate ($28.85 trillion) and currency derivatives ($17.7 trillion). In addition to OTC derivatives, intermediaries that were involved in the survey coordinated by the BIS reported that they were engaged in a further $16.6 trillion of exchange-traded derivatives. In aggregate, therefore, respondents to this survey of users of derivatives in 26 countries revealed that (after adjusting for double counting) they were involved in about $64 trillion, by notional principal, of derivative contracts. To put this in perspective, the aggregate market value of all bonds, equities, and bank assets in Japan. North America, and the 15 European Union countries totaled $68.4 trillion at end-1995, which is about 7 percent larger than the size of derivative markets as measured by the above survey.

    Considering the sheer size of the derivative markets, sustained growth rates of the magnitudes reported above are unprecedented in global financial markets. This fact clearly points beyond purely cyclical influences on derivative volumes, and in particular to the importance of structural factors fueling growth of these markets. Hedging and position taking associated with cyclical conditions are surely key motives for the use of derivative financial instruments to begin with—as reflected clearly in the sharp rise in the volume of interest rate products during the 1994 global bond market correction, for example—but the sustained growth of these markets does not derive from cyclical influences per se. Rather, the key structural factors influencing the growth of these markets are an increased understanding by financial and nonfinancial institutions of the capabilities these instruments offer for repackaging and reengineering major cyclical and balance-sheet risks, in tandem with technological, analytical, and numerical advances in pricing and evaluating the risks of derivative contracts.

    Although the markets for derivative financial instruments, particularly in the major industrial countries but increasingly also in the developing world, are large by any measure, it is clear that sustained growth of the global derivative markets is not likely to abate soon. This momentum derives both from the major countries—where derivative markets for some of the largest risks, such as credit risk, are far from mature—as well as from the newer markets in the developing world—which have recorded the most dramatic growth in exchange-traded derivatives activity in the 1990s. Further growth in the global derivative market will present challenges to private risk management technologies and to supervision and regulation. Nonetheless, these markets, and the instruments traded in them, are increasingly better understood by dealers in the over-the-counter markets and by the financial and nonfinancial institutions that are the end users of derivative products.

    Structural Changes in Derivative Markets

    Three key structural changes have accompanied the rapid growth of derivative markets. First, derivatives have increasingly become a low-margin, high-volume business. This commoditization of derivative markets does not necessarily imply that the markets are becoming less personalized, or more centralized, but rather simply that the products traded in these markets have become familiar to market participants, and these markets have also become concentrated in well-understood, or “standard,” instruments. Some have attributed this standardization of products to the huge losses incurred by financial and nonfinancial enterprises associated with large derivative positions—including Orange County, Procter & Gamble. MG Corporation, and Barings. Although these losses were not always associated with positions in exotic products, the magnitude of the losses provoked an awareness and reevaluation of the purposes and risks of derivatives in general. The immediate consequence was a widespread withdrawal of demand for exotic, highly leveraged structures and a shift toward simpler structures—especially currency and interest rate swaps—and a refocusing on the risks and benefits of using these instruments. Some structures that have traditionally been regarded as exotic—notably, digital and barrier structures—have become mainstream products. This category of derivative products has its payoff tied—often in a binary fashion—to whether an underlying asset price reaches some trigger level; it has become mainstream because it facilitates the trading of volatility of asset prices directly. Products in this category accounted for an important share of activity in the currency and interest rate segments of the over-the-counter markets in 1996 and 1997.19

    Second, derivative markets are consolidating, planting the seeds for an integrated, concentrated global market. In the over-the-counter market, the concentration of activity is already large. For instance, U.S. commercial banks had $20 trillion notional principal of derivatives on their books at the end of 1996, of which 8 banks accounted for 94 percent and the top 25 banks accounted for 98 percent.20 This concentration of derivatives activity among a small number of institutions in the United States mirrors the reality in global derivative markets, in which these same U.S. institutions account for a large share of OTC derivatives activity. Although it remains to he determined which, and how many, institutions are able to establish themselves as truly global financial institutions, it is clear that OTC derivatives activity has increasingly become concentrated among the handful of institutions that arc best positioned to be global institutions, as these institutions are best able to intermediate risk management needs on a worldwide basis.

    This globalization of derivative markets is also apparent in the exchange-traded segment of derivative markets. This is reflected most clearly by the proliferation of trading links among both major and smaller exchanges in all regions of the world. For example, the major U.S. derivatives exchanges—the Chicago Mercantile Exchange and the Chicago Board of Trade—have established links with foreign exchanges, and discussions continue between the two main U.S. exchanges over the possibility of a direct link between them. In 1996 alone, the Chicago Mercantile Exchange unveiled alliances with the Marché à Terme International de France (MATIF). London International Financial Futures Exchange (LIFFE), and Deutsche Terminbörse (DTB), the three major derivatives exchanges in Europe. For instance, the link between the Chicago Mercantile Exchange and the DTB permits DTB trading screens to be placed directly on the floor of the Mercantile Exchange for trading in German stock index (DAX) futures, while in the case of MATIF and LIFFE, the Chicago Mercantile Exchange is permitted to trade short-term European interest rate products after closing hours in Europe (an arrangement that mirrors the alliance between the Chicago Board of Trade and LIFFE for longer-term interest rate products).

    Within Europe, consolidation of exchange-traded derivative markets has occurred as a result of trading links as well, but it has also been reflected in mergers and closures of exchanges. For instance, the Swiss (Soffex) and German (DTB) exchanges announced in late 1996 a strategic alliance that will create a common technical platform for trading derivatives and integrate the two clearing and settlement systems, and in 1997 the Danish and Swedish exchanges announced plans to merge their dealing systems. In London, LIFFE and the London Commodity Exchange have recently been merged, and in Ireland the Irish Futures and Options Exchange was closed in August 1996.

    As discussed in Annex IV, EMU threatens to eliminate much of the trading in exchange-traded derivative products in Europe, and thus the three major derivatives exchanges—DTB, LIFFE, and MATIF—are aggressively trying to capture market share ahead of the introduction of the euro. Specifically, the major exchanges have accelerated the introduction of new products in an attempt to establish leadership in the interest rate instruments that are likely to be at the core of fixed-income markets with a single currency. On the heels of the launch in late 1996 of Euromark futures by LIFFE and DTB, a working party of MATIF members suggested in late 1996 that a three-month contract in euros should be introduced by April 1998 and suggested as well the introduction of one-month. 5-year, and even 30-year euro contracts in order to establish a position in a broad range of the euro yield curve.

    A third structural change in derivative markets is that OTC markets are fast becoming the cornerstone of global derivative markets. In 1987, the notional principal of outstanding OTC interest rate and currency swaps was 44 percent larger than the global exchange-traded market. By 1990, the relatively faster growth of the OTC market pushed its size to 51 percent larger, and by 1995 it was 65 percent larger. In 1995, trading volume on the major North American, European, and Asia-Pacific derivative exchanges actually contracted, whereas in the OTC markets currency and interest rate swap activity soared—in terms of notional principal outstanding, it expanded by over 40 percent, reaching $15.2 trillion. The combination of renewed focus on EMU and intense focus on the direction of interest rates in several of the major countries contributed to a slight rebound in activity on exchanges in 1996—particularly in Europe—but the major arena of growth in global derivative markets in 1996 was once again the OTC markets.

    There are at least four factors underlying the growth in the OTC markets’ share of the derivative business. First, the flexible, personalized nature of the OTC market gives these markets natural advantages in terms of arranging suitable packages of products for customers. Second, the OTC markets hove adopted those features of exchange-traded markets that are valuable. For example, in October 1996, 11 of the major players in the swaps market established a swaps collateral depository—the Chicago Mercantile Depository Trust Corporation—which will standardize and automate the process of managing collateral and manage payments netting, trade valuation and administration, and global reporting to dealers involved in OTC transactions. Third, the large risks for which derivative instruments have not yet been fully developed are most likely to be successful in OTC markets. Most important, the market for credit risk derivative contracts—in which banks in particular could trade loan credit risk—is inherently a highly heterogeneous product market (see Appendix I to Annex III), which exchanges are not conducive to handling efficiently. Finally, the OTC derivative markets have an important regulatory advantage over the exchange-traded markets. Specifically, whereas the Commodity Exchange Act of 1974 gives the Commodities Futures Trading Commission regulatory authority over the derivative exchanges, the so-called Treasury Amendment effectively exempted from CFTC oversight certain financial futures traded off of U.S. exchanges. There has been a great deal of uncertainty about the extent of the amendment’s reach, which has in turn prompted numerous lawsuits and legal uncertainty. Legislation introduced in early 1997 would, among other things, limit the CFTC’s oversight of OTC markets only with respect to foreign currency products21 and would also permit the exchanges to establish separate, unregulated markets that are restricted to institutional investors. If passed into law, this latter feature would reduce the competitive advantage of the OTC markets, but the significance of this effect is unclear.

    The increasingly central role of the OTC markets is reflected also by the approach that market participants took in establishing convergence positions in 1996 in Europe. As noted previously, these convergence positions were heavily concentrated in the swaps market, which pushed deutsche mark-denominated swaps activity up 44 percent in the first half of 1996. Earlier in 1996 these positions entailed paying deutsche mark and receiving a higher-yielding currency. Perhaps the most notable positioning in the context of EMU occurred later in 1996 and focused on the French franc versus the deutsche mark yield spread. Earlier in 1996, convergence positions in this market were mounted on the belief that this yield spread would narrow. However, the French forward curve actually traded below deutsche mark’s later in the year, which in turn caused very large deconvergence position taking—estimates and commentary from market participants suggest that the magnitude of these deconvergence trades made them some of the largest speculative positions ever mounted in international capital markets, with some estimates putting positions amassed by U.S. investors alone in excess of $50 billion at end-1996.

    Developments in Systemic Risk Management

    As financial markets become more integrated and increases in technology and telecommunications permit risks to be unbundled and managed on a more centralized basis, regulation will need to adapt to the changing environment of the institutions under its purview. This trend calls for a more centralized, or at least a more coordinated, form of regulation—or as Federal Reserve Board Chairman Greenspan put it, “regulation must Fit the architecture of what is being regulated.”22

    A trend toward the removal of regulatory barriers separating financial institutions and toward more integrated regulatory structures is being considered in a number of countries. In the United States, the Treasury Department has proposed a financial sector restructuring that would permit further involvement of commercial banks in the securities and insurance businesses. The Federal Reserve Board has made it known that it would favor a repeal of the Glass-Steagall Act and has already relaxed a number of restrictions placed on commercial banks’ Section 20 subsidiaries, the organizational structures originally permitted to carry out limited nonbanking businesses. In the United Kingdom, the new Labor government has proposed to house all financial institution regulation, including bank supervision, under one roof, within the Securities Investment Board. Japan, too, is introducing financial sector reforms that will allow linkages among banking and securities market activities and establish an independent agency, the Supervisory Agency for Financial Entities, charged with overseeing bank supervision as well as supervision of other financial institutions.

    Along with proposals for consolidated new regulatory structures are efforts to increase the coordination among existing regulatory and supervisory entities. The number of bilateral memoranda of understanding (MOUs) signed by various regulatory agencies continues to increase dramatically. Recently a number of countries have signed multilateral MOUs assuring that their information sharing and emergency procedures are mutually consistent. Along these lines is the ongoing work of the Joint Forum, a group made up of bank, securities firm, and insurance regulators previously called the Tripartite Group. The Joint Forum is charged with the improved regulation of financial conglomerates in a global environment. While the forum has made some headway over the last year in facilitating the exchange of information among the groups of supervisors, the establishment of a definition for a “lead regulator” has eluded them so far. In many countries, the regulation of banks, securities firms, and insurance companies is executed in an isolated way and fears of losing influence in a merged regulatory entity or to another regulator have long dominated the political debates in these countries. The forum’s latest progress was reported to the G–7 Summit Meeting in Denver.

    As the international regulatory community attempts to find more efficient ways to ensure a systemically sound financial system, more reliance is being placed on self-regulation and market discipline. As public disclosure is the cornerstone for market discipline, efforts to harmonize accounting standards and increase meaningful disclosure are accelerating world wide.23 For instance, to enhance the functioning of market discipline by setting accounting and disclosure standards for all corporations, the International Accounting Standards Committee (IASO) has agreed to complete a core set of International Accounting Standards by March 1998, 15 months ahead of its scheduled release. If acceptable to the International Organization for Securities Commissions (IOSCO) and the various national regulatory bodies represented on its committees, this set of standards will pave the way for increased cross-border offerings and listings and other international capital flows. Note that IOSCO endorsement would permit companies outside of North America and Japan to have access to those capital markets without the cost or confusion caused by restating their accounts.

    The U.S. Financial Accounting Standards Board (FASB) is pursuing a longer-term project that is devoted to studying the conceptual and measurement problems associated with establishing a “fair value” for all financial assets and liabilities appearing on the balance sheet. Although the United States usually leads developments, both the International Accounting Standards Committee and the United Kingdom’s Accounting Standards Board (ASB) are hot on the heels of this development. By March 1998, the IASC intends to have in place, as part of its comprehensive set of standards, a standard that requires all financial instruments to be carried on the balance sheet at their current value. The ASB has also promulgated a discussion paper on the topic. The move toward fair value accounting, whereby financial instruments on the balance sheet will be carried at their current values, is now broadly based, with the FASB, the ASB, and IASC all on board. While it will not happen overnight, balance sheets are likely to provide a fuller picture of the financial health of corporations in the future.

    Private sector financial institutions have already responded to increased pressure, both from regulators, but more important from other market participants, to publish additional information about their risk-taking activities. The Basle Committee and IOSCO jointly released a study in November 1996 examining the latest round of derivatives disclosures of banks and securities firms. Last year, for instance, a number of banks voluntarily disclosed more than is legally required of them. Financial institutions recognize that well-informed investors are more willing to provide funding and capital than poorly informed ones.

    Of course, to report reliable information regarding earnings and the risks undertaken to obtain them, an institution must have the means to calculate the information. Information systems and risk measurement techniques are thus continuing to absorb an increasing share of the resources of most financial institutions. One U.S. accounting firm, utilizing the expertise of a number of private sector experts and regulators, has introduced a set of “Generally Accepted Risk Practices” to help coordinate an assessment of a firm’s risk management system. The comprehensive framework comprises 89 core principles, grouped into the following categories; risk management strategy; risk management function; risk measurement, reporting, and control; operations; and risk management systems. The intention is to provide a robust risk management framework for banks, securities houses, and other financial institutions that will address the whole range of risks faced by these firms. The proposed practices could be used as a benchmark against which private sector auditors would judge firms’ risk management systems, perhaps endorsing or certifying those that exceeded the benchmark. This is one type of self-regulation, potentially enhancing market discipline, that is part of a more general trend toward increased risk disclosure.

    In sum, supervisory and regulatory developments mirror the trends occurring in global capital markets—increased consolidation of regulatory structures accompanied by increased coordination. Moreover, there is a movement toward more reliance on mechanisms and rules that are “market friendly,” that is, ones that encourage self-regulation and reinforce market discipline.

    Appendix 1 Extracting Information from Options Prices

    Forward rates have long been used as indicators of market expectations. Adjusted for risk, these rates can be interpreted as the market’s assessment of the mean of the distribution of possible future values for the underlying price. New techniques based on options prices offer a refinement on this information by providing market-based indications of the probabilities associated with different ranges for the future value of the underlying price, thus lending some context to the mean interpretation of the forward rate.

    Options prices can reveal this added information because of the unique way their ultimate value depends on the price of the underlying security. An option has value at its expiration date (expiry) only if, on that date, the price of the underlying security (S) falls within a particular range. If S is outside this range at expiry, the option is worth nothing. The range of positive value is determined by the option’s type—put (sell) or call (buy)—and its strike price. For a call option to have value at expiry, S must lie above the strike price. For a put to have value, S must lie below the strike price. If S is inside this range at expiry, the value of the option is always positive and equal to the difference between S and the strike price.

    Prior to expiry, a call option’s price will reflect the market’s current assessment of the probability that 5 will lie above the call’s strike price and the market’s assessment of how far above the strike price S is likely to be, assuming that at expiry it does lie somewhere above the strike price. Analogously, a put’s price is determined by the probability that S will lie below the put’s strike price and the market’s assessment of how far below the strike price S is likely to be, assuming that it is somewhere below the strike price.

    There are now several techniques to extract market-based probabilities from options prices and they all rely on the fact that options with different strike prices reflect the probabilities associated with S falling within different ranges. By comparing the prices of several options that differ only in strike price, one can infer the probabilities associated with S being within these different ranges at the options’ expiration date.24

    When there are many options on a given instrument, so that the strikes are relatively close together and they range over a significant interval of the possible outcomes, the inferences one can draw about the distribution become quite detailed. Nonetheless, even when there are many options, the fact that the strikes do not range over all possible outcomes means that some a priori assumptions must be made about the distribution before it can be estimated from observed option prices. One such assumption is that the distribution comes from a particular family of distributions with unknown parameters. The unknown parameters are estimated by finding those that best explain the observed options prices—often by least squares.25

    Figure 50 gives an example for the yen-dollar exchange rate for early September 1997 as estimated on May 20, 1997. It was estimated from options on the futures traded on the Chicago Mercantile Exchange, under the assumption that the distribution can be described by a weighted average of three lognormal distributions.26 From this estimated distribution one can compute market based probabilities associated with the exchange rate being within specific ranges. For example, on this day the market priced the options as though it assigned a 15 percent probability to the yen-dollar rate being below 104 in early September and an equal probability to the rate being above 118.3. The mean of the distribution, at 111.3, is below the peak of the distribution, reflecting a skew toward relatively large appreciations of the yen against the dollar.

    Figure 50.Distribution for Yen-Dollar Exchange Rate in Early September 1997 Implied by Options Prices on May 20, 1997

    Source: Bloomberg Financial Markets L.P; and IMF staff calculations using data from the Chicago Mercantile Exchange.

    There are two important caveats attached to any technique that extracts probabilities from options prices. The first concerns inferences about the shape of the distribution in the range above the highest strike price and below the lowest strike price. For the day plotted, the lowest available strike price was 98. The only information the options data provide about the distribution below 98 is the probability of S being below 98 and the mean of this portion of the distribution (E[S | S ≤ 98]). If the mean of this portion of the distribution were 96, for example, then the options data alone could not distinguish between a distribution that assigned this probability evenly over the range from 95 to 97; a distribution that assigned this probability evenly over the range of 94 to 98; and any other distribution below 98 that has a mean of 96. For many purposes, such as the computation of the 15 percent confidence limits given above, the distinction between these “observationally equivalent” distributions is irrelevant. Nonetheless, it is important to note that the particular shape of the distribution above the highest strike and below the lowest strike is largely determined by the assumptions one makes about the functional form of the distribution.

    The second caveat pertains to the interpretation of the estimated distribution. As noted above, the distribution implied by options prices can provide a context for the mean interpretation of the forward rate. As with forward rates, option prices incorporate market participants’ preferences, or attitudes toward risk, as well as their beliefs about the possible future values of the underlying price. Thus, the probabilities calculated from options do not reflect market participants’ beliefs alone. Instead, they reflect how much market participants are willing to pay to insure against certain outcomes, which incorporates both the probabilities attached to these outcomes as well as the costs associated with them. Without detailed information about the preferences and portfolio holdings of market participants, it is impossible to disentangle the influence of preferences and beliefs. (This difficulty arises with any inference made from financial market prices and is not confined to forwards and options.) Nonetheless, the information in options prices provides a glimpse at the range of possible outcomes that market participants consider possible and how much they are willing to pay to insure themselves against these various outcomes.

    Appendix 2 Have Securities Markets Become More Volatile?

    As a measure to dampen volatility in the U.S. equity market, the New York Stock Exchange begins to limit computer-guided trading once the Dow Jones Industrial Average moves by 50 points during the day. In 1996, the Dow Jones Industrial Average changed by more than 50 points from the previous day on 56 days. Compared with previous years, 1996 was truly unusual in that 50-point changes in the Dow Jones index were much more common. While this phenomenon is partly due to the rising index level, the large market moves along with worldwide financial market deregulation and the increase in international capital flows have fostered a widespread belief that volatility has increased in recent years. Critics have pointed to the introduction of derivative instruments with complex, nonlinear payoffs, and uncertain macroeconomic policies for additional causes of increasing securities market volatility.

    Can one conclude that equity markets, and in particular, the U.S. equity market, have become more volatile recently.’ Similarly, the “bond market massacre” of 1994 has lead some observers to conclude that bond markets have become increasingly volatile in recent years. Is that a reasonable assessment?

    Consider first the volatility of the U.S. equity market, as measured by the standard deviation of weekly percentage changes in the Standard & Poor 500 index (top left panel of Figure 51).27 Comparing recent volatility levels to the postwar record clearly shows that current volatility by this measure is well within the historical range of variation. The U.S. stock market actually displays less volatility now than any time since the early 1970s, and the period following the first oil crisis remains the most volatile since the 1930s (not shown). Any long-lasting effects from the most recent upsurge in volatility following the 1987 crash seem to have all but vanished.

    Figure 51.Historical Stock and Bond Price Volatility in the United States, Japan, and Germany1

    (In percent)

    Sources: IMF staff calculations based on data from Bloomberg Financial Markets L.P.; and The WEFA Group.

    1Weekly standard deviation in stock and bond price changes, computed as follows: standard deviations are calculated from a weighted moving average of past weekly squared returns. The weights decline exponentially starting with a coefficient of .01 producing relatively smooth curves designed to highlight long-run changes. Returns are weekly percentage changes in prices. Bond prices are calculated assuming a coupon rate equal to the yield.

    In Japan and Germany, current volatility in the equity markets is not large compared with historical levels, and in both countries the recent trend in volatility is downward sloping. The recent boom and bust of asset prices in Japan have increased volatility but only from a historical low in the early 1980s to a level now just slightly above the average for the last 45 years. Stock market volatility in Germany peaked around the 1987 crash and again in the aftermath of the reunification and the Gulf War of 1990 but has clearly tapered off in recent years.

    Next, consider bond markets, where volatility is generally lower. Events that change investors’ beliefs about future inflation typically trigger bond market volatility; oil price shocks and changes in monetary policy are prominent examples. The volatility in weekly long-term U.S. government bonds (top right panel) topped around the shift in monetary policy in 1980, and bond price volatility has fallen markedly since. The bond market turbulence in 1994 associated with the tightening of policy by the Federal Reserve Board halted a further decrease in volatility. The subsequent bouts of turbulence in 1995 and 1996, associated in part with heightened uncertainty about the strength of the U.S. economy—and thus the possible course of Federal Reserve policy—have also kept volatility from decreasing to the pre-l980s level. This points to the key role of money market volatility in explaining recent bond market volatility, while conventional economic fundamentals such as inflation and economic growth seem to have played only minor roles.28

    Comparing volatility of weekly long-term government bond yields across the United States, Japan, and Germany shows that the recent volatility has not been unusual in a historical context in any of the countries. The relatively high volatility in recent years in Japan reflects the fluctuating yen and money market rates, arising from heightened uncertainty about the collapse in asset prices, about the health of the financial system, and about the strength of economic growth. With zero inflation and short-term interest rates near zero, signs of a substantial strengthening of economic growth could have quantitatively important effects on expected inflation and official short-term interest rates, which would then filter into the long end of the yield curve. While bond market volatility in Japan is higher now than in the unusually tranquil 1970s, it is only slightly higher than in Germany and the United States, and the most recent trend in Japan seems to be downward sloping. In Germany, bond market volatility is now close to a historical low by the measure applied here.

    The increased popular concern with securities market volatility is not warranted in general. Current stock and bond market volatilities are close to their historical averages in the major economies, and the folklore effects on volatility of the bond market sell-off in 1994 persisted only in the Japanese market and are dissipating now. The Japanese experience does illustrate that economic policies that successfully achieve stable goods prices do not necessarily accomplish (he more difficult task of asset-price stability.29

    Appendix 3 Current U.S. Equity Prices Compared with the 1987 U.S. and 1989 Japanese Bubbles

    A natural approach to gauging whether stock prices are overvalued is to compare them with the theoretical price calculated from the discounted (expected) future stream of dividend payments that the stock has claim to. There are significant practical difficulties associated with this approach because of uncertainty about future dividends and the appropriate risk-adjusted discount rates. An alternative approach is simply to compare characteristics of equity price valuations with features of well-known bubbles in equity markets to determine whether there are clear differences between bubbles and current valuations. This appendix takes the latter approach by comparing some key aspects of the current U.S. equity price rally with the 1987 U.S. and the 1989 Japanese price bubbles.

    Figure 52 shows the behavior of key variables for the three-year periods (–36 to 0 months) leading up to the 1987 U.S. and 1989 Japanese market crashes and the subsequent behavior for 12 months (0 to +12 months), compared with the behavior of the same indicator for the past three years in the United States. First, consider the behavior of equity prices themselves in the three cases. In the past three years U.S. equity prices have risen 95 percent, compared with about 100 percent in the United States over 1984–87 and in Japan over 1986–89. Thus, the current market rally is similar in magnitude to the earlier ones. It is noteworthy that the climb in equity prices recently has been much less volatile than in the two historical cases.

    Figure 52.United States and Japan: Developments in Equity, Bond, and Money Markets Surrounding Significant Stock Market Increases

    Sources: Bloomberg Financial Markets L.P.; and The WEFA Group.

    Note: Stock markets peaked on August 25, 1987, in the United States and on December 29, 1989, in Japan. The figures show developments taking place in the respective markets 36 months prior to and 12 months after these dates. Earnings per share data for the United States are available only on a quarterly basis and are not available for Japan.

    Second, consider earnings per share. Here there is clear evidence that the current market rally is much better supported by earnings. Earnings growth in the United States during 1984–87 provided no support for equity price gains. There is some indication that earnings are currently leveling off, however, and this raises concerns. Indeed, the recent upward pressure on the price-earnings ratio (as discussed in the text) reflects this fact.

    Third, consider bond yields, which can be important for equity prices both because they are a competing asset and also because they are associated with the cost that firms must incur for servicing their debt. On this score, there is a rather marked difference between the current experience and the two historical experiences. Bond yields had been increasing sharply in the six months or so leading up to the 1987 U.S. and 1989 Japanese market corrections, whereas recently they have been trendless.

    Finally, consider the lone of monetary policy in the three cases, as measured by short-term interest rates. The evidence is clear that monetary conditions had been loosened considerably in the earlier stages of the 1987 and 1989 market rallies, which differs from the current experience. Further, policy was tightened sharply beginning about 12 months prior to the U.S. and Japanese market corrections. By contrast, short-term interest rates have currently been trendless for over 24 months. The explanation for these differences in monetary policy across the three cases is easy to understand. Namely, real GDP growth was about 4 percent at the 1987 and 1989 markets’ peaks, and inflation had increased sharply in the year leading up to the markets’ peaks; by comparison, currently real GDP growth has been below the levels in the other episodes and inflation has not increased.

    In summary, the doubling of U.S. equity prices in the past three years is of the same order of magnitude as in the United States prior to the crash in 1987 and in Japan prior to the collapse of asset prices in 1989. But, in contrast to these historical experiences, some of the critical underlying fundamentals—including corporate earnings and low volatility—provide considerably more justification for recent equity price gains. The key question in assessing the susceptibility of U.S. equity prices to sharp corrections—as occurred in 1987 in the United States and in Japan beginning in 1989—is whether corporate earnings can continue to grow well above the historical average. The answer to this question hinges on whether the recently high rates of corporate profitability derive primarily from possibly temporary cyclical influences or from more fundamental corporate restructuring.

    Annex III Developments in International Banking

    The major international banking systems, including those in many emerging markets, have been engaged in a process of restructuring and consolidation in recent years that is profoundly changing the nature of the banking industry. Changes in regulation and technological developments have increased competition in traditional banking activities while simultaneously opening up new markets for expansion. In many countries, banking crises or the failure of important individual banks have provided additional impetus for restructuring. Continuing advances in information technology and the ongoing globalization of capital markets and the risk management business have resulted in consolidation in the investment banking industry. These issues are discussed with examples of recent important transactions in industrial countries and emerging markets in the first section of this annex.

    Many of the catalysts of financial system restructuring have led to a reassessment of supervisory and regulatory practices. A consensus is emerging in policy circles that a functional approach must give way to supervision and regulation on a legal-entity basis to align the reporting requirements and inspection systems with the management structures of financial conglomerates. However, the most efficient method of implementing this approach remains undetermined. In many countries, the architecture of supervision itself would have to be changed, and the question of if or how to retain a role for the central bank in the supervision of money center banks considered. At the same time, bank supervision is moving away from the auditing of financial condition reports toward an emphasis on assessing the adequacy of systems, including those for risk management and internal controls. There is also increasing discussion of the role that could be played by disclosure and other market-based systems of supervision. Developments in supervision and regulation in industrial countries, and the international effort to improve financial supervision and regulation in emerging markets, are discussed in the second section.

    These structural changes do not occur in isolation from the condition of the banking systems. Indeed, profitability and asset quality will affect the speed and nature of such developments. Banking systems that are dealing with the urgent resolution of asset-quality problems will have a restructuring dynamic that is very different from that in countries where such immediate pressures are less important, resulting in different patterns of mergers and acquisitions and perhaps a delay in responding to the longer-term structural issues. Recent developments in the banking systems of most of the systemically important industrial countries and emerging markets are discussed in the final two sections. Table 30 gives one indicator of relative strength of banking systems around the world, the Bank Financial Strength Ratings assigned by Moody’s. The industrial country banking systems are deemed to be stronger on a stand-alone basis than those in the emerging markets, but there is considerable diversity of conditions within the two groups, with some of the emerging market banking systems faring well in comparison with the industrial country group, while some of the latter are considered to be in serious difficulty.

    Table 30.Bank Financial Strength Ratings for Selected Countries, June 2, 19971
    Industrial countries
    United Kingdom348563100C+
    United States32068129639300C+
    Emerging markets
    Hong Kong, China002072000C
    Taiwan Province of China000144000C
    Czech Republic000003210D
    Slovak Republic000000300D
    Latin America
    Puerto Rico000001100D-D+
    Middle East and Africa
    Saudi Arabia000124410D+
    South Africa000033000D+−C
    United Arab Emirates000002110D
    Source: Moody’s Investors Service.

    Restructuring and Consolidation of International Banking Systems

    The restructuring and consolidation that are under way in international banking systems have been motivated by a number of developments in the past decade or so, among which four stand out: (1) the deregulation of international and domestic financial markets; (2) improvements in communications and computational technology; (3) significant asset-quality-driven problems in many banking systems; and (4) a growing recognition of the costs and distortions associated with official support for banking institutions. These mutually reinforcing developments have both provided the impetus for banking sector restructuring and in turn been affected by this restructuring.

    Changes in the supervisory and regulatory framework have been an important source of pressure for industry consolidation and restructuring. Such changes include the liberalization of domestic and cross-border banking activities, the easing of segmentation barriers within national financial systems,1 and the reorientation of the emphasis of bank supervision toward capital adequacy. The deregulation of financial markets allowed for greater disintermediation from banking, as depositors searched for higher yields in investment funds and securities, and corporate borrowers found access to financing at competitive terms in the securities markets by selling bonds to institutional investors. Table 31 provides some evidence for these trends by presenting data on two indicators of disintermediation, the ratio of institutional investors’ financial assets in total financial assets of domestic financial institutions, and the share of household assets represented by claims on institutional investors. Both ratios have increased sharply since the mid-1980s for most industrial countries. For example, in France the assets of institutional investors rose from 11 percent of total financial system assets in 1985 to 23 percent in 1995; the share of household assets held by institutional investors also rose, from 16 percent in 1985 to 27 percent in 1995. Disintermediation is most advanced in the United States, where in 1995 institutional investors accounted for 55 percent of financial system assets and held 45 percent of household assets. These indicators suggest the difficulties banks have faced in responding to the deregulation of financial markets.

    Table 31.Distintermediation in Selected Industrial Countries
    Assets of Institutional

    Institutionalization of

    Household Savings2
    United Kingdom626.731.652.0
    United Slates43.854.633.445.3
    Source: Organization for Economic Cooperation and Development.

    The move toward risk-based capital requirements in the late 1980s marked a significant change in the supervision of banks and led to a reorientation of bank management objectives toward maximizing risk-adjusted returns on capital, rather than increasing gross measures of performance such as total revenues, profits, or market share. The need to increase capital as the scale of bank operations expands or as its activities become more risky reinforces the incentives to maximize shareholder value in order to maintain access to sources of capital. Mergers and acquisitions (M&A) have been viewed as one way to increase shareholder value, although it is notable that there is little evidence to support the existence of economies of scale or scope for banks or for improvements in efficiency resulting from mergers or takeovers.2

    It may be, however, that recent technological developments will make improvements in shareholder value arising from M&A transactions more significant, even for so-called mega-mergers of very large banks. Developments in computational and communications technologies have been an important catalyst for restructuring both by possibly changing the scale and scope economies within banking and by increasing the effectiveness of competition from nonbank financial institutions. These developments have already revolutionized the products and services offered by banks, and the ways in which these are delivered. They have also increased the efficient scale of a number of operations, including check processing and other payment operations, the processing of loan applications (e.g., through the adoption of credit scoring models for mortgage loan applications), risk management and treasury operations, trading, and the centralization of most data back-office activities. More recently, and most visibly, the employment of electronic banking technology—automated teller machines, point-of-sale debit machines, telephone banking, and Internet banking—is more cost effective if the bank has a very large customer base. It is possible, therefore, that consolidation among very large banks can allow a more efficient adoption of these new technologies than is possible for smaller banks.

    For many banks, the most effective incentive for restructuring has been a sudden decline in profitability, and this is true also for banking systems as a whole. The past decade or so has seen the emergence of serious banking difficulties in a large number of industrial countries and emerging markets, and their resolution has often provided an opportunity for consolidation or acquisitions by outside investors. Takeovers by strong banks of relatively weaker ones is a common pattern in bank mergers and acquisitions and can hasten the resolution of asset-quality problems by improving bank management, reducing the constraint on cash flow posed by nonperforming assets by combining them in a portfolio with a higher average quality, and providing greater income flows or capital with which to set aside loan loss reserves or to write off problem loans; for these reasons such takeovers are often encouraged by bank supervisors.3

    Another factor that has contributed to the dynamics of restructuring in banking systems has been the declining role of the state in financial systems due, in part, to the growing pressure for fiscal consolidation and the allocation of scarce budgetary resources to other priorities, including pension finance, for example. As it becomes more difficult for governments to justify the expenditures necessary to bail out failed banks4—government owned or not—the direct ownership stake that governments have had in individual institutions has gradually been withdrawn in many countries.

    Banks have responded to these pressures with a variety of strategies, including concentrating on core activities in which they have a strong comparative advantage and expanding vertically (to exploit economies of scale) or horizontally (to capture economies of scope). The expansion strategy has been in many ways the most visible, as the past few years have seen a seemingly unprecedented wave of mergers and acquisitions within the international financial markets. One strategy that some large institutions appear to have adopted is to become a globally competitive actor. The liberalization of international financial transactions over the past two decades made most markets contestable to firms willing to expand geographically. With the asset-quality problems associated with the 1980s debt crisis and then the early 1990s real estate crisis behind them, many of the large banks in the major industrial countries have begun to compete for business on an international level.

    The pace of consolidation in industrial country banking systems slowed slightly in 1996 after a hectic year. In the United States, where mergers and acquisitions are far more common than in other mature systems, activity slowed significantly in 1996 after a record volume of transactions in 1995. A total of 442 deals were announced in 1996, with a total value at announcement of $44 billion, compared with 537 deals for $73 billion in 1993.5 More important than the volume of transactions, however, has been the types of deals announced. In the past two years, the five largest U.S. bank mergers ever arranged have been announced—Wells Fargo and First Interstate ($13.7 billion); Chemical Bank and Chase Manhattan Bank ($13.3 billion); NationsBank and Boatman’s Bank-shares ($9.5 billion); First Bank and U.S. Bancorp ($8.4 billion); and First Union and First Fidelity ($6.1 billion)—and a hid for the third-largest savings and loan institution (Great Western Financial Corp.) appears to have been won by the second-largest thrift (Washington Mutual) in a deal that might cost $6.7 billion. The banking industry has embarked on an almost unprecedented process of consolidation leading to the creation of a small number of institutions that are dominant on a national scale.

    Most of this M&A activity has been truly consolidation of commercial banking within or between geographical areas as restrictions on interstate banking have been progressively relaxed, culminating in their widespread elimination under the Riegle-Neal Act of 1994. However, with the liberalization of banking regulations on sales of insurance and on securities underwriting and sales, the focus of attention has recently turned toward acquiring complementary skills as in, for example, the 1995 Mellon Bank acquisition of Dreyfuss, the $7.3 billion Bank One acquisition of First USA, or the more recent takeover of Alex. Brown by Bankers Trust. Increasingly, in response to the gradual relaxation of constraints on their nonbanking activities, banks are expanding their securities and asset management capabilities. Nor has such activity been restricted to the banking industry, as reflected in the $10.2 billion merger between Dean Witter. Discover and Morgan Stanley (which had itself acquired Van Kempen America in 1996) creating the largest credit card issuer and fifth-largest asset manager in the world.

    The U.K. banking system also saw some important repositioning by individual institutions in 1996, but of a more varied nature than is possible in the United States. Lloyds Bank and TSB Bank merged in 1996, creating the third-largest bank in the country. NatWest sold its U.S. and Australian banking subsidiaries, but acquired Gartmore, a U.K. asset management company. Three building societies (Alliance and Leicester, Halifax, and Woolwich) renounced their mutual status and took bank licenses, and Halifax also acquired a life insurance company. Canadian banks have been engaged in acquiring securities and trust subsidiaries since they were permitted to do so under the 1987 and 1992 revisions of the Bank Act. Similarly, some of the banks have acquired U.S. financial institutions: in 1996, the Bank of Montreal acquired a U.S. thrift, Household Bank, and Toronto Dominion Bank bought Waterhouse Securities in New York. In addition, Canadian banks continued to invest in emerging markets in 1996–97, as discussed below.

    French banks also have responded to recent asset-quality difficulties and highly competitive markets in part by engaging in heightened merger and acquisition activity within the past year. This has included the mergers of Banque Française du Commerce Extérieure and Crédit National, and of Crédit Local and Crédit Communal de Belgique; the takeover of Banque Indosuez by Crédit Agricole: and the takeover of Banque Hydro-Energie by Crédit Commercial de France in 1996. In early 1997, Société Générale had acquired a majority stake in é du Nord, which had last year acquired Banque Laydemier. Unlike mergers elsewhere, however, the elimination of redundancies is not usually a major objective in French bank mergers, since strict labor laws and high unionization within French banks make it difficult to reduce staff. M&A activity is geared more toward identifying complementary specializations or risk concentrations.

    German banks have been perhaps the most active among continental banks recently in expanding into international investment banking and asset management (e.g., Deutsche Bank acquired Morgan Grenfell in 1989, and Dresdner Bank acquired Kleinwort Benson in 1995). This activity continued in 1996, as Commerzbank acquired American Martingale Asset Management, and Bayerische Hypotheken-und Wechsel Bank acquired management control over U.K.-based Hypo Foreign & Colonial Management Holdings by increasing its equity stake to 65 percent, and entered into a cooperation agreement with a U.S.-based asset management company, Massachusetts Financial Services.

    In Italy, bank restructuring has involved consolidation among the smaller institutions, often through mergers with larger banks, and through changes in bank ownership. Banco di Napoli was formerly 71 percent owned by a private charitable foundation. Such foundations had been important owners of banks in Italy, but under the Amato Law of 1990, incentives were given to these foundations to sell their stakes. This has been achieved gradually, most recently with the sale of 45 percent of the shares in Istituto Bancario San Paolo di Torino by its owners, Italy also has a large number of very small banks, many of which have low capital levels. While such institutions are gradually being merged with or taken over by larger banks, the labor laws and practices in Italy make it very difficult for banks to realize large efficiency gains from mergers. Banks still seem to view these mergers as mechanisms for increasing market share rather than increasing profits. It is not uncommon for merged banks to continue to operate under their own names in competition with each other, with almost no apparent effort to rationalize their structures or products.

    Japan has seen a large number of bank takeovers in the last two years as this has been an integral element in the official response to asset-quality problems. Since the emergence of the current banking sector problems, at least 20 institutions, mostly credit unions and credit cooperatives, have been merged with healthier institutions, in many cases with assistance from the Deposit Insurance Corporation (D1C). The major institutions have not been left out of this process. In October 1994, Mitsubishi Bank took over Nippon Trust in an assisted transaction of the troubled bank, and in April 1997, Hokkaido Takushoku Bank and Hokkaido Bank agreed to merge. The only recent merger that does not appear to have been motivated by portfolio weakness in either of the participants was the merger between Bank of Tokyo and Mitsubishi Bank that took effect in 1996, creating the largest commercial bank in the world. By combining two banks with different traditional emphases—Bank of Tokyo had a relatively small branch network and did little domestic lending in Japan—this transaction appears to have been motivated more by international developments than by domestic asset-quality concerns.

    The restructuring of the banking sectors in each of the Nordic countries continued in 1996. In Norway. Den Norske Bank acquired Vital Försikring, the second-largest life insurance company in the country, and Christiania Bank acquired Norgeskreditt. In Sweden, Svenska Handeslbanken acquired 98 percent of the shares in Stadshypothek, a leading housing finance institution, and in February 1997, a merger of Swed bank and Föreningsbanken—the central banks of the cooperative bank system and the savings bank system respectively—was announced. Further consolidation is to be expected. The Nordic banks are experiencing the same increased competition, disintermediation, and narrower interest margins as the other banking systems in Europe.

    The emerging market banking systems have also experienced an increase in the pace of restructuring and consolidation. In fact, banking crises have precipitated a remarkable restructuring of banking systems in some Latin American countries. Among most of the Asian emerging market countries, until very recently, the absence of a crisis atmosphere had prevented any sense of urgency about the need for restructuring, although many supervisory authorities had actively encouraged consolidation. An important element in the restructuring of Latin American and European emerging market banking systems has been the use of M&A transactions by foreign banks as a means of penetrating the market. Domestic shareholders, often lacking the means to recapitalize the banks, or seeking to reorganize the corporate structure of mixed financial/nonfinancial conglomerates, have increasingly been willing to sell controlling stakes to foreign financial institutions. Also, governments in many countries have reconsidered the benefits of public ownership of financial institutions and have privatized large proportions of financial sector assets. While most of the foreign purchasers have been from Europe (especially Spain) and North America, it is significant that there are a number of Latin American financial institutions that have regional ambitions (for example, Chile’s Infisa group, and Argentina’s Banco de Galicia). Foreign investment not only brings new competition, and therefore greater pressure for consolidation, but their more advanced practices and technology, broader range of products, and deeper capital base put pressure on the local banks to modernize and become more efficient.

    The 1995 liquidity crisis and ongoing restructuring in response to the stabilization of inflation since the early 1990s have resulted in a significant reorganization of the Argentine banking system. Since the end of 1994, 11 government-owned banks and 36 privately owned banks and cooperatives have closed or been privatized or merged with other institutions.6 As in other countries in the region, there has been a surge in foreign investment in the banking system in recent months. In October 1996, Banco Bilbao Vizcaya acquired a controlling interest in Banco Frances del Rio de la Plata, and in May 1997 the latter took over Banco de Credito Argentino. Also in May 1997, Banco Santander acquired control over Banco Río, the Bank of Nova Scotia took control of Banco Quilmes, and a group of investors including Chile’s Infisa, Chase Manhattan Bank, and National Bank of Canada was approved to take over Banco Union Comercial Industrial, which had been experiencing capital adequacy problems, in December 1996. In June 1997, HSBC Holdings increased its 30 percent stake in the holding company of Banco Roberts to 100 percent. Further consolidation and foreign investment in the banking system are expected—including the privatization of more government-owned banks, among them the national mortgage bank. Banco Hipotecario Nacional.

    The restructuring of the banking industry in Brazil is also well under way. Since July 1994, some 30 financial institutions have been closed, merged, or liquidated, including some quite large private and publicly owned banks (e.g., Banco Econômico, Banco National, and Banerj). Foreign participation in the industry has increased, both in terms of new entry and in investment in existing banks. For example, in March 1997. HSBC Holdings bought the domestic banking operations of Banco Bamerindus, and Banco Santander acquired a controlling stake in Banco Geral do Comercio. These developments are likely to continue since a large number of small private banks are thought likely to need assistance or want to increase their chances of survival by merging or attracting new investors (six small commercial banks had been shut down by the central bank by end-July 1997). This process will likely be a costly one, as long as the central bank is prepared to make liquidity funding available and as long as the federal government is prepared to reschedule the state governments’ debts as a means of ensuring that their banks are restructured.7

    The Mexican banking system has also seen its share of foreign investment, associated with the sales of banks that were taken over by the central bank and efforts by the surviving banks to increase their capital and improve their management and product availability. Since early 1996, foreign banks have acquired important slakes in eight Mexican banks. In 1996, the Bank of Montreal acquired a 16 percent stake in Bancomer, the Bank of Nova Scotia increased its stake in Inverlat. Banco Santander acquired Grupo Financiero Invermexico (the holding company of Banco Mexicano). Banco Bilbao Vizcaya took over Multibanco Mercantil Probursa and acquired the branch network of Banco Oriente, and Banco Comercial Portugues and Banco Central Hispanoamericano both invested in Banco Internacional. In 1997, this process has continued, with HSBC Holdings taking a 20 percent stake in Grupo Financiero Serfin and GE Capital acquiring Banco Alianza. Domestic banks have also recently taken an interest in growing by acquisition, especially as a means of diversifying into insurance and other nonbanking activities. For example. Grupo Financiero Banorte acquired two insurance subsidiaries of Banco Obrero and will participate in the auction for Banpais in July 1997.

    The immediate future for Venezuelan banks is also strongly influenced by the sudden emergence of foreign competition. For the past two decades foreign banks have not been permitted in Venezuela, but that ended in December 1996 when foreign financial institutions were allowed to acquire controlling stakes in three of the four largest Venezuelan banks—two of which were acquired in privatizations. Banco Santander successfully bid for Banco de Venezuela when the latter was privatized in December 1996, and Banco Bilbao Vizcaya reached agreement to purchase 40 percent of Banco Provincial (including a 17.6 percent stake held by Crédit Lyonnais), the largest private bank. At the same time, a Chilean investment group, Infisa CA (which controls Chilean bank Banco Concepcion), won control over Banco Consolidado, Infisa plans to sell pail of its 93 percent slake to a group of investors including Chase Manhattan Bank and National Bank of Canada. Most recently. Banco República was sold in June 1997 to a Colombian savings bank. With these sales, there remain two banks to be privatized: Banco Andino Venezolano and Banco Popular.

    The Czech banking system has already undergone a fundamental restructuring in the transition from a highly centralized, exclusively government-owned system to a market-oriented system. However, this transformation is not complete. The industry is highly concentrated, and many banks rely heavily on the interbank market and on institutional funds for liquidity. Two new foreign banks were given licenses in 1996 (the first new banks of any kind since 1994), joining the 23 existing foreign banks and joint ventures. In addition, the privatization of two banks, Investicni a Postovni Banka and Ceskoslovenska Obchodni Banka, is under way.

    Restructuring has also continued in the Hungarian and Polish banking systems, where privatization continues to dominate the structural changes. In Hungary, Magyar Hitel Bank, one of the three banks that were carved out of the National Bank of Hungary in 1987, was privatized, with 89 percent of its shares sold to ABN Amro, joining Budapest Bank, which was privatized in 1995. Both banks are being restructured by their foreign majority shareholders into retail-oriented banks, which will involve a significant reduction in size, especially for Magyar Hitel Bank, which has been dominant in lending to large enterprises. The state privatization agency (APV) has also announced plans to privatize the last of the government-owned banks, Kereskedelmi es Hitel-bank (KHB), in a two-round sale in 1997. (KHB itself acquired a smaller Hungarian bank. Ibusz Bank, in 1996.) In the first round, a 25 percent stake will be sold, after which the remaining shares will be distributed to social security funds, management, employees, and small shareholders. In April 1996, the Deutsche Genossenschaftsbank received approval to purchase 61 percent of the shares of the central bank of the cooperative banks in Hungary, Takarekbank. Finally, the APV has announced that it will reduce its 25 percent ownership stake in the National Savings and Commercial Bank (OTP) to 10 percent in 1997, and the National Bank of Hungary has announced that it will seek to sell its 34 percent stake in Central European International Bank in 1997.

    After a relatively slow start, in which only five of the Polish state-owned banks were privatized over as many years, the government in 1996 adopted a revised bank privatization strategy envisaging the consolidation and privatization of most remaining state-owned banks by the year 2000. As a first step, in late 1996 three of the remaining state-owned regional banks were merged into Bank PkO-SA, with a view to privatizing the consortium in 1998. In 1997, three banks are being privatized: first, in June, Bank Handlowy was sold in a complex transaction involving a public offer to individuals (29 percent of share capital), institutional investors (30 percent), three major “core shareholders” (foreign financial institutions expected mainly to strengthen the bank’s know-how in different areas of financial services, 26 percent), employees (7 percent), and the Polish Treasury (8 percent). In addition, the bank is to issue so-called convertible bonds, initially to the treasury, for the equivalent of about 35 percent of its current share capital. These bonds are to be used later to help finance the reform of Poland’s pension system. Second, the National Bank of Poland is finalizing the sale of its 100 percent stake in the Polish Investment Bank; and third, the government will shortly complete the sale of a 65 percent stake in the Warsaw-based Powszechny Bank Kreditowy SA (PBK). After that, the last of the nine state-owned regional banks (Bank Zachodni), the State Agricultural Bank (BGZ), and the Polish Development Bank are slated for sale in 1998–99. This would leave only two banks in state hands: the large domestic savings bank (PKO-SA), to be privatized eventually, and Bank Gospodarska Komunalna (BGK), the only bank to remain government-owned in the future.

    Supervisory and Regulatory Developments

    Regulatory Structure

    The technological advances noted above have allowed an increase in the sophistication of finance and the introduction of new products and innovative delivery mechanisms, and have also reinforced the blurring of divisions between different segments of the financial system in most countries. Improvements in risk management have allowed the portfolios of larger, more complex corporate organizations to be managed centrally. These trends call for a change in the structure of supervision over financial firms to reflect their changing internal organizations. Supervision along functional lines is not efficient in the presence of centralized risk management by financial conglomerates.

    The problem of how best to organize supervision and regulation of financial conglomerates in a global marketplace is being examined by the Joint Forum, a group made up of bank, securities firm, and insurance regulators, previously called the Tripartite Group. The Joint Forum has made some headway over the last year in facilitating the exchange of information among the groups of supervisors and in outlining the responsibilities of a regulatory “coordinator” for conglomerates. A number of smaller countries, including Hungary, Norway, and Sweden, have implemented a centralized system of supervision, in which one agency, independent of the central bank, is responsible for supervising all types of financial institutions. Such a structure has been proposed too in Korea. This has not yet been implemented in any of the major industrial countries. However, in May 1997, the Chancellor of the Exchequer in the United Kingdom proposed a reorganization of financial supervision in which eventually all financial supervisory and regulatory authority will reside in the Securities and Investments Board (SIB). In addition to reducing the role of the self-regulatory organizations and their oversight agencies, the government proposes to remove responsibility over bank supervision and regulation from the Bank of England and transfer it to the SIB.

    As the U.S. financial system evolves toward one in which there may be few restrictions on banks’ activities in the securities and insurance businesses (and reciprocally for securities and insurance firms), there is now considerable debate on the need for a change in the regulatory structure. While a bid to consolidate the supervision of financial institutions in one agency did not win approval in 1996, banks’ powers to sell insurance and to engage in the underwriting and trading of “ineligible” securities have been expanded.8 The main issue being debated is the form of corporate organization that will be preferred. While in the European Union, for example, banks often have the choice of undertaking insurance or securities business in-house or through bank subsidiaries, the U.S. Federal Reserve Board has proposed allowing such activities only through bank holding company subsidiaries, suggesting that such a structure facilitates the supervision of insured deposit taking and related banking activities separately from other activities. The Office of the Comptroller of the Currency has argued that the holding company structure may not be the most efficient and that banks should have the option of engaging in securities dealing within the bank rather than through subsidiaries.

    During 1996, the Japanese authorities announced a number of important policy initiatives. In June, three financial laws were passed by the Diet that implemented a U.S.-style bank resolution framework built around (I) increased powers for regulators to intervene in problem banks, including declaring them insolvent; (2) prompt corrective action (PCA) measures for intervening in weak banks; and (3) increased resources for the deposit insurance corporation to pay off depositors of failed banks, including a fourfold increase in deposit insurance premiums (to 0.048 percent of insured deposits), the temporary addition of a special premium of 0.036 percent, and a ¥2 trillion line of credit from the Bank of Japan. The special premium, implemented by a provisional amendment to the Deposit Insurance Act, was introduced to provide insurance for alt deposits, including those in excess of the statutory ¥10 million limit. This additional guarantee, and an all-encompassing official guarantee that none of the banks with international activities would close, terminate at the end of the 2000/2001 fiscal year.

    The PCA measures, which take effect April 1, 1998, require banks to classify their loan portfolios more rigorously by repayment risk and to set aside adequate reserves thereby providing a more accurate measure of economic capital in the bank; they also allow the authorities to intervene by forcing banks to take corrective measures or ultimately by closing them down based on their risk-weighted capital ratios. While these rules represent a move in the direction of market discipline of the banking system, the measures prescribed for the supervisory authorities are less forceful, and allow for a greater deterioration of capital, than those implemented in the United States by the Federal Deposit Insurance Corporation (FDIC) Improvement Act of 1991 (Table 32). For example, while the FDIC is required to intervene in a bank by demanding recapitalization plans and restricting asset growth and new activities for banks once their total risk-weighted capital ratio falls below 8 percent, similar restrictions in the Japanese system would be introduced only after the capital ratio fell below 6 percent. Restrictions on deposit taking and managerial compensation are introduced in the U.S. system when the capital ratio falls below 6 percent, but under the Japanese regulations these are introduced only when the capital ratio falls below 4 percent. Finally, regulators have more discretion under the Japanese PCA rules than they do under the U.S. rules. For example, in Japan suspension of activities can be avoided if the bank’s net income is expected to be positive as a result of implementing a restructuring plan, while in the U.S. system, once a bank is “critically undercapitalized” the regulators are required to close it down. More generally, in the U.S. PCA system, regulators have discretion to strengthen the required response to a decline in capitalization, while under the Japanese rules, they have discretion to scale down the required response.

    Table 32.Japan and the United States: Summary of Prompt Corrective Action Provisions
    United States
    Capital levels1ActionsCapital levels2MandatoryDiscretionary
    n.a.n.a“Well capitalized” Total ≥10 percent, and Tier 1 ≥6 percent, and Leverage ratio ≥5 percent.NoneNone
    n.a.n.a.“Adequately capitalized” Total ≥8 percent, and Tier 1 ≥4 percent, and Leverage ratio ≥4 percent.Disallow brokered deposits, except with I-’DIC approval.None
    International capital

    ratio <8 percent:

    National capital

    ratio <4 percent.
    Order formulation and implementation of management improvement plan.“Undercapitalized” Total <8 percent, or Tier 1 <4 percent, or Leverage ratio <4 percent.Suspend dividends and management fees.

    Require capital restoration plan.

    Restrict asset growth.

    Require approval for acquisitions, branching, and new activities.

    Disallow brokered deposits.
    Order recapitalization.

    Restrict interaffiliate transactions.

    Restrict deposit interest rates.

    Order other measures necessary to carry out prompt corrective action.
    International capital ratio <4 percent: National capital ratio <2 percent.Order recapitalization plan.

    Impose restraints on asset growth.

    Impose ban on new activities and branches and limits on current activities.

    Impose ban on new subsidiaries and overseas affiliates and limits on the current activities of such entities.

    Limit payment of dividends.

    Limit payment of bonuses to directors and management.

    Limit deposits, interest rates.
    “Significantly undercapitalized” Total <6 percent, or Tier 1 <3 percent, or Leverage ratio <3 percent.Same as above.

    Order recapitalization.

    Restrict interaffiliate transactions.

    Restrict deposit interest rates.

    Restrict pay of officers.
    Same as above.

    Order conservatorship or receivership if bank fails to submit or implement a plan to recapitalize.

    Impose any provision for “critically undercapitalized” banks if necessary.
    International capital ratio <0 percent: National capital ratio <0 percent.Suspend whole or part of banking business. This order can be replaced with lesser actions if: (1) the net value of assets, including unrealized gains, is positive; (2) the net value including unrealized gains is negative but expected to be positive after considering: (a) the implementation of management improvement plans and other specific measures; (b) business income and profitability: (c) the bad assets ratio. A business suspension order can be issued at any time when the net value of assets, including unrealized losses is, or is expected to be, negative.“Critically undercapitalized” Tangible equity to total assets ratio of ≥2 percent.Same as lot “Significantly undercapitalized” banks.

    Order receivership/conservatorship within 90 days.

    Order receivership if critically undercapitalized for four quarters.

    Suspend payments on subordinated debt.

    Restrict certain other activities.
    Sources: Japan, Ministry of Finance; and United States, Federal Deposit Insurance Corporation.

    A change in the structure of financial supervision has also been initiated in Japan. The responsibilities for bank, insurance, and securities supervision currently assigned to the Japanese Ministry of Finance will be transferred to a new agency, the Supervisory Agency for Financial Entities (provisional translation) in mid-1998 (the Bank of Japan will retain its bank examination powers). The ministry will retain responsibility for the formulation of bank regulation policy and will be consulted by the supervisory authorities in cases of systemic importance. The Securities Exchange Surveillance Commission will be merged into the new agency, creating a consolidated banking/securities supervisory agency.

    A more far-reaching set of reforms was announced in November 1996. Described as Japan’s “Big Bang” these measures include (1) the elimination of most foreign exchange controls and ex ante reporting requirements, and the abolition of the authorized foreign exchange bank system; (2) the acceptance of financial holding companies; (3) the abolition of fixed commissions on securities transactions; and (4) the elimination of restrictions segregating securities, trust, and banking activities. The latter three measures are expected to lead to a significant restructuring and consolidation of the financial services industry, as they allow financial holding companies to be established combining all types of banking, securities, and insurance activities, while at the same time making each type of activity more competitive and responsive to market forces. It is the first measure, approved in May 1997 and to lake effect in May 1998, however, that is potentially the most significant. Liberalization of the foreign exchange controls may allow foreign financial institutions to compete more effectively and to offer more services to Japanese customers. Indeed, this is the central element of the reform plan, as increased foreign competition is expected to force domestic firms to be more innovative and efficient and to provide better services and higher returns to retail investors.

    Capital Requirements

    Following a lengthy debate about the appropriate form of regulatory capital requirements for market risk, bank supervisors in the major industrial countries are taking a step back to evaluate the entire concept of regulatory capital. The Basle Committee on Banking Supervision (“Basle Committee”) has formed a working group to examine the issue from a fresh perspective. Part of the impetus for doing so is a recognition that the complexities and interrelationships among the various sources of risk (credit, market, operational, legal, and so on) in bank portfolios makes it difficult to prescribe specific rules for the calculation of capital that appropriately capture the risks against which it is meant to insure. For example, even before the latest amendment to the Basle Capital Accord to incorporate market risks had been finalized, members of the Basle Committee were cognizant that even with fairly strict initial assumptions the use of a value-at-risk (VAR) model could result in different measures of regulatory capital for different banks. This “implementation risk” has important implications for regulators; clearly, it is insufficient for supervisors to vet the theoretical version of a bank’s internal model, yet it would consume enormous resources if supervisors were to attempt to verify each bank’s model by running a benchmark set of portfolios. Backtesting may emerge as the only way to gauge the effectiveness of a bank’s VAR model.

    Put forth as a possible solution to the problems associated with the inflexibility of the Basle Committee’s rule-based approach to measuring market risk, the precommitment approach, in which each bank agrees not to violate its “precommitted” level of market risk capital, continues to elicit heated discussions. A one-year pilot study involving some 10 banks started on October 1, 1996, under the auspices of the New York Clearing House whereby these banks will, on paper, pre-commit to maintaining a certain minimum amount of capital and any violations will be recorded. Despite its potential lack of realism (there are no penalties applied to banks whose capital falls below their pre-committed amount), the pilot study has at least forced those banks involved formally to allocate an amount of capital and to think carefully about whether to use their VAR model or some other technique for calculating that amount. In fact, several banks have augmented their VAR number to take account of risks that are not included in their VAR models.

    While a discussion of the management and measurement of market risk continues to dominate the lives of many regulators and practitioners, attempts to measure and manage other risks, most notably operational risk, are emerging as the next major issue. Banks are beginning to take a hard look at the operational risks in their business, and some of them are developing methods by which such risks can be measured and capital allocated against them. In the forefront is Bankers Trust which, after the collection and examination of appropriate data, discovered that their hierarchy of risks has been altered to first, credit risk, followed by operating risk, and lastly, market risk. Other banks have also found that unexpected losses can result from the application of sophisticated pricing systems, including recently NatWest Bank and Tokyo-Mitsubishi Derivative Products. The importance of internal controls and the measurement of risks incurred due to their failure is only now becoming fully appreciated.

    Another important element in the discussion of capital and its purpose has been spurred by the development of credit derivatives (see Appendix 1 at the end of this annex) and the new progress achieved in analyzing credit risk. The international regulatory community has arrived at a consensus that these developments warrant a rethinking of credit risk capital requirements. The Bank of England has requested comments on a discussion paper regarding a possible supervisory approach to credit derivatives, and the Securities and futures Authority has provided its first formal pronouncement regarding credit derivatives. In the United States, the three federal regulatory agencies have issued guidance notes on credit derivatives. However, all these documents stress the preliminary nature of the guidance regarding regulatory treatment and the quickly evolving characteristics of tradable credit risks. In addition to the impetus from credit derivatives, international regulators are also aware that the credit risk categories assigned within the original Basle Accord were based on relatively crude, qualitative assessments of relative risk and these should be reevaluated in light of the newer analysis.9

    Supervision and Regulation in Emerging Markets

    The soundness of the financial system, and especially the banking system, is increasingly recognized as an important part of any evaluation of the economic prospects of an emerging market. The role of the financial sector in propagating the crisis in Mexico illustrated the complicated implications for the conduct of macroeconomic policy that newly liberalized financial sectors raise. Investors in emerging markets are now provided with much more analysis of the health of the financial systems and of the potential relationships between financial fragility and macroeconomic performance in these countries.

    In addition, the official international community has stepped up efforts to support improvements in financial infrastructure and to incorporate analysis of financial market developments in the monitoring of developing countries’ macroeconomic performance. Thus, international attention has turned to the economic consequences of inadequate bank supervisory and regulatory capacity in emerging markets. Such concerns motivated a broad-based effort by the international financial community in 1996–97 to provide a basis for support for emerging market governments in this area. The Basle Committee released a set of Core Principles for Effective Banking Supervision (April 1997) and a G-10 Working Party on Financial Stability in Emerging Market Economies released its own paper (Financial Stability in Emerging Market Economies) the same month. The latter report called for a “concerted international strategy to promote the establishment, adoption and implementation of sound principles and practices needed for financial stability” in emerging markets, in which the IMF would promote the adoption and implementation of these principles and practices. An early proponent of this approach, Goldstein (1997), has called for the establishment of International Banking Standards, based on the principles developed by the international agencies, to which bank supervisors would voluntarily adhere.

    Developments in Profitability and Asset Quality in Selected Industrial Countries

    The banking systems in the major industrial countries performed strongly in 1996: most of the banks in the G-7 countries reported higher earnings and improved asset quality. The banks have benefited from a sustained period of declining interest rates and from an increase in economic growth rates that have raised loan volumes and provided the means to finance loan loss provisions and write-offs. Aside from a few individual institutions in certain countries, the problems that banks in these countries face are medium-term structural problems, not the immediate solvency threats they confronted in the early to mid-1990s.

    Performance by Country

    The performance of commercial banks in the industrial countries in 1996 differed markedly between groups of countries depending upon their relative position in the credit cycle. For banks in Canada, Norway, Sweden, the United Kingdom, and the United States, which had resolved their asset-quality problems relatively quickly in the early 1990s, profitability and capitalization levels remained at or near historic high levels in 1996, with nonperforming loan ratios at correspondingly low levels. Asset quality may have peaked in these countries, and some deterioration can probably be expected. However, the banks have high loan loss reserves and are well capitalized, so they are generally well equipped to deal with such a development.

    For banks in Finland and France, there is now reason to believe that the worst of their asset-quality problems are over. The overall condition of the banking systems in these countries has improved, and the emphasis is increasingly on the resolution of problems in individual institutions and in France on the longer-term structural issues that confront the industry, such as the effects of European monetary union on competition within European banking systems, as is discussed in Annex IV. The situation in Italy, however, remains difficult, especially for banks in the central and southern regions, while the Japanese banks continue to face serious asset-quality problems. The German banking system, alone it seems among the major industrial countries, has not experienced a serious decline in asset-quality due to real-estate-related loans, although some segments of the property market in Germany have weakened.

    Supported by a strong increase in noninterest earnings and a 24 percent decline in provisions, the Schedule I banks in Canada reported a 22 percent increase in net income in the fiscal year ending October 1996, an average return on equity of 15 percent.10 Canadian banks compensated for a tightening in net interest margins—the fifth consecutive year of tightening—by increasing the share of consumer lending and other relatively high-margin lending in their loan portfolios (including the contribution of some of the banks’ Latin American subsidiaries). To date, this has not led to a deterioration of average loan quality. At the end of the fiscal year, only 0.6 percent of loans were nonperforming, down from 1.2 percent in 1995 and a peak of 3.2 percent in 1992. Reserve coverage is generally very high, in excess of 270 percent of nonperforming loans, although the ratio of reserves to gross loans fell to 1.6 percent, the lowest level in at least eight years. Canadian banks are also well capitalized by international standards, with an average total risk-weighted capital ratio of 9.4 percent (a Tier I ratio of 6.8 percent) even after share repurchases by at least four of the banks.

    Most banks in the United Kingdom also enjoyed record profits in 1996, bolstered by a sharp reduction in loan loss provisions, and despite a slight narrowing of the net interest margin. Aggregate net income for the five largest commercial banks rose 2 percent, for an average return on equity of 19 percent.11 Banks compensated for declining interest margins on corporate loans by increasing both the volume of loans and the share of consumer loans and mortgages in their loan portfolios. After having peaked in 1992, asset-quality problems eased significantly, with the end-1996 incidence of impaired lending representing 2.5 percent of loans (compared with 9.8 percent at end-1992)12 and loan loss reserves covering 85 percent of impaired lending. Capitalization remains high, with a total risk-weighted capital ratio of 10.9 percent, 7.26 percent of which was Tier I capital.

    Banks in the United Stares continued their string of record earnings levels in 1996, as total net income of FDIC-insured commercial banks increased by more than 7 percent for a return on equity of 14 percent. Despite slightly lower net interest margins, net interest income increased by 5 percent, supported by strong loan growth. The diversification of banks’ income sources continued in 1996, with noninterest income contributing 36 percent of gross operating income—the highest proportion ever for U.S. banks. While overall asset quality continued to improve in 1996—total noncurrent loans (i.e., loans that are 90 days or more past due or on nonaccrual status) fell to 1.05 percent of gross loans—consumer lending and domestic syndicated lending have been a source of concern. At end-1996, 3.6 percent of loans to individuals were past due, reflecting in particular a continuing deterioration of credit card loans.13 Concerns about potential asset-quality problems were reflected in an increase in loan loss provisions for the second consecutive year. Overall, banks are well covered against potential losses on existing problem assets—loan loss reserves rose to 182 percent of noncurrent assets—although the ratio of reserves to total loans declined to 1.9 percent, the lowest ratio since 1986. While reserves may not provide a significant buffer against a sudden increase in loan losses, U.S. banks are well capitalized, with a core capital ratio of 7.6 percent at end-1996, its highest level in at least nine years, despite a relatively high volume of share repurchases by many of the banks.14

    For the first time in six years, all of the major banks in France reported positive net profits for 1996.15 However, this outcome was due mainly to exceptionally high income from capital market activities and from lower loan loss provisions. The core banking operations continue to suffer from low loan demand, poor asset quality, and aggressive competition, which have limited net interest income and reduced the capitalization of the banks. Thus, while noninterest income rose by 23 percent in 1996, net interest income declined by 2 percent. A 10 percent reduction in loan loss provisions allowed net income to rise by 140 percent, albeit from a relatively low base, for a return on equity of 7.7 percent. Not all French banks disclose their nonperforming assets, but it is generally believed that the worst of the asset-quality difficulties is past. While a large number of loans to small and medium-sized firms are nonperforming, the banks are believed to have set aside sufficient reserves to cover anticipated loans. With the caveat in mind that it is difficult to judge the asset-quality position of the banks, the published figures on capitalization indicate that the leading banks are reasonably well capitalized, with total risk-weighted capital ratios of between 8.7 percent and 11.4 percent.16

    Property markets, a key source of nonperforming loans in France, continued to deteriorate in 1996, with the vacancy rate on prime office space in the center of Paris, for example, rising to 9.2 percent at end-1996 from 8.4 percent a year earlier, and rents falling by 3.8 percent over the same period.17 Since the market peaked in 1990–91, commercial real estate prices have declined by about 50 percent. As a result, some of the banks have had to seek assistance from their shareholders either to move real estate assets off their balance sheets or for capital injections.18 Until the end of 1995, these operations had not resulted in significant sales of real estate. However, in 1996, some of the French banks began in earnest to reduce their exposure to property markets after having opted in earlier years simply to increase provisions against potential losses. In 1996, sales of real estate by bank shareholders helped to boost the volume of property investment transactions to F 12 billion, more than twice the 1995 turnover. The banks also began selling off large portions of their real estate loan portfolios, much of it to U.S. investors. The process began with the sale of F 870 million in real estate loans by Barclays Bank in late 1995, followed in 1996 by Créditsuez (F 4.75 billion), and UAP (F 3.2 billion), Créditsuez has announced its intention to sell off a further F 4.9 billion in real estate loans by 2001, and the Consortium de Réalisation (CDR) has proposed to sell F 1 billion. The first securitization of real estate loans in France was accomplished in January 1997 with the sale of F 1.5 billion in commercial mortgage backed securities (CMBS) backed by mortgage loans of Banque SOFAL, a subsidiary of Union lndustrielle de Crédit.19

    The French government announced a three-point resolution plan for Credit Foncier de France (CFF) in July 1996: (1) the Caisse des Depots et Consignations (CDC) would launch a public offer on behalf of the government for a sale of at least two-thirds of CFF; (2) a new public entity, the Caisse National du Credit Foncier, wholly owned by the state, would be established and acquire these shares from the CDC and wind down CFF’s business over the next 10 years; and (3) Credit Immobilier de France, a mutual institution, would take over the management of CFF’s portfolio of F 110 billion in subsidized home loans as well as its branch network and 1,500 of its 3,300 employees.20 The government has estimated the costs of this plan at F 2.5 billion. While the share issue went ahead as planned, with the government now owning more than 90 percent of CFF, the creation of Caisse National du Credit Foncier and the transfer of CFF assets to Credit Immobilier de France have not. CFF remains in business and reportedly earned F 1 billion in profit in 1996, but has virtually no capital (its total capital adequacy ratio at end-1995 was 0.5 percent) and its loan portfolio has fallen by half.

    The other institution that was forced to turn to the French government for assistance as a result of asset-quality problems. Crédit Lyonnais, returned to profitability in 1995 and improved on those results in 1996, earning a net profit of F 1.5 billion. However, this was possible only after the French Ministry of Finance agreed to neutralize the effect of the negative interest rate spread on a F 135 billion loan to the Etablissement Public de Financement et de Réstruciuration—used to finance the purchase of loans from Crédit Lyonnais by the CDR. The carrying cost of the loan, estimated at F 3 billion in 1996, has been assumed by the government. As part of the EU approval for the recapitalization plan in 1996, Crédit Lyonnais has begun selling its domestic and foreign subsidiaries and investments, including Banque Laydernier, Crédit Lyonnais Bank Sverige, Woodchester Investments, Banco Portugues de Investimento, and Banco Provincial. The total cost of official support to Crédit Lyonnais was recently estimated by the ministry of finance at F 100 billion.

    The largest private commercial banks in Germany reported substantially higher profits in 1996, owing mainly to strong trading profits and an increase in holdings of securities and other investments. Net income for the five largest commercial banks rose 19 percent (for a return on equity of 9.8 percent), as non-interest income rose by 25 percent.21 German banks alone among those of the seven major industrial countries have not experienced a severe deterioration in asset quality in recent years, although data on nonperforming loans and specific loan loss provisions are not generally disclosed. Disclosed provisions—which include gains on securities held in the liquidity reserve—declined by 21 percent in 1996. Real estate may become more of a problem in 1997–98, however, with the removal of tax concessions on development in the eastern Länder and a softening of the office rental markets in some key markets, including Berlin and Hamburg.

    The banking system in Italy continues to struggle with worsening asset quality, increasing labor costs, and an unfavorable tax regime that has contributed to weak profitability. However, the Italian Banking Association and trade unions have recently agreed to a reduction in labor costs to the EU average as measured by the ratio of labor costs to gross income. The costs associated with labor shedding will be borne by the banks without any support from the government. Furthermore, a recent reform of the corporate tax code and an increase in the deductability of loan loss provisions will lower the tax burden on banks. The continuing weakness of the economy, particularly in the southern region, has slowed loan growth, while increasing competition has narrowed interest margins. The steady decline in the level of interest rates contributed to a resurgence of activity in the securities market, resulting in a 61 percent increase in noninterest income, which allowed operating profit to increase by 7 percent.22 However, asset quality continued to deteriorate in 1996, albeit at a slower pace than in 1994 and 1995. The stock of bad loans (defined as sofferenze and protested bills) increased by 11 percent to Lit 128 trillion, or 10 percent of total loans, up from 9 percent at end-1995. The estimated loss rate on these loans also increased, to 38 percent from 33 percent at end-1995.

    The relative severity of the economic recession in southern Italy has led to a higher concentration of bad loans in that part of the country,23 with the result that a number of banks based in the south have experienced serious capital depletion because of loan losses. One of these, Banco di Napoli, incurred net losses of Lit 4.3 trillion during 1994–95. In the first half of 1996, the bank lost a further Lit 700 billion, and by the end of the year, it had lost Lit 1.6 trillion. In July 1996, the government announced a recapitalization package involving (1) the elimination of existing equity and the issue of Lit 2 trillion in new capital; (2) the transfer of about 30 percent of the loan portfolio to a special purpose company (Societa per la Gestione di Attivita); and (3) the sale of a 60 percent stake, which ultimately went in January 1997 to a combined offer from government-owned Banca Nazionale del Lavoro and a private insurer, Istituto Nazionale delle Assicurazioni. In addition, Banco di Napoli has sold much of its performing medium- and long-term loan portfolio; its largest subsidiary will be liquidated; 50 branches in the north have been sold; and agreement has been reached with trade unions to lower personnel costs. However, half of the capital increase obtained by the new share issue has already been drawn down to cover the losses incurred during 1996.

    The provision of extensive government support for the Finnish, Norwegian, and Swedish banks in the early 1990s prevented what may have been the complete collapse of the banking systems in these countries. By the end of 1996, the recovery from the crisis was more or less complete in Norway and Sweden—marked by the removal of the government guarantee for Swedish banks in July 1996. Surging loan growth and plummeting loan loss provisions—indeed, many of the banks have been writing back significant amounts of provisions—and a supportive capital market environment have provided the stimulus for recovery. For the Finnish banks, however, important weaknesses remain, and the recovery has been uneven. The Danish banks, while avoiding an outright crisis, nevertheless endured a serious deterioration in asset quality and earnings in the early 1990s.

    Net income for the four largest commercial banks in Denmark declined by 2 percent in 1996, for a return on equity of 16 percent.24 The stock of nonperforming loans fell for the second consecutive year, by 31 percent, accounting for 1.5 percent of total loans. Despite a decline in provisioning, loan loss reserves at end-1996 equaled 207 percent of the stock of nonperforming loans. The banks in Finland, of all the Nordic banks, have experienced the slowest recovery. After five years of net losses, the three major banks25 recorded a net profit in 1996 of FM 1.6 billion, a return on equity of 8 percent, due mostly to a 37 percent increase in noninterest income and a 23 percent decline in provisions. Net nonperforming loans declined by 41 percent to 2.5 percent of gross loans, or 26 percent of equity. The four major banks in Norway suffered an 11 percent decline in net income, but nevertheless recorded a return on equity of about 20 percent.26 Net nonperforming loans declined by 20 percent, to 1.4 percent of total loans, or 20 percent of equity. The five largest commercial banks in Sweden reported sustained high income and lower nonperforming loans.27 Aggregate operating profit rose 28 percent, owing mostly to strong noninterest income and a 43 percent decline in loan loss provisions and write-offs. Net problem loans fell by 30 percent, to 1.6 percent of aggregate loans. The extent of the recovery was highlighted by the removal on July 1, 1996, of the blanket guarantee that the government had established for the banks.

    The resolution of asset-quality problems in Japan continued to dominate developments in the banking system. Net interest income rose by 9 percent owing mainly to special circumstances affecting the interest expenses of the long-term credit and trust banks;28 the city banks’ net interest income declined by 6 percent. Without the benefit of declining interest rates boosting the valuation of the banks’ large investment bond portfolios, noninterest income declined by 31 percent, while commissions and trading income were flat overall. At the same time, the banks wrote off ¥2,267 billion in unrealized losses on their equity holdings, which offset much of the ¥3,488 billion they had realized in order to finance provisions. As a result, net profits on equity of ¥1,128 billion were used to offset part of the ¥5,555 billion in loan loss provisions and write-offs, leading to a net loss for the 20 banks of ¥146 billion.

    At end-March 1997, the 20 major banks had ¥ 13.193 billion in core nonperforming loans (loans six months or more past due and loans to bankrupt borrowers), ¥3,247 billion in loans restructured at below the prevailing official discount rate, and ¥2,890 billion in loans made in support of customers. This total of ¥19.331 in problem loans represents a 25 percent decline over the previous year’s ¥25,663 billion. Total problem loans represented 4.9 percent of gross loans at end-March 1997, compared with 5.4 percent at end-March 1996, but most of this decline was due to writing off loans to the housing loan corporations (jusen), which had been fully reserved but not written off in the previous year. Excluding the jusen loans, problem loans fell only 15 percent, while the core nonperforming loans remained unchanged.29

    While the definition of problem loans has been gradually widened in the last two years it is still less encompassing than the U.S. model for example. The Japanese definition does not include loans that have been sold to the Cooperative Credit Purchasing Company (CCPC) or special purpose vehicles and excludes loans restructured at interest rates above the official discount rate.30 Most, if not all, of these loans would likely be considered nonperforming loans of a bank under U.S. practices. Applying a broader definition of problem loans to the Japanese banks yields an estimated aggregate problem loan figure somewhat higher than the official estimate.31

    Five depository institutions failed in 1996—including two regional banks, Taiheyo Bank and Hanwa Bank—some of which required intervention by the deposit insurance corporation. Also, 2 of the 20 major banks reported significant restructuring packages in April 1997. On April I, 1997, Nippon Credit Bank (NCB) announced that it would: (1) write off ¥460 billion in nonperforming loans, which would result in a pretax loss of ¥350 billion; (2) cut salaries (by 10–30 percent for most personnel, and 50 percent for managers) and personnel (by 900, or 31 percent), and sell off all of its own real estate holdings, including its headquarters; (3) not pay dividends for the 1996/97 fiscal year; (4) cut assets by ¥6 trillion to ¥10 trillion (a 37 percent decline); (5) close its foreign branches and subsidiaries (five branches and seven representative offices); and (6) seek an injection of capital of ¥300 billion to prevent insolvency. This capital injection would come from shareholders and the other long-term credit banks (for a combined ¥70 billion), subordinated debt holders (mostly insurance companies, ¥140 billion), and the New Financial Stabilization Fund (¥80 billion), which was originally established to assist in the resolution of the jusen last year. It also announced that its three nonbank affiliates—Crown Leasing, Nippon Total Finance, and Nippon Assurance Finance Service—would file for bankruptcy rather than be bailed out by the NCB. These three institutions had total loans of ¥1.8 trillion, of which only ¥300 billion was owed to NCB. All other creditors would have to write off their loans to these nonbanks (it was reported that agricultural cooperatives had total loans to the nonbanks of ¥240 billion and the seven trust banks had lent a combined ¥430 billion). On April 10, 1997, NCB announced an agreement with Bankers Trust involving collaboration in securitization and international operations. Also on April 1, HokkaidoTakushoku Bank (HTB) announced that it would merge with Hokkaido Bank (the twenty-second largest of the first-tier regional banks) at the end of the 1997/98 fiscal year, forming a new institution, tentatively called the New Hokkaido Bank (NHB). HTB announced that it would also close or transform to representative offices all of its 20 overseas branches and subsidiaries, and the merged entity would close 100 of its combined 340 branches. HTB will also cut 2,000 employees, resulting in expected cost savings equivalent to 30 percent of operating expenses.

    The NCB resolution strategy combined elements of the old approach to bank failures—shareholder banks, the long-term credit banks, and the Bank of Japan (through the New Financial Stabilization Fund) are to provide capital for NCB—and some new ideas, and holders of subordinated debt were also asked to inject capital. More important, the bankruptcy of NCB’s nonbank affiliates was a significant departure from the principle of parent responsibility. However, at end-March 1997, NCB had ¥1.3 trillion in problem loans, still 7 percent of total loans. In addition, NCB’s total capital ratio has been reduced to 3 percent and its hidden reserves on listed and unlisted securities have been reduced by 86 percent, covering only 3 percent of problem loans. More fundamentally, NCB, like all of the long-term credit banks, suffers from a decline in franchise value. The market for long-term corporate loans has been eroded by competition from other banks, foreign institutions, and capital market funding.

    The Next Challenge: Core Profitability

    As the major banking systems recover from their asset-quality problems, the focus is turning toward their underlying strength and their ability to respond to the changes in the financial landscape outlined at the beginning of this annex by increasing noninterest income and yielding greater returns to equity holders. Table 33 shines some light on these issues by comparing the income generation and profitability of the banking systems of the major industrial countries over the 10-year period 1985–94, During that time, almost all of these countries have experienced a significant deterioration in asset quality and profitability at some time, which shows up in the reported returns on equity and preprovision profits (banks with higher nonperforming loans incur greater expenses in managing their loan portfolios). Nevertheless, data reveal some major trends.

    Table 33.Major Industrial Countries: Commercial Bank Profitability
    Net Interest IncomeNoninterest IncomeOperating ExpensesProvisionsReal Return on Equity1
    (In percent of total assets)(In percent)
    United Kingdom3.
    United States3.
    Sources: International Monetary Fund, World Economic Outlook database; OECD (1996); and IMF staff estimates.

    Over the 1985–94 period, the Canadian, U.K., and U.S. banking systems were consistently the most profitable in terms of core, preprovision earnings among the G-7 countries. The average preprovision return on assets over that period was 1.8 percent for U.S. banks, 1.7 percent for Canadian and U.K. banks, 1.1 percent for German banks, 0.6 percent for French banks, and 0.5 percent for Japanese banks (not shown). The relatively strong recent performance of the Canadian, U.K., and U.S. banks compared to those in, for example, France and Japan, is not, therefore, simply due to their having gone through the recent asset-quality cycle more quickly; banks in the three leading countries have simply been fundamentally more profitable. These profits, however, do not reflect relatively wide intermediation spreads.32 As Figure 53 shows, loan rate spreads over deposits have tended to be higher in France, Germany, and Italy compared with Canada, the United Kingdom, and the United States. A relatively higher proportion of low-margin securities and other assets, and relatively worse overall asset quality during the period, appear to explain most of the difference.

    Figure 53.Major Industrial Countries: Intermediation Spreads

    (Average lending interest rate minus average deposit interest rate)

    Source: International Monetary Fund, International Financial Statistics.

    While most banks have responded to the increased competition in lending activities by expanding their noninterest earnings, banks in Germany and Japan have actually seen a decline in importance of such income. Noninterest income for the Japanese and three continental European banking systems, which is decreasing, remains a much less well developed source of income than it is for banks in Canada, the United Kingdom, and the United States. This lack of diversification in income has meant that the decline in asset quality was particularly costly, since these banks relied much more heavily on interest income. With increasing loan loss provisions and declining interest income, the banks’ profitability dropped. The real return on equity, for example, has fallen sharply in Japan and in Germany, and less so in the United Kingdom, while it has risen significantly in Canada and the United States owing to a combination of increasing underlying profitability and declining provisions. As the data in Table 33 indicate, however, it is relatively weak earnings generation that hampers the French, German, Italian, and Japanese banking systems, and which has made their recovery from their asset-quality difficulties in some cases quite difficult.33

    The issue of core profitability is inextricably linked to the structure of the banking system and, particularly, the advantages that certain types of institutions may have because of different regulatory regimes, subsidies, or ownership structures that place less emphasis on returns to capital. In an environment in which banks are encouraged by the regulatory regime to maximize returns to equity capita), such features of a financial system can make it more difficult for some banks to compete. To be sure, certain of these differences can be to the detriment of the special institutions—a contributing factor to the savings and loan (S&L) crisis in the United States was the more relaxed supervisory and regulatory environment in which they operated, which allowed the S&Ls to run up very large loan losses. Similarly, the lower level of official oversight over the credit unions and credit cooperatives in Japan contributed to difficulties in those sectors. However, often the special treatment of certain types of institutions works to the detriment of the larger, internationally active, commercial banks, in part by reducing the domestic profits on which they attempt to leverage their international activities. (For example, since profit margins on retail banking services tend to be wider than on wholesale transactions, restrictions on competition in retail markets can prevent access to profitable business by commercial banks.) While few, if any, countries have a perfectly level playing field—for instance, savings institutions often enjoy advantages over commercial banks—some of these distortions have been more prominent in recent years, including the government ownership or mutual ownership of banks in France, Germany, and Italy which, by apparently downplaying the incentive for maximizing shareholder returns, has arguably allowed these institutions to compete aggressively against the private commercial banks in the same countries. Private bankers from France and Germany have argued to the European Commission on Competition that government-owned banks in these countries have been unfairly subsidized by the manner in which they were recapitalized.

    In France, a number of the largest commercial banks have criticized the terms of the official assistance given to Crédit Lyonnais. The EU Competition Commissioner approved the recapitalization only after it was agreed that Credit Lyonnais would divest itself of 35 percent of its foreign affiliates. The private banks in Germany have also questioned the government support that had been extended to some of the Landesbanks in the late 1980s. More generally, the Landesbanks have received higher credit ratings than many of the private banks because they are guaranteed by the government. For example, Moody’s rates the major commercial banks’ long-term foreign currency debt between Aaa and Aa2 and nine Landesbanks’ debt between Aaa and Aal—with the average Landesbank rating slightly higher than that of the commercial banks—but the standalone Bank Financial Strength Ratings are much higher for the commercial banks (B+) compared with the Landesbanks (C+). This public guarantee allows the Landesbanks to raise funds domestically and on international markets at more favorable rates than private banks can. At the same time, the Landesbanks earn an interest margin approximately half of that of the major banks and earn a correspondingly small return on equity (averaging 4 percent over 1990–95).

    The issue is not necessarily that government-owned or cooperative banks are less efficient than privately owned banks, although that is often the case, but that official guarantees, subsidies, or regulatory advantages that segregate markets—for example, by giving certain institutions exclusive rights to offer certain types of retail deposit instruments or bonds, or certain types of loans—are inherently inefficient. By restricting the scope of competition, such structures can end up supporting an inefficient allocation of capital in the financial system, and worse, allow the accumulation of large losses that ultimately become a claim on the official sector.

    Banking System Developments in Selected Emerging Markets34

    Developments in Asian Emerging Markets

    Macroeconomic developments strongly influenced the performance of banking systems in Asian emerging markets. Growth slowed (and current account deficits worsened) partly because of cyclical factors but also because of longer-term fundamentals. Growth in countries such as India, Indonesia, Korea, Malaysia, and Thailand slowed in 1996 compared with 1995 and is forecast to slow further in 1997.35Granted that even these slower growth rates surpass all but the highest growth rates in Latin America, the slowdown, nonetheless, has had serious repercussions for the financial systems in some Asian emerging markets, because it has revealed the underlying illiquidity in the corporate sectors and the lack of preparedness among some financial institutions for the slowdown and the resulting worsening of asset quality. Unless growth and export performance improve in 1997 the banks in some Asian countries may face significant challenges to their profitability and possibly their solvency. Given the history in many Asian emerging markets of providing broad support to the financial system by ensuring the survival of individual financial institutions, this vulnerability of the banking sector to further deterioration of macroeconomic fundamentals may become an increasingly important constraint on fiscal and monetary policies.

    Property market developments also figured prominently in Asian banking developments in 1996 as key real estate markets deteriorated. Table 34 shows the recent trends in vacancy rates and rents on prime office space in some of the key property markets in Asia. The well-publicized weakness in the Bangkok market appears in the table as a slight increase in the vacancy rate, from an already high 13.2 percent at end-1995 to 13.9 percent at end-1996, and virtually flat rents. Moreover, record amounts of commercial real estate and office space are due to become available in 1997–99, which will further depress values. In Jakarta, a significant increase in vacancy rates was offset at least partly by an increase in rents, but there too the market has underlying weakness. In Kuala Lumpur and Manila (Makati), conversely, vacancy rates are extremely low.

    Table 34.Selected Asian Property Markets: Vacancy Rates and Changes in Rents1(In percent)
    Vacancy RateChange in

    March 1996March 1997
    Hong Kong, China4.75.16.3
    Kuala Lumpur2.82.90.0

    Korea’s accession to OECD membership in 1996 focused international attention on its banking system and in particular raised some concern that increasing competition from foreign banks and a surge in foreign capital inflows could put pressure on the system. However, foreign banks already have a presence in Korea and operate on an equal basis with domestic banks (there were 77 foreign branches at end-1995); and while capital account liberalization may have a more important effect, the gradual approach to liberalizing inflows provides an opportunity to strengthen the domestic financial sector. Instead, the vulnerabilities in the Korean banking system have their roots in past practices. Explicit government-directed lending (Industrial Rationalization Loans)36 has given way to directed lending of a different kind—for example, banks are required to allocate a certain proportion of marginal loans to the small and medium-sized enterprise sector—while political influence on lending decisions appears to continue. Consequently, many Korean banks have not yet developed sophisticated internal credit evaluation systems, placing greater emphasis on collateral than on an analysis of the project being financed.

    In the current environment, many banks have built up large exposures to individual corporate groups (chaebols), many of which are highly leveraged.37 The reduction in economic growth and export prices in 1996 heightened the illiquidity of many of these groups. In addition, many Korean firms have taken on increasing amounts of foreign-currency-denominated debt—both by borrowing in the international markets directly and by borrowing in foreign currencies from Korean banks. It is believed by many market analysts that very little of this currency exposure is hedged (the onshore forward foreign exchange market is very illiquid and would provide cover for only about 12 months in most cases), so that as the won depreciated these debts became more expensive to service. In the past year at least four large corporations or groups have defaulted on their debts, and other large companies are rumored to be in difficulty.

    The growing liquidity problems among the large corporations have already put pressure on the banks’ asset quality. At end-1995, nonperforming loans (those six months or more past due) among the eight largest commercial banks, for example, amounted to 6 percent of total loans.38 By the end of 1996, the reported non-performing loans had declined to 4.3 percent of loans for these eight banks, but if the exposures to Hanbo Iron and Steel Company and Sammi Steel are included, this ratio rises to 6.0 percent.39 Net of reserves, these nonperforming loans equate to 68.9 percent of the banks’ equity. However, since only loans six months past due are included, and restructured loans are not included at all, the true asset-quality situation of the Korean banks may be much worse (see Appendix 2 at the end of this annex for a description of asset-quality accounting practices in 20 emerging markets).

    As is true among emerging markets elsewhere, much of the collateral that secures problem loans is real estate or specialized fixed capital that may be difficult to repossess or to sell at book value. Also, there are other potential sources of problems for the Korean banks in addition to the loans to the chaebols. Consumer loans, including credit cards, have been an important source of expansion for Korean banks in the past few years, and these have emerged as having high ratios of nonperforming loans. In addition to declining loan quality, the banks have had to deal with the collapse of Korean equity prices in 1996, which has imposed large revaluation losses on their extensive equity portfolios.

    The Korean authorities have responded to the declining asset quality and the losses on banks’ equity portfolios in part by engaging in regulatory forbearance. After raising the loan loss provisioning requirement for doubtful loans to 100 percent at the beginning of 1996, the requirement was later lowered to 75 percent. In addition, the banks were allowed to provision for only 30 percent of the securities revaluation losses in 1996, rather than 50 percent as would otherwise have been required. Despite these measures, the 15 largest commercial banks reported a 3.7 percent decline in net income in 1996.

    The commercial banking system can be viewed as comprising three groups, the six large, older commercial banks, two government-owned banks, and a number of newer commercial banks.40 Performance differs greatly between the first two groups and the third. The newer banks, with a shorter history of policy lending, generally have much better asset quality, are believed to have superior credit risk management skills, and are more efficient. The group of older banks have seen net income decline for two years in a row, with the return on equity falling from 5.9 percent in 1994 to 1.8 percent in 1996. With the share of loans at fixed low rates declining, net interest margins have increased slightly, but are still very low, at 2.3 percent in 1996. As an indicator of their relative inefficiency, the cost-to-income ratio for this group of banks rose from 82 percent in 1994 to 94 percent in 1996. Finally, 5.1 percent of their loans were nonperforming at end-1996 (6.9 percent with the exposures to the Hanbo and Sammi groups). For the group of newer banks, net income rose 34 percent in 1996 after a decline in 1995, for a return on equity of 8.8 percent. The net interest margin has fluctuated around an average of 3.1 percent since 1994, and the cost-to-income ratio in 1996 was 84 percent. On all of these measures, they outperform the older banks. Finally, for the two newer banks for which asset-quality data are available, nonperforming loans at end-1996 reached 2.4 percent of loans (only 2.5 percent with the loans to the Hanbo and Sammi groups), marginally higher than in 1995 (since they had fewer policy loans to begin with, and repayments on previously nonperforming policy loans were the main source of improvement in the asset quality of the older banks).

    The Philippines has been the recipient of a surge in capital inflows since 1991–92 as foreign investor sentiment about the economic fundamentals—and expectations of ratings upgrades (S&P raised its rating for the Philippines in February 1997, and Moody’s followed suit in May 1997)—have led to increases both in foreign direct investment and in portfolio investment. The latter has partly fueled a rise in equity prices of almost 200 percent since early 1992. Also, in the last few years, property prices have surged, which has raised concern about the sustainabilily of asset prices, although as Table 34 indicates, at least some of the increase in prices was due to declining vacancy rates in office buildings.

    Partly as a result of the inflows of capital, liquidity in the banking system has been high in recent years, fueling a rapid expansion in bank lending—up by 44 percent a year during the past two years. This doubling of loans in the past two years raises questions about asset quality in the future. While nonperforming assets accounted for only 3.3 percent of total bank assets in 1996, down from 3.6 percent in 1995, some analysts have raised concerns about the expansion in bank lending. The first concern is that, as in other countries in the region, commercial banks in the Philippines have a relatively high direct and indirect exposure to real estate. The central bank has estimated the direct exposure of commercial banks through real estate loans at about 10 percent in March 1997, up from 9 percent in March 1996. The true exposure may be higher since property is a common form of collateral, and the banks have other exposures through investments in property developers. In April 1997, the central bank imposed a limit on real estate loans of 20 percent of a bank’s total loans and reduced the maximum loan-to-value ratio to 60 percent from 70 percent. Analysts have also expressed concern that consumer lending, including credit card debt, has increased as a share of total loans.

    As is the case elsewhere in the region, foreign currency exposure, particularly of the corporate sector, has been expanding. Philippine banks’ Foreign Currency Deposit Unit (FCDU) loans expanded by 110 percent in 1996 alone. Prudential regulations require banks to keep balanced FCDU books, so the direct foreign currency exposure should be small. Moreover, 74 percent of FCDU loans at end-1996 were extended to exporters, oil companies, and public utilities for whom currency risk is not thought to be significant. However, if the remaining borrowers are not hedged, their currency risk could be translated into credit risk for the banks. In addition, much of the expansion in peso-denominated lending has been financed by offshore foreign-currency-denominated borrowing by the banks, whose net foreign liabilities had increased to $6 billion at end-1996 from near zero at end-1995. In response, the central bank introduced a 30 percent liquidity requirement on foreign-currency-denominated assets in July 1997.

    The condition of financial institutions in Thailand figured prominently in the exchange rate pressures in 1996 and 1997, as the financial system was seen as a source of potential weakness by foreign investors. The banks were thought by some investors to be vulnerable to a depreciation of the baht as a consequence of their own net foreign-currency-denominated liabilities or those of their customers. At the same time, however, the strength of the baht and the high interest rates needed to maintain the exchange rate, combined with a heavily overbuilt Thai property market, have contributed to asset-quality problems for the banks.

    In many respects, the performance of the major Thai commercial banks in recent years has been good.41 While earnings, as measured by the return on equity, have dipped slightly, the return in 1996 was over 20 percent, and the decline has been due to an increase in equity rather than a decline in earnings. Net income has increased steadily for the last eight years, driven by average loan growth of more than 20 percent a year and a gradual widening of net interest margins, although there has been some retrenchment in margins in the past two years. The banks have, however, sharply increased their net foreign borrowing. The Bank of Thailand reports that all commercial banks’ net foreign liability increased from B 51 billion (35 percent of capital) at end-1991 to B 1,035 billion (twice capital) in October 1996, but these are not necessarily the banks’ true foreign currency positions. Some market participants report that the Thai banks had hedged most of their net foreign liabilities, so that the devaluation had perhaps relatively little direct effect on their balance sheets. The opposite is believed to be true for the Thai corporate sector: market participants report that until doubts about the sustainability of the exchange rate policy grew in late 1996, Thai firms had made little attempt to cover their foreign currency exposure. As a result, the depreciation of the baht could result in an increase in non-performing foreign currency loans.

    The incentive to borrow abroad over the past few years reflects the illiquidity in the domestic markets. The Thai banking system has grown at a very rapid pace in the past six years—the average annual growth rate of credit to the non-financial private sector over 1990–95 was more than 23 percent—and the loan-to-deposit ratio increased from 103 percent at end-1990 to 141 percent in October 1996. Liquidity concerns were exacerbated by the high interest rates used to prop up the baht, the inability to manage short-term liabilities in the domestic markets actively because these markets are illiquid or underdeveloped, and more recently, the emergence of wider spreads on Thai credit in international markets, including in the international interbank markets at times.

    The illiquidity of the Thai corporate sector has created difficulties for the banks in terms of declining asset quality. The Bank of Thailand reported in December 1996 that total non-performing (doubtful and substandard) loans amounted to 6.92 percent at end-1995 and 7.73 percent at end-June 1996.42The worsening non-performing loan problem is frequently attributed to the weakness in the property market.43Developments in the property market in Thailand have been an important source of vulnerability both because the banks have lent to this sector—already this year two property developers have defaulted on debts—and because property has often been used as collateral for loans. While the official estimate is that 10 percent of bank loans are to the property sector, one private estimate puts the figure at 20 percent (equal to 33 percent of end-1996 equity). The Government Housing Bank has estimated that 40 percent of the housing stock built during 1992–96 was unoccupied at the beginning of this year. While property prices have not declined significantly, the data in Table 34 show that even the highest-quality office space in Bangkok has a relatively high vacancy rate and flat rents, and the supply of office space is projected to grow by 35 percent by the end of 1998. The excess supply is even worse for apartments and hotels.

    A potentially greater danger to the banking system, however, is the possibility that the banks may be expected to supply financial support to affiliated finance companies, even if they are not majority shareholders. Many of the large commercial banks have invested in non-bank financial institutions, which are generally in weaker financial shape than the banks because of their much higher exposure to the property market. Since the Bank of Thailand has already stated its willingness to recapitalize nine very small unaffiliated finance companies through the Financial Institutions Development Fund (FIDF), some market participants suspect that the banks would be expected to do the same for the larger finance companies.

    The Thai authorities’ response to the latest concerns about the health of the financial system have focused on improving accounting and disclosure for asset quality and on rehabilitating the property market. On the first front, as of July 1997, banks are required to begin disclosing non-performing loans and provisions and to report restructured loans as well. Also, by end-June 1999 the banks will be required to have set aside reserves equal to 15 percent of substandard loans. On the second front, the authorities have introduced a number of measures to try to support the property market including (1) the creation of a secondary mortgage market by the Government Housing Bank, which will also provide low-cost mortgages to government employees; (2) increases in the foreign-ownership limit on some types of property; (3) a reduction in the land transfer tax from 2 percent to 0.01 percent; (4) proposed legislative measures to allow the establishment of real estate investment trusts and the securitization of property related loans; and (5) the establishment in March 1997 of the Property Loan Management Organization (PLMO). With a mandate similar to that of the CCPC in Japan, the PLMO was capitalized with B 1 billion from the fiscal budget and authorized to borrow up to B 100 billion (the first B 1 billion bond has already been issued) to purchase bank loans to property developers (or the collateral) at fair market value and restructure the loan or the project.44 The banks that sell the loans would have to write off any difference between the purchase price and the book value of the loan and guarantee repayment of 50 percent of the loan.

    Finally, the Bank of Thailand is encouraging mergers among financial institutions, especially finance companies, as a way to weed out the weak and inefficient firms. In April, a group of seven finance companies announced plans to merge. However, merger talks between the largest finance company. Finance One, and Thai Danu Bank broke down in June 1997. Subsequently, in late June, the Thai Ministry of Finance ordered 16 finance companies to suspend operations and to draw up recapitalization plans within two weeks. Five large finance companies have agreed to lake over the good assets of those of the 16 that are unable to comply.

    Developments in Latin American Emerging Markets

    The economic recovery in Argentina in the second half of 1995 continued in 1996 and allowed the banks to recover from the liquidity crisis that developed in the first quarter of 1995 and to proceed with the longer-term process of restructuring in a less crisis-charged atmosphere.45 This recovery in economic activity was reflected in the growth of bank loans, particularly in the second half of 1996. After declining by just over 1 percent (in nominal terms) in 1995, and rising by less than half a percent through the first half of 1996, loan growth averaged 7 percent for the second half of the year.46 Deposit growth was even more impressive: the stock of deposits grew by 21 percent in 1996 after falling by 4.5 percent in 1995. Moreover, the growth in deposits in 1996 has been evenly balanced between dollar deposits and peso deposits, and the “flight to quality” that was observed in 1995—wherein the larger banks’ deposits actually increased while total banking system deposits declined—was not in evidence through most of 1996: deposits of the 20 largest banks grew only marginally faster than total systemic deposits.

    The return to more normal levels of liquidity in the banking system allowed the level of interest rates to fall, fueling loan demand. At the same time, however, net interest margins declined on average, reaching 4.95 percent for 1996 compared with 5.75 percent in 1995. Despite a modest increase in lending, the narrower margins and higher loan loss provisions (which increased by 14 percent in 1996) resulted in only a marginal increase (0.4 percent) in net income for the 20 major banks, with a return on equity of only 6.8 percent compared with 7.3 percent in 1995.

    However, a tiering of banks is observed in Argentina, consisting of (I) a group of large private banks that perform better than most other banks in terms of asset quality and profitability; (2) a group of smaller private banks and cooperative institutions; and (3) the government-owned banks (including the two largest banks) whose poor asset quality has necessitated high provisions and write-offs, resulting in poor profitability. The differences between these sectors are immediately apparent in their respective asset-quality data. At end-September 1996, the ratio of loans past due more than 90 days to total loans was 19 percent for the federally owned banks, 27 percent for the provincial banks, 9 percent for the domestic private banks, and 15 percent for the cooperative institutions.47 Moreover, the return on average equity for the government-owned banks was only 3.9 percent in 1996, compared with 10.9 percent for the large private banks, and the respective overhead costs as a proportion of total revenues were 70 percent and 65 percent.

    The banking system in Brazil came under pressure during the transition to a post-hyperinflationary economy under the Real Plan, which exposed weaknesses in asset quality and the regulatory and supervisory structures. The Real Plan succeeded in sharply reducing inflation but at the expense of a significant slowdown in economic growth—an outright contraction in activity for much of 1995. Loans in arrears and in liquidation increased from 7.25 percent at cnd-1994 to 13.4 percent at end-1995 and 14.4 percent at end-1996.48 These developments have resulted in a number of bank failures and mergers under distress over the past two years, and more consolidation is expected.

    The transformation of the economy after June 1994 has created a three-tiered commercial banking system in Brazil. There are, first, the publicly owned banks, including the banks owned by the state governments, which have in the past often been used simply as extensions of state treasuries. These banks did not develop a credit culture upon which to build franchise value from alternative lending, and they accumulated large stocks of non-performing loans. A second group of institutions includes a large number of small, private commercial and multiple banks with limited branch networks and relatively undiversified product lines that tended to fill niches in the system—such as treasury operations and wholesale banking. Having relied upon “float income” for much of their profit, and lacking the economies of scale necessary to compete in a low-inflation environment, these institutions too have found it difficult to adapt to the new economic reality. Finally, there are a small, but growing, number of well-capitalized and well-run large private commercial banks that are expanding to fill the gaps left by the retreating government-owned banks.

    Some insight into the condition of the government-owned banks can be gained by looking at the 12 state-owned banks for which balance sheets and income statements for 1995 and 1994 are available.49 Between 1994 and 1995, net interest income declined by 24 percent and margins fell by more than half (to 10.4 percent), non-interest income declined by 52 percent, and loan loss provisions tripled, all of which contributed to a net loss of R$862 million in 1995, after a net loss of R$151 million in 1994. The weakest of this group, Banerj, had a negative net capital position at end-1995 of R$ 1,859 million, or 79 percent of assets. For the seven banks that reported non-performing loans to the private sector, these amounted to 10.8 percent of private sector loans and leases in 1995, up from 4.4 percent in 1994, and reserve coverage declined from 107 percent to 77 percent. Hence, even excluding non-performing loans to the public sector (about 5 percent of total loans), asset quality deteriorated seriously in 1995.

    The situation is similar among the four federally owned commercial banks. As a group, they earned a net loss in 1995 of R$4.96 billion, although this was mostly owing to the loss incurred by Banco do Brazil. Excluding that bank, the three other federal banks for which data are available saw net interest income fall by 30 percent, interest margins fall to 13.1 percent, but noninterest income double, which compensated almost exactly for the shortfall. However, loan loss provisions rose to 13.7 percent of private sector loans and leases at end-1996 from 8.6 percent at end-1995.

    A key development in the government-owned banking sector was the R$12.3 billion loss by Banco do Brazil over 1995–96. In April 1996, the government announced a recapitalization plan for Banco do Brazil in which the government would underwrite a capital injection of R$8 billion in new capital. In the event, the government itself injected R$3.9 billion in new capital and transferred R$2.9 billion in shares of state-owned companies, while the bank’s pension fund injected R$1.2 billion.

    The group of small private banks is another source of stress in the Brazilian banking system, albeit perhaps less of a systemic threat. Aggregation of a sample of small banks’ financial statements for 1994–96 reveals a pattern of declining profitability and capital and rising levels of nonperforming loans.50 The return on average equity has declined for two successive years, from 42 percent in 1994 to 19 percent in 1996 (the return on assets has declined from 6 percent in 1994 to 3 percent in 1996), while the net interest margin has fallen from 13.7 percent to 8.6 percent. Net interest income has increased significantly since 1994, reflecting the banks’ reorientation toward lending and away from the securities trading and interbank lending activities that were so profitable during the high-inflation period. However, operating expenses have also increased, and loan loss provisions have tripled, which contributed to a decline in net income in 1996. At the end of 1996, these banks had non-performing loans equal to only about 5 percent of total loans (up from 2.8 percent in 1994). While these institutions are still well capitalized, with an equity-to-total-assets ratio of 15 percent at end-1996, they face growing competition from larger competitors, which have expanded to fill the niches these smaller institutions once occupied.

    The third group of institutions in Brazil are the larger private banks. Overall, profitability improved slightly in 1996 despite a narrowing of the net interest margin (to 9.8 percent, from 12.3 percent in 1995).51 Banks offset the declining margins by expanding loan volume by 12 percent and by increasing their noninterest earnings by 39 percent. However, a key source of improvement in the banks’ earnings was a reduction in loan loss provisions made possible by a significant improvement in asset quality, after two years of high charge-offs. While differences clearly exist between the banks, asset quality appears at least to have stabilized. The ratio of private sector loans in arrears and liquidation to total gross loans fell from 5.1 percent to 3.7 percent. Improving asset quality allowed room for banks to make lower loan loss provisions, while still increasing reserve coverage to 233 percent of nonperforming loans. This improvement in asset quality may simply be due to the expansion in banks’ loan books, but since much of this expansion reportedly went to consumer lending (almost all of it short term), where loan losses are usually relatively higher, it may only be temporary.

    Since the inception of the Real Plan, a number of measures or programs have been implemented to improve the process of removing problem banks from operation and to restore confidence in the remaining banks. In August 1995, a private deposit insurance scheme, funded and operated by the larger banks, was established as a temporary mechanism until a formal deposit guarantee fund (FGC) could be established in November 1995. This fund—financed by premiums levied on all financial institutions, charges on returned checks, and the possibility of special levies (up to 50 percent of the ordinary premium) or advances from member financial institutions or the central bank—guarantees repayment of a maximum of R$20,000 to each depositor at each bank.

    Also in November 1995, the government announced the PROER program to help promote consolidation in the banking sector by providing fiscal and financial incentives to banks that merge with or acquire all or part of another bank.52 Funding through PROER has been used on a number of occasions (most notably in the cases of Bamerindus, Banco Economics Banco Nacional, Banco Mercantil de Recife, Banorte, Banco United, and Banco Martinelli).

    In August 1996, the federal and state governments agreed to a program to reschedule the state government’s debt in return for the privatization, liquidation, or transformation (to development agencies that would not accept deposits) of the state-owned banks. The first to be privatized was Banco do Estado do Río de Janeiro, which was sold in June 1997. In 1996, the federal government established a program to restructure up to R$7 billion in debt of small agricultural producers (individual loans up to R$200,000).

    The banking system in Chile is considered to be one of the strongest among the emerging markets. This is reflected mainly in the asset quality of Chilean banks. At end-1996, only 1.03 percent of the total loans of the major Chilean banks were nonperforming, somewhat below the 1992–95 average.53 This low ratio is not thought to be due to tax accounting practices. On the contrary, the Chilean authorities have one of the most conservative accounting standards among all emerging markets regimes, and they are believed to crosscheck the information they receive from the banks on loan performance against information from the tax authorities. On the issue of accounting for bad loans, therefore, the Chilean banks are widely considered to be among the most conservative banks in the world. Like their counterparts elsewhere in Latin America, Chilean banks maintain very high loan loss reserves relative to the stock of bad loans, although at 180 percent at end-1996, this ratio has declined rapidly since end-1993 (when it was almost twice as high).

    Chilean banks’ earnings have come under pressure from increasing competition in recent years. The net interest margin has declined gradually over the past four years to just over 3 percent, and this shrinkage is not due to a large stock of nonperforming assets. Rather, it reflects declining interest rates, increasing competition, and the concentration of the banks’ activity in corporate lending. Only about 24 percent of end-1996 loans were to the consumer and mortgage sectors, where margins are higher than in the corporate sector. The banks are gradually changing this orientation—consumer loans grew at three times the overall growth rate in loans in 1996—and over time this reorientation will tend to boost net interest margins. Similarly, noninterest income has tended to be relatively unimportant for Chilean banks, but such income has increased significantly in the last few years. Faced as they are with relatively low income, Chilean banks are forced to be efficient providers of banking services, as indicated in the low overhead ratio (63 percent at end-1996).

    While the Chilean banking sector is fairly efficient and free of the asset-quality problems that plague banking industries elsewhere, it is a relatively capital-poor industry. With the subordinated debt situation essentially resolved (the last bank has just announced an agreement with the central bank to repay its debt), the banks’ capital levels have declined (subordinated debt was part of secondary capital). The average equity-assets ratio was only 5.3 percent at end-1996 and has declined steadily since at least 1992. Moreover, the ratio of liquid assets to total assets was only 8.9 percent at end-1996.

    Notwithstanding the improvement in the underlying economic environment, banks in Mexico endured a very difficult year in 1996. Commercial banks, excluding banks that were under central bank intervention or in other special situations, recorded an aggregate net loss of MexN$6.9 billion in 1996, after a profit of MexN$2.5 billion in 1995.54 The deterioration in net income stemmed mainly from a 27 percent decline in net interest revenue, reflected in a fall in the net interest margin to 3.83 percent from 6.54 percent in 1995. This contraction in interest income is attributable to the decline in interest rates and an increase in nonperforming or low-yielding assets on the banks’ balance sheets. The bulk of the banks’ total loans of MexN$698 billion were loans to FOBAPROA. UDI-restructured loans, loans to the government due to the ADE program, and nonperforming loans.55 These assets, which more than doubled in 1996, earn relatively low yields, which drags down the banks’ income.56 Loan loss provisions, which had been mainly responsible for the deterioration in income in 1995 over 1994 levels, rose by only 20 percent in 1996, to MexN$30 billion.

    After the serious disruption to the economy in 1995, asset quality remains a key concern for the Mexican financial system, and recent indications are that the situation has not improved. Nonperforming loans increased by 2.5 percent in 1996, to MexN$47.5 billion, despite the sale of MexN$l24 billion in (mostly non-performing) loans to FOBAPROA by virtually all of the important banks in Mexico in 1996.57 At the end of the year, nonperforming loans represented 6.8 percent of total loans, compared with 7.8 percent at end-1995. (Of the total classified portfolio of MexN$495 billion, medium-risk, high-risk, and irrecoverable loans represented 17.5 percent at end-1996, compared with 13.8 percent at end-1995.)

    The transfer of such a large stock of loans to FOBAPROA (approximately 39 percent of the end-1994 loan portfolio) and the associated recapitalization commitments have placed the system on what appears to be a much sounder footing.58 At end-1996, the capital adequacy ratio calculated by Banco de Mexico for the banking system was 13.1 percent, up from 12.1 percent at end-1996. However, most of the increase in capital has been in the form of revaluation gains. For example, of the total equity of MexN$70 billion at end-1996, only just under half (MexN$32 billion) was paid-up capital. Revaluation gains on equity and fixed assets contributed almost as much to equity (MexN$25 billion).

    After two turbulent and difficult years—in which 17 banks holding 54 percent of end-1993 deposits were closed or taken over by the deposit guarantee fund, FOGADE, and official assistance to the banking sector amounted to Bs. 1.6 trillion (18 percent of 1994 GDP)—the banking system in Venezuela recovered significantly in 1996. There were no further actions taken against any bank in 1996, and the authorities began toward the end of the year to privatize the banks that had been taken over. While important weaknesses and vulnerabilities remain, the crisis that began with the failure of Banco Latino in January 1994 appears to have eased.

    The major Venezuelan banks’ net income increased by more than a factor of three in 1996, resulting in a return on average equity of 74 percent.59 While this fell short of the average inflation rate for the year, much of the profit was earned in the second half of the year, during which inflation was considerably lower, which suggests that the banks have recently begun earning positive real profits. The three main sources of higher profits were net interest earnings from lending—loans increased by more than 80 percent in 1996—income from securities holdings, and extraordinary gains. The latter includes for some banks profits from long dollar positions (mostly in Brady bonds) that they had at the time of the devaluation of the bolivar in April 1996.

    Asset quality also improved markedly in 1996. At the end of the year, the banks reported past-due loans and loans in litigation equal to 4 percent of the gross loan portfolio, down from 10 percent at end-1995. (If restructured loans are included, the figures are 6.5 percent of the gross loan portfolio, down from 13 percent.) The ratio of loan loss reserves to past-due loans and loans in litigation likewise rose from 121 percent to 196 percent during 1996. The Venezuelan banks are similarly well capitalized. As a result of capital injections by shareholders, the equity-assets ratio has increased to 13 percent, from 8 percent at end-1995.

    All is not entirely well, however, with the Venezuelan banks. They continue to invest a large proportion (nearly 50 percent) of their earning assets in securities (mostly government bonds)—only 34 percent of total assets is represented by net loans, although lending has picked up significantly in 1997. Moreover, their current profitability is supported by still very wide net interest margins—19 percent in 1996, compared with 16 percent in 1995 and 14 percent in 1994, Finally, banks do not yet provide a reliable store of value for depositors, who are believed to keep only a small fraction of savings in Venezuelan banks.

    Developments in European Emerging Markets

    The banking system in the Czech Republic endured an eventful year in 1996, in which eight banks were intervened in by the Czech National Bank (CNB), including the fifth-largest bank, which has been supported by an irrevocable blanket guarantee of all liabilities by the CNB. At the end of the year, five banks, accounting for less than 4 percent of banking system assets, were under special CNB conservatorship and two had been liquidated—joining the four other banks that had failed since 1994.

    The situation is not as bleak as these facts suggest, however, in part because the authorities had known of these banks’ difficulties for a few years—this was not a situation in which a large number of banks suddenly became insolvent. Serious asset-quality problems began to emerge in 1993—94 among the newer private commercial banks that had been established in the Czech Republic during 1991–93 when licensing rules were fairly tax. Two banks were closed because of asset-quality problems in 1994 (and another because of alleged fraud). In 1995, similar problems emerged at a large number of the smaller banks, and the authorities responded by closing the worst affected, placing others under special monitoring, and ensuring that the CNB had sufficient powers to intervene in weak banks. When the financial results for 1995 revealed how weak these institutions were, the CNB exercised these powers in seven banks and closed them or placed them in receivership. An eighth bank required the support of the CNB, including a blanket guarantee for all domestic and foreign creditors, because it had a common shareholder with one of the failed banks and had therefore suffered a serious decline in liquidity. A scheme for restructuring small banks was announced in October 1996.

    Although the share of impaired loans remains high, the official data on classified assets indicate a slight improvement in asset quality in 1996. Excluding the Consolidation Bank, classified loans declined from 33.4 percent at end-1995 to 30.1 percent at end-1996 and 29.7 percent at end-March 1997. In addition, the larger banks have seen a decline in nonperforming loans, are adequately (if not fully) reserved against nonperforming loans, and have investment grade ratings; foreign-owned or joint-venture banks are also in reasonably good condition.

    However, poor asset quality is still a concern. Even after taking CzK 140 billion in nonperforming loans off the banks’ balance sheets (by selling them to the Consolidation Bank) and transferring bonds worth CzK 57 billion to banks to boost their capital, some CzK 339 billion in classified assets remained on banks’ balance sheets at end-September 1996. The ongoing provisioning requirements to meet the required coverage ratio and to maintain it as the stock of classified loans rises consumes a third of operating income (although this fraction is declining). The history of high loan toss provisioning and high nonperforming loan ratios has resulted in a system that is not highly profitable or capitalized. While most of the banks exceed the 8 percent minimum capital adequacy requirement, the average capital adequacy ratio at end-1996, excluding the Consolidation Bank, was 10.3 percent.

    The banking systems in both Hungary and Poland continue to recover from the asset-quality problems that plagued them in the late 1980s and early 1990s. Improvements have been due in large part to extensive restructuring in the context of bank recapitalization and privatization. At end-1996, classified loans in Hungary represented 11.6 percent of total loans, down from a peak of 28.5 percent at end-1993. Similarly, in Poland, at the end of the first quarter of 1996, irregular loans represented 18 percent of the portfolio, compared with 31 percent at end-1993.

    Hungarian bank regulations have gradually been brought up to EU standards, including the requirement that banks must publish their financial statements in both International Accounting Standards formal and in Hungarian accounting format. The adjustments to the supervisory and regulatory structure accelerated in 1996, with a complete overhaul of the architecture of supervision and important changes to the content of bank regulations. The banking and securities supervisory agencies have been combined into a joint supervisory agency, and approval for banks to engage in securities activities was given in the new banking law implemented at the beginning of 1997, In addition, the new law doubled the minimum capital requirement to Ft 2 billion;60 and tightened restrictions on ownership, related party transactions, and credit and market risk exposures.

    Appendix 1

    Credit Derivatives

    While credit risk has vexed financial markets since money lending began, the ability to trade and transfer credit risks with ease is a new phenomenon. The first deals done with such structured products, called credit derivatives, were executed in 1992. Steady growth since then has brought the credit derivative market to what is viewed as a critical mass, catching the attention of potential customers and regulators alike. Although official statistics are not yet available, industry estimates put the size of the credit derivative market at about $40 billion of outstanding transactions.61 While still only a drop in the bucket when measured against the $10 trillion of notional principal outstanding in the over-the-counter derivative market in 1996, the credit derivative market is expected to grow quite quickly.

    Participants, Liquidity, Types of Products

    Currently, the 10–15 dealers in credit derivatives consist of investment and commercial banks, which act as both intermediaries and end users. Investment banks, in particular, can be constrained by credit exposures obtained within the huge bond and derivatives portfolios they maintain, requiring them to free up credit lines. Large money center banks, with both their traditional expertise in evaluating and managing credit risks and their extensive customer bases, have been in the forefront of the development and pricing of credit derivatives.62 Moreover, commercial banks are also large holders of credit risk (although it is typically more diversified) naturally leading them to consider altering their credit risk via derivative products.

    To date, the end users in this market are predominantly banks. They have an appetite for over-the-counter products and understand credit evaluation, making them easily educated customers. Other institutions, such as corporate treasuries and hedge funds, are making their way into the market. The corporate entities are interested in trading their trade credit concentrations and in lowering their credit exposures to various derivative counterparties. Hedge funds are also interested since credit risk derivatives provide them with an efficient method of obtaining credit exposures. Institutional investors, such as pension funds and insurance companies, are expected to be interested in finding new types of credit exposures other than the ones they have traditionally undertaken through their predominantly fixed-income portfolios.

    While the growth in the market has been led by U.S. entities, European institutions have recently become participants. Since European markets are heavily bank intermediated, and long-run client relationships are a large part of banks’ franchise value, credit derivatives are likely to provide a particularly apropos method of unbundling credit risk from the client relationship. The Asian market is thought to be the next area of significant growth for credit derivatives as this market is now adopting more sophisticated credit risk extension policies and Asian institutions are becoming more conscious of the credit risks that characterize their balance sheets.

    As yet, the market cannot claim to be liquid, as many of the transactions are “one-off deals that are tailored to the specific needs of a customer. Despite the absence of a generic product, four or so of the most common product designs have been widely adopted, and most derivatives can be marked-to-market, or at least marked-to-model, providing at least some comfort as to their ability to be offset. The liquidity of the credit derivatives is higher for derivatives written on more liquid underlying instruments, such as Brady bonds, whose outstanding stock had reached $156 billion by the end of 1996.

    There are four principal types of credit derivatives: credit default swaps, total rate of return (TROR) swaps, credit-linked notes, and credit spread options. All credit derivatives transfer credit risk between a credit-risk seller, interested in shifting the credit risk to another party in exchange for paying a premium, and a credit-risk purchaser, interested in obtaining credit risk along with receiving the premium for taking on such risk.

    In a credit default swap the buyer of protection (the hedger) pays a fee, which effectively represents an option premium, in return for the right to receive a conditional payment if a specified “reference credit” defaults. The reference credit is the party whose credit performance determines whether payments are made. The amount to be paid is negotiated between the counterparties and may be determined prior to the default event or may be determined based on the observed prices of similar obligations after a default.

    A total rate of return swap is structured so that the buyer swaps the “total return” on the reference asset for a regular floating-rate payment (in general based on LIBOR). For example, the buyer agrees to pay the total return on an emerging market Brady bond, consisting of all contractual payments as well as any appreciation in the market value of the bond; the seller agrees to pay the buyer LIBOR plus a spread and any depreciation in the value of the Brady bond. The TROR swap differs from the credit default swap in that a default event need not occur nor be verifiable: the TROR swap protects the buyer against a deterioration of credit quality, which can occur even without a default.

    A credit-linked note is an on-balance-sheet structured note in which a credit derivative is embedded in the structure. Often these notes are issued by a special purpose trust vehicle, which is collateralized with high-quality assets to assure payment of the contractual payments due. A purchaser of a credit-linked note assumes the credit risk of the reference credit and the underlying collateral. To illustrate: a special purpose vehicle, rated AAA, issues a credit-linked note based on the credit risk of Corporation XYZ. If Corporation XYZ defaults on its debt, the credit-linked note is no longer redeemable at par value, but note holders receive, say, 60 percent of the par value.

    A more recently developed credit derivative is the credit spread option. A credit spread option provides a payout to the buyer when the spread on two underlying assets exceeds a predetermined level. The buyer pays a premium for such protection and the seller pays out based on the spread. Since the credit risk of many fixed-income securities is often measured as a spread over a comparable-maturity “risk-free” security, this derivative product is highly sensitive to the market’s assessment of credit risk in these securities and is especially tailored to holders of emerging market debt and other high-yielding debt instruments.

    Outstanding Issues

    As yet, valuation and risk management methods for credit derivatives are not as analytically developed as those for other financial derivatives, in part, because (he information required is more difficult to obtain. Generally, defaults are rare and severe—historical data are spotty, not fulfilling the normal preconditions for modeling: a continuous data series with few large observations. These data constraints are most severe for the relatively illiquid loan market: data are slightly more plentiful regarding the default experience on publicly issued debt for rated companies. However, these data focus on U.S. corporate entities, whereas about half the outstanding credit derivatives are written on emerging market Brady bonds, which have yet to experience a single default. The current paucity of data on the default experiences of emerging market debt instruments may be hindering accurate assessments of credit risk—and there may be a rude awakening for holders of such debt and the related credit derivatives if the economic environment changes for Brady bond countries.

    Additionally, effective risk sharing is enhanced when adequate information about the reference credit is provided to the end user who is taking on the credit risk. Adequate information is more easily acquired for reference credits whose securities trade publicly and are subject to reporting requirements. But where no liquid security is traded and financial disclosures are absent, credit-risk purchasers may be at an informational disadvantage relative to the sellers. This may be especially true for credit derivatives related to bank loans, where the banks that originate such loans are thought to have superior information about the credit risk of the borrower. Further, there may be an incentive for a bank to rid itself of its more poorly performing loans when they cannot be easily distinguished from the better-performing loans. Moreover, once the credit risk of a loan has been transferred to a different party, the incentive for the bank to monitor the loan may be diminished, and the safety and soundness of banks may be negatively affected.

    Some market participants believe the market’s growth is hindered by the lack of regulatory clarity concerning the potential regulatory capital requirements and accounting treatment of credit derivatives. While many practitioners have strong views about the appropriate capital treatment, the debate so far can be characterized as one of intellectual exploration by both practitioners and the bank supervisory community. As of June 1997, regulators had offered little clear guidance regarding capital requirements, although some preliminary rules had been discussed in both the United States and the United Kingdom. The main issues are whether the hedging benefits of a credit derivative will lower the credit risk capital required on the combined position (derivative plus the underlying credit risk) and whether a credit derivative should be accounted for as product within a marked-to-market trading book or a “held to maturity” instrument within the so-called banking book, an issue that will take on added importance when market risk capital requirements are imposed on January 1, 1998.

    At an industry level, issues regarding how the development of credit derivatives will affect banks’ main business, the intermediation of savings by taking on credit risk, are yet unanswered. Will the credit derivative market overshadow the secondary loan market? If so, will this development mean more or less efficient credit risk sharing? Many dealers argue that the growth in credit derivatives allows those most willing and able to hold credit risk the opportunity to do so. Does this imply that the presumed special expertise within banks to measure and control credit risk will diminish relative to other institutions? Will bank loan products diminish in importance since increased information and pricing mechanisms for credit risk may permit bank customers direct access to public debt markets?

    From a systemic point of view, the growth of credit derivatives could be expected to increase the diversification of credit risk across more types of institutions. This appears, at first glance, to be a net benefit since the cascading of defaults during a crisis period would appear less likely when credit risks are held more broadly. However, credit derivatives also permit more concentrated holdings of credit risk as well—and there is little transparency about the institutions that may be holding the risk. One could imagine a situation in which U.S. pension funds, for example, hold Latin American sovereign credit risk through the use of credit derivatives. How would a debt restructuring be implemented in such a situation, especially if the underlying reference instrument is not owned by those holding the credit risk? Alternatively, would end users attempt to unload their credit derivative positions with the dealers in periods of stress and would this exacerbate an already turbulent period, putting dealers in a precarious position? Put another way, is it reasonable to assume the new participants in credit risk markets are as knowledgeable about the risks they are undertaking as the more seasoned banking institutions? While the answers to these questions depend on future developments in the market, a periodic appraisal of such issues is warranted as credit derivative markets expand.

    Appendix 2

    Accounting for Nonperforming Loans

    Table 35 provides information on loan classification systems and exposure limits in 20 emerging markets. Clearly there is considerable diversity. While some countries have relatively strict classification and provisioning requirements, others have systems that allow for much more discretion on the part of banks to assess asset quality. What cannot be represented in such a table, however, is the monitoring of asset-quality accounting and provisioning regulations. In banking systems where the banks generally have sophisticated internal risk management systems, where these are monitored by auditors and bank examiners, and where there is no presumption of official support in the event of bank failure, it may be less important to have detailed requirements specifying how to classify loans and what level of provisions to hold. Even if banks are subject to rigorous auditing that verifies application of these prudential requirements, few of the systems described in Table 35 address issues such as recapitalization or “evergreening” of loans or the treatment of restructured loans. However, such conditions do not generally apply to most countries. Hence, in the absence of rigorous monitoring by bank supervisors, disclosed asset-quality figures may not accurately reflect the true extent of problem loans, and therefore the true capitalization of banking institutions.

    Table 35.Loan Classification in Selected Emerging Markets
    Loan Classification SystemProvisioning Requirements
    Hong Kong, ChinaPerformingBorrowers are current in meeting commitments and full repayment of interest and principal is not in doubt.There are no requirements as to provisioning other than that individual banks have their own internal guidelines for maintaining adequate provisions. Interest must be accrued to a suspense account if loans arc substandard and not fully secured or overdue by more than 6 months; interest accrual ceases altogether for substandard loans past due more than 12 months and for loans classified as doubtful or loss. Loans must be written off after they are deemed irrecoverable.
    Special mentionBorrowers are experiencing difficulties; ultimate loss is not expected but could occur.
    SubstandardBorrowers displaying definable weakness: loan losses or rescheduling at concessional terms are possible.
    DoubtfulCollection in full is improbable: loss of principal and/or interest is expected, taking account of collateral.
    LossUncollectible after exhausting all collection efforts, including realization of collateral.
    IndiaNonperformingLoans on which interest is overdue for at least six months.None.
    SubstandardLoans that have been nonperforming for up to two years, term loans on which the principal has not been reduced for more than one year, and all rescheduled debts.10 percent.
    DoubtfulLoans that have been nonperforming for two to three years and term loans on which the principal has not been reduced for more than two years.100 percent of unsecured assets; for secured assets: 20 percent if doubtful for less than one year: 30 percent if doubtful for one to three years; 50 percent if doubtful for more than three years
    LossAll other assets deemed irrecoverable, where the loss has been identified by internal or external auditors or by the Reserve Bank of India inspectors, but where the amount has not been written off.100 percent.
    Interest accrual stops once loans arc nonperforming. For loans with balances below Rs 25,000, banks must set aside reserves equal to at least 10 percent of the balance.
    IndonesiaCurrentInstallment credit with no arrears, other credit in arrears less than 90 days, overdrafts less than 15 days.0.5 percent.
    SubstandardGenerally, loans with payments in arrears between three and six months.10 percent.
    DoubtfulNonperforming loans that can be rescued and the value of collateral exceeds50 percent.
    75 percent of the loan, or loans that cannot be rescued, but are fully collateralized.100 percent.
    LossDoubtful loans that have not been serviced for 21 months: credit in process of bankruptcy/liquidation.
    Loans must be written off 21 months after litigation indicates the loan will not have to be repaid.
    KoreaCurrentBorrower’s credit conditions (including collateral) are good and collectibility of interest and principal are certain.0.5 percent.
    Special mentionPayments are past due for between three months and six months, but collection is certain.1 percent.
    SubstandardLoans covered by collateral but borrower’s credit conditions arc deteriorating and payments are more than six months past due.20 percent.
    DoubtfulUnsecured portion of the loans that are more than six months past due and losses are expected.75 percent
    Estimated lossUnrecoverable amounts due net of collateral.100 percent.
    Loans must be written off within six days of being declared unrecoverable; write-offs in excess of W500 million require Bank of Korea approval
    MalaysiaFor loans less than RM 1 million:
    SubstandardMore than a normal risk of loss due to adverse factors: past due for between 6 and 12 months.0 percent.
    DoubtfulCollection in full is improbable and there is a high risk of default: past due for between 12 and 2–1 months.50 percent of net (of collateral) outstanding value.
    BadUncollectible; past due for more than 24 months.100 percent of net outstanding value.
    Loans must be written off when bankruptcy hearings have finished and/or partial or full repayment is unlikely.
    A general provision of at least 1 percent of total loans net of interest in suspense and specific provisions is also required.
    PhilippinesUnclassifiedBorrower has the apparent ability to satisfy obligations in full; no loss in collection is anticipated0 percent of net (of collateral) exposure.
    Special mentionPotentially weak due, for example, to inadequate collateral, credit information, or documentation.0 percent.
    Substandardloans that involve a substantial degree of risk of future loss25 percent.
    DoubtfulLoans on which collection or liquidation in full is highly improbable, substantial losses are probable.50 percent.
    LossUncollectible or worthless.100 percent
    Interest is not accrued on past-due loans, which are loans or other credit not paid at the prescribed maturity date or, in the case of installment credit, in arrears by more than a prescribed amount depending upon the frequency of installments
    SingaporeSpecial mentionAccounts with evidence of potential weakness in creditworthiness, such as untimely repayment.A provision of 50 percent of the loan value for defaults of over a year, for defaults of 3 to 6 months provision is the difference between the loan amount and 80 percent of collateral: for 6 to 12 months, the difference between the loan amount and 70 percent of the collateral. In aggregate. 100 percent of substandard, doubtful, and bad loans must be provided for, with those graded doubtful to have al least 50 percent provision.
    SubstandardNormal repayment may be jeopardized by continuing adverse trend of severe financial weakness
    Doubtful BadRepayment of outstanding debt appears questionable; expectation of loss. Outstanding debt is uncollectible.
    BadOutstanding debt is uncollectible.
    Loans must be written off in the year that they are recognized as a loss. The Monetary Authority of Singapore has established a minimum (tax exempt) general provision of 2 percent of outstanding loans (including accrued interest) net of specific provisions.
    Taiwan Province of ChinaNonperforming loans (on which interest is not accrued) are:Only provisions (general and specific combined) up to 1 percent of loan balance are tax deductible Specific provisions in excess of that amount arc made on a quarterly basis. Interest is no longer accrued after 180 days. Loans must be written off after all legal proceedings have finished.
    1. Short-term loans with principal payments three months past due.
    2. Loans with interest payments (or installments) six months past due.
    3. Loans to companies for which legal proceedings by the hank have commenced.
    ThailandLoans are nonperforming (substandard or doubtful) if they are 12 months past due but fully collateralized or secured, or if they are 6 months past due but not fully secured.
    Substandardloans is in arrears, but there is sufficient security to ensure that full recovery of the debt will be possible.15 percent (by end-June 1999; at least 75 percent by end-June 1998)
    DoubtfulLoan is in arrears, but there is insufficient collateral.100 percent
    IrrecoverableLegal enforcement has been initiated and has been unsuccessful.100 percent.
    Czech RepublicWatchAccounts overdue by 30–90 days.5 percent of net (after collateral) exposure
    SubstandardAccounts overdue by 91–180 days.20 percent of net exposure.
    DoubtfulAccounts overdue by 181–360 days.50 percent of net exposure.
    LossAccounts have been overdue for more than a year, there is little likelihood of repayment, and assets are not adequately secured.100 percent of net exposure.
    Restructured loans must be classified as substandard for six months after restructuring and then as watch for three scats
    HungaryPerformingAssumption that interest or principal will not be more than 15 days overdue.0 percent.
    To he monitoredNo loss is assumed, but management is of the opinion that the exposure requires separate monitoring.0–10 percent.
    SubstandardRisks are higher than average or some loss may he assumed at the time of classification.11–30 percent.
    DoubtfulA loss will be incurred but the size of the loss is uncertain or where payment is at least 90 days past due or payment delay becomes regular.31–70 percent
    BadThe loss will exceed 70 percent or the company is in bankruptcy71–100 percent.
    General reserves (out of net income) must amount to 1.25 percent of the balance sheet total plus 1 percent of guarantees
    PolandStandardNo arrears or doubts about the borrower’s financial strength: receivables guaranteed by the state.None.
    SubstandardLoans in arrears by more than a month, or loans to a borrower with weakened financial standing.20 percent.
    DoubtfulLoans in arrears by more than three months, or loans to a borrower with deteriorating financial standing50 percent.
    LossLoans in arrears by more than six months or the subject of legal dispute, or loans to a borrower who is either in liquidation or whose location is unknown, or whose financial standing makes repayment impossible.100 percent.
    Approval for lower provisions may be given if the loans are adequately collateralized. General reserves may be set up without limit, although reserves equal to only the first 1 percent of impaired loans are tax deductible.
    TurkeySpecial follow-upLoans to uncreditworthy borrowers (defined as borrower whose capital is insufficient to pay the debt when due, or borrower lacks the ability to pay the debt, or the borrower’s working capital is insufficient to meet its operating needs).Initial 15 percent provision. Increased to 50 percent by the end of the first year. 100 percent after two years.
    Increased to 50 percent by the end of the first year. 100 percent after two years.
    Administrative follow-upLoans classified as overdue of one mouth in arrears15 percent provision is required after two months (i.e‥ when 90 days past due)
    Legal follow-upLoans in arrears for three months.25 percent provision is required after 6 months, rising to 50 percent after one year. 75 percent after 18 months, and 100 percent after two years. Banks must cease accruing interest on loans in legal follow-up.
    Loans to state entities (including state-owned enterprises) are not included in the classification system, and provisions arc not required for these loans.
    Latin America
    ArgentinaConsumer loansCommercial loansLiquid guaranteePreferred guaranteeWithout guarantee
    NormalLess than 31 days overdueNo doubt exists.1 percent.1 percent.1 percent.
    Potential risk31–89 days overdue.Performing, but sensitive to changes, or more than 30 days overdue.1 percent.3 percent.5 percent.
    Problem90–179 days overdue.Problems meeting obligations: or 90–179 days overdue.1 percent.12 percent.25 percent.
    High risk180–365 days overdue or subject to judicial proceedings for default.Highly unlikely to meet obligations; or more than 180 days overdue.1 percent.25 percent.50 percent.
    IrrecoverableMore than 365 days overdue.Obligations cannot be met; more than 365 days overdue.1 percent.50 percent.100 percent.
    Irrecoverable for technical decisionBankruptcy/liquidation/insolvency.Bankruptcy/liquidation/insolvency.100 percent.100 percent.100 percent.
    BrazilConsumer loansCommercial loansUnsecuredPartially/fully securedExport/import
    0–29 days overdue.0–29 days overdue.0 percent.0/0 percent.0/100 percent.
    30–59 days overdue.30–59 days overdue.0 percent.0/0 percent.100/100 percent.
    60–180 days overdue60–180 days overdue.100 percent.50/20 percent.100/100 percent.
    181–360 days overdue.181–360 days overdue.100 percent.100/20 percent.100/100 percent.
    More than 360 days overdue.More than 360 days overdue.100 percent.100/100 percent.100/100 percent.
    ChileConsumer loansCommercial loansAllowance
    ACurrentProbability of default: 0 percent.0 percent.
    B1–29 days overdue.Probability of default: less than 5 percent.1 percent.
    B30–59 days overdue.Probability of default: 5–40 percent.20 percent.
    C60–119 days overdue.Probability of default: 40–80 percent.60 percent.
    DMore than 120 days overdue.Probability of default: 80–100 percent.90 percent.
    ColombiaConsumer loansCommercial loansUnsecured principalInterestSecured principal
    A (Normal)Current.Current.0 percent.0 percent.0 percent.
    B (Subnormal)30–59 days overdue.30–119 days overdue.1 percent.1 percent.0 percent.
    C (Deficient)60–89 days overdue120–179 days overdue20 percent100 percent0 percent.
    D (Doubtful)90–79days overdue.180–359 days overdue.50 percent.100 percent.Q percent.
    E (Unrecoverable)180–360 days overdue.360–719 days overdue.100 percent.100 percent.0 percent.
    E (Unrecoverable)More than 360 days overdueMore than 720 days overdue.100 percent.100 percent100 percent.
    MexicoConsumer loansCommercial loansAllowance
    AMinimal risk.Minimal risk.0 percent
    BLow risk.Low risk.1 percent.
    CModerate riskModerate risk.20 percent.
    DHigh risk.High risk60 percent
    ENoncollectibleNoncollectible.100 percent.
    Loan loss reserves should be at least equal to the greater of (1) reserves calculated according to the above classification: (2) 4 percent of total loans; or (3) 45 percent of past-due loans. The entire amount of an amortizing loan (including past due interest) is considered past due if any payment is 90 days overdue (180 days for mortgages). Nonamortizing bullet loans are past due if more than 30 days overdue. Credit cards are past due when two minimum payments have been missed. Loans restructured into UDIs are transferred to trusts (consolidated into the bank’s financial statements) and attract a 15 percent loan loss reserve.
    PeruConsumer loansCommercial loansUnsecured consumerUnsecured commercial
    A (Normal)Current.Current with no doubts.0 percent.0 percent.
    B (Potential problem)10–29 days overdue.Demonstrated difficulties.3 (on total balance) percent.1 (on total balance) percent.
    C (Substandard)30–59 days overdue.Serious weaknesses.30 percent.25 percent.
    D (Doubtful)60–120 days overdue.Making payments, but less than contracted.60 percent.50 percent.
    E (Loss)More than 120 days overdue.Unrecoverable.100 percent.100 percent.
    VenezuelaConsumer loansCommercial loansAllowance
    A (Normal)Fully performingFully performing.0 percent.
    B (Potential risk)1–3 monthly payments overdue.Performing, but showing signs of potential future problems (e.g‥ deterioration of financial condition, inadequate documental ion).5 percent.
    C (Real risk)4–6 monthly payments overdue.Experiencing delays in interest and/or principal payments with estimated losses.10 percent.
    D (High risk)7–12 monthly payments overdue.Interest and/or principal payments three months or more past due and where legal recovery proceedings have been initiated.50 percent.
    E (Irrecoverable)More than 12 monthly payments overdue.Interest and/or principal payments are 12 months past due or where legal proceedings indicate very scarce possibility of recovery.100 percent.
    Once a loan is 30 days past due it is placed on nonaccrual status, and a reserve equal to 100 percent of accrued interest must be created immediately. Loans must be written off after 36 months. Provisioning requirements do not apply to credits guaranteed by the Venezuelan public sector.
    Consumer Loans

    Provisioning based on the amount of principal overdue for A–D loans. For E loans, the reserve must take into account the entire outstanding balance.
    Sources: IBCA Ltd.; ING Barings Securities; and J.P. Morgan.
    Background Material—Part II Selected Issues

    Annex IV European Monetary Union: Institutional Framework for Financial Policies and Structural Implications

    This annex begins with a brief description of the potential size of the domestic euro capital markets in a European Economic and Monetary Union (EMU) and the role of existing European currencies in international capital markets. An analysis of the institutional framework for financial markets in EMU, including the payments system, the European System of Central Banks (ESCB), and the framework for other financial policies (financial supervision and regulation, lender-of-last-resort functions, and deposit insurance) follows. The next section discusses the catalytic role of the euro. The structural implications of EMU for European and international securities markets, including the possible evolution of EMU markets for repurchase agreements (repos), interbank funds, bonds, equities, and derivatives are then evaluated. The annex concludes by examining implications for wholesale and retail banking markets and the remaining impediments to cross-border competition in banking and financial services.1

    Potential Size of EMU Financial Markets

    In absolute terms, and compared with any reasonable benchmark, the introduction of the euro has the potential for creating the largest domestic financial market in the world. At end-1995, the market value of bonds, equities, and bank assets issued in EU countries amounted to more than $27 trillion (Table 12), roughly the same order of magnitude as world GDP (94 percent of world GDP).2 By comparison, the market value of assets in North America—with roughly the same population and GDP as the European Union—amounted to about $25 trillion ($23 trillion in the United States). If the initial union includes only Austria, Belgium, France, Germany, Luxembourg, the Netherlands, Ireland, and Finland (the EU-8), the domestic euro market would equal the size of Japan’s domestic market ($16 trillion). If the union includes in addition Italy, Portugal, and Spain (EU-11), it would roughly equal the size of the U.S. domestic market. An interesting aside is that the value of bonds, equities, and bank assets is roughly three times the respective GDPs in the European Union, the United States, and Japan (about 320 percent in the European Union and Japan and about 315 percent in the United States).

    EU private entities overwhelmingly have tended to finance their activities through bank loans rather than through bond and equity financing, and U.S. entities have relied more heavily on bond and equity financing. In the EU-11, bank assets represented 54 percent of all outstanding financial assets at end-1995. By contrast, U.S. bank assets accounted for only 22 percent of total assets outstanding.

    In contrast to government securities markets, European private debt securities markets are segmented, with all but the largest firms borrowing solely from a domestic investor base. In the EU-11, for each dollar of bank borrowing, private firms borrowed, on average, only 50 cents through private securities issues. By contrast, in the United States, for each dollar of borrowing from banks, U.S. firms borrowed slightly more than two dollars through debt securities issues. Japanese private entities were much closer to their EU, than to their U.S., counterparts.

    Although the amount of EU private bonds outstanding appears to be sizable enough to suggest a reasonably large market for corporate bonds (roughly three-fourths the size of the U.S. market), the bulk of these bonds were issued by European financial institutions. From the point of view of corporate balance sheets, as of end-1994, bonds accounted for a relatively small share of the total liabilities of nonfinancial firms in France (5.7 percent) and in Germany (less than 1 percent); by contrast, they accounted for 18.8 percent of the total liabilities of U.S. nonfinancial firms.3 The low share of debt financing by European companies extends to the short end of the maturity spectrum as well, because European companies tend to rely on bank financing for short-term funds. U.S. corporate entities tend to rely more heavily on short-term financing because of their access to the very liquid and highly developed commercial paper market, which accounts for more than half of the world’s outstanding commercial paper. These observations about the use of debt securities reflect the greater historical reliance by firms in the United States on direct intermediation through the corporate debt securities markets, the heavy reliance in Europe on bank financing, and the relatively undeveloped European corporate securities markets.

    Another way of assessing the potential importance of the euro from a purely quantitative perspective is to examine the use of existing European currencies as currencies of denomination in international financial transactions. In international bond markets, 35 percent of the outstanding stock of international debt securities was denominated in EU currencies at end-September 1996 (Table 36). Although this is a substantial share of international issues outstanding, and is a close second to the amount of dollar international issues outstanding, EU countries themselves issued more than 45 percent of all international bonds outstanding. In addition, in the five-year period ending in December 1995, only a minor share of developing country debt was issued internationally in EU currencies.

    Table 36.Amounts Outstanding of International Debt Securities by Currency and Country of Nationality, March 19971(In billions of U.S. dollars)
    Amounts Outstanding
    By currency
    U.S. dollar1,301.4
    Japanese yen499.6
    Currencies of European Union (EU) countries21,107.9
    By country of nationality
    EU countries1,478.2
    United Kingdom275.5
    North America647.2
    United States418.1
    All countries3,240.7
    Source: Bank for International Settlements, International Banking and Financial Market Developments (May 1997).

    Still another way to gauge the potential role of the euro is to examine daily turnover in the global foreign exchange markets. According to the most recent Bank for International Settlements (BIS) survey, as of April 1995 the dollar was involved in at least one side of a transaction about 42 percent of the time, the deutsche mark 18.5 percent, the yen 12 percent, and the pound sterling 5 percent. EMS currencies combined were involved in at least one side of a transaction about 35 percent of the time, including European cross-currency trading (Table 37). In related derivative markets, the dollar, EU currencies, and the yen accounted for shares of trading that are roughly equivalent to the relative sizes of their economies (in terms of GDP), but most of this activity actually involved U.S. and U.K. financial institutions. Transactions involving currency swaps were clearly tilted toward the dollar, reflecting its now dominant position in international finance and as a reserve currency (Table 38).

    Table 37.Use of Selected Currencies on One Side of Foreign Exchange Transaction, April 1989, April 1992, and April 19951(As a percentage of global gross foreign exchange market turnover)
    CurrencyApril 1989April 1992April 1995
    U.S. dollar908283
    Deutsche mark2274037
    Japanese yen272324
    Pound sterling151410
    French franc248
    Swiss franc1097
    Canadian dollar133
    Australian dollar223
    European currency unit (ECU)132
    Other European monetary system (EMS) currencies3913
    Currencies of other reporting countries332
    Other currencies1988
    All currencies200200200
    Memorandum item:
    EMS currencies including ECU487070
    Source: Bank for International Settlements, Central Bank Survey of Foreign Exchange and Derivatives Market Activity 1995 (May 1996).
    Table 38.Notional Principal Value of Outstanding and New Interest Rate and Currency Swaps, 1995(In billions of U.S. dollars)

    New Swaps
    Interest rate swaps12,810.78,698.8
    U.S. dollar4,371.72,856.5
    Japanese yen2,895.92,259.3
    Currencies of European
    Union (EU) countries14,620.93,160.9
    Of which:
    Deutsche mark1,438.9984.5
    French franc1,219.91,113.5
    Italian lira405.4217.3
    Netherlands guilder101.862.3
    Pound sterling854.0433.4
    Spanish peseta163.791.9
    Of which:
    Swiss franc331.7159.2
    Currency swaps22,394.8910.2
    U.S. dollar837.8307.9
    Japanese yen400.0164.5
    Currencies of EU countries 1684.7248.1
    Of which:
    Deutsche mark238.078.1
    French franc81.441.6
    Italian lira72.618.5
    Netherlands guilder28.113.0
    Pound sterling91.523.4
    Spanish peseta27.522.4
    Of which:
    Swiss franc150.629.7
    Source: Bank for International Settlements, International Banking and Financial Market Developments (November 1996).

    In summary, although the EU currencies command a significant share of activity in international financial markets, they do not now command shares in line with either the size of the EU economy or the relative size of their domestic financial markets.

    Institutional Framework for Financial Markets

    Between now and the start of EMU, countries of the European Union will implement a new institutional framework for EMU financial policymaking. The main parts of this framework are the new EU-wide payments system, the institutional framework for conducting the single EMU monetary and exchange rate policy, and a still-evolving institutional framework for implementing and coordinating financial supervision and regulation across European financial markets, including the management of systemic risk. Each of these important elements of the new framework is discussed in this subsection.

    TARGET Payments System

    TARGET (Trans-European Automated Real-Time Gross Settlement Express Transfer) is a payments system designed to process cross-border transactions denominated in euros after the start of Stage III of EMU on January 1, 1999, TARGET has two main objectives. The first objective is to provide a safe payments mechanism within the euro area based on real-time gross settlement (RTGS) procedures that will insulate the payments system across Europe from the effects of liquidity and payment difficulties experienced by a single institution.4 The second goal is to create an efficient system of cross-border payments that will integrate the money markets of the participating countries and support the implementation of the single monetary policy in Stage III.5


    The TARGET system is composed of one RTGS system in each of the EMU countries and the payments mechanism of the European Central Bank (ECB) connected by common infrastructures and procedures forming the Interlinking system (a communications network) (Figure 54).6 Only the ECB and national central banks (NCBs) will be allowed to use the Interlinking system, but any participant in any RTGS system connected to TARGET will be allowed to send payments via TARGET. Because TARGET is designed to process only euro transactions, RTGS systems of EU countries not in EMU will be allowed to connect to TARGET only if they are able to process euros. Remote access to domestic RTGS systems will be granted on a nondiscriminatory basis to credit institutions licensed in other EU states either through their local branches or directly from another EU country. (At the start of EMU, however, remote access to monetary operations will not be available.) To facilitate the operations of large-value net settlement systems working in euros through TARGET, net settlement systems will be allowed to open a special account with the ECB or a national central bank that must be used exclusively for settlement purposes and must have a zero balance at the beginning and at the end of the day.

    Figure 54.Cross-Border TARGET Payment


    TARGET is designed as a decentralized system in which payments messages are exchanged on a bilateral basis among national central banks, according to the “central banking correspondent model,” without any central counterparty. It remains to be decided whether the ECB will have its own payments mechanism connected to TARGET. This may not be necessary because the national central banks will implement most monetary policy operations, in agreement with the principle of decentralization underlying monetary policy in EMU, Even if the Governing Council of the ECB decides to retain the execution of fine-tuning operations and foreign exchange intervention, the settlement of transactions for both operations may remain decentralized and the ECB may still not need to access the payments system.7

    The ECB will neither monitor nor receive information on inter-NCB payments orders during the day. At the end of the day, the ECB will perform specific control operations with the aim of checking the correctness of cross-border payments exchanged during the day and the resulting inter-NCB balance positions. The European Monetary Institute (EMI) has not yet decided on the clearing and settlement modalities (frequency of settlement, degree of centralization, means of payment) of outstanding balances among national central banks.

    In the U.S. Federal Reserve System, the Board of Governors, like the ECB, does not monitor the settlement positions of each federal reserve bank during the day. At the end of each business day, the reserve bank’s Integrated Accounting System settles the cross-district financial transactions by debiting or crediting as appropriate each reserve bank’s Interdistrict Settlement Account. This daily clearing process is known as the “gold wire process.” The board coordinates once a year (in April) the settlement of the balances on the Interdistrict Settlement Accounts by means of transfer of Gold Certificate assets among reserve banks. The amount settled is equal to the daily average balance in the Interdistrict Settlement Account over the previous year. No such clearing process has been decided upon in the European System of Central Banks, and this opens up the possibility of one national central bank accumulating large claims against another national central bank with no mechanism for settling them.8

    Transactions Processed

    In accordance with the objective of facilitating the implementation of a single monetary policy, credit institutions will be required to use TARGET for payments directly connected with monetary policy operations. Furthermore, large-value net settlement systems are likely to use TARGET to perform their settlement operations because they are bound to settle in central bank money9 and therefore in euros. Credit institutions will decide whether to use TARGET for other categories of payments, and there will be no upper or lower limits to the amounts transferred besides those in the domestic RTGS systems. Nevertheless, the European Monetary Institute has indicated that TARGET is expected to process mainly large-value payments between credit institutions, whereas private systems are expected to process small-value payments.10

    Intraday Liquidity

    Participants in RTGS systems may experience a liquidity shortfall whenever they need to send a payments order before receiving one. In this instance, payments may be blocked or queued until sufficient funds become available either through incoming payments or by borrowing in the market; in the limit, settlement may be delayed and gridlock may take place with systemic implications (i.e., payments cannot be processed because of a lack of sufficient funds). To avoid such events, EMU national central banks will allow intraday mobilization of reserve requirements and will provide participants in their RTGS systems with fully collateralized intraday credit in the form of daily overdrafts or repurchase agreements.

    No decision has been made on whether non-EMU national central banks will be allowed to grant intraday credit in euros to participants in their RTGS systems linked to TARGET. The Governing Council of the ECB will have to choose one of the three mechanisms currently being prepared by the European Monetary Institute with the aim of preventing intraday credit granted by non-EMU national central banks from spilling over into overnight credit and thus from having a monetary impact. The first mechanism would set a limit—possibly zero—to the intraday credit in euros that the ECB would provide to non-EMU national central banks (for participants in their RTGS systems) and would impose penalty rates on spillovers. The second would just impose penalty rates. The third would require non-EMU participants to complete their operations before the closing time of TARGET, so that they would have time to avoid spillovers by borrowing euros in the money market.

    If non-EMU national central banks are not granted access to intraday credit or are penalized, institutions making cross-border payments to the euro area could adapt their behavior in a number of ways. In some instances they would still channel payments through the TARGET system; in others they would not. First, non-EMU national central banks could borrow euros in the market to provide intraday credit to participants in domestic RTGS systems for cross-border payments to the euro area; in this instance, systemic risks could be reduced as much as they would be reduced with direct access to ECB’s intraday credit. Second, non-EMU banks could channel cross-border payments in euros through branches and subsidiaries in the euro area that have access to both intraday and overnight credit; the potential risk reductions associated with TARGET would be fully captured in this second instance. Third, non-EMU institutions could decide to make cross-border payments to the euro area through private net settlement systems, thus reducing the number of transactions across TARGET; in this instance, some of the systemic risk reductions that could be achieved through TARGET would not be realized.

    Operating Hours and Pricing Policies

    The operating hours of TARGET will he from 7:00 a.m. to 6:00 p.m. and domestic RTGS systems will be allowed to open earlier to process domestic payments. One hour before closing time, participants in RTGS systems will stop processing customers’ payments in euros and only interbank payments will be allowed. These hours will allow for a longer overlap between TARGET and the payments systems in North America and the Far East in an effort to reduce cross-currency settlement risk.

    TARGET pricing policy will be directed at cost recovery but also at (1) maintaining a level playing field between participants; (2) contributing to risk-reduction policies by preventing institutions from using a less secure payments mechanism; and (3) avoiding transaction charges that would discourage interest rate arbitrage and hinder the integration of the money market.

    Framework for EMU Monetary Policy

    Decentralization is the key principle underlying the operational framework for monetary policy in Stage III. According to the European Monetary Institute, “the ECB should have recourse to the NCBs to carry out operations ‘to the extent deemed possible and appropriate’” in accordance with Article 12 of the statute of the ESCB. The agreed goal is to “rely as much as possible on the existing infrastructure and on the NCBs’ experience, provided that the application of this principle does not conflict with the other guiding principles.” The latter include operational efficiency; conformity to market principles; equal treatment to all financial institutions accessing the ESCB’s facilities; simplicity, transparency, and cost efficiency; conformity with the decision-making process of the ESCB, which requires the Governing Council of the ECB to be able to control the overall stance of monetary policy at all times; and harmonization of the instruments across countries to the extent necessary “to ensure a single monetary policy stance across the euro area, as well as the equal treatment of counterparties and the avoidance of regulatory arbitrage.”11

    Monetary Policy Instruments and Procedures

    Open market operations will be the main monetary policy instrument of the ESCB. In addition, there will be standing facilities, and in particular a marginal lending and a marginal deposit facility. The option has been left open to rely on minimum reserve requirements, and a final decision on this will be taken by the ECB.

    Open market operations are expected to lake mainly the form of reverse transactions (repos), but four other instruments are envisaged: outright transactions, issuance of debt certificates, foreign exchange swaps, and collection of fixed-term deposits. To conduct open market operations, the ECB will be able to choose between three procedures; standard tenders, quick tenders, and bilateral procedures. These operations will be executed by the national central banks, which—in the case of tenders—will collect all the bids and transmit them to the ECB; the latter will then sum them up and select the winning bids. Most refinancing to the financial sector will be provided through regular weekly reverse transactions (repos) with a maturity of two weeks (Table 39).

    Table 39.European System of Central Banks: Open Market Operations and Standing Facilities
    Monetary Policy OperationsTypes of Transactions
    Provision of liquidityAbsorption of liquidityMaturityFrequencyProcedure
    Open market operations
    Main refinancing operationsReverse transactions (repos)n.a.Two weeksWeeklyStandard tenders
    Longer-term refinancing operationsReverse transactions (repos)n.a.Three monthsMonthlyStandard tenders
    Fine-tuning operationsReverse transactions (repos)Reverse transactions (repos)NonstandardizedNonregularQuick tenders
    Foreign exchange swapsForeign exchange swapsReverse transactions (repos)
    Collection of fixed-term depositsBilateral procedures
    Outright salesn.a.NonregularQuick lenders
    Structural operationsReverse transactions (repos)Issuance of debt certificatesStandardized/nonstandardi/edRegular and nonregularStandard tenders
    Outright purchasesOutright salesn.a.NonregularBilateral procedures
    Standing, facilities
    Marginal lending facilityReverse transactions (repos)n.a.OvernightAccess at the discretion of counterpartiesAccess al the discretion of counterparties
    Deposit facilityn.a.DepositsOvernightAccess at the discretion of counterpartiesAccess at the discretion of counterparties

    To steer interest rates in the event of unexpected liquidity fluctuations, the ESCB will use fine-tuning operations. These will be executed primarily as reverse transactions but they may also take the form of outright transactions, foreign exchange swaps, or collection of fixed-term deposits. The European Monetary Institute established that “fine-tuning operations will normally be executed by the NCBs through quick tenders or bilateral procedures. The ECB Governing Council will decide if, under exceptional circumstances, fine-tuning operations may be executed in a centralized or decentralized manner by the ECB.”12

    Longer-term refinancing operations with a monthly frequency and a maturity of three months are also foreseen, but they would not be used to send signals to the market. Finally, reverse or outright transactions and debt certificates will allow the ECB to affect the structural liquidity position of the system.

    Standing facilities (a marginal lending and a marginal deposit facility) will allow counterparties to obtain overnight liquidity or make overnight deposits with EMU national central banks. The interest rates on these two facilities should determine the ceiling and the floor of a corridor within which overnight rates are expected to fluctuate. Under normal circumstances, the access to these two facilities will not be restricted so that any eligible counterparty will be able to obtain an unlimited credit from the lending facility as long as it has enough eligible collateral.

    The European Monetary Institute has indicated three possible rationales for the introduction of minimum average reserve requirements. First, average requirements would help to stabilize short-term interest rates. Second, reserve requirements could be used to create or enlarge a structural liquidity shortage in the money market. Third, they could help to stabilize monetary aggregates. By stabilizing short-term rates, average reserve requirements would reduce the amount and frequency of fine-tuning operations, which in a decentralized operational framework could become cumbersome. The institute indicated that terms and conditions for reserve requirements would be harmonized in the euro zone, but it did not specify whether reserve requirements would be remunerated.

    Eligible counterparties of the ESCB for monetary policy operations will be either institutions established in the euro area subject to at least one form of EU supervision or branches of non-EMU institutions that have their head office in an EU or European Economic Area (EEA) country. These institutions must be financially sound and the ESCB will have the authority to suspend temporarily or permanently their access to monetary policy instruments on prudential grounds. Branches of institutions from third countries could be counterparties only in bilateral outright operations involving purchases or sales of securities.

    All ESCB liquidity-providing operations will be based on adequate collateral as required by Article 18.1 of the statute of the ESCB. Both public and private assets denominated in euros will be eligible as collateral. Tier I collateral will include assets that fulfill eligibility criteria specified by the ECB for the whole euro area; Tier II collateral will include other assets that EMU national central banks may consider eligible in accordance with ECB guidelines (Table 40). Both Tier I and Tier II assets will be eligible in the whole euro area, but, whereas the default risk related to Tier I paper will be borne by the ESCB as a whole, default risk related to Tier II paper will be borne by the EMU national central bank that proposed it.13 To avoid the “cheapest to deliver” problem (counterparts delivering the lowest-quality collateral), the ECB could impose margins (“haircuts”) or additional guarantees on Tier II assets with a lower credit standing. A list of Tier II assets was deemed necessary because several national central banks have traditionally accepted sizable amounts of nonmarketable private bills and loans as collateral; to assess the related counterparty risk, some national central banks employ a considerable number of people (about 500 in France, 300 in Germany, and 100 in Austria).

    Table 40.European System of Central Banks: Eligible Assets
    CriteriaTier ITier II
    Type of assetEuropean System of Central Banks debt certificates. Other marketable financial obligations.Marketable financial obligations. Nonmarketable financial obligations. Equities traded on a regulated market.
    Settlement proceduresAssets must be centrally deposited in book-eniry form with a national central bank or a Central Securities Deposit fulfilling European Central Bank minimum standards.Assets must be easily accessible to the national central bank that has included them in its Tier 11 list.
    Type of issuerEuropean System of Central Banks.

    International and supranational institutions.

    Public sector.

    Private sector.
    Public sector

    Private sector.
    Financial soundnessThe issuer (guarantor) must be financially sound.The issuer/debtor (guarantor) must be financially sound.
    Location of issuerEuropean Economic AreaEuro area.

    Location in other European Economic Area countries can be accepted subject to European Central Bank approval.
    Location of assetEuro areaEuro area.

    Location in other European Economic Area countries can be accepted subject to European Central Bank approval.
    Currency of denominationEuroEuro.

    Other European Economic Area or widely traded currencies can be accepted subject to European Central Bank approval.
    Memorandum item:
    Cross-border useYesFor “domestic” assets: yes.
    For “foreign” assets: possibly restricted.

    The ESCB will have the capacity to conduct foreign exchange intervention from the start of Stage III by means of reserves transferred from the EMU national central banks to the ECB, totaling a maximum amount of 50 billion euros (Article 30 of the statute of the ESCB). The management of foreign reserves that remain with the EMU national central banks will be subject to guidelines issued by the ECB (Article 31.3) to ensure that such operations will not interfere with the monetary and exchange rate policies of the ECB. Exchange rate policy cooperation between the euro area and other EU countries is envisaged within the framework of a new exchange rate mechanism called ERM2 (see Box 7). The ECB will make decisions related to foreign exchange intervention, but it has not yet been decided whether the ECB or the EMU national central banks will implement them; this decision is left to the Governing Council of the ECB. Counterparties for foreign exchange intervention will need to satisfy a number of prudential and efficiency criteria.

    Monetary Policy Operating Procedures in Other Industrial Countries

    Monetary policy operating procedures in industrial countries seem to be guided by two alternative paradigms (Table 41).14 On the one hand, the central banks of the United States, Japan, the United Kingdom, Canada, and Australia play an active role in their domestic money markets by intervening daily. This reflects a relatively volatile demand for liquidity, owing in part to their more developed securities markets. On the other hand, most continental European central banks intervene infrequently, relying mainly on average reserve requirements to smooth liquidity shocks.15

    Table 41.Permissible Banking Activities and Bank Ownership in the European Union and G–10 Countries, 1995
    European Central BankAustriaBelgiumFranceGermanItalyNetherlandsSpainSwedenUnited KingdomAustraliaCanadaJapanSwitzerlandUnited States
    Main operation Maturity (days) FrequencyRTRPRP1RPRPRTCLRPRTOTRTRTRTFXSRT
    Maturity (days)1477–15714≤302–81071–33av.711–908–201–15
    Fine-tuning operationsYesNoYesYesNoYesYesYesYesYesYesYesYesYesYes
    Long-term refinancing
    Standing facility
    Below marketNoYesYesNoYesYesYesYesNoNoNoNoYes5NoYes
    Reserve requirements6YesNoYesYesYesYesYesNo7YesYes8YesYesYes
    Source: Borio (1997).Notes: RT = reverse transaction (repo or reversed repo): RP = repo 1 reversed purchase); OT = outright transaction, secondary market; CL = collateralized loan; FXS swap (purchase or sale).

    Like the ECB, most central banks use reverse transactions, in the form of repos or reverse repos, as their main monetary policy instrument. Only in Canada are reverse transactions not the main monetary policy instrument; there the central bank transfers government deposits between its balance sheet and that of clearing banks. In the United Kingdom, since 1994 the Bank of England has increasingly used repos alongside the traditional outright purchases of commercial bills; this trend has continued with the opening of the private repo market in January 1996.

    Box 7.ERM2

    The Treaty of Maastricht does not specify the exchange rate arrangement between EMU and the EU countries that are not initial members. To eliminate this uncertainty, in December 1995, the European Council in Madrid announced that the current ERM will be replaced by a new exchange rate mechanism, called ERM2, whose main features were agreed on in the Resolution of the Amsterdam European Council in June 1997.

    The main objective of ERM2 will be to support the single market by avoiding the disruption of trade flows resulting from real exchange rate misalignments or excessive nominal exchange rate volatility. Participation will be voluntary but expected, especially by countries planning to join EMU with a delay. To allow for different degrees and strategies of convergence, the structure of ERM2 will be flexible. Target fluctuation bands vis-à-vis the euro will be wide: plus or minus 15 percent. Narrower bands between the ECB and non-EMU national central banks are foreseen, but they will be “without prejudice to the interpretation of the exchange-rate criterion” of the Maastricht Treaty Also, bilateral fluctuation bands and intervention arrangements between two non-EMU national central banks will be possible. Intervention at the margin should be automatic and unlimited, but the ECB and the EMU national central banks will be entitled to suspend intervention if the primary objective of price stability is threatened. Intramarginal intervention will remain discretionary. The Very Short Term Financing Facility (VSTF) of the current ERM will be available also in ERM2 “broadly on the basis of the present arrangements.”

    The main uncertainty about the functioning of ERM2 regards the commitment of the ECB to support a currency of the system under attack. This commitment seems to be limited by the provision that intervention could be suspended “if this were to conflict with the primary objective of price stability.” Threats to price stability, however, are likely to be much rarer than in the present ERM because the large scale of EMU will allow easier sterilization of any ERM2-related intervention; in addition, the latter will have a much more limited impact on the liquidity of the euro area. At the same time, intervention by a non-EMU national central bank will not be very effective in stabilizing its parity with the much larger euro zone. Thus, non-EMU countries can reasonably be expected to exercise their obligation for stabilizing ERM2 parities primarily through the maintenance of appropriate monetary, fiscal, and structural policies, rather than through foreign exchange market intervention.

    The two-week maturity and the weekly frequency selected for the ECB’s operations are identical to those of the reverse transactions in Germany. The maturity of reverse transactions is shorter in most other countries. There is a clear-cut distinction between the higher frequencies of intervention (up to three times a day in the United Kingdom) in the United States, the United Kingdom, Canada, Australia, and Japan, and the lower (generally weekly) frequencies in all other countries, especially those that are likely to be inaugural members of EMU. Additional irregular fine-tuning operations are used in every country with the exceptions of Germany and Austria. Also fairly common are long-term refinancing operations, though these are not used in Canada, Australia, Spain, and Sweden.

    Most countries also have marginal lending and marginal deposit facilities. Where a formal standing facility does not exist, similar arrangements are in place. In the United Kingdom, there are several facilities charging escalating rates aimed at limiting the rise in the overnight rate. In Canada, discretionary reverse transactions operate as quasi-standing facilities. In Germany, issuance of short-term paper plays the role of a deposit facility. Although some countries still maintain a subsidized below-market facility (discount window), it has generally not been used in recent years for liquidity management purposes.

    Average reserve requirements exist in Australia, Austria, Canada, France, Germany, Italy, Japan, the Netherlands, Spain, Switzerland, and the United States, but they arc remunerated only in Australia, Italy, the Netherlands, and Switzerland. To reduce the volatility of the overnight rates, some countries without reserve requirements have introduced averaging provisions. In Canada, for example, there is a “zero” reserve requirement with averaging and banks are penalized when they have negative average settlement balances on a one-month period. In the United Kingdom, reserve requirements have been replaced by a small cash deposit ratio, but without averaging.

    Although frequent interventions have not been ruled out, the announced framework for the ECB’s monetary policy appears much closer to the continental European model than to that of one of the other industrial countries. Key decisions remain, however, and events could force the ECB to play a more active role.

    Framework for General Financial Policies

    Banking Supervision and Functions of Lender of Last Resort

    Among the industrial countries, there is no clear tendency to combine banking supervision functions with monetary policy functions (Table 42). About half of the countries combine the two functions within the central bank. The other countries separate these functions and assign supervisory responsibilities to another agency, usually under the control of the ministry of finance. In some instances the distinction is blurred. In France, for example, the Banking Commission (Commission Bancaire) is chaired by the Governor of the Bank of France with representatives of the French Treasury; the commission supervises compliance with regulations, but the Bank of France carries out inspections on behalf of the commission.16

    Table 42.Monetary and Supervisory Agencies
    Monetary AgencySupervisory AgencyNotes
    AustraliaReserve Bank of Australia (CB)Reserve Bank of Australia (CB)C
    AustriaNational Bank of Austria (CB)(Federal) Ministry of Finance (MF)S
    BelgiumNational Bank of Belgium (CB)Bank and Finance CommissionS
    CanadaBank of Canada (CB)Office of the Superintendent of Financial Institutions (MF)
    DenmarkDanmarks Nationalhank (CB)Financial Supervisory Agency (MEA)S
    FinlandBank of Finland (CB)Financial Supervision Authority (CB)S
    Bank of Finland ICB)
    FranceBank of France (CB)Bank of France (CB) Banking CommissionC
    GermanyDeutsche Bundesbank (CB)Federal Banking Supervisory Office Deutsche Bundesbank (CB)S
    GreeceBank of Greece (CB)Bank of Greece (CB)C
    Hong Kong, ChinaHong Kong Monetary Authority (CB)Hong Kong Monclary Authority (CB)C
    IrelandCentral Bank of Ireland (CB)Central Bank of Ireland (CB)C
    ItalyBunca d’Italia (CB)Banca d’Italia (CB)C
    JapanBank of Japan (CB)Ministry of Finance (MF)S
    LuxembourgLuxembourg Monetary Institute (CB)Luxembourg Monetary Institute (CB)C
    NetherlandsDe Nederlandsche Bank (CB)De Nederlandsche Bank (CB)C
    New ZealandReserve Bank of New Zealand (CB)Reserve Bank of New Zealand (CB)C
    NorwayNorges Bank (CB)Banking, Insurance and Securities Commission (MF)S
    PortugalBanco de Portugal (CB)Banco de Portugal (CB)C
    SpainBanco de Espana (CB)Banco de Espana (CB)C
    SwedenSveriges Riksbank (CB)Swedish Financial Supervisory AuthorityS
    SwitzerlandSwiss National Bank (CB)Federal Banking CommissionS
    United KingdomBank of England (CB)Bank of England (CB)C
    United StatesFederal Reserve Board (CB)Office of die Comptroller of the Currency (MF)S
    Federal Reserve Board (CB)
    State governments
    Federal Deposit Insurance Corporation
    VenezuelaBanco Ceniral de Venezuela (CB)Superintendency Of BanksS
    Source: Goodhart and Schoenmaker (1995).Note: The sample covers all industrialized countries (OECD); Hong Kong, China; and Venezuela. C = combined: CB = Central Bank; MEA = Ministry of Economic Affairs; MF = Ministry of Finance; and S = separated.’In May 1997, the U.K. government announced plans to move responsibility for banking supervision from the Bank of England to the Securities and Investments Board.

    There does not seem to be any clear-cut correspondence between monetary operating procedures and banking supervision models. Industrial countries outside continental Europe do not share the same model. Some countries (Australia, New Zealand, the United Kingdom,17 to some extent the United States) combine monetary and supervisory functions within the central bank, whereas other countries (Canada, to some extent the United States) separate them. Continental European countries are also split as to how to allocate these responsibilities. Germany, some of its close neighbors (Austria, Switzerland, Belgium, Denmark), and three Scandinavian countries (Sweden, Norway, and Finland) separate the two functions, whereas the other EU countries combine them.

    Current plans suggest that EMU is likely to follow the German model of separating monetary and supervisory responsibilities. The Treaty of Maastricht limits the role of the ECB in the area of prudential supervision to “specific tasks” that the EU Council may confer to it on a proposal of the European Commission. Specifically. Article 105(6) of the treaty states: “The Council may, acting unanimously on a proposal from the Commission and after consulting the ECB and after receiving the assent of the European Parliament, confer upon the ECB specific tasks concerning policies relating to the prudential supervision of credit institutions and other financial institutions with the exception of insurance undertakings.” The commission has not yet taken any initiative in this direction.

    The treaty makes clear that the role of the European System of Central Banks is subordinate to that of the competent supervisory authorities by indicating that the ESCB is expected “to contribute to the smooth conduct of policies pursued by the competent authorities relating to the prudential supervision of credit institutions and the stability of the financial system” (Article 105(5)). Accordingly, the statute of the ESCB assigns the ECB only an advisory function by indicating that “the ECB may offer advice to and be consulted by the Council, the Commission and the competent authorities of the Member States on the scope and implementation of Community legislation relating to the prudential supervision of credit institutions and to the stability of the financial system” (Article 25(1)).

    Central banks of industrial countries with highly securitized and liquid financial markets, such as the United Slates and the United Kingdom, have acted as lender of last resort in order to satisfy their respective mandates to ensure financial market stability.18 In contrast, central banks of countries where credit is mainly intermediated by banks, such as Germany and other continental EU countries, have generally not taken up the role of lender of last resort for which they rarely have a statutory mandate.19

    The treaty follows the German model in not attributing any lender-of-last-resort role to the ESCB. In fact, no mention is made of this function in either the treaty or in the statute of the ESCB. This implies that the ECB is not expected to inject liquidity into the system to deal with liquidity or insolvency crises of the banking system. In addition, it is yet to be determined how crises of this nature will be detected, monitored, and resolved. Although this arrangement may reduce moral hazard and enhance the credibility of the ECB, which would be less influenced by considerations of financial system stability when deciding monetary policy, it may be at odds with other functions assigned to the ECB by the Treaty of Maastricht, such as promoting “the smooth operation of payments systems” (Article 105(2)). Given that a central bank usually remains the only immediate source of funding in the system, close coordination between the ECB and supervisory agencies in participating countries will be essential for the ECB to have enough information to carry out its refinancing operations.

    Clear and unambiguous mechanisms for managing liquidity crises are crucial to the smooth functioning of TARGET. There may be situations in which the ECB will have to extend a sizable credit within hours of being presented with an institution unable to meet its payments obligations. In this instance, the ECB should have all the supervisory information needed to assess whether it is facing a liquidity crisis or a solvency crisis. As the U.S. experience shows, the likelihood and the systemic consequences of liquidity crises are bound to increase as the volume of transactions in securities markets grows. Given that the rapid expansion of these markets is a widely anticipated consequence of EMU, it is of concern that no clear EMU-wide mechanism to deal with a liquidity crisis has been agreed upon in a context in which supervisory functions are decentralized nor has the ECB been given any supervisory or lender-of-last-resort role.

    No additional agreement has yet been announced on the flows of supervisory information between the ECB and the competent authorities—not even in the event of a banking crisis. Information sharing is likely to be regulated by the so-called BCC1 Directive (Directive 95/26/EC of June 29, 1995), which removes all legal obstacles to the exchange of information between the authorities supervising credit institutions, investment firms, or insurance companies and the staff of central banks or “other bodies with a similar function in their capacity as monetary authorities”—including the ECB. The implementation of this directive remains ambiguous, however, because it neither specifies the information that could be exchanged nor creates an obligation to provide it. Further arrangements between supervisory authorities and the ECB will be needed to make sure that the relevant information for the smooth functioning of the payments system and the conduct of monetary policy operations will be exchanged in a timely manner in the event of a crisis.

    Deposit Insurance Schemes

    The Directive on Deposit-Guarantee Schemes (May 1994) required all EU countries to introduce a deposit insurance scheme by July 1995 with the following main features: (I) a minimum coverage of ECU 20,000 for each depositor (ECU 15,000 until December 31, 1999); (2) insurance of deposits at foreign branches according to the home country scheme,20 unless the foreign branch joins a more favorable host country scheme; (3) a possibility of excluding from coverage the deposits of financial institutions and insurance companies, as well as bonds issued by banks.

    The directive notwithstanding, the structure of deposit insurance schemes in the EU is far from being harmonized (Table 43). Deposit insurance administration is the responsibility of the government in five EU countries, of the banking system in six, and of both in the remaining four. Funding is provided ex ante (i.e., a reserve fund is established before the occurrence of a bank failure) in two-thirds of the countries and ex post (i.e., funds are obtained after the occurrence of a bank failure) in the remaining ones, but no country seems to make explicit the source of funding for catastrophic losses; among ex ante funding schemes, only those of Denmark and the United Kingdom specify a minimum reserve level for the fund. Deposit insurance premiums are risk based only in Italy, Portugal, and Sweden, and the basis on which the premium is calculated varies considerably across the European Union. The extent of coverage is uneven, ranging from a low of about $12,000 in Spain to a high of some $118,000 in Italy. In Finland, each depositor is insured in full; full insurance exists in Germany but only up to 30 percent of the bank’s capital per depositor. Coinsurance schemes, in which depositors share part of the losses, exist in the United Kingdom and Ireland and to some extent in Portugal, where depositors are fully covered up to a limit and only partially for additional amounts.

    Table 43.Permissible Banking Activities and Bank Ownership in the European Union and G–10 Countries, 1995
    Administration of System: Government or IndustryExtent or Amount of CoverageEx Ante or Ex Post FundingFund Minimum Reserve LevelBase for PremiumRisk-Based Premiums
    AustriaIndustryS 260,000 (per physical person depositor)Ex post; system organized as an incident-related guarantee facilityn.a.The deposit guarantee system shall obligate its member institutions, in case of paying out of guaranteed deposits, to pay without delay pro rata amounts that shall be computed according to the share of the remaining member institution at the preceding balance sheet data as compared to the sum of such guaranteed deposits of the deposit guarantee systemn.a.
    BelgiumGovernment/industry (joint)ECU 15,000 until Dec, 1999,

    ECU 20,000 thereafter
    Ex ante, but in case of insufficient reserves, banks may be asked to pay, each year if necessary, an exceptional additional contribution up to 0.04 percentNoTotal amount of customer’s deposits that qualify for reimbursement and that are expressed either in BF, ECU, or another EU currencyNo
    CanadaGovernment (Crown Corporation)Can $60,000 (per depositor)Ex anteNoInsured depositsNo
    DenmarkGovernmentDKr 300,000 or ECU 42,000 (per depositor)Ex anteYes, 3 billion DKrDepositsNo
    FinlandIndustry100 percent (per depositor)Ex anteNoTotal assetsNo
    FranceIndustryF 400,0001 per depositor)Ex postn.a.The contribution consists of two parts: (1) A fixed part, irrespective of the size of the bank, equal to 0.1 percent of any claim settled and with a F 200.000ceiling: and 12) a proportional part, varying according to a regressive scale relative to the size of the bank contributing, based on deposits and one-third credits.n.a.
    GermanyIndustry100 percent up to a limit of 30 percent of the bank’s liable capital (per deposit]Ex ante: however, additional assessments may be made if necessary to discharge the fund’s responsibilities. These contributions are limited to twice the annual contributionNoBalance sheet item “Liabilities to Customers”No
    GreeceGovernment/industry (joint)ECU 20,000 (per depositor)Ex anteNoTotal depositsNn
    IrelandGovernment90 percent of deposits: maximum compensation is ECU 15,000Ex anteNo, but minimum Premium Rate of £20,000Total deposits excluding interbank deposits and deposits represented by negotiable certificates of depositNo
    ItalyIndustry100 percent of first Lit 200 million (per depositor)Ex post: banks commit ex ante: however, contributions are ex postNoMaximum limit for funding the whole system: Lit 4,000 billion. Contributions are distributed among participants on the basis of deposits plus loans minus own funds with a correction mechanism linked to deposit growth.Yes
    JapanGovernment/industry (joint)¥10 million yen (per depositor)Ex anteNoInsured depositsNo
    LuxembourgIndustryLux F 500.000 (per depositor), only natural personsEx postn.a.Banks’ premiums based on percentage of loss to be metn.a
    NetherlandsGovernment/industry (joint)ECU 20.000 (per depositor); compensation paid in guildersEx postn.a.Amount repaid in compensation to insured is apportioned among participating institutions. However, the contribution in any one year shall not exceed 5 percent per an institution’s own funds and per all institutions own fundsn.a.
    PortugalGovernment100 percent up to 15,000 ECU 75 percent: 15,000-30,000 ECU 50 percent: 30,000-45,000 ECU (per depositor)Ex ante. However, (he payment of the annual contributions may be partly replaced, with a legal maximum of 75 percent, by the commitment to deliver the amount due to the Fund, at any moment it proves necessaryNoGuaranteed depositsYes
    SpainGovernment/industry (joint)Ptas 1.5 million (per depositor): to be increased to ECU 20.000Ex anteNoDepositsNo
    SwedenGovernmentSKr 250.000 (per depositor)Ex anteNoCovered depositsYes
    SwitzerlandIndustrySwF 30.000 (per depositor)Ex postn.a.Two components: fixed fee in relation to gross profit: variable fee depending on share of total protected deposits of an individual bankn.a.
    United KingdomGovernment90 percent of protected deposits, with the maximum amount of deposits protected for each depositor being £20,000 (unless the sterling equivalent of the most an individual can collect in a bank failure is £18.000 (per depositor) or ECU 20.000 if greaterEx ante: banks make initial contributions of £10,000 when a bank is first authorized, further contributions if the fund fall below ECU 22,222 is greater!. Thus falls below £3 million not exceeding £300.000 per bank based on the insured deposit base of the banks involved, andYes, the fund is required by law to maintain a level of £5 million to £6 million but the Deposit Protection Board can decide to borrowow to meet its needsAll deposits in European Economic Area currencies less deposits by credit institutions: Financial institutions, insurance undertakings, directors, controllers and managers, secured deposits, CDs, deposits by other group companies and deposits that are part of the bank’s own fundsNo
    United StatesGovernment$100,000 (per depositor)Ex anteYes, 1.25 percent of insured depositsDomestic depositsYes
    European Union (EC Directive on Deposit Guarantee Schemes)Only directs that each member state shall ensure within its territory one or more deposit guarantee schemes are introduced and officially recognizedThe aggregate deposits of each depositor must be covered up to ECU 20.000. Until Dec. 31, deposits are not covered up to ECU 20.000 may retain the maximum amount laid down on their guarantee schemes, provided that this amount is not less than ECU 15.000 (per depositor)Determined within each member stale 1999, member states in whichDetermined within each member stateDetermined within each member stateDetermined within each member state
    Sources: IMF country desks; and Barth, Nolle, and Rice (1997).

    The lack of harmonization of deposit insurance schemes may become a source of concern. Various degrees of deposit insurance protection could trigger regulatory competition between banking systems in the European Union, with funds flowing toward countries offering the most protection. Furthermore, given that foreign branches can join a host country scheme, situations may arise in which foreign branches obtain “insurance coverage in a country even though that country has no authority to regulate the risk-taking behavior of those branches because of mutual recognition.”21

    Financial Regulation, Capital Standards, and Supervisory Practices

    There are considerable differences in the regulation of banks’ activities and their ownership structure across EU countries. Table 44 classifies EU and G-10 countries according to the extent to which they are allowed to engage in securities, insurance, and real estate activities, and to own or be owned by non-banks.22 Unless further harmonization takes place, banking regulations grant considerably different powers to banks in each country, ranging from the “very wide powers” given to British, French, Dutch, and Austrian banks to the “somewhat restricted powers” of Italian, Swedish, Belgian, and Greek banks; the banks in the remaining EU countries (Germany, Spain, Portugal, Ireland, Denmark. Finland, and Luxembourg) fall in an intermediate group with “wide powers.”

    Table 44.Permissible Banking Activities and Bank Ownership in the European Union and G–10 Countries, 1995
    Securities1Insurance2Real Estate3Commercial Bank

    Investment in

    Nonfinancial Firms
    Nonfinancial Firm

    Investment in

    Commercial Banks
    Banks given very wide powers
    United KingdomUnrestrictedPermittedUnrestrictedUnrestrictedUnrestricted
    Banks given wide powers
    Banks given somewhat restricted powers
    ItalyUnrestrictedPermittedResti ictedRestrictedRestricted
    Banks given restricted powers
    United StatesRestrictedRestrictedRestrictedRestrictedRestricted
    Source: Barth, Nolle, and Rice (1997).Definitions: Unrestricted: a full range of activities in the given category can be conducted directly in the bank. Permitted: a full range of activities can be conducted, but all or some must be conducted in subsidiaries. Restricted: less than a full range of activities can be conducted in the bank or subsidiaries. Prohibited: the activity cannot be conducted in either the bank or subsidiaries.

    Of all these possible banking activities, securities operations are the most uniformly regulated across the European Union: they are “unrestricted” in all EU countries except Belgium (where a bank may not underwrite stock issues) and Greece (where dealing and brokerage must be conducted through subsidiaries).23 Firewalls (i.e., restrictions designed to maintain securities and insurance operations separate from affiliated banks) are mandated only in Italy. Denmark, and Greece. Insurance activities by banks are also “permitted” in most countries if they are conducted through subsidiaries, but they are “restricted” in Germany, Finland, and Greece (i.e., less than a full range of activities can be conducted in the bank or subsidiaries), and they are “prohibited” in Ireland. Real estate activities are restricted in more than one-third of the EU countries; permitted in Germany, France, the Netherlands, Denmark, and Finland; and unrestricted only in the United Kingdom, Ireland, Austria, and Luxembourg. Commercial bank investment in nonfinancial firms is unrestricted in two-thirds of the EU countries, permitted in Portugal, and restricted in Denmark, Italy, Sweden, and Belgium. Similarly, nonfinancial firm investment in commercial banks is unrestricted in 11 EU countries, permitted in Spain, and restricted in Italy and Luxembourg.

    Most securities activities are on the list of bank activities subject to mutual recognition in the European Union, included in the Second Banking Directive, which took effect on January 1, 1993 (Table 45). This means that the single EU passport will allow any EU bank to follow its home country regulations on securities activities when it operates in another EU country even if the host country regulations are different. As a result, lack of harmonization of the regulations on securities activities may hamper the competitive position of some banking systems by causing outflows of funds toward countries permitting the widest range of activities, but it cannot be an obstacle to cross-border competition. This may explain the greater harmonization of securities regulations. In contrast, insurance and real estate activities are not included in the list of activities subject to mutual recognition so that whether banks are allowed to engage in them depends on both home country and host country regulations. Differences in these regulations can create opportunities for regulatory arbitrage and be an obstacle to cross-border competition.

    Table 45.List of Bank Activities Subject to Mutual Recognition in the European Union1
    Acceptance of deposits and other repayable funds from the public.
    Financing leasing.
    Money transmission services.
    Issuing and administering means of payment (e.g., credit cards, traveler’s checks and banker’s drafts).
    Guarantees and commitments.
    Trading for own account or for account of customers in:
    Money market instruments (checks, bills, certificates of deposit)
    Foreign exchange
    Financial futures and options
    Exchange and interest rate instruments
    Transferable securities.
    Participation in share issues and the provision of services related to such issues.
    Advice to undertakings on capital structure, industrial strategy, and related questions and advice and services relating to mergers and the purchase of undertakings.
    Money brokering
    Portfolio management and advice.
    Safekeeping and administration of securities.
    Credit reference services.
    Safe custody services.

    The implementation of several EU directives24 and of the Basle Accord has not fully harmonized capital standards, which still differ somewhat across EU countries owing to the different lists of items that banks can use to meet capital requirements (Table 46). Likewise, supervisory practices vary in terms of procedures for examinations and inspections, disclosure of regulatory information, lending limits (on borrowers, sectors, countries, and large exposures), and limits on bank activities abroad (Table 47). Whereas a single currency will increase pressures for harmonization, decentralized supervisory functions may well allow these differences to persist long enough to affect the location of the banking industry within EMU.

    Table 46.Components of Capital for Meeting (he Capital Standards or Requirements in the European Union and G-10 Countries
    Noncumulative Perpetual Preferred StockCurrent Year Profit Added (or Loss Deducted)Intangible Assets Other than GoodwillGoodwillUndisclosed ReservesHybrid Capital Instruments (Including Cumulative Perpetual Preferred Stock)Subordinated Term DebtLimited Life Redeemable Preference SharesFixed-Asset Revaluation ReservesLatent, or Hidden Revaluation ReservesGeneral Loan/Loss ReservesInvestment in the Capital of Other Banks and Financial Institutions
    AustriaYesYesNoNoYes, but limitsYes, but limitsYes, but limitsNoYes, but limitsNoYesNo
    BelgiumYesYesNoNoYes, but limitsYes, but limitsYes, but limitsYes, but limitsYes, but limitsNoYesNo
    DenmarkNo, doesYesNoNoNoYes, but limitsNoNo, does not existNo, does not existNo, does not existNo, does not existNo
    FinlandYesYesNoNoNoYesYesNot applicableYesNoYesNo
    FranceNo, issues not permitted in domestic marketsYesNo, except lease renewal rightsNoNoYesYesYes, but not issuedYesNoYesYes, but limits
    GermanYesNoNoNoYes, but limnsYes, but limitsYes, but limitsNoNoYes, with limitsYes, with limitsNo
    GreeceYesYesYesYesNoYes, but limitsYes, but limitsYes, but not utilized at presentYes, but limitsNoYesNo
    IrelandYes, no limitsYesNoNoNoYes, but limitsYes, but limitsYes, but limitsYes, but limitsNoYes, but limitsNo
    ItalyYes, but limitsYesYesYesNoYes, but limitsYes, but limitsNo, does not existYes, but limitsNoYes, but limitsNo
    PortugalYesYesYesYesNo informationYesYes, but limitsNo informationYesNo informationYesNo
    SpainYesNoNoNoNoYes, but limitsYes, but limitsYes, but limitsYes, but limitsNoNoN0
    SwedenYesYesNoNoYes with approvalYesNoYes, with approvalNoNoNo-
    United KingdomYesYesNoNoNo.Yes, but limitsYes, but limitsYesYes, with cautionNo.Yes, but limitsNo
    CanadaYesYesYesNoNoYesYesYesNoNoNoYes, but back-to-back issues are deducted
    JapanYesYesYesNoNoYes, but not prevalentYesYes, but not issuedNoYesYesNo, it sole purpose is to raise capital ratio
    SwitzerlandYes, but limitsYesNoNoYes, but limitsYes, but limits and not including cumulative perpetual preferred stockYes, but limitsNoYes, but limitsYes, but limitsYes, no limitsNo
    United StatesYesYesNo, with limited exceptionsNoNoYes, but limitsYes, but limitsYes, but limitsNoNoYes, but limitsNo
    Table 47.Commercial Bank Supervisory Practices in the European Union and G-10 Countries, 1995
    Domestic Bank Activities Abroad
    Information Publicly DisclosedLimits or restrictions placed on Domestic bank’s foreign activitiesLending Limits on:
    Examinations and/or

    Required External AuditsBank examinations or inspectionsEnforcement actionsSpecific authorization requiredA single borrowerPersons connected with the bankParticular sectorsCountry risk exposureLarge exposures
    On-siteBank pay exam
    BelgiumYesNoYesNoYesNo, only notificationNo, only notificationyesYesNoNoYes
    DenmarkYes, usually every 3 yearsYesYesNoNoNoNoYesNoNoNoYes
    FinlandYes, not regularlyYesYesNoNoNoNoYesYesYesYesYes
    FranceYesNo informationYesNo informationNo informal ionNoNoNo informationNo informationNo informationNo informationNo information
    GreeceYes, generally every 2–3 yearsNoYesNoNoYesNoYesYesNoNoYes
    IrelandYes, usually every 18–24 monthsNoYesNoNoYesNoYesYesNoNoYes
    ItalyYes, usually every 4–8 yearsNoYes, for banks quoted on the stock exchangeNoYesYesNoYesYesYesNoYes
    LuxembourgYes, on an ad hoc basisYesYesNoNoNoNoYesYesNoNoYes
    NetherlandsYes, depends on size/risk profileNoYesNoNoNoYesYesYesYesYesYes
    PortugalYes, usually annuallyNoYesNoYesYesNoYesYesNoNoYes
    SpamYesNoYesNoNoYes, but only

    branches outside EU
    United KingdomYes, but limited and usually bienniallyNo, not directlyYesNoYes, but not explicitly naming institutionsNoNoYesYesNoNoYes
    CanadaYes, annuallyYesYesNoNoNoNoYesYesNoNoYes
    JapanYesNo informationNo informationNo informationNo informationNo informationNo informationYesNo informationNo informationNo informationNo information
    SwitzerlandNoYesYes, official part of supervisory systemNoNoNo, only notificationNoYesYesNoNo but provision requirements per countryYes
    United StateYesYesYes, for Banks with assets exceeding $500 millionNoYesNoYesYesYesNoNoNo

    Euro as a Catalyst: Incentives for Continued Structural Change

    Driven by financial deregulation, changing opportunities for investment, and bank disintermediation, European securities markets have become more highly integrated and liquid. These changes have been associated with the placement of large sovereign debt issues, which provided strong incentives to develop liquid and efficient secondary bond markets, and with the accumulation of large stocks of public debt, which raised yields on government securities thereby making them an attractive alternative to bank deposits. Facilitated by the recent convergence of macroeconomic policies, greater capital mobility has contributed to market integration by linking national securities markets, reducing bond spreads, and increasing co-movements in bond and equity returns across EU countries.25

    Against the background of these ongoing structural changes, the introduction of the euro will alter incentives in such a way so as to encourage the further securitization26 of European finance, greater uniformity in market practices, more transparency of pricing, and increased market integration.27First, by eliminating separate currencies, the introduction of the euro reduces the direct cost of spot transactions and eliminates a relatively volatile element of market risk—foreign exchange risk—in longer-dated real and financial contracts between entities in EMU member countries. While foreign exchange risk between some ERM currencies may have diminished recently (as measured by implied volatilities, for example), the costs incurred by market participants—including central banks—during the violent disruptions in the ERM crisis in 1992–93 will long be remembered as will the frequent realignments, often preceded by speculative attacks, in the early years of the EMS and in the less formal exchange rate arrangements before the EMS.

    Second, the elimination of currency risk increases the relative importance of other elements of risk, including credit, liquidity, settlement, legal, and event risks. Credit risk is likely to be the most important component of securities pricing within EMU, with the implication that the “relative value” of underlying credits rather than judgments about the stability and volatility of currency values will drive securities prices.

    Increased attention will be paid to other elements of risk. Bond issues of two otherwise identical credit risks—say, a German company and a French one producing the same goods and having similar balance sheets—may be priced differently if issuing techniques, clearing and settlement procedures, and legal procedures are different in the respective countries. The impact of these remaining and less volatile components of risk on the cost of raising funds will provide incentives to suppliers of securities to narrow further their interest rate spreads by increasing transparency and by improving issuing techniques and financial infrastructures to attract investors. This competitive process, if allowed to run its course, could lead to the sufficient harmonization of market practices within the euro zone to eliminate the advantages a particular geographical market may now have. In this way, the elimination of currency risk could lead to greater uniformity and transparency of market practices, with the benefits of more uniform pricing and a breakdown of market segments within Europe.

    The elimination of currency risk and its costs, the convergence of credit spreads, and more uniformity in market practices together can be expected 10 increase the depth and liquidity of European securities markets. In short-term markets (money, swap, and short-term treasury bill markets, for example), contracts denominated in individual currencies will be redenominated in euros and could be traded across national markets, even if small credit spreads remain. For securities with multiple exchange listings, competition among exchanges could lead to a consolidation of trading in a single location. Even in markets that remain some-what segmented (because of higher credit spreads or restrictions), lower transaction costs (elimination of commissions on foreign exchange transactions and costs of hedging exchange rate risk) and the removal of trading restrictions (e.g., on institutional investors) will add liquidity. Moreover, competition among issuers—no longer based on the strength of the currency—will encourage sovereign borrowers to introduce market reforms.

    Third, the euro will directly reduce the number of existing barriers to cross-border investment and eliminate some restrictions on currency exposures of various pools of capital (pension funds, insurance companies, other asset managers). To begin with, all intra-EMU foreign exchange restrictions on the investments of pension funds and insurance companies will become irrelevant within the EMU area (see the appendix at the end of this annex). The EU matching rule (liabilities in a foreign currency must be 80 percent matched by assets in that same currency) for insurance companies, which has been extended to pension funds in some countries, will also cease to be binding within EMU since insurance companies will be able to invest their assets in any country of the euro area as long as their liabilities are denominated in euros. The size and country diversification of assets managed by institutional investors in the European Union, say mutual funds—still far smaller than in the United States-could rapidly increase together with their share of foreign investments (Table 48). Finally, the “anchoring” principle, restricting lead managers of issues to full subsidiaries domiciled in the issuing country, will become irrelevant and will thereby increase the potential for intra-EMU market penetration.

    Table 48.Mutual Funds, June 1996

    Net assets (in billions of U.S. dollars)
    European Union1366.74533.94496.321,396.99
    United States1,532.46741.78817.753,091.99
    Number of funds (in units)2
    European Union17,1364,4361,91213,484
    United States2,6112,3909955,996
    Source: Investment Company Institute.

    Fourth, portfolio diversification will change along with volatilities and correlations of assets in the EMU area, although some “home bias” could remain (sec Box 8). Moreover, the advantages of currency diversification will be lost to the extent that business cycles have been asynchronous and shocks asymmetric. This will encourage investors and financial institutions to search for, and find, new opportunities for portfolio diversification within EMU repo, government securities, and corporate securities markets, but it may also encourage them to seek diversification outside the euro area as well.

    European securities markets will also be shaped by other important factors. Technological progress will soon make fully integrated EU-wide securities and derivative markets unavoidable, by making the location of trading, clearing, and settlement largely irrelevant. Continued fiscal consolidation—as part of the Stability and Growth Pact—is likely to reduce the volume of new government bond issues, providing room for private entities to issue new equity shares and debt securities. Finally, if the role of the unfunded social security system diminishes, the stepped-up activities of institutional investors (e.g., insurance companies and private pension funds) will increase the demand for public and private paper of various maturities and types, perhaps including corporate bonds.

    Structural Implications for Securities Markets: Further Securitization of European Finance

    As just discussed, the euro has the potential for catalyzing and enhancing the impact of EU financial directives, increasing transparency in credit evaluation, accelerating the processes of financial market integration, and further expanding Europe’s institutional investor base. This section examines prospects for the development of EMU-wide securities markets, including repo, bond (public and private), equity, and derivative markets.

    Box 8.Volatility and Correlation of Asset Returns in EMU

    The relation between exchange rate stability and the volatility of asset prices has been one of the most debated issues in the economic literature. One view is that a fixed exchange rate regime—hence EMU—increases the volatility of securities prices. According to this view, when the exchange rate is not allowed to change, shocks to productivity, consumer preferences, or other real shocks of domestic origin will be reflected to a larger extent in securities prices (“volatility transfer hypothesis”).

    Several arguments have been put forward to counter or qualify this view. First, the volatility transfer hypothesis holds unambiguously only when real domestic shocks prevail; if domestic or foreign money demand shocks prevail, a fixed exchange rate regime would have, instead, an opposite, dampening, effect on the volatility of securities prices. Furthermore, for foreign real shocks, the consequences of fixing the exchange rate become ambiguous. Second, if the volatility of the exchange rate is created by uninformed “noise traders” or “chartists” responding to nonfundamental factors, then credibly fixing the exchange rate would eliminate the excess volatility without transferring it to other sectors of the economy. Finally, if the fixed exchange rate regime is imperfectly credible and stochastic shocks may trigger a speculative attack, then the volatility of interest rates is higher than it would be with a perfectly credible parity or a single currency, as in EMU; in this case, the impact of a fixed-rate regime on the volatility of interest rates provides no indication of what would happen with a perfectly credible fixed exchange rate regime or EMU.

    The question can only be settled empirically. A recent study by Flood and Rose (1995) of various episodes of fixed and flexible exchange rates over the 1960–91 period for OECD countries concludes that there is little evidence that “reducing exchange rate volatility compromises the stability of other macroeconomic variables” (p. 36). Similar results are obtained for EMS countries by Artis and Taylor (1994) and Fratianni and von Hagen (1990). Following a methodology similar to Mussa (1988). Bodart and Reding (1996) compare the volatility of bond and equity market returns across different exchange rate regimes. They use high-frequency data (daily returns between January 1989 and December 1994) for Belgium, France, Germany, Italy, Sweden, the United Kingdom, and the United States. They find that the countries with the lowest foreign exchange volatility (Germany, France, and Belgium) have the lowest volatility of bond returns also. In these countries, the volatility of equity prices is also lower than in Sweden and Italy. Furthermore, after breaking up the sample into subperiods, they find that, as long as the EMS regime was credible, the low volatility in foreign exchange markets was associated with a low volatility in bond markets. When foreign exchange volatility increased, bond market volatility did also. Analogous—although weaker—results were obtained for equity prices. Frankel (1996) conducts a similar experiment on stock prices and reaches similar conclusions. This evidence suggests that lower—not higher—volatility of securities prices is associated with lower exchange rate variability.

    There are two main reasons why securities prices could be correlated across countries: a common fundamental factor or contagion effects. In both instances, the correlation is likely to be affected by EMU. First, if EU securities prices share a common fundamental. EMU can increase their correlation because it reduces the variance of idiosyncratic shocks due to independent monetary policies. EMU might also reduce the correlation of securities prices by increasing the variance of the credit risk component. In the government bond market, this may happen because EMU eliminates the possibility of using the inflation tax to resolve country-specific budgetary difficulties. Similarly, in the corporate bond market, EMU eliminates the possibility of using the exchange rate instrument to compensate for real idiosyncratic shocks. In stock markets, EMU is expected to have a lower impact on price correlations because of the much higher potential for idiosyncratic shocks. A higher cross-country correlation of equity prices should, however, also be expected because EMU eliminates idiosyncratic monetary policy shocks and is likely to increase the correlation of business cycles.

    Second, international correlations of securities prices can also be explained by contagion effects due to noise trading or herd behavior unrelated to fundamentals. In this case, cross-country correl