WORLD BANK LATIN AMERICAN 44 ^AND CARIBBEAN STUDIES Work in progress for public discussion VA ), ') . < \_____________ _ jCurrency Risks (unhedged currency Lax Prudea mismatches of:banks/corps.) / \c\ Lax Prudential /&. \ FrinEcag Regulation & Supervision Sca \rc / LiquidityCriss Interest-rate/Slowdown Risks Partial & II-Sequenced l Financial & Capital Account (credit booms and Liberalization asset-price bubbles) LE S S ONS FROM T HE As IAN C R I SE S F O R LATI N AME R I CA * 7 ities in these economies, which are compared to tion of the current account balances for all other Asian economies that were not severely the crisis-Asian countries (Indonesia, Korea, affected by financial crises, and to major Latin Malaysia, the Philippines, and Thailand) dur- American economies. ing 1994-1995-see Table Al. From this However, these factors are more aptly point of view, Asian vulnerabilities began to defined as "symptoms" than as the root causes of emerge a few years before 1997.3 Since financial vulnerability; we must explain how they items 2-4 (below) produced fragile banking came about. The key question, thus, is what were and corporate systems, which reduced the the underlying causes of the emergence of these capacity of governments to raise interest symptoms.We present an explanation in Figure rates, the rigid exchange-rate systems in 1, that integrates several elements of the alterna- effect increased the likelihood of speculative tive explanations of the crises: attacks against the Asian currencies by pro- viding a "one-sided bet" for speculators. 1. Rigid exchange-rate regimes that in practice tightly linked domestic currencies to the US 2. There were moral hazards in domestic finance, dollar, created an environment in which arising from implicit or explicit deposit economic agents perceived that the risks of insurance schemes, and from the general devaluation were extremely low. Domestic perception that the "government stood economic agents, both banks and corpora- behind" banks and large corporations, tions, had an incentive to borrovv from for- which, according to this view, would not be eigners in foreign currency at interest rates allowed to go bankrupt. This perception led that were lower than domestic interest rates, bankers and investors to over invest in real without covering themselves against estate, stocks and other financial assets, exchange-rate risks. On the other side of the without paying sufficient attention to the transaction stood foreign lenders that based risks associated with such investments. As a their decisions on the presumption that gov- result, credit to the private sector rose ernments would rescue troubled banks and rapidly (especially after financial liberaliza- corporations in times of distress. Given the tion was undertaken) accompanied by the precedent of the Mexican bailout of 1995, rise of asset prices. These booms were fol- this domestic moral hazard may have been lowed by the closure of financial institutions exacerbated by an expectation that govern- that revealed that not all financial interme- ments, in turn, would be rescued by interna- diaries were going to be rescued, which tional financial institutions. lead to a sell-off of assets and asset-price In addition, the use of the US dollar as busts, leaving some insolvent financial inter- the anchor currency led to a progressive real, mediaries behind (Krugman 1998). effective appreciation of these Asian curren- cies, because the dollar was appreciating vis- 3. There was a lack of transparency in financial a-vis other "hard" currencies, including the dealings, combined with corporate governance JapaneseYen. Consequently, some of these structures based on family-control of con- countries experienced slowdowns in their glomerates, with weak minority shareholders dollar export revenues, and widening cur- rights. These conglomerates did not main- rent account deficits in the years leading up tain consolidated balance sheets, but relied to the crises. Competitiveness problems on intra-conglomerate guarantees to secure were intensified by the depreciation of the access to credit for their firms, which were Chinese currency on January 5, 1994, and also favored by government policies, and the subsequent devaluation of the Mexican were not disciplined by market forces. The Peso on December 22, 1994. These pres- media has referred to these systems as "crony sures were then reflected in the deteriora- capitalism."4 This situation contributed to 8 * F I NAN C IAL VU LNE RAB I L I T Y, S P I LLOVE R EFFECTS, AND CONTAGION the booms (and subsequent busts) of credit reveals that while some Latin countries and asset prices by obscuring the financial (Brazil, Chile, Colombia) discouraged short- situation of the private sector, thus covering term capital inflows, some Asian countries, up highly leveraged and unprofitable busi- especially Indonesia, Malaysia, Thailand, and ness transactions. Korea discouraged FDI and long-term bor- rowing. This particular sequencing of finan- 4. In addition, these economies suffered from cial and capital account liberalization adopted lax prudential regulation and supervision of its in these Asian countries seems to be a good banking and corporate systems, which facil- recipe for problems, especially in the pres- itated financial transactions and investments ence of moral hazard and weak prudential (induced by the perverse incentive structure regulation and supervision. described above) that were not properly assessed in terms of their potential risks. The ill-sequenced liberalization of the capi- This situation was aggravated by the fact tal account exposed and magnified the effects of that some Asian banking systems suffered the perverse incentive structure described in from a lack of reliable information about points 1 to 4 above. Such perverse incentives had the soundness of bank loan portfolios; for been present for a long time, but had not led to example, Table A2 shows that none of the financial crises of the magnitudes of 1997, crisis-Asian countries required banks to seek although they must have created microeconormc credit ratings (as of March 1997). That is, inefficiencies and vulnerabilities. Partial financial there were not only strong moral hazards, liberalization happened precisely at the time that but also weak market and regulatory/super- international capital flows were blooming. Thus, visory discipline. the ill-sequenced liberalization of the capital account facilitated the credit booms by opening 5. Credit and asset-price booms, and currency a channel for cheap (short-term) external funds exposures, were fed by partial and ill-sequenced to finance domestic investment (directly or financial and capital account liberalization pro- through the protected domestic financial system), grams that were implemented in several which aggravated the tendency to accumulate Asian countries during the late 1980s and unhedged foreign liabilities (by both banks and early 1990s.5 In several Asian countries, capi- corporations), increased rapidly the accumulation tal account liberalization took the form of of short-term external liabilities, and contributed easing restrictions on external borrowing by to the process of currency appreciation as foreign domestic banks and/or corporations, while at funds flowed into these economies. the same time restricting foreign entry and As mentioned, Figure 1 presents a schematic ownership (foreign direct investment, FDI) in illustration of the dynamics of the aforemen- the domestic banking sector. Appendix 1 tioned causes of financial vulnerability, its symp- contains an inventory of the controls prevail- toms, and two characteristic maladies experienced ing in Asia and Latin America prior to crises by economies after the economic downturns. As of 1997, focusing on the policies that reflected by the intersecting arrows, it may be dif- imposed restrictions on foreign ownership ficult to assert unequivocally that one of the and direct investments in the financial-bank- causes dominated in a given instance, as the ing sectors, and on domestic credit opera- symptoms of a gestating crisis may result from tions in foreign currency. A quick read different combinations of these factors. X~~~~G GE EVI E OFTHE SYMPTOMS OF REGARDING THE SYMPTOMS leading up to the crises, here we present some suggestive evidence showing the emergence of the symptoms described in Figure 1, for the crisis-Asian countries, compared to other Asian countries. Later, in sectionV, we assess the extent to which Latin American economies were financially vulnerable by 1997, in terms of the same performance indicators. * Real appreciation and its consequences in 1996; nevertheless its current account deficit remained high for international standards. Figure Al shows that the real effective exchange rates (REERs) of Asian economies appreciated, * Credit booms especially between April 1995 and mid-1997.6 Only Korea seems to have experienced a some- Figure A3 shows the evolution of bank credit to what stable REER during this period.The Philip- the private sector in real terms, during 1992- pines experienced the greatest appreciation among 1997. By early-1997, the crisis economies Asian countries, but it also did not experience a (Indonesia, Korea, Malaysia, the Philippines, and slowdown in the growth rate of export revenues, Thailand) had experienced increases in bank and its current account deficit had remained stable credit to the private sector in amounts exceeding at around 4.5 percent of its GDP since 1994. In by 50 percent or more the level observed in late- contrast, as shown in Figure A2, Malaysia, Thai- 1993. In contrast, the rise of credit in China, land, and Korea did experience notable slow- Hong Kong, and Taiwan was clearly more mod- downs of export growth during the first quarter erate than in the crisis-Asian econoiues. of 1996, a year that ended with a record current account deficit for Korea, and a comparatively * Asset-price bubbles high deficit in Thailand-see Table Al. Malaysia did not experience a deterioration of its current The Boom: Figure A4 shows the stock market account deficit in 1996, which actually declined price indexes in US dollars for the five crisis- from 10 percent of its GDP in 1995 to 5 percent Asian economies during the 1990s. Between 9 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ o o C1~~~~r -PI o) co ND 41 O) c O ° 1992 JAN .1992JAN --0 021992APR .1992APR . 0 0 Z (0 1992 JUL 1992 JUL _ co O 1992 ACT 1992 OCT CD 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4- 1993 JAN ; 1993 JAN 'ND .\ ,@ <: co 1993 APR . .1993 APR ' $.3m 00 j993JUL . . 0 .1993JUL .Z 2 1992 JU LI 1993 0CT 1993 0CT ' " 0 (D ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~ 3 1994JAN 1994JAN ' 3 1994 APR 4/19AR t. , E 1995APR 1995APR , o~- 1995JUL 1995JUL > ° 1994 OCT 1994 OCT 1995 JAN 1995 JAN *> v 1995 APR 1997APR 0 1995 JUL 1995 J U L 1994OCT 1995 OCT . -| 2 - 19965JAN 19965JAN 1995 APR c~1995 APR -O 195UL? 1996 JU L 0 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~0h 1996 OCT 1996 OCT o 1996 JAN 1~99 JAN o 19976APR 19976APRC 19976JUL 019976J UL 0 19976OCT 19976OCT Z -Cz C - C C CD CD C 2 - CO CC 1992JAN . . _ 1 1992 JAN 1 _ 1 I 1992APR , 1992APR D Z -C1992JU I U L 1992JUL ) c'Z 1992 OCT ,l o 1992 OCT ' I w I 193 JAN 1993 JAN . l _m 1993 APR '8U | A1993 APR 0 o1993 JUL .1993JUL C DC c 19930 / 1993 0CT I 1994JAN 1994JAN Z 1994APR 1994APR 1994JUL 1994JUL x CD> 1994 0CT 1994 OCT ' 1995 JAN l 1995 JAN 3 o- I00'-FA\R J_ 0 95P t 1995 OCT 1995 OCT | c/ 1996 JAN fii 'S1996 JAN ' , .o O 1996APR 1996APR 1996JUL 1996JUL >< 0 1996 0CT 1996 0CT 0 } - D s197JAN C 1997JAN 1 ' 1997APR 1997APR APR 1997JUL - 1997JUL ' > 1997OCT ' 1997OCT ._ C 12 * FI NANCIAL VULNERAB ILITY, SPILLOVE R EFFECTS, AND CONTAGION Table 1. International Bank Loans and Foreign Currency Exposure (End-June 1997) Debt over Reserves by Maturity and Sector Maturities Sectors Over Up to one year and up to Over Non- including two two Un- bank Un- Total one year years years allocated Banks Public private allocated Total Crisis Asia Indonesia 2.9 1.7 0.2 0.8 0.2 0.6 0,3 2.0 0.0 2.9 Malaysia 1.1 0.6 0.0 0.3 0.1 0.4 0.1 0.6 0.0 1.1 Philippines 1.4 0.8 0.0 0.4 0.2 0.6 0.2 0.7 0.0 1.4 South Korea 3.0 2.1 0.1 0.5 0.4 2.0 0.1 0.9 0.0 3.0 Thailand 2.2 1.5 0.1 0.5 0.1 0.8 0.1 1.3 0,0 2.2 Non-Crisis Asia China 0.5 0.2 0.0 0.2 0.0 0.2 0.1 0.2 0.0 0.5 Hong Kong 3.3 2.7 0.1 0,4 0.1 2.1 0.0 1.1 0.0 3.3 Singapore 2.6 2.4 0.0 0.1 0.0 2,2 0.0 0.4 0.0 2.6 Taiwan 0.3 0.2 0.0 0,0 0,0 0.2 0.0 0.1 0.0 0.3 Latin America Argentina 2.3 *1.2 0.1 0.8 0.2 0.4 0,5 1.3 0.0 2.3 Brazil 1.3 0.8 0.0 0.4 0.1 0,4 0.3 0.6 0.0 1.3 Chile 1,0 0.4 0.0 0.5 0.0 0.2 0.1 0,8 0,0 1.0 Colombia 1.7 0.7 0.1 0.9 0.0 0.5 0.3 0,9 0,0 1.7 Mexico 2.6 1.2 0,1 1.0 0.3 0.5 0.9 1.2 0.0 2.6 Peru 0.8 0.5 0.0 0.2 0.0 0.2 0.1 0.5 0,0 0.8 Venezue a 0.8 0.5 0,0 0.2 0.0 0.2 0.1 0.5 0.0 0.8 Source: Bank for Irternational Settlements (1998); Reserves from MF as of end-June 1997. Data comes from B S-report ng banks only. *The data on reserves does not nclude do lar deposits held by subsidiaries of BIS-reporting banks. June 1992, and January 1994, most Asian coun- turns. However, after July 1, 1997, all of these tries experienced rapid increases in stock prices, Asian markets plunged; Hong Kong's decline which was followed by a downturn in early Jan- began after October 20. uary, and a subsequent period until 1996, when most sock markets fluctuated within a price * Foreign currency exposure and official range that was still much higher than when the reserves rise began in 1992. During the boom period June 1992-January 1994), stock prices in Thai- Figure A6 shows that the net foreign asset posi- land increased by over 130 percent; in Malaysia tion of Thai banks suffered a continuous deterio- by over 110 percent; in the Philippines and ration since early-1994. Korea's deterioration Indonesia by over 85 percent; and in Korea by began in November 1994, and the Philippines in over 60 percent. Of the non-crisis-Asian early 1995. In contrast, Indonesia did not experi- economies, only Hong Kong experienced a rise ence a noticeable deterioration of the foreign of stock prices of similar magnitude. asset position of its banking system, but some The Bust: The decline of the Thai stock deterioration may have occurred after the Thai market was already underway at the beginning of devaluation in July 1997. Figure A6 also shows 1996, and Korea's decline began by April 1996, as that among the non-crises Asian countries only shown in Figure A5. The stock markets of Singapore experienced a noticeable deterioration Indonesia, Malaysia, and the Philippines muddled of its banks' foreign asset position, beginning through 1996, without apparent drastic down- after November 1994. LESSONS F ROM THE AS IAN CR ISES FOR LATIN AMERICA * 13 These trends in net foreign asset positions economies. This table shows the corresponding should be juxtaposed against the trends in the ratios of debt to foreign reserves as of end-June levels of international reserves to assess the 1997. It is remarkable that the bulk of foreign capacity of the monetary authorities to support debt accumulated in the crisis economies had the banks in case of a run on their foreign cur- maturities of less than one year. In Indonesia, the rency liabilities. Figure 2 shows that Indonesia, ratio of short-term debt to reserves was 1.7, in the Philippines, Korea, and Thailand maintained South Korea this debt was more than twice the relatively low and stable ratios of official reserves level of international reserves, and Thailand's was to M2, despite the seeming deterioration of their slightly higher than its reserves. Only Malaysia banking systems' net foreign asset positions. seems to have had an exposure to foreign liabili- Malaysia experienced greater volatility of its ties with maturities of less than one year signifi- reserves-to-M2 ratio during 1994, but declined cantly below the level of reserves.While all to the level of its Asian partners in the years countries had a substantial accumulation of for- leading up to 1997. Singapore held very high eign debt in the financial and corporate sectors, reserves to M2 ratios that covered well the dete- there were some differences across countries. rioration of its banking sector's net foreign assets South Korea's foreign debt seems to have been position. due mostly to its banking sector, while Indone- sia 's was due primarily to its non-banking private o Short-term external liabilities sector, as shown in Table 1. The net foreign asset position of the banks It should be noted, however, that these data on and the high levels of foreign debt in these the foreign asset positions of the banks may economies do not necessarily mean that all these underestimate the extent of the foreign currency exposures were unhedged. Documenting this exposure of the five crisis-Asian economies, would require information on the positions taken where private corporations are known to have by the banks and the private sector in derivatives been highly exposed in terms of their accumu- markets, which is not readily available. However, lated foreign liabilities. Table 1 presents data on the fact that these economies were driven into the level, maturity structure, and allocation of severe financial crises may indicate that an impor- foreign debt across sectors in the five crisis-Asian tant share of these exposures were unhedged. PART II: THE EFFECTS OF THE ASIAN CRISES ON LATIN AMERICA X~~~~E CHS s SPILLOVER EFFECTS SPILLOVER EFFECTS AND CONTAGION are different issues from vulner- ability. A vulnerable country can experience a crisis, but some countries that are not vulnera- ble can suffer the consequences of another's misfortunes, and can even become vulnerable once other dominoes have fallen. Figure 3 shows the complex relationships among various channels and consequences of spillover effects and contagion, and the policy responses that they may provoke. The following discussion explains the various items in Figure 3. A. FINANCIAL EFFECTS of effects of informational shocks flight to safety and demonstration effects-that differ 1. Informational shocks. News about an according to their impact on economies erupting crisis in one country may change that are not similarly vulnerable to those the perception of international investors suffering a financial crisis. about the profitability of investing in other emerging markets. As elaborated in Calvo * Flight to safety under uncertainty. (1995), when investors do not have suffi- Domestic and foreign investors may cdent information to distinguish among the prefer to move their funds into finan- fundamentals of different emerging markets, cial safe-havens during periods of high a currency crisis in one country can lead to market uncertainty or volatility. Figure speculation against other emerging market A7 provides a measure of the volatility currencies, even when the financial condi- or uncertainty associated with several tions are different.When investors have suf- emerging stock markets in Asia and ficient information, an informational shock Latin America; it shows a 20-day mov- associated with a financial crisis in one ing variation coeflicient of the stock country may lead investors to reassess their price indexes denominated in US dol- perceived risks of countries with similar lars. It is noteworthy that the volatility vulnerabilities. Below we discuss two types of these markets increased markedly in 17 18 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION Asia during 1997, especially after July. particularly important in relation to the Similarly, the volatility of major Latin investment of so-called "dedicated" funds, emerging markets increased in the that specialize in emerging markets, perhaps aftermath of the tequila crisis at end- even regional emerging market funds. 1994, and again in the second half of 1997. The corresponding flight to Notice that these financial effects would safety is reflected in Figure A8, which cause both a fall in asset prices and capital out- shows the evolution of the yield of the flows from emerging markets, leading to upward 10-year US Treasury bond in the pressures on exchange rates, declines in stock 1990s. It is evident that the yield fell market prices, and reduced access to interna- after major shocks, including the inva- tional capital markets for affected economies. sion of Kuwait, the onset of the tequila Table A3 shows the evolution of the monthly crisis, the beginning of the Thai crisis, average of gross capital flows to developing and the crash of the Hong Kong stock countries during 1997. Developing countries, market in October 1997, which including Latin American countries experienced reflects increasing US treasury bond a sharp reduction in gross private capital flows prices during periods of uncertainty. In during the last few months of 1997.Whether this case, all "emerging markets" suf- these financial effects, especially the reduced fered from the investors inclination to access to international capital, will continue to be move funds into US treasury bonds. felt through 1998, and beyond, remains an open question. However, some reduction in capital Demonstration effects. This effect inflows to developing countries compared to the implies that economies that are simi- levels of 1996 can be expected in the near future larly vulnerable will suffer from attacks (World Bank 1998b). against their currencies. Below we dis- cuss the stock market reactions in Asia B. INTERNATIONAL "REAL" and Latin America after the Thai EFFECTS devaluation of July 2,1997. Figure A5 clearly shows that there was an intra- We can distinguish between the following two Asia contagion effect, that could be types of real-side effects: denominated the "Thai drag effect," a tribute to the fall of Asian stock mar- 1. Demand contraction in the economies ket prices in July 1997, which was led suffering from the financial crises may lead by the Thai devaluation. to declines in the quantities of imports in the crisis countries, and to declines in 2. Institutional investors and the cash-in world prices of some commodities. In effect. When clients withdraw their invest- other words, Asia may be economically ments in diversified portfolio funds, the large enough to affect world prices, espe- fund managers must then liquidate portions cially for commodities that they import or of their portfolios. In doing so, managers export heavily. However, it should be may prefer to liquidate fund investments in noted that the corresponding declines in assets that have not yet been affected by the commodity prices, such as oil, copper, financial crises, especially from those that wheat, iron ore, and others, may produce seem more vulnerable-partly a demonstra- an increase in the financial vulnerability of tion effect-but that have not yet fallen, countries already running relatively high thus offering opportunities to reduce port- current account deficits, even if they are folio risk while enhancing the funds' liquid- not expected to have a significant direct ity position. The cash-in effect would be impact on the rate of (real) economic LESSONS FROM THE ASIAN CRISES FOR LATIN AMERICA * 19 Figure 3. Contagion and Spillovers from Asia: Channels and Consequences Channels Consequences Financial: Asset-Price Busts * Flight to Safety _ _ _'_- * Demonstration Effects (against similarly vulnerable) * Cash-in Effects \ Reduced Access to Int'l Capital Markets - ----' (& rising r) Policy Responses (Fiscal and/or Monet.) - Exchange-Rate Pressures &/or Real: - Current Account Deterioration * Demand Contraction (Quantity and Price Effects) * Competitiveness (Substitution Effects) Growth Slowdown ________________________________ (Expected and Actual) growth. In addition, the Asian crises may effects were very strong as there were similar affect indirectly overall exports of Latin symptoms of financial vulnerability in the other American countries due to its effect on the regional economies. The Thai devaluation ofJuly growth rates of advanced economies-an 1997, created strong competitiveness pressures on expected slowdown of approximately 0.4 its Asian neighbors and increased their vulnerabil- percentage points in the growth of GDP, ity. The currencies of Malaysia, Indonesia, and the according to the IMF (1997). Philippines then came under attack. As the crises deepened in each country, both the financial con- 2. A competitiveness or substitution tagion and the real-side spillover effects led to a effect due to real depreciation of the cur- vicious circle of currency depreciations, economic rencies of suffering economies. This effect contraction, and declines in other asset prices will be felt strongly by countries that: (i) throughout Asia. The symptoms of financial vul- export goods to Asian markets with domes- nerability in Korea were evident since at least tic competitors; (ii) countries that import 1996, and were subsequently strongly affected by goods from Asian producers, that compete the events surrounding its neighbors, which pre- with domestic industries; (iii) countries that cipitated Korea's crisis. In turn, the fall of Korea export goods that compete with Asian affected the other Asian economies. All this hap- exports in third country markets. pened against the backdrop of the Japanese eco- nomic slowdown in 1996-97, and the weaknesses The combination of financial and real conta- of its own financial system. gion channels led to an Asian domino effect. After The degree to which economies were and the crisis erupted in Thailand, financial contagion will be affected by the Asian financial crises 20 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION depends, of course, on their policy responses. For tion rates) and monetary and fiscal contraction example, authorities may wish to avoid widening (leading to lower growth rates). Some countries current account deficits, in the context of uncer- that have suffered strong speculative attacks have tainties about access to international capital mar- avoided a crisis (though at the cost of reduced kets, through different combinations of currency growth) through swift policy responses, as wit- depreciation (leading to some increases in infla- nessed by Hong Kong and Brazil. NA ERICA BEFORE AND DURING T I AN CRISES IN THIS SECTION, we first examine the degree of financial vulnerability in Latin American countries in the period prior to the 1997 crises. We then examine the spillover effects of the Asian crises on these econormies during 1997 and beyond. FINANCIAL VULNERABILITY IN and October 1996, perhaps due to a fall of the LATIN AMERICA COMPARED AND price of copper.Venezuela's export revenues did CONTRASTED WITH AsIAN not experience a significant slowdown until early COUNTRIES 1997, when the price of crude oil began to fall. Colombia's export revenues did not show an In order to analyze the effects of the Asian crises apparent slowdown during 1994-1997. In con- on the Latin American economi'es, it will be use- trast to some of the crisis-Asian econormies ful to assess whether Latin American countries (Malaysia, Thailand, Korea), the apparent lack of a showed sim-ilar symptoms of financial vulnerabil- systematic slowdown in export revenues in Latin ity in the period preceding the Asian debacle of America in the face of some R-EER appreciation, 1997. As explained below, the main differences might be explained partially by a variety of fac- between Latin America and crisis-Asian coun- tors: (i) the performance of their main export tries appear in the degree of foreign-currency markets in previous years (i.e., the US economy exposures and the health of the financial systems. has been boorming while Japan has been in reces- sion); (ii) by a higher elasticity of export and *Real appreciation and its consequences import volumes in Asia with respect to exchange- rate variations (which may be explained by the Brazil's R-EEk appreciation actually took place fact that Latin American countries depend more mostly between imid-1994 and early-1995, but its heavily on the export of commodities rather than merchandise export revenues remained virtually manufactures); and, maybe, (iii) by higher produc- stable between January 1996 and January 1997- tivity growth in Latin America during the last see Figures Al and A2. There was much R-EER few years. Also, Latin American countries did not appreciation in Chile, Colomnbia, and Venezuela experience as high current account deficits as between 1996 and 1997. The growth of Chile's Thailand or Malaysia after 1994 (with the excep- export revenues slowed between October 1995 tion of Peru in 1995). However, it is clear that 21 22 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION there has been a trend towards deterioration of in June of 1997. Only Argentina, Brazil, and current account deficits in Latin American coun- Mexico had reserves-to-M2 ratios below the 0.5 tries during 1995-96, except for the cases of mark, but still their ratios were slightly higher Mexico, Peru, andVenezuela. than those of Indonesia, Malaysia, Thailand, and the Philippines-see Figure 2. In the case of Credit booms and asset prices Argentina, these figures do not include the repo agreement between its Central Bank and foreign In contrast to the crisis-Asian countries, the private banks that was negotiated after the 1995 major Latin American countries, except Peru, had crisis (see below), and thus underestimate the not experienced strong credit booms in recent foreign currency reserve coverage this country years. In particular, Argentina, Brazil, Mexico, and had in 1997. Venezuela actually experienced slowdowns in In addition, bank restructuring, including credit growth after the Mexican crisis of 1994, as significant foreign entry, and significant seen in Figure A3. Also, asset prices (proxied by improvements in prudential regulation and stock market prices in US dollars) had increased supervision were implemented after the 1982- by less than in Asian countries in most Latin 83 in Chile and Colombia, and in the early American countries, with the exceptions of Brazil nineties in Peru. Upgrades of the supervisory and Peru.These two countries actually experi- frameworks were undertaken in all countries enced more rapid increases in their stock prices after the 1994-95 financial crises in Argentina than the crisis-Asian economies between 1992 and Mexico, which led to healthier financial and 1994, and their rise resumed in mid-1995 sectors than what existed in the crisis-Asian (after the Mexican crisis) and continued until countries in 1997. miid-1997. It should be pointed out, however, that The fact that most Latin American coun- both of these economies were coming out of tries (except Argentina and Brazil) have had periods of high inflation, which may have led to a flexible exchange-rate systems during recent fast re-monetization of these economies, and to a years suggests that the prevalence of unhedged fast growth of their financial sectors. currency mismatches in the corporate sector may not be as pronounced in Latin America as - Foreign currency exposure and official it was in Asia. In some highly dollarized reserves economies, however, a high share of lending through their domestic financial sectors takes There is also evidence that some Latin Ameri- place in foreign currency, suggesting the need can economies had improved the net foreign for closer surveillance.8 Other countries, such as asset position of their banking sectors after fac- Chile, Colombia, and Brazil, have been actively ing the "tequila" crisis in late 1994 and 1995, discouraging short-term foreign indebtedness of especially Chile, Argentina, and Mexico, as their corporate sectors through taxes and reserve shown in Figure A6. By June 1997, only Brazil requirements.9 had a level of net foreign assets in its banking sector comparable to the Asian economies that * Short-term external liabilities suffered financial crises. Colombia and Peru had shown a trend towards deterioration of the net Although the data on the debt structure of devel- foreign asset position of banks before the onset oping countries reported by the Bank for Inter- of the Thai crisis, but these figures are still bet- national Settlements (BIS) is not as accurate as ter than they were in some of the crisis-Asian would be desirable, available evidence indicates countries. that most Latin American countries did not have Another important difference is that Latin high ratios of short-term debt to reserves when American economies had higher reserves-to-M2 contrasted with their Asian counterparts. Table 1 ratios. For example, Peru had a ratio of over 0.7 indicates that by end-June 1997, only Mexico LESSONS FROM THE ASIAN CRISES FOpR LATIN AMERICA * 23 and Argentina had short-term debt (with matu- the demand contraction and substitution rity of less than one year) to reserves ratios above effects to a greater extent than other regional one, but still lower than Thailand, Indonesia, and economies-Table 2 shows the share of total especially Korea.10 Also, most Latin American exports that go to the developing countries of countries (especially Argentina, Mexico, and Asia. Moreover, both economies had relatively Brazil) had improved significantly the maturity high current account deficits-see Table A2- structure of their external public debt through and therefore the expected loss of export rev- very active debt management policies during the enues may have been expected to be accom- nineties. panied by higher borrowing requirements In synthesis, Latin American economies did from abroad, if no policy responses were to be not show strong symptoms of rising financial implemented. Brazil was also immediately vulnerability between 1994 and 1997, when affected by the baht devaluation and the ensu- contrasted to the crisis-Asian countries. In par- ing exchange-rate depreciations in Asia, ticular, their financial sectors were in much bet- because of its similar level of the current ter shape, due to the avoidance of credit booms, account deficit, which is symptomatic of its to past and recent improvements in prudential real exchange-rate appreciation that has been regulation and supervision, and to recent bank built-up since mid-1994, its well-known restructuring in some countries. Though some problems of fiscal sustainability, and its consid- other indicators of vulnerability were apparent in erable trade links with Asia. some Latin American countries, they were not, in general, as severe as those of their Asian coun- * Dynamics during October 21-November 14 terparts, and therefore, they did not experience severe contagion (demonstration) effects, as As shown in Figure A9, the fall of Hong Kong's shown below. However, Latin markets suffered stock market during October 22-24 was accom- spillovers from the Asian financial crises, and may panied by stock market declines in the US, Japan continue to feel the consequences in the near and the UK, thus reflecting a worldwide reassess- future. Overall, such effects may reduce Latin ment of future growth prospects for the world America's growth rate in 1998 by over one per- economy. Figure 4 shows that most Latin Ameri- centage point, as discussed below. can stock markets were severely affected during this period, though there were significant differ- THE DYNAMICS OF SPILLOVERS ences among them. FROM THE ASIAN CRISES The deepest falls took place in Brazil and Argentina. Brazil's currency, the real, came under Dynamics during July 1-October 21 strong attack and gross international reserves declined by about US $9 billion in a few days. Figure 4 shows that the stock markets in The Brazilian authorities responded swiftly to Brazil, Chile, and Peru immediately felt the these developments, initially by doubling effects of the Thai crisis, which was punctu- overnight nominal interest rates to over 40 per- ated by the baht devaluation of July 2, though cent per annum. On November 7, the govern- in a very modest way when contrasted to the ment complemented its interest-rate defense crisis-Asian countries. Other Latin stock mar- with an announcement of an emergency fiscal kets experienced declines only in the very adjustment program, which, as a whole, was pro- short-run. Our interpretation of these events jected to improve the primary surplus by about is that Chile and Peru were immediately 2.5 percentage points of GDP.This package was affected by the turmoil in Asia, because they approved by the Congress (with some amend- are the Latin American economies that have ments which did not alter the aggregate fiscal the most significant trade links with that effect) by mid-December. Moreover, the legisla- region, and thus were expected to suffer from ture also responded positively to President 24 * FINANCIAL VULNERABILITY, SPILLOVE R EFFECTS, AND CONTAGION Figure 4. Stock Price Indexes in Latin America and Crisis Asia, July 1997-January 1998 130 - Oct. 21 Nov. 14 120 - 110 - H 100 - C- 100 ---\- <-*- * < "'-th- --- - -IGPA (Chile) 80_ / "-^ General (P ru) Bovespa 60 - - CC-- r r > > > > r r > r r > r > > > > > > r r __ co co OD co Z 0) F- LO CD J (D CO Lr O LO CO C -) ar r-- ' CD ) 9 CD D 0 D a) C ) C > r- N :n N N 2 C 2 N o :o-N2- N N, 52 N- ~ - N 1-- oo co cO D OD) 0)~ O === 130 120- . ... ;...- ,.IPCMxc; . -~~~ ~ ~ ' \7,0a - ICMexico renin 60 r C- >- r- C- r- C- r- C- C- r- C- C- C- C- C- C- C- C- C- C- C CD CO CO CO O CO CO CO cO cz O cO CO CO CO C CD CO CO) C CO) CO C CD CO 0) 05) 0) 0) 0 CO C- 10 N O CD CO 10) 10 Cl) 1- 1-- O N CO CD CO CO CO m D CO CO C ~~~O - - NO o o N- N - N o CoN - CO CO N - N - -CN goI V- C CO CO eO O O O O CO os 70D Merval-A'genti n 60 -O C C C-- C-- LO CO--- ---CO L O Lf Cl CO t c- CO CO CO CO - - C Z~~ ~~--L 2~~~~L 8- O - -- Source: Datastream, BankeCr's Trust Economics and Strategy Grou.p LESSONS FROM THE AsIAN CRISES FoR, LATIN AMERICA U 25 Figure 4. (cont.) Stock Price Indexes in Latin America and Crisis Asia, July 1997-January 1998 130 - Oct. 21 Nov,.14 120 - 110 ---- - - 100 - > 80 - ------- 60-Philippines - 20 - 10- ~~~~ Thil 50 ~ ~ ~ _______ x~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~c Indonesia~~~~~~~~~~~~~~~~a Note: Thescale is :tierent rom the at.n Amercan markes' scale Cardoso's call to accelerate the consideration of Table 2. Regional Export Structure in Latin constitutional reforms to modernize public America and Asia (percentages of exports to developing Asia) admi'nistration and the social security system, 1994 1995 1996 which would be crucial in strengthening the Argentina 8.4 1i.0 io.9 public finances on a lasting basis. Brazil 10.4 10.8 12.7Aretnwaaloatcdafrthfllf Chile 15.7 16.4 17.0Aretnwaasoatceafrthfllf Colombia 1.6 2.3 1.0 Hong Kong, presumably because of its currency Mexico 0.7 1.2 1.3 board exchange-rate system, and its close com- Peru 16.7 16.9 19.1 mrilte ihBai Agniesokmre Venezuela 0.5 0.5 0.6 mrilte ihBai Agniesokmre prices followed closely the movement of their China 40.3 40.0 373Brazilian counterparts). Initially, interest rates Hang Kong 45.6 46.6 47.2 Indonesia 29.4 32.2 26.8 increased markedly, but, in contrast to the devel- Korea 32.2 35.1 37.8 opments after the tequila crisis, there was no sig- Malaysia 44.2 43.8 44.4 nfifcant deterioration of its international reserve Philippines 22.7 25.8 25.0 n Singapore 50.4 51.4 51.3 position. Moreover, Argentine bank deposits con- Taiwan 38.2 40.9 40.8 tinued to grow, albeit with increased dollariza- Thailand 34.3 36.0 36.4 ~tion. Interest rates subsequently declined some- Source. Direct ion of Trade Statistics database, International Mona- what, and by early December, Argentina was the tary Fund. Note: Deveioping Asia inc[udes China and India, first developing country to have regained access _________________________________ I to the dollar bond market. 26 * FINANCIAL VULNERABILITY, SPILLOVE R EFFECTS, AND CONTAGION Latin American countries with more flexible Hence, it seems that Latin American countries exchange-rate arrangements, such as Mexico, that were expected to suffer the most from real- Chile, Colombia and Peru, absorbed the external side spillovers, also suffered deteriorations in asset shocks by some depreciation of their currencies prices (exchange rates and stock prices) during and some increases in interest rates-Figure 5 the first month of 1998. shows the evolution of Latin American nominal exchange rates duringJuly 1997-January 1998. SUMMARY OF THE EFFECTS OF Mexican stock prices experienced initially a drop THE ASIAN CRISES ON LATIN as severe as those in Brazil and Argentina, but they AMERICA showed afterwards the fastest recovery, after the exchange rate depreciated by about 7 percent in Although the impact of the Asian crises on October 27, 1997.Venezuela also experienced a Latin American countries has been relatively sharp drop in stock prices, but its currency limited so far, prospects for growth in the remained relatively stable during this period, at the region have been revised downwards. Our pro- expense of a significant loss of reserves.The declin- jections for economic growth in Latin America ing trend of Chile and Peru stock market price before and after the Asian crises are shown in indexes accelerated somewhat.The smallest decline Table 3. These projections assume that there in stock prices during this period took place in will be no further shocks in the international Colombia, where the authorities had facilitated a markets during 1998. nominal devaluation of the currency of 16 percent As shown, we now expect that the region between July 1 and October 21, 1997. will grow by 3.1-3.5 percent in 1998, compared to 5.2 percent in 1997. Some of the slowdown * Dynamics after November 14 between 1997 and 1998 was to be expected even without the crises in Asia (as reflected in our The immediate financial spillovers discussed pre-crises projection of 4.4 percent growth for above receded somewhat since mid-November, the region), given the fact that the region but international real effects have been increas- (excluding Brazil) recorded in 1997 its fastest ingly felt and recognized by the capital markets. rate of economic growth in more than a decade. Interest rates in Argentina and Brazil have fallen The rest of the slowdown (the difference a bit since November, and their stock prices between 4.4 and 3.1-3.5 percent) would be due showed some recoveries since that time. In con- to the Asian crises, plus other unanticipated and trast, Chile, Peru, andVenezuela experienced unrelated factors, such as the weather phenome- additional significant falls in their stock prices non known as el Niiio and exogenous develop- between mid-December andJanuary-see Fig- ments in the international oil market (additional ure 4. These countries are perhaps the most to the crisis itself), and also to the authorities' severely affected by falls in the prices of copper response to the new situation. (Chile and Peru), oil (Venezuela) and other com- The impact of Asian crises will be partly due modities. More recently, Mexico and Colombia, to real effects, especially through reduced com- also oil-producing countries though not as modity prices and lower demand for exports by dependent on this commodity for their export Latin America's trading partners, plus some lin- revenues, have also experienced downward pres- gering financial effects in the form of reduced sures on their stock markets. Chile, Colombia, access to international capital markets (and higher and Mexico experienced further depreciations of spreads). In addition, Latin American authorities their currencies during January, as shown in Fig- have responded and are responding to the ure 5. The Chilean monetary authorities reacted changed international economic environment to these exchange-rate pressures in January 7, by with policies that further reduce the rate of eco- raising sharply interest rates, which led to some nomic activity in the short-run, as they are com- appreciation of the Chilean peso in late-January. bating both the pressures towards increased cur- LESSONS FROM THE ASIAN CRISES FOR LATIN AMERICA * 27 Table 3. Macroeconomic Projections before and after the Asian Crises Before After 1997 1998 1999 1998 1999 GDP Growth Rate (percent) Argentina 8.4 5.5 5.5 5,0 5.5 Brazil 3.0 3.0 4.5 0.5-1.5 3.0 Chile' 7.1 6.5 6.0 5.0 6.0 Colombia 3.2 4.5 4.0 4.0 3.5 Mexico 7.0 5.5 5.2 4.8 4.6 Peru 7.5 6.02 6.0 4.9 5.5 Venezuela 5.1 4.5 5.0 3.5 4.0 Mean 4.4 4.1 4.5 3.7-3.8 4.3 GDP-weighted mean 5.2 4.4 4.9 3.1-3.5 4.2 Inflation (Dec.-Dec. change in CPI) Argentina 0.5 2.5 2.6 2.0 2.5 Brazil 7.9 8.0 8.0 5.8 4.7 Chile 5.5 5.0 5.0 5.5 5.0 Colombia 17.6 17.0 15.5 18.0 17.0 Mexico 15.7 13.0 11.0 14.0 12.0 Peru 8.2 7.5 6.0 7.5 5.0 Venezuela 37.6 25.0 15.0 35.0 30.0 Mean 10.0 8.7 6.9 9.2 7.5 GDP-weighted mean 10.8 9.7 8.1 9.7 8.1 Current Account Balances (as percentages of GDP) Argentina -3.7 -3.3 -3.0 -3.9 -3.7 Brazil -4.1 -4.9 -4.2 -3.7 -3.9 Chile -4.0 -4.3 -3.5 -5.9 -5.0 Colombia -5.9 -4.3 -4.2 -5.2 -4.4 Mexico -1.8 -2.8 -3.3 -2.6 -2.9 Peru -5.1 -5.2 -4.9 -5.5 -5.1 Venezuela 6.9 3.0 2.7 1.5 1.0 Mean -5.1 -4.9 -4.5 -5.1 -4.6 GDP-weighted mean -3.1 -3.6 -3.4 -3.5 -3.5 Non-Financial Public Sector Balances (as percentages of GDP) Argentina -1,6 -1.0 -0.7 -1.2 -1.0 Brazil3 -5.9 -5.2 -4.4 -5.5 -4.1 Chile 1.8 2.0 2.0 1.0 0.9 Colombia -3.3 -3.4 -2.4 -3.6 -3.2 Mexico -0.5 -1.6 -1.8 -1.6 -1.1 Peru -0.3 -0.6 0.0 0.1 0.0 Venezuela 1.8 1.0 1.7 -3.0 -2.5 Mean -1.3 -1.5 -1.0 -1.9 -1.5 GDP-weighted mean -2.7 -2.6 -2.2 -3.1 -2.3 Source: World Bank staff projections. I Before changes to the national accounts. 'Without the effect of El Nino. The numbers are for the Public Sector Borrowing Requirements (PSBR). 28 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION Figure 5. Evolution of Nominal Exchange Rates in Latin America, July 1, 1997-January 28,1998 Oct. 21 Nov. 14 125- 120 - 105 -_ _ __ _ Venezuela == 100 ___I --- + Brazil Argentina CD a as aD as as a as as a aD as a as as a as as as 0 0 5 as as aC r- 05 aEl Cs N5 N 05 C\5 05 05 us o 0N as s- C 05 5 0c Cco0 c 05 sN LO CD co CO 5 as as - =s CS 05 125 - 120 - 115- .. ._ . Chile, O2 110- . .. / . .. . ... . M z Mexico 105---- - l __ loo- G _ ., ............................................. , Peru~~~~~~~r-co co 00 o as as rs Ls as as a C a as LO aL i as s a a Ca CO as as as a a- as as Ns C s a s a cr: , as Ns Ns O Os N = N as N N- as N as- as cO Oas as a as - a - - - - = Source: Datastream, Banker's Trust Economics and Strategy Group. Note: A rise in the index reflects a nominal depreciation. LESSONS FROM THE ASIAN CRISES FOR LATIN AMERICA * 29 rent account deficits (anticipating reduced capital fers less than other countries from reduced inflows) and, some of them, towards depreciation demand for its exports and reduced competitive- of exchange rates (to avoid or mitigate the infla- ness. Also, Brazil may actually benefit from tionary consequences of devaluations), through improvements in its terms of trade, because it different combinations of fiscal and monetary imports oil and other commodities. tightening. Such policy responses, designed to The downward revision in Brazil's pro- reduce vulnerability and/or inflationary pressures, jected inflation rate for 1998 reflects the amplify significantly the growth slowdown effects assumption that the current exchange-rate of the spillovers from the Asian crises. regime (which pre-set the rate of annual nomi- The ultimate effects of the Asian crises on nal depreciation of the currency at 7 percent) Latin America in 1998, and from 1999 onwards, will hold, thus the rate of inflation of tradeable will depend also on the relative importance of goods will remain low. In addition, the domestic increased Asian competitiveness. It is too early to prices of some Asian and commodity imports assess fully such effects, because there is consider- will (and have) decline(d). Moreover, the mone- able uncertainty on the level at which Asian cur- tary and fiscal contractions will contribute a fur- rencies will stabilize in real terms, and how soon ther decline in the rate of inflation of non-trad- their export sectors (now affected by reduced able prices. access to foreign finance) will react to their On the other extreme, Chile, Peru, and depreciated exchange rates. Venezuela, are the countries that may be the At this point, it is self evident that the effects most affected through real channels. All three of of the Asian crises on Latin America has been these countries are major exporters of com- quite different across countries. In terms of the modities, such as copper, fishmeal, and oil, that typology of channels of spillover effects presented are heavily imported by the affected Asian econ- in Figure 3, some Latin American economies suf- omies. Consequently, our projections for the fered financial spillovers, while others will suffer current account balances of these economies primarily from the real-side effects. have been revised downward. In the case of In the case of Brazil, we expect that financial Chile, the authorities have responded to the spillovers and the policy responses of the authori- changed international environment, and to the ties will dominate over real-side effects. As men- realization of very fast growth in aggregate tioned earlier, Brazil responded decisively to the demand during the last quarter of 1997, by rais- attack on the real through a combination of a ing interest rates. They will also probably reduce doubling of interest rates and fiscal adjustment, public spending, though not as by much as the which is currently being implemented. Conse- drop in fiscal revenues, as they have in place a quently,Table 3 shows that we now expect Brazil copper-price stabilization fund. Hence Chile's to experience a lower rate of economic growth, growth rate has also been revised downward. In lower inflation, and a lower current account addition, both Chile and Peru are expected to deficit than what we had projected before the suffer contractions in the foreign demand for Asian crises. Despite the significant fiscal mea- some of their non-traditional exports to Asia, sures, public sector borrowing requirements will also leading to lower growth rates than what had increase slightly in comparison to previous fore- been projected before the Asian crises. Peru is casts, due to the increase in interest payments, and also suffering the dramatic consequences of el the effect of the slowdown on fiscal revenues. It Nino, which has dampened investment, destroyed should be noted, however, that this set of projec- infrastructure, and lowered the rate of econonmc tions is highly sensitive to the evolution of inter- growth. Both countries will mitigate growth est rates, hence we provide a range of possible effects by allowing some depreciation of their growth outcomes. In contrast, real-side effects are currencies, which may increase inflationary pres- not expected to affect Brazil significantly, because sures by more than the beneficial effect of falling it is not a very open economy, and therefore suf- import prices.Venezuela will have to reduce its 30 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION non-oil fiscal deficit, just when oil revenues are the economy does not slowdown, perhaps due to falling sharply, and oil-related investments are a renewed inflow of capital since December, the also diminishing, thus reducing its rate of current account deficit may even surpass our growth. In the end,Venezuela's economic per- projection of 4 percent of GDP in 1998. As of formance will depend on the authorities policy now, we expect that growth will be reduced responses. from 8.4 percent in 1997, and a previous projec- Mexico and Colombia have an intermediate tion of 5.5 percent, to around 5 percent in 1998. situation. They are affected by the drop in oil At the same time, inflation will be less than pre- prices, though not as severely as Venezuela, as viously expected, due to falling import prices. they are both increasing their oil exports (a fact In summary, financial spillovers were that was not considered in our previous proj ec- stronger in countries with higher relative finan- tions for Mexico) and they represent a much cial vulnerability (especially in Brazil), as lower share of export and fiscal revenues. Mexico expected. Brazilian authorities responded vigor- has already responded by tightening fiscal expen- ously and timely, avoiding serious contagion at ditures and increasing oil taxes. Colombia will the expense of a significant slowdown in activity. probably also reduce expenditures, though not by In most other countries financial spillovers were as much as the fall in oil revenues (from previous short-lived and less intense, except for a general- projection) as it has an oil-price stabilization ized expected reduced access to international fund. Both countries have allowed some depreci- capital and higher spreads. ation of their currencies and have increased Trade-price effects were quite significant in moderately interest rates. Also, Mexico will prob- countries dependent on commodities for which ably suffer the strongest competitive impact of Asian countries were large and dynamic Asian devaluations (estimated in around one per- importers (Chile,Venezuela, and Peru being the cent of GDP), but will probably react to it by most affected; Mexico and Colombia more mod- allowing further depreciations. All in all, we erately affected). Trade-volume effects appear sig- expect in both countries slight increases in infla- nificant-though not large-in countries with tion rates, slight deterioration of current account higher Asian export-shares (Chile and Peru) or deficits (in Mexico with respect to 1997 levels, in with high export-shares in countries that suffer Colombia just with respect to previous projec- slowdowns (Argentina vis-a-vis Brazil). All coun- tions, as the sharp increase in the volume of oil tries will suffer some effects from the expected exports was already considered in previous pro- global economic slowdown, depending on their jections). The effect on growth in both countries degree of openness. Competitive effects will will be slight compared to previous projections. probably be more severe for Mexico (up to 1 Colombia will continue its recovery from the percent of GDP in 1999) than for other major slowdown in 1996, though at a more moderate Latin American countries, due to similarities in pace (4 percent instead of 4.5 percent, compared Mexico's export structure with those of the cri- to 3.2 percent in 1997) and Mexico will proba- sis-Asian economies. bly experience a sharper slowdown than previ- However, the net effects of the Asian crises ously expected (from 7 percent in 1997 to 4.8 depend critically on policy responses.The pres- percent in 1998). sures towards increased current account deficits Argentina is a unique case. Direct effects in most countries, coupled with reduced access have been and will be small. However, the effect to capital markets, have led authorities to mone- of the slowdown in Brazil will imply a higher tary and fiscal tightening, thus amplifying the than expected current account deficit and lower growth slowdown. Such effect is more severe in growth, as the currency board arrangement pre- countries with rigid exchange-rate systems cludes changes in the nominal exchange rate, and (Brazil and Argentina), which, on the other hand, the rise of interest rates between October and will benefit from reduced inflationary pressures. December will contribute to lower growth. If Countries with more flexible exchange-rate LESSONS FROM THE ASIAN CRISES FOPR LATIN AMERICA * 31 regimes have allowed and will continue to allow reducing the net growth slowdown, but at the some depreciation of their currencies, thus expense of a slight increase in inflation rates. L ING ARKS THE EXPERIENCE OF THE CRISIS-ASIAN economies highlights the importance of institutions (rules and their enforcement) in preventing the emergence of financial vulnerability, especially in the areas of financial and banking regulation, and corporate governance. Indeed, the main difference between past crises in Latin America and the events of 1997 is that in the latter financial vulnerability emerged primarily as a consequence of private behavior, rather than public macroeconomic policies. Regarding specific elements of the legal and regulatory frameworks, Latin America still has much to do in the areas of share- holders rights, bankruptcy laws, accounting standards (including consolidated balance sheets), information disclosure, and the like. Enforcement of prudential regulations also needs to improve, especially with regard to consolidated supervision, offshore on-site inspections, and supervision of off-balance sheet liabilities, particularly regarding the use of derivatives. In addition, countries should avoid offering exces- ting the development of deeper and more liq- sive deposit insurance, either explicitly or uid capital markets. implicitly. Most countries thus need to con- In addition, the region must not forget old tinue the institutional reforms in the financial policy lessons, which were painfully learned as a sector that were initiated in recent years; they consequence of past currency crises experienced should also pay attention to issues related to by Latin American countries. Prudent fiscal and corporate governance, including accounting monetary policies are important for maintaining and disclosure standards for corporations, a stable macroeconomic environment. Policies which will have the added benefit of permit- should aim in particular to prevent prolonged 33 34 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION credit booms, sharp currency appreciations, high exchange-rate management may help to prevent current account deficits, and large build ups of excessive foreign currency exposure by the pri- short-term debt, public or private. Discouraging vate sector and facilitate the response to market short-term capital inflows, as Chile and Colom- pressures. Countries that wish to fight inflation bia have done, or at least not encouraging them, with an exchange-rate anchor may need to as some Asian countries did, may be part of the establish very strong institutional commitments repertoire of prudential policies. Finally, an addi- to this policy, such as a currency board, in order tional lesson that has been reinforced by the to enhance its credibility. events of 1997, is that some flexibility in APPENDIXES APPENDIX 1: STATISTICAL APPENDIX Table Al. Banks and Credit Ratings Number out of the ten largest domestically owned banks receiving Banks required Does a local credit- ratings in February 1997 from to seek credit rating rating agency exist? local agency1 international agency2 Hong Kong No No - 33 Korea No Yes 0 10 Singapore No No - All 6 Taiwan No Since May 1997 - 6 Indonesia No No - 10 Malaysia No Yes 74 2 Thailand No Yes 35 9 Argentina Yes Yes 105 10 Brazil No Yes na. 9 Chile Yes7 Yes 10 10 Colombia No8 No - 59 Mexico No Yes 5 10 Peru Yes Yes 10 (6) Venezuela No Yes All 7 Source: Bank for nternational Settlements (1997, p.116). 1 For local currency deposits. 2 For long-term foreign currency debt. f only available for loca currency deposits, this is shown in brackets, Domestically owned banks only. Five banks have received ratings for their long-term foreign currency debt, Banks receive ratings based on overall banking operations. Rated for their long-term local currency bonds; none rated for local currency deposits. 6 One bank whose liabilitites are guaranteed by the Federal Government is not required to seek a credit rating. At least two agencies must rate bank securities. B However, a rating is required if certain operations are undertaken (e.g. securitisation and bond issuance). 9 For issuance of ADRs and GDRs. 35 36 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION Table A2. Current Account Balances in Latin America and Asia, 1991-1997 (as percentages of GDP) 1991 1992 1993 1994 1995 1996 1997 Argentina -1.5 -3.7 -2.9 -3.7 -0.9 -1.4 -3.7 Brazil -0.4 1.6 -0.1 -0.3 -2.5 -3.3 -4.1 Chile 0.3 -1.6 -4.6 -1.2 0.2 -4.1 -4.0 Colombia 5.6 2.0 -4.4 -4.8 -5.4 -5.6 -5.9 Mexico -5.1 -6.7 -5.8 -7.0 -0.6 -0.6 -1.8 Peru -3.0 -4.5 -5.2 -5.3 -7.3 -5.8 -5.1 Venezuela 3.2 -6.2 -3.3 4.4 3.0 12.2 -6.9 China 3.5 1.5 -2.7 1.4 0.2 0.9 2.5 Hong Kong 7.1 5.7 7.4 1.6 -3.9 -1.3 -1.5 Indonesia -3.4 -2.2 -1.5 -1.7 -3.3 -3.3 -2.9 Korea -3.0 -1.5 0.1 -1.2 -2.0 -4.9 -2.9 Malaysia -8.8 -3.8 -4.8 -7.8 -10.0 -4.9 -5.8 Philippines -2.3 -1.6 -5.5 -4.6 -4.4 -4.7 -4.5 Singapore 11.2 11.3 7.4 17.1 16.9 15.0 14.0 Taiwan 6.7 3.8 3.0 2.6 1.9 5.2 4.2 Thailand -7.7 -5.6 -5.0 -5.6 -8.0 -7.9 -3.9 Source: Latin American countries: World Bank, World Development Indicators database and staff estimates; East Asian countr es: World Economic Outlook database, IMF's International Financial Statistics. and IMF staff estimates. Note: 1997 figures are preliminary. Table A3. Gross Private Capital Flows to Developing Countries in 1997 (US$ billions, monthly averages) Region Jan.-July Aug.-Oct. Jan-Oct. Nov. Dec. All developing countries 17.5 18.1 17.7 12.0 East Asia 6.3 5.1 5.9 2.5 East Asia 5 6.7 5.6 6.3 2.2 East Asia 5 excl. Korea 4.2 2.9 3.8 1.5 Latin America 7.3 8.9 7.8 4.2 Europe & Central Asia 2.2 2.5 2.3 3.0 Source: World Bank, GEP 1998 (forthcoming), LESSONS FROM THE ASIAN CRISES FOR LATIN AMERI CA * 37 Figure Al. Real Effective Exchange Rates, 1992-1997 (Jan. 1992=100) Asia 180 - Nov. '94 June'97 160_ ° 140_ Cd a ~~~~~~~~~~~~~ ~ ~~~~~Philippines /\ c7 120_ \ -. .+ co a) xx -W100_ _a)l Korea Indone ia\i Thailand 180- 60- C\j C\j c l C ,) Ct Cl) C0 't It "zi LO US) LO LO co CD CO C r r ou au a o) ~C~ C o CD 0) a CD CP 0) CD CP 0~c) CD ) q)0 CD l G) Cz C o I 1997APR 1996JUL . 1996 JUL Z 1997OCT . g . \ E 1996OCT | H CD~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~S 1997 JAN~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 3 1997 JAN ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1997 JAN 1991 APR ~~~~~~~~~~~~1997 APR 1997JUL - I ~~~~~~~~~~~~1997 JUL . 19970OCT _______________ 97C __ __ Merchandise Export Indexes (Jan. 1992=100) Merchandise Export Indexes (Jan. 1992=100) E CD -~~~~~~~~ -' 0 -~~~~~~~~~~~ C.]DNDN COD < @~~~~~~~~~~~~~~~~~~~~~0 CD O n o cn CD l CD cn CD ol CDn oD CD I 01 C 0 CD CD 01 0D 0 0CD CD CDC CD 1992 JAN I I 1992 JAN Iz o 1 1992 APR 1992 APR 0 . ,w 1992 JUL 1992 JUL > u, 1992 OCT 1992OCT 51 1993JAN : 1993 JAN (\D C o CD ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.~~~~~~~( % DC I F[\1993APR 1993APR : im 1993 JUL . 1993 JUL . :. CD 1993 OCT 1993 OCT |, 1994 JAN 1994 JAN = 1994 APR 1994 APR 'o DC 1994 JUL \ 1994 JUL U)0 1994 OCT 1994 OCT o , .c 1995 JAN 1995 JAN CDB 1995 APR 1995 APR . CD~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C 1995 JUL 1995 JUL CD C: 19950CT (D 19950CT 1996 JAN 1996 JAN 1996 APR . 1996 APR 1996 JUL 1996JUL 19960CT ( I 19960CT . CD~~~~~~~~~~~~~~~~~~C 1997 JAN . '1T 1997 JAN . . Co C 1997 APR v_D 1997 APR 'D 1997 JUL 1997 JUL .. 1997 OCT 1997 OCT Indexes (Dec. 1993=100) Indexes (Dec. 1993=100) (ThCD N) N) CO -~~~~~~ -N ND ND co CD CDO o C) C o oO CD > o 1992JAN I _ I L I I 1992JAN I __ I _ I _ 1992 APR 1992 APR 1992JUL 1992JUL 0 1992 OCT 1992 OCT 1993 JAN 1993 JAN °. 2- 1993 APR 1993 APR (o Cr Z~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C D2 O1993 JUL I 1993 JUL > ~- 1993 OCT 1993 OCT . 1994 JAN 1994 JAN co 4 < 1994APR 1994APR X > 1994JUL 1994 JUL iz 1994 OCT . 1994 OCT D> -~~~~~~~~~~~~~~~~~~~ I ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ 1995 JAN . &.19JN. t z 1995 APR . ).1995 APRI ' 1995JUL y 0 . ~~~~~~~~ ~~1995 JUL |\8 19954OCT 1995 OCT . 1996JAN . 1996JAN . 1996 APR I;I, 1996 APR C\ - 1996 JUL 1996 JUL C 1996 OCT 1996 OCT 1997 JAN 1997 JAN Cw 1997 APR 1997 APR Cc 1997 JUL C 1997 JUL 1997 OCT 1997 OCT . I -0~~~~~~~~~~~~1 0 9 Indexes (Dec. 1993=100) Indexes (Dec. 1993=100) CThCD N\) N) co- - ) ) 0 3- o ° ° @° ° m° ° o o o Oo °o m\o o Co_ C) C) (31 C) (31 C) (31~~~~~~~~~~~~U CD 3 CD 0 CD ) C ) C C) ) C C) CD CD CD CD CD CZ)- o 1992JAN 1992 JAN z 0 Ln 1992APR 1992APR . . 1992JUL 1992JUL > 0 n 51992 OCT 1992 0CT 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1993 JAN 1993 JAN o 1993APR 1993APR 0 m 1993JUL 1993JUL -'o 1993 OCT.. 1993 OCT . 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ DC > 1994JAN , 8 199JAN . - .@ >71994 JUL /.1994 JUL. .z _ 1994 ACT 1994 ACT 0 ,o c 1995 JAN 1995JAN4 1995 APR |.'1995 APR . ,, 1995JUL 1995JUL m 1995 OCT 1995 OCT 1996 JAN 1996JAN J c (D E, ' : x,' I: ; 1996 APR (D 199APRo 1996 JUL tD | 1\7-; ,(D 1996 JUL CD~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C 1996 OCT: 1996 OCT.z 1997 JAN 0 1997 JAN (D/> 1997 APR l 7 ' 1997 APR : ; 1997 JUL D199 JUL 1997 OCT 1997 OCT .~~~~~~~~~~~~~~~~3 H , 'o" Indexes (June 30, 1992=100) Indexes (June 30, 1992=100) (D ND CD 4~ Cl C) -J CO (0 N) CD) -P- (" 0) -1- 0) CO C) C) C) CD CD CD CD CD C) C) D CD) C) CD C) C) CD CD CD C) CD ) C) CD CD CD CD CD CD (D C) C CD C) CD C) CD CD CD CD /1/90 [- ! 1/1/90 1 __ _ 1 , 1-I -. 3 4/17/90 0 f 8/1/90 D 8/1/90 (D D 611/15/90 - f 11/15/90 IDI 3/1/91 3- 3/1/91 oD 6/17/92 6/17/91 0 10191 f )1/31/91 _ O 3 10/1/92 10/1/92 S 14/30/92 4 4/30/92 6 / 3 / 9 f _ _ _ _ _ _ _ _ f_ 3 6 / 3 / 9 o CD ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~(D 8/14/92 8/14/92 CO 11/30/92 11/30/92 3/16/93 3/16/93 6/30/93 6/30/93 ff 10/14/93 10/14/93 f 1/28/94 1 1/28/94 5/16/94 5 5/16/94 f 8/30/94 8/30/94 12/14/94 12/14/94 -f 3/30/95 3/30/95 0 7/14/95 7/14/95 10/30/95 10/30/95 2/13/96 I2/13/96 ~ z 5/29/96 5/29/96 f 9/12/96 9/12/96 12,/27/96 12/27/96 - 4 4/14/97 I4/1 4/97 c ( 7'/29/97 7/29/97 11/12/97 I _______ ___11 / 12/97 U 44 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION Figure A4. (cont.) Stock Market Prices in US$, 1990-1998 Asia 900 June'92 Jan.'94 June '97 800 - . 700 - . . . . 600 - c'J OD 0 500 - co) 53 400 - ox 300 z JW . ~~~~~~~~~~~~~~~~Hong Kongv -20 00 CD CD C( 0 C (0 \l C\l (0 (0 CO (0 (0 ( (0 L() ( 0 (0 D CO (0 (0 r( (0 (0 CD~~~~~~~~0 ) O) 0) O) 0) 0) O- a)) a) cn ou 0) q m CD) O)) CZ 0) arO a n)ol 0 (- =0 i0 LD O (O (0 O ( OD (0 Ot (0 t (0 (0 ( : r- t (0) ( Z( co 0 ) ,- ( z- Q- (0 ,- (03 :- ( (0 2 - (0 \- \0 ,It ., (0 o- O = - O = Source: Datastream, Banker's Trust Economics and Strategy Group. Indexes (June 30,1992=100) Indexes (June 30,1992=100) D 1/1/90 g 01 0 1, 01 -4 01 1/1/9 C 1 ,- -4 11 01 -4 C1 0 ~3 4/17/90 4 . ± - 1/17/90 0 , t ____o Xw 8/1/90 b ¢ 1i } 08/1/90 . @-11/15/90 .. 11/15/90 0 O B 3/1/91 3/1/91 6/17/91 6/17/91 10/1/91 10/1/91 }DZ C 1/15/92 1/15/92 0 , 4/30/92 _ s 4/30/92 D t tCe -H~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~( :) 81 4/92 . 8/14/92 i ' o G)11/30/92 11/30/92 5.*1|00|:0 C 3/16/93 3/16/93 6/30/93 r 6/30/93 , 10/14/93 10/14/93 'D 1/28/94 1/28/94 3 8 /30/92 ___ ___ __ |__ __ 4/30/92 ___ ___ ___ - _ ___ ___ _ I _ 12/14/94~~~~_ _ ._-_ _ 1/494W0 _ _ _ _ + 0 8/14/92 8/14/92 4 I 0I 81/30/92 11/30/92 12/16/93 32/16/93 6/30/93 6/30/93 5 1/14/93 10/14/93 80/30/94 80/30/945 .-J 12/14/94 12/14/946 D 3/29/95 3/30/95 7/14/95 7/14/95 10/095 10/30/95 2/13/96 4 /13/96 5/29/96 5/29/967 11/12/97 11/12/97 Un 46 * FINANCIAL VULNERABILITY, SPILLOVER EFFECTS, AND CONTAGION Figure A4. (cont.) Stock Market Prices in US$, 1990-1998 Latin America 900_ -Jan. '94 June'97 800 - 800 700 - Coj co CD) T QQ Co 1 00 0) O~~~~~~~~~\ OI O" O NN CID CID 't < t ';l U LO) L O CD C!:) (D ~Or- r- r- CD CDO C5 O> C a) ou or o or ) or O cnor s nC uo ) o,0, ou0 Om ou a ZZ- =4 L21 DOtO OOtO D t ON>tO -~~~~~~~ cD 4. cD N. Cr) ID0 NNN CO CO CN D O L:7 -_ co CD 1z -2 z-02~Z~2 Nz 2LS O): N~ zz 1- 7 100rce: Datastream, Banker's Trust -cnmc and Statg Group. Source: Dtastream Banker' TrList Eonomics nd Stratgyenrona Indexes in U.S. $ (July 1, 1997=100) Indexes in U.S. $ (July 1, 1997=100) U g C °D C C° g Ogg 1/1/96 . 1/1/96 '- 1/29/96 1 1/29/96 CO C 2/26/96 2/26/96 3/25/96 3/25/96 0) 2-- 4/22/96 t K 4/22/96 e 0 ' 5/20/96 .'715/2096 < } (9 m~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0)0 6/17/96 6/17/96 j a 07/15/96 7/15/96 .i 8/12/96 8/12/96 ia 9/9/96 9/9/96 10/7/96 10/7/96 11/4/96 11/4/96 co 12/2/96 12/2/96 12/30/96 i 12/30/96 1/27/97 R .,0 ; i , 1/27/97 I 2/24/97 - 2/24/97 ,| 3/24/97 3/24/97 4/21/97 4/21/97 5/19/97 5/19/97 0 6/16/97 6/16/97 7/14/97 7/14/97 8/11/97 8/11/97- 9/8/97 19/8/97 10/6/97 10/6/97 '1 11/3/97 11/3/97 12/1/97 12/1/97 12/29/97 $ 12/29/97 1/26/98 . ____ ______ 1/26/98 C) U~~~~~~~~~C (3 0) NI c) C) Cq Co) N C) C n 0 1992JAN I I 1992 JAN / 1992APR | 1992 APR Z 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ , 1992JUL K 1992JUL L _- 1992OCT 1992OCT 1993JAN 1993JAN 1993APR 1993APR .. 1993JUL 1993JUL ... 1993OCT 0 1993OCT . 1994JAN 0 1994JAN / _ _ 1994APR 1994APR 1994 JUL 1994JUL 1994 OCT 1994 OCT . 1995 JAN 1995JAN /:F 1995 APR 1995APR 1995 JUL 1995 OCT 0 1995 OCT . 1996 JAN 1996 JAN 0 1996 APR 1996 APR 1996 JUL 1996 JUL 1996 OCT 1996 OCT 1997 JAN 0 1997 JAN 1997 APR 1997 APR 1997JUL 1997| J U L 1997 OCT 0 1997 OCT p cm p ~~~~CD CD CD9 CD CD C 3 N - ) CDo \ D C D - ND (P CC (1 cnl °P C) l (7 D C 01 O C cn c m ( P * 1992JAN I 1992JAN I- I _ ! o- 1992APR 1992APR o , 1992JUL 1992JUL , U o 1992OCT . 1992OCT . > 1993JAN 1993 JAN 0 1993APR 1993APR o CT~~~~~~~~~~~~~~~, R 1993JUL g 1993JUL , CD 1993 OCT . X S \a. &',1993OCT / - 1994JAN 1994JAN CD > 1994APR , 1994APR 'I' 0 C iT hi O 1994JUL |D 1994JUL L PD 1994 0CT 1994 0CT 0 1995JAN 1995JAN z 1995 APR 1995APR 1995 JUL . 1995JUL Cd 1995 OCT 1995 OCT 1996 JAN 1996 JAN . 1996APR 1996APR 1996 JUL 1996JUL . 1996OCT : 1996 OCT 1997JAN 1997JAN 0)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r 1997 APR 1997 APR . . - c 1997JUL 1997JUL ,o 1997 OCT 1997 OCT . . I B , S . t ' j . wWo~~~~~~~~~~~~~~~~~~~~~~~~~~~~~( 50 * FINANCIAL VULNERABILITY, SPILLOVE R EFFECTS, AND CONTAGION Figure A7. Volatility of Stock Market Prices (moving 20-day variation coefficient) Asia Indonesia 0.30 - Nov. 30, '94 July 1, '97 0.25- 0.20 0.10 0.05 (C O (c (c ` m t L O L( L( L( LO (C (C ( co co (c (CD r- rF- r- r- r- n- co (c C (C C) ( (C (C ( O C (CO CD co (C C () CA (C () LC (C r- (C (c L (C C- -C (C -C ce (C( ( C c) ( ) -C ' (C C (C (C m (C (C (C (C m- co O C C t- (D C O 0O CO C L r- ( C Malaysia 0.30 - 0.25 - - 0.20- 0.15 ---- - ---- ------- -- ------- co CO Co I t It 't (t L( Uh- L( (C (C (C (c (C Co (C (c (D > (C5 (C (C (C OD (Cl (Cl (Cl (C (C N - CD CO O (C (C (C (C' (C r-- ~ , - - (C iZ (C (C 0 i - ( (C( C( FC C( C- - (C coC rr- (C co (C,(C CO ta Cm Bk Trust Eni aC n = LoStt eg ) rCO up. Source: Datastream, Banker's Trust Economics and Strategy Group. LESSONS FROM THE AsIAN CRISES FoR LATIN AMERICA 51 Figure A7. (cont.) Volatility of Stock Market Prices (moving 20-day variation coefficient) Asia Korea Nov. 30, '94 July 1, '97 0.30 - 0.25 - - 0,15- 0.10 - ~-- ~ - -- ------ ---- co CO CO It 'I tIt ILO IL) ILO ILO ILO ILO 04 0c 04 Q0 CD 0 - 0- 0- 0- 0-0- CD 0l- co 0) (0) - Q0 CO 0J 04 0- (04 0) cj 04 Q0 CD 0) 04d 04 IL) ~ CO 04i 04) IL) "Z 0404040404, Z Z - :E - z - R----040 0404 04, 2 -. : ,- :.:- Z - 2- Thailand 0.30 - _ _ _ _ _ _ _ _ _ _ _ _ 0.20- 0.15- 0.10 - ---------------. 0.05 - ~. 04i 04j 0\4 04j 041 z - zz -4 0 0 40 --- ~ ~ Source: Datastream, Banker's Trust Economies and Strategy Group. 502 * FI NANC IAL VULNE RAB ILITY, S PILL OVER E FFECTS, AND C ONTAG ION Figure A7. (cont.) Volatility of Stock Market Prices (moving 20-day variation coefficient) Latin America Argentina 0.30 - Nov. 30, '94 July 1, '97 0.25 0.20 --. 0.15- 0.00 Co co co ' 't t t C (C LO) (C LCO (C O L) (CD (O C C ( C >- - - C O- C C C- (C Co ,- -) C- (C -: -D - -C Co CD (C ( (C NO CD Co CO - z O> (C - ; - CD C -D C.D CD 00 CD DC CD CL C N- 0 - - L - oo ( CD CD ( coco(C~at CD ,-C - -C - rC Co = Chile 0.30 - ____________ ________________________________ 0.25- 0.20- 0.15- 0.10 Co Co Co t t (C (C (C LO LO LO (C (D (C (CO CD (C - C- C-- C-- C- r- Co (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C co (cCo (C r-- (c o C oC C C ( (C (C Co) CDi C (C co CDj (C C Co (C C\D CDl Nt (C C (C D C o (C C\l- (C CI (C Source: Datastream, Banker's Trust Economics and Strategy Group. LESSONS FROM THE AsIAN CRISES FoR, LATIN AMERICAU 53 Figure A7. (cont.) Volatility of Stock Market Prices (moving 20-day variation coefficient) Latin America Brazil1 0.30 - Nov. 30, '94 July 1,9] 0.25 - 0.20- - - - --------- 0.15 - 0 .00 - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ (C (C (C ~~~~~~~~~ ~ ~ ~ (C~L LO IC I) (CL (C) (c c (C (C (C (C n- r- r- h-- n-- n-- ( (C C C (0 CO CD (C (C (CI (CO( (C ( (C (CI (C ( (0Co (C (C (C LO (Cl (CI (C ( Lo 'IT 2-( C~ C n-( C ( C n- (C (C ( (C (C ( (C, (CL (C (C 7 (C ( (C (C Mexico 0.30- 0.25. 0.20 ----.-- 0.15- 0.10 (co (CO (C -t - - T T LO LO Lo Lo ( LO (C (C (C (C (C (C r- r(- h- r-- 0-- ( (CL (C- (C (C (C - 2-Lz 2.--. 2-. (C (C -, (C (C ~ -- - (co (C (C (C 'IT (C0 (C (CD (C - - (C0, (C F- (C - (C LCo Source: Datastream, Banker's Trust Economics and Strategy Group. 54 * FINANCIAL VULNERAB ILITY, SPILLOVER E FFE CTS, AND CONTAGI ON Figure A8. Ten-Year U.S. Treasury Bond Yield Rate Invasion of Kuwait Tequila Crisis Begins Thai Crisis C rash of '97 (.. .C ( - (C -- -C -- h- -C -0 - -- (C -C - (C (-( - n- n- ... C! O 5 _ ~ ~ ~ ~ ~- (C~ - - -- Source: Datastream, Banker'sTrust Economics and Strategy Group. Figure A9. Stock Price Indexes in Hong Kong, the U.S., Japan, and the U.K., July 1997-January 1998 120- 9, 0 8 60 50-_ r r r- r- r r r- n-r 0 - r o- r- n- - o- - r- 0- - 0- (o ( co OD (C (C (C (C (C (C (C (C (cr (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C (C ( 0i - (C) t (C (C (C (C ( (C (C 0- r- C NC (C (CD (DC t , (C (C r~~C- ,- (C (C N- NC 7 N o (C N- (C (Cl N N - ,- NC >~~~~\ 1\ ODj ODj Ol CD CID LO Z LO N. Q0 (00 Source: Datastream, Banker's Trust Economics and Strategy Group. APPENDIX 2: INVENTORY OF CONTROLS ON CAPITALTRANSACTIONS IN ASIA AND LATIN AMERICA PRIOR TO THE 1997 CRISES Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits Indonesia The purchase of shares is A foreign investment com- Resident entities, especially Banks are permitted to lend The limits are: (1) the average (4/30/97) limited to a maximum of pany may be established as nonbank private sectors, locally in foreign exchange, total NOP, both on and off bal- 49% of total shares issued a straight investment, may borrow from nonresi- subject to requirement that ance sheet, in a week must not by an individual company which means that 100% of dents; however, they have 80% of the foreign exceed 25% of the bank's capital; 0 listed on the Indonesian the shares may be owned to submit periodic reports exchange loans must be and (2) the average NOP off bal- z Stock Exchange. by a foreign citizen and/or to the Commercial Offshore provided for export activity. ance sheet in a week must not entities. However, some of Loan Team (COFLT) of the Banks can also purchase exceed 25% of the bank's capital. m the company's shares must Bank of Indonesia on their locally issued securities o be sold to an Indonesian borrowing. denominated in foreign citizen and/or entities exchange, subject to the through direct placement requirement that the secu- and/or indirectly through rities must be investment the domestic capital mar- grade and should not be > ket no later than 15 years issued by their groups. To after commencement of do this, banks should take commercial operations. into account other z exchange regulations, namely the regulation on C net open position (NOP) aimed at prudential control. See next column. Korea Nonresidents may freely Equity participation is pos- Foreign exchange banks There are sector-specific The limits are as follows: (1) o (3/31/97) acquire listed stocks up to sible by increasing the may borrow abroad. Bor- loan ceilings. overall overbought position, 15% O 5% individually and 20% amount invested in newly rowing by high-technology, of the total equity capital at the collectively of the local established or existing foreign-financed manufac- end of the previous month; (2) > number of shares issued. enterprises. Direct invest- turing companies is allowed overall oversold position, 10% of - For convertible bonds ment by means of mergers up to 100% of the foreign- the total equity capital at the end issued by small-and and acquisitions is not invested capital. Also, of the previous month or $20 mil- medium-sized enterprises, allowed. For the establish- maturity is limited to 3 years lion, whichever is greater; and (3) > they may acquire up to 5% ment and extension of a or less, and limitations are spot oversold position, 3% of the individually and 30% collec- domestic branch of a for- imposed on the use of total equity capital or $5 million, tively of the total value of eign enterprise, approval funds. Foreign borrowing whichever is greater. convertible bonds. For other from the MOFE is required repayable within 3 years is purchases, including public for financial institutions. governed by the Foreign bonds designated by the Exchange Act. The maxi- Securities Exchange Com- mum maturity period per- Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises a Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange z information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits > z mission, nonresidents mitted for deferred pay- n require prior approval from ments is 180 days for > the Ministry of Finance and imports of raw materials for Economy (MOFE). export production. Malaysia No exchange controls apply The following inward A resident is permitted to Authorized dealers and The criteria in determining (12/31/96) to a resident for the sale of investments require prior obtain total credit in foreign merchant banks are banks' net overnight foreign cur- z securities abroad, but approval from the Foreign currency of up to RM 5 allowed to lend n foreign rency open position limits are approval is required. Investment Committee rnillion from licensed currency to residents. based on a matrix that takes into > Approval is given if the pro- (FIC): (1) acquisition of arny banks, licensed merchant account their shareholders' ceeds are used to finance substantial fixed assets by banks, and nonresidents. funds and dealing capacity. domestic product ve activi- foreign interests: (2) acqui- Any larger amount would ties, particularly for pro- sition of assets or interests, require the prior approval of jects that generate foreign mergers, and takeovers of the Controller of Foreign exchange earning or save companies and businesses Exchange. Residents are on the future outflow of for- in Malaysia by any means not allowed to obtain loans eign exchange through the that will cause ownership in ringgit from nonresidents. productions of import sub- or control to pass to foreign r stitution goods. interests; (3) acquis tion of < 15% or more of the voting m power (equity interests) by P any foreign interest or asso- M ciated group or by a foreign interest in the aggregate of m 30% or more of the voting n power of a Malaysian com- pany or business; (4) con- trol of Malaysian compa- > nies and businesses z through any form of joint- venture agreement, man- agement agreement, or o technical assistance z arrangement. Philippines Registration with the None, As a generai rule, all for- The following foreign cur- Depository banks operating - (3/31/97) Bangko Sentral ng Pilipinas eign loans contracted by rency loans may be gr-anted FCDUs or Expanded Foreign Cur- (BSP) or designated custo- nonbank residents with by Foreign Currency rency Deposit Units (EFCDUs) dian bank is required. Reg- guarantees from the public Deposit Unions (FCDUs) of need to maintain full cover for istration is necessary only sector or from local com- commercial banks without their foreign currency liabilities it the source of the foreign mercial banks require prior prior BSP approval: (1) pri- at all times. For FCDUs, at least Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange inforrnation) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits exchange needed for cap - BSP approval or registra- vate sector loans, if they 70% of the said cover must be tal repatriation and remit- tion. Those not covered by are to be serviced using maintained in the same currency tance of div dends, profits guarantees are required to foreign exchange to be of the liability and up to 30% may and earnings which accrue be registered only to make obtained outside the bank- be denominated in other accept- thereon is purchased from them el gible for subse- ing system; (2) short term able foreign currencies. Long and the banking system. quent debt servicing using loans to financial institu- short foreign exchange positions foreign exchange from the tions for normal interbank of banks must not exceed 20% banking system. transactions; and (3) short and 10% respectively, of their 0 z term loans to commodity total unimpaired capital. Any and service exporters, and excess beyond the limit must be producers/manufacturers, settled daily. provided that the loan 0 proceeds are to be used to finance the import costs of goods and services neces- sary in the production of goods. Commercial banks > may also grant foreign currency loans to residents involving trade z transactions, n Thailand Foreign equity participation Foreign capital may be There is no restriction on Commercial lending to par- The limit imposed is between (12/31/96) is lim ted to 25% of the brought into the country these credits. Repayment of ticular industries denomi- -15% (for a negative balance) paid-up registered capital without restriction, but pro- financial credits to nonresi- nated in foreign currencies and 20% (for a positive balance). of locally incorporated ceeds must be surrendered dents can be made freely can be partially (50%) banks, finance companies, to authorized banks or as long as residents have included as foreign assets o credit finance companies, deposited in foreign cur- an obligation to pay to non- in order to recognize the and asset management rency accounts with autho- residents in foreign potential risk that banks companies. The combined rized banks in Thailand currency. may not be fully repaid as > share holdings of an individ- within 15 days of receipt. exchange rate risk H ual and his/her related fam- heightened. z ily members must not exceed 5% of a bank's paid > up registered capital and 10% of that of finance com- panies and credit foncier companies. Foreign equity > participation is limited to 49% for otherThai corporations. U' Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises U Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial m (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange z information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits > z Hong Kong No exchange control None. None. None. Authorized institutions are r (12/31/96) requirements are imposed required to report to the Hong > on capital receipts or pay- Kong Monetary Authority ments by residents or non- (HKMA) their foreign currency residents. A license or an positions (including options) authorization is required for monthly. Locally incorporated companies, whether incor- institutions are required to m porated in Hong Kong or report their consolidated foreign elsewhere, to conduct bank- currency positions. The aggre- > ing, insurarice, securities, gate net open position (calcu- :d and futures dealings. Oth- lated as the sum of net long/ erwise, all overseas compa- short positions of individual cur- nies are required only to rencies) should normally not register with the Compa- exceed 5% of the capital base of nies Registry within 1 the institution, and the net open month of establishing a position in any individual cur- business in Hong Kong. rency should not exceed 10% of the capital base. For subsidiaries o of foreign banks, where the par- ent consolidates the foreign m exchange risk on a global basis and there is adequate home t supervision, the HKMA may accept higher limits. For branches of foreign banks, the HKMA reviews and monitors the internal limits, which are usually set by their head offices and > home supervisory authorities. z Singapore None. None. None. None. None. (1 (12/31/96) 0 z Argentina None. None. None. None. None. (12/31/96) 0 Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits Brazil The direct purchase of Investments in commercial The proceeds of financial All contracts, securities, or Limits differ according to the (5/31/97) shares of Brazilian compa- banks are limited to 30% of credits granted to residents other documents, as well as exchange market in which the nies by nonresidents basi- the voting capital, if there must be kept within the any obligations executable transactions take place as fol- cally occurs through direct are restrictions on the oper- country, and the resources in Brazil that require pay- lows: (1) banks authorized to investments and portfolio ations of Brazilian banks in must be used for invest- ments in foreign currency, conduct foreign exchange opera- investments made by insti- the markets where their ment in economic activi- are null and void. Conse- tions in the free exchange rate tutional investors through main offices are located. ties. Exchange contracts quently, banks are prohib- market may hold long positions the managers of the The establishments in involving the entry of for- ited from granting foreign of up to US$5 million, including 0 respective portfolios. Braz I of new branches of eign exchange in connec- currency loans within all currencies and all of each Z Depository receipts (DRs) financial institutions domi- tion with borrowing are Brazil. However, this bank's branches. Amounts constitute another method ciled abroad is prohibited. subject to prior approval by restriction does not apply exceeding this ceiling must be of acquiring shares through Also, any increase in the the Central Bank of Brazil to the onlending of external deposited with the CBB in U.S. o stock exchanges. They pro- percentage of equity partic- (CBB). foreign currency loans. dollars. The ceiling on banks' vide a mechanism for the ipation in financial institu- short exchange position is con- placement of shares of tions headquartered in A 15% income tax rate is tingent upon each bank's Brazilian enterprises in the Brazd by natural or juridical levied on remittances of adjusted net worth; and (2) on international markets. persons resident or domi- interest and other income the floating exchanqe rate mar- > Earnings from variable- ciled abroad is prohibited, associated with foreign ket, the following ceilings have income investments are except for authorizations loan operations, except been set: a) for licensed banks, subject to a 10% income tax resulting from international when bilateral agreements the long exchange position is z withholding and those from agreements, from reciproc- to avoid dual taxation spec- US$1 million (any amount in fixed-income investments ity arrangements, or in the if y another rate or when the excess of this ceiling must be to a 15% income tax with- interest of the Brazilian borrower or lender is tax deposited with the CBB), and holding. Remittances of government as expressed exempt. the short exchange position is capital gains are subject to by presidential decree. contingent on the institution's income tax at a rate of 15%. The federal government, adjusted net worth; b) for states, municipalities, the licensed dealers (brokerage o federal district, and their firms; securities distributors; foundations and agencies, and credit, financing, and invest- as well as multilateral orga- ment enterprises), the ceiling on > nizations and foreign gov- the long exchange position is ernment agencies located US$500,000 and no short z abroad, are exempted. exchange position is allowed. Colombia The purchase of 10% or Up to 100% ownership in Yes (not specified). None. The limit is 20% of net worth. (12/31/96) more of the shares of a any sector of the economy, There are no regulations govern- Colombian financial institu- except in defense, waste ing the net foreign exchange tion requires the prior disposal, and real estate, is positions of exchange houses; approval of the Superinten- allowed, Special regimes they may sell their excess foreign dency of Banks. Foreign remain in effect in the holdings to authorized financial investments in the form of intermediaries because they do Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises l Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange z information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange pos tion limits > z placement of shares in a financial, petroleum, and not have access to the Banco de n fund established to make mining sectors. Ia Republica. > investments in the stock exchange and in debt papers issued by the finan- c cial sector are permitted. r Chile Nonresidents can invest in Capital contributions to Financial credits can be Banks are only permitted to The limit is 20% of capital and m (4/30/97) domest c securities in 3 new establishments or contracted with foreign grant foreign exchange reserves, including derivative > ways: shares in existing ones are banks and financial enti- credits associated with for- and spot instruments, foreign 1) acquisition by foreign subject to a 1-year mini- ties, subject to a 1 -year 30% eign trade, but may grant investment, assets and liabilities investment funds. Invest- mum holding period and a reserve requirement, a 4% loans or acquire securities issued abroad or denominated in ments cannot exceed 5% of minimum amount of tax on interest, and a stamp denominated or expressed foreign exchange. There is a for- social capital in any 1 com- US$10,000. Projects of sig- tax of 1.2%. in foreign exchange pro- eign financial investment ceiling pany and 10% of the funds' nificant size may be under- vided they remain within of 25% of the capital and total assets. Funds cannot taken; there is a minimum the open position limits, reserves of each bank, a maxi- invest more than 40% of holding period of 1 year, mum investment in subsidiaries their portfolios in equities and the investor enjoys and branches abroad of 20% of of the same holding period, favorable taxation treat- capital and reserves of domestic and all funds as a group ment with regard to the banks, and a requirement that may not hold more than choice between the general the balance of all acquisitions 25% of the equities of the income tax law or the guar- and sales of foreign exchange for same open society. Other anteed payment profit tax a bank be positive. requirements for Foreign of 42%. There is also guar- Investment Funds (FICEs) anteed access to the for- n are a 5-year minimum hold- mal exchange market for ing period, a profit tax of repatriation. 10%, and some portfolio > restrict ons that vary with the duration of the holding period; 2) purchase of fixed n income securities and equi- O ties subject to a reserve 7 requirement of 30% for 1 year, and a minimum hold- ing period requirement of 1 year and to the general o income tax law, Some equi- z ties can be acquired in the country and converted into American Depository Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits Receipts (ADRs) subject to the 30% reserve require- ment. The issuance of pri- mary ADRs is exempt from the above restrictions, but the issuers are subject to minimum international rat- ng requirements and there 0 is a minimum amount to be Z issued; and 3) loans can be used to purchase securities m in the country, subject to the 1 year reserve require- ment of 30%, a 4% tax on interest payments, and a 1.2% stamp tax. m Mexico None. Investments in credit There are limits on credits Banks are permitted to lend The liabilities of comrnercial (2/28/97) unions and development denominated in foreign locally in foreign currencies banks denominated in foreign banks are reserved currency. but the debt may be dis- currency must not exceed the z exclusively for Mexican charged by delivering the larger amount of either (1) 10% of nationals or Mexican corpo- equivalent amount in accor- the daily balances of their liabili- n rations with a foreign dance to the rate of ties in the previous quarter plus exclusion clause. Acquisi- exchange in domestic and additional 4% of such liabili- tion of more than 49% of currency. ties if those resources are used the equity in a Mexican cor- to finance foreign trade transac- poration requires prior tions or capital goods purchases, o authorization if the total or (2) 1.6 times the net capital of value of assets exceeds $2 the institution. Commercial million, Ceilings on foreign banks must balance their posi- > ownership are applied to tions in foreign currency on a e financial institutions. daily basis subject to a covered z exchange rate risk, irrespective of the residence of the lender. > Short and long positions are acceptable as long as they do not exceed 15% of the bank's net Cc capital. U ON Appendix 2. (cont.) Inventory of Controls on Capital Transactions in Asia and Latin America Prior to the 1997 Crises Controls on Foreign Ownership Controls on Credit Operations Country Commercial or financial (date of On capital market securities Foreign direct investment in credits to residents from Lending locally in Open foreign exchange z information) purchased by nonresidents financial/banking sectors nonresidents foreign exchange position limits z Peru None. None for financial/banking None. None. None. n (12/31/96) sectors. Venezuela Foreign investors are New investments do not None. Local banks can make The net position cannot exceed (1/31/97) allowed to purchase corpo- require prior authorization loans denominated in for- 25% of bank's capital. rate stocks in the Caracas from the SIEX but must be eign exchange. Stock Exchange but must registered with the SIEX z inform the Superintendency after the fact, and approval of Foreign Investment is automatically granted if > (SIEX) of such purchases at the new investment is con- the end of each calendar sistent with national year. legislation. S nf H Source: IMF (1997a)~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ LESS ON S FROM THE AS IAN C RP I SE S FOpR L AT IN AMEPR ICA * 63 APPENDIX 3: DATA DEFINITIONS AND SOURCES Table 1. International Bank Loans and Table A2. Current Account Balances in Foreign Currency Exposure Latin America and Asia, 1991-1997 Debt by maturity and sector are in billions of US The current account balances for the Latin $ and as a ratio of total reserves, where total American countries were obtained from the reserves are those prior to July 1997.Total World Development Indicators database of the reserves are the sum of the US $ value of mone- World Bank; 1997 estimates were provided by tary authorities' holdings of SDRs, reserve posi- World Bank staff. In the case of East Asian coun- tion in the Fund, and foreign exchange. Total tries, current account balances are from the reserves don't include gold. World Economic Outlook database and the IFS The source was the Bank for International of the IMF; 1997 estimates are from IMF staff. Settlements (1998). Table A3. Gross Private Capital Flows to Table 2. Regional Export Structure in Developing Countries in 1997 Latin America and Asia The source was the World Bank's Global Eco- Exports to developing Asia (includes China and nomic Prospects, 1998 issue (forthcoming.) India). The figures are percentages and were cal- culated by taking the ratio of each of the coun- tries' exports to developing Asia to the total Figure 2. Total Reserves minus Gold over value of exports. M2 in Asia and Latin America, 1992-1997 The source was the Direction of Trade Sta- Total reserves minus gold is the sum of the US $ tistics database of the IME Exports in US $ f.o.b. value of monetary authorities' holdings of SDRs, (line 70dzf). reserve position in the Fund, and foreign exchange. M2 or broad money is defined as money (the sum of currency outside banks and Table 3. Macroeconomic Projections for demand deposits other than those of the central Latin America before and after the Asian banks) plus quasi-money (time deposits, savings Crises deposits, and foreign currency deposits other The following macroeconomic indicators are than those of the central banks). included in this table: 1) GDP growth rate: pro- The source was the IFS database, IMETotal jections of annual percentage growth rates of reserves minus gold (line 1l.d); M2 (hne 351). GDP at market prices in local currency at con- stant prices; 2) inflation: projections of December to December change in consumer price indexes; Figure 4. Stock Price Indexes in Latin 3) current account balances: projections of cur- America and Crisis Asia, July 1997-January rent account balances as percentages of GDP; 1998 and 4) projections of the non-financial public The following stock price indexes were used: sector balances. The sources were World Bank staff Argentina Merval projections. Brazil Bovespa Chile General (IGPA) Colombia IBB Table Al. Banks and Credit Ratings Mexico IPC The source was the Bank for International Peru General Settlements (1997, p. 116). Venezuela IBC 64 * FI NAN C IAL VULNERAB ILI TY, SPI LL OVE R EFFE CTS, AND C O NTAG I ON Indonesia Jakarta Composite derive the real effective exchange rate is as Korea Korea Composite (KOSPI) follows: Malaysia Kuala Lumpur Composite Philippines Philippines Composite n Thailand Bangkok S.E.T. REERJ (CPI1 ERIk)/EXP (WT i*Log(CPI1*ERIi)) * 100 The local currency indexes were divided by i= their respective nominal exchange rates and then where, the indexes were converted to base July 1, 1997=100 by using the following formula: REER is the real effective exchange rate index. ERI is a nominal exchange rate index. It is com- PI = (Pi / PiBase) * 100 puted from a monthly time series of exchange where, rates expressed as US $ per domestic currency. The ERI series is the arithmetic average of PI is the stock price index rebased to July 1, the monthly rate of exchange between the 1997=100, national currency and the US $. PiT is the daily price index at day t, and CPI is the series of consumer price indexes. 'iBase is the price index of the day used as base: j is the index for reporting country July 1, 1997. n is the number of partner countries to j i is the index of partner country, i=1,...,n The source was Banker's Trust Economics and WT.. is the weight that country j attaches to Strategy Group, based on data from Datastream. country i. Currently, weights are based on trade flows averaged over 1980-82. Figure 5. Evolution of Nominal Exchange The source was the IFS database, IME Rates in Latin America, July 1, 1997- January 28, 1998 Nominal exchange rates were converted to base Figure A2. Merchandise Export Revenues July 1, 1997=100 by using the following formula: in US$ Merchandise exports f.o.b. in US $. They are e = (et / eBase) *100 customs statistics reported under the general where, trade system. A twelve-month moving average was calculated. The twelve-month moving aver- e is the nominal exchange rate with base July 1, age was then based to January 1992=100 and the 1997 =100, following formula was used: et is the nominal exchange rate at day t, and eBase is the nominal exchange rate of the day used ME = (MEt / MEBase) * 100 as base:July 1, 1997. where, The source was Banker's Trust Economics ME is the index of Merchandise exports with and Strategy Group, based on data from base Jan. 1992=100, Datastream. MEt is the merchandise export value of month t, and ME Base is the Merchandise export value for Janu- Figure Al. Real Effective Exchange Rates, ary 1992. 1992-1997 The methodology used by the Information The source was the IFS database, IMF Mer- Notice System (INS) department of the IMF to chandise exports f.o.b. in US $ (line 70dzf). LESSONS FROM THE ASIAN CRISES FOR LATIN AMERICA * 65 Figure A3. Claims on the Private Sector, The local currency indexes were divided by 1992-1997 the nominal exchange rate. In turn, the price Claims on the private sector includes those insti- indexes were converted to base June 30, tutions that accept transferable deposits and insti- 1992=100 by using the following formula: tutions that do not accept transferable deposits but that engage in financial intermediation. The PI = (PIn / PiBase) * 100 data for Argentina, Brazil, Chile, Colombia, Mex- ico, Peru,Venezuela, and Malaysia are consoli- where, dated claims on private sector of the whole banking system. In the case of China, Indonesia, PI is the stock market price index rebased to Hong Kong, Taiwan, and Thailand, claims on pri- June 30, 1992 = 100, vate sector refers to those of deposit money Pi is the daily stock price index at day t, and banks. And in the case of Singapore, Philippines, PIBase is the stock price index of the day used as and Korea, claims on private sector of deposit base: the base year for all the countries is money banks plus other banking institutions are June 30, 1992, except for Argentina and used. Venezuela where December 31, 1993 is the The source was the International Finan- base. cial Statistics (IFS) database, International Monetary Fund (IMF). Line 52d was used in The source for the data is Banker's Trust the case of the first group of countries above; Economics and Strategy Group, based on data line 22d was used in the case of the second from Datastream. group; and line 22d plus line 42d was used for the last group. Figure A5. Stock Market Prices in US$, 1996-98 Figure A4. Stock Market Prices in US$, In this figure, the price indexes of Hong Kong 1990-98 (Hang Seng), Indonesia (akarta Composite), The price indexes used for each of the countries Korea (Kospi), Malaysia (Kuala Lumpur Com- were the following: posite), Philippines (Philippines Composite), and Thailand (Bangkok S.E.T.) were rebased to July China Shanghai Composite '97= 100. Hong Kong Hang Seng Indonesia Jakarta Composite Korea Korea Composite (KOSPI) Figure A6. Net Foreign Assets over M2 Malaysia Kuala Lumpur Composite (Banking System) Philippines Philippines Composite Net foreign assets refers to the difference Singapore Singapore DBS 50 between foreign assets and foreign liabilities of Taiwan Taiwan SE weighted price deposit money banks. Net foreign assets were index then divided by M2 or broad money, which is Thailand Bangkok S.E.T. defined as money (the sum of currency out- side banks and demand deposits other than Argentina Merval those of central banks) plus quasi-money Brazil Bovespa (time deposits, savings deposits, and foreign Colombia SE (IBB) currency deposits other than those of central Chile General (IGPA) banks). Mexico IPC (Bolsa) The source was IFS database, IMF. Foreign Peru SE General (IGBL) assets (line 21); foreign liabilities (line 26c); and Venezuela SE General M2 (line 351). 66 * F I NAN C I AL VU LNE R AB I L I T Y, S P ILL OVER E F FE C T S , AND C O NTAG I O N Figure A7.Volatility of Stock Market Prices The local currency indexes were divided by A moving 20-day standard deviation and a mov- their respective nominal exchange rates (except ing 20-day average were calculated from the the U.S. S&P 500) and then the indexes were stock market price indexes. A variation coeffi- converted to base July 1, 1997= 100 by using the cient was then calculated by dividing the moving following formula: 20-day standard deviation by the moving 20-day average. PI = (Pit/ PiBase) * 100 The source of the stock price indexes was Banker's Trust Economics and Strategy Group, where, based on data from Datastream. PI is the stock price index rebased to July 1, 1997 =100, Figure A8. Ten Year US Treasury Bond PIt is the daily price index at day t, and Yield Rate PiBase is the price index of the day used as base: This is the yield rate of the ten-year U.S. July 1, 1997. Treasury bond. The source was Banker's Trust Economics The source was Banker's Trust Economics and Strategy Group, based on data from and Strategy Group, based on data from Datastream. Datastream. Figure A9. Stock Price Indexes in Hong Kong, the U.S., Japan, and the U.K., July 1997-January 1998 The following stock price indexes were used: Hong Kong Hang Seng Japan Nikkei U.S. S&P 500 U.K. FTSE 100 NOTES 1 On the relation among institutions, finan- 7 On the evolution of commodity prices see cial development, and growth see La Porta et al. World Bank (1998c). (1997) and Levine (1997). On the relation 'This is true in highly dollarized economies between institutions and growth see Knack and such as Bolivia, Paraguay, and Peru. The fact that Keefer (1995) and Keefer and Knack (1997). they have flexible exchange rates suggest, how- 2 For example, see Dornbusch et al. (1995) ever, that corporations may have been careful in and Sachs et al. (1996). taking unhedged exposures. 3 In fact, most Asian stock markets fell after 9 See Appendix 2. China's devaluation in January 1994, see Figure A4. 10 In the case of Argentina, as for other dol- 4Young and Kwon (1998) also use the term larized, rigid-exchange-rate economies such as "crony capitalism." Hong Kong and Singapore, the data on reserves 5 On the liberalization of the capital does not include the dollar deposits and reserves account and financial deregulation undertaken in held by subsidiaries of foreign banks. The debt the late 1980s and early 1990s by Indonesia and data does include the claims in foreign currency Thailand, see Bhattacharya and Pangestu (1997, made by these subsidiaries in the domestic mar- 419-420) and Kawai (1997, 8-9), respectively. ket. According to data from the Central Bank of 6 It must be recognized, however, that this Argentina, when we subtract the claims of the appreciation was not as pronounced as that expe- subsidiaries, the ratio of short-term claims to rienced by some Latin American countries prior reserves is cut in half. to the "tequila" crisis of December 1994. 67 REFERENCES Bank for International Settlements. 1997. 67th Annual Knack, Stephen, and Philip Keefer. 1995. "Institutions and Report. Basle, Switzerland: BIS. Economic Performance: Cross-Country Tests Using .1998. The Maturity, Sectoral and Nationality Distribu- Alternative Instituitional Measures." Economics and Poli- tion of International Bank Lending, First HaW 1997. Basle, tics 7(3):207-227. Switzerland: BIS. Krugman, Paul. 1998. "What Happened to Asia?" Mimeo- Bhattacharya, Amar, and Mari Pangestu. 1997. "Indonesia: graphed, Department of Economics, Massachusetts Development Transformation and the Role of Public Institute of Technology, Cambridge, Mass. [http://web. Policy." Chapter 7 in Lessons from East Asia, edited by D. mit.edu/krugman] Leipziger. Ann Arbor: The University of Michigan Press. La Porta, Rafael, Florencio L6pez-de-Silanes, Andrei Calvo, Guillermo. 1995. "Varieties of Capital Market Shleifer, and Robert W Vishny. 1997. "Legal Determi- Crises." Mimeographed, Center for International Eco- nants of External Finance."Journal of Finance 52: nomics, University of Maryland at College Park. 1131-1150. Dornbusch, Rudiger. 1998. "Asian Currency Crises." Levine, Ross. 1997. "Law, Finance, and Economic Mimeographed, Department of Economics, Massachu- Growth." Mimeographed. Department of Economics, setts Institute of Technology, Cambridge, Mass. University of Virginia, Charlottesville. Dornbusch, Rudiger, Ilan Goldfajn, and Rodrigo O. Sachs, Jeffrey D., Aaron Tornell, and Andres Velasco. 1996. Valdes. 1995. "Currency Crises and Collapses." Brookings "Financial Crises in Emerging Markets:The Lessons Papers on EconomicActivity 2: 219-270. from 1995." Brookings Papers on Economic Activity 1: International Monetary Fund. 1997a. Exchange Arrangements 147-215. and Exchange Restrictions Annual Report 1997. Washing- World Bank. 1998a. Global Development Finance. ton, DC: IME Forthcoming. .1997b. World Economic Outlook: Interim Assessment, . 1998b. Global Economic Prospects, 1998. Forthcoming. December 1997. Washington, DC: IME .1998c. Commodity Markets and the Developing Coun- Kawai, Masahiro. 1997. "East Asian Currency Turbulence: tries:A World Bank Business Quarterly. Washington, DC: Implications of Financial System Fragility." Mimeo- The World Bank, February. graphed, Institute of Social Science, University of Young, Soogil, and Jae-Jung Kwon. 1998. "Korean Econ- Tokyo,Japan. omy under the IMF Program." Mimeographed, Korea Keefer, Philip, and Stephen Knack. 1997. "Why Don't Poor Institute for International Economic Policy, Seoul, S. Countries Catch Up? A Cross-National Test of an Insti- Korea. tutional Explanation." Economic Inquiry 35: 590-602. 69 WORLD BANK LATIN AMERICAN AND CARIBBEAN STUDIES VIEWPOINTS SERIES Determinants of Crime Rates in Latin America and the World:An Empirical Assessment Latin America after Mexico: Quickening the Pace by Pablo Fajnzylber, Daniel Lederman, and Norman by Shahid Javed Burki and Sebastian Edwards Loayza forthcoming Poverty, Inequality, and Human Capital Development in Latin America, 1950-2025 Beyond the Washington Consensus: Institutions Matter by Juan Luis Londofio by Shahid Javed Burki and Guillermo E. Perry available in English and Spanish forthcoming in English and Spanish Dismantling the Populist State: the Unfinished Revolution in Latin America and the Caribbean PROCEEDINGS SERIES by Shahid Javed Burki and Sebastian Edwards Currency Boards and External Shocks: How Much Pain, Decentralization in Latin America: Learning through How Much Gain? Experience Edited by Guillermo Perry by George E. Peterson Annual World Bank Conference on Development in Latin Urban Poverty and Violence inJamaica America and the Caribbean 1995:The Challenges of by Caroline Moser and Jeremy Holland Development available in English and Spanish Edited by Shahid Javed Burki, Sebastian Edwards, and Sri-Ram Aiyer Prospects and Challenges for the Caribbean by Steven B. Webb Annual World Bank Conference on Development in Latin America and the Caribbean 1996: Poverty and Inequality Black December: Banking Instability, the Mexican Crisis Edited by Shahid Javed Burki, Sebastian Edwards, and and Its Effect on Argentina Sri-Ram Aiyer byValeriano Garcia Annual World Bank Conference on Development in Latin The Long March: A Reform Agenda for Latin America and America and the Caribbean 1997:Trade: Towards Open the Caribbean in the iNext Decade Regionalism by Shahid Javed Burki and Guillermo E. Perry Edited by Shahid Javed Burki, Guillermo E. Perry, forthcoming in Spanish and Sara Calvo forthcoming Dealing with Public Risk in Private Infrastructure Edited by Timothy Irwin, Michael Klein, Guillermo E. Perry, and Mateen Thobani Crime and Violence as Development Issues in Latin America and the Caribbean by Robert L. Ayres Financial Vulnerability: Spillover Effects, and Contagion: Lessons from the Asian Crises for Latin America by Guillermo E. Perry and Daniel Lederman THE WORLD BANK 18181I Sti-cL .N .X Vashiliigton, I).C. 20433 'S.A 'I'ctlepilonc: 2(02-477-1234 FIcsinilel: 202-477-639 1 Tclex: XICI 64145 \\ ()ORII)13A.Nk ,M(CI 248423 WOL)I ANK NVorld Uid. U.'h: hittp:!/www N .orldbhnk.orig/ FI-mail: books((t % orldbhnk.or.- ISBN 0-8213-4228-2