175(.'4 v.l Moard, jqq s Glolgal~Y SI Development m"lance . | 199 A WD S UMMA RY _, TAB LE S 1 9 98 A W O R L DB A N K B O O K Cl Ma co P. =0 Global Development Finance 1998 ANALYSIS AND SUMMARY TA B L E S The World Bank Washington, D.C. Copyright © 1998 by the International Bank for Reconstruction and Development/The World Bank 1818 H Street, N.W, Washington, D.C. 20433, USA All rights reserved Manufactured in the United States of America First printing March 1998 Global Development Finance was formerly published under the title World Debt Tables. This publication has been compiled by the staff of the Development Prospects Group of the World Bank's Development Economics Vice Presidency. The World Bank does not accept responsibility for the accuracy or completeness of this publication. Any judg- ments expressed are those of World Bank staff or consultants and do not necessarily reflect the views of the Board of Executive Directors or the governments they represent. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when reproduction is for noncommercial purposes, without asking a fee. Permission to photocopy portions for classroom use is granted through the Copyright Clearance Center, Suite 910, 222 Rosewood Dr., Danvers, MA 01923, USA. ISBN 0-8213-4117-0 ISSN 1020-5454 Contents Preface ix Acronyms and abbreviations x Part 1. External finance for developing countries Overview 3 1. Development in private capital flows 7 Developments in capital markets in 1997 8 Spreads narrow, then widen 13 Increasing financial integration of emerging markets 15 Foreign direct investment has surged in the 1990s 19 Annex Private capital flows and domestic bond markets 24 Notes 25 2. East Asia's financial crisis: causes, evolution, and prospects 29 The crisis and its causes 30 The response of the international community 43 Prospects for private capital flows in 1998 46 Notes 48 3. The changing role of official finance 49 Shifting focus of aid 49 Progress under the HIPC Debt Initiative 54 Growth and innovation in official guarantees 57 Notes 61 Part I. Appendixes 1. Debt burden indicators and country classifications 65 2. Official external debt restructuring 79 3. Commercial debt restructuring 83 4. Progress on privatization 103 5. Trends in flows from international capital markets 111 6. Regional trends in debt and flows 121 References 137 v vi Part Ill. Summary tables Methodology 143 Sources and definitions 147 Country groups 155 Summary tables 159 All developing countries 160 East Asia and the Pacific 162 Europe and Central Asia 164 Latin America and the Caribbean 166 Middle East and North Africa 168 South Asia 170 Sub-Saharan Africa 172 Severely indebted low-income countries 174 Severely indebted middle-income countries 176 Moderately indebted low-income countries 178 Moderately indebted middle-income countries 180 Other developing countries 182 Low-income countries 184 Middle-income countries 186 Special Program of Assistance 188 Tables 1 Net long-term resource flows to developing countries, 1990-97 3 1.1 The international economic environment, 1981-97 8 1.2 Net long-term resource flows to developing countries, 1990-97 9 1.3 Developing countries' short-term debt stocks, 1994-97 9 1.4 Bond issues and loan commitments to developing countries, 1996 and 1997 11 1.5 Changes in stock market and exchange rate indexes, 1997 12 1.6 The rise of emerging market hedge funds, 1992-97 17 1.7 Performance of stock market indexes and mutual funds, 1992-97 18 1.8 FDI flows to the top ten recipient developing countries, 1991, 1994, and 1997 20 1.9 FDI flows to developing countries by income group, 1990-97 20 1.10 FDI in selected developing countries, by sector 21 1.11 FDI outflows from developing countries, 1991-96 23 lA.1 Domestic bond markets in developing countries 25 2.1 Fiscal balances and gross national savings in East Asian countries, 1993-96 30 2.2 Net private capital flows to East Asia, 1994-96 31 2.3 Ratios of short-term debt to total debt and to reserves in East Asia and Latin America, mid- 1997 35 2.4 Real effective exchange rate indexes, December 1996-December 1997 35 2.5 Credit growth in various countries, 1990-96 37 2.6 Macroeconomic indicators in Southeast Asia, 1993-96 38 2.7 Credit ratings for East Asian countries, 1996-97 39 2.8 Composition of private flows, 1990-97 46 2.9 Private flows to Argentina and Mexico, 1994-96 47 vii 2.10 Current account deficits, changes in reserves, and long-term amortization among major developing country borrowers, 1997-98 48 3.1 Net official flows of development finance, 1990-97 50 3.2 Recipient's share of official development assistance by policy performance, 1990 and 1995 53 3.3 Heavily indebted poor countries relative to all developing countries, 1996 54 3.4 Economic performance of the heavily indebted poor countries, 1980-95 55 3.5 Countries with a decision point in 1997 under the HIPC Debt Initiative 57 Figures 1.1 Net flows from international capital markets to developing countries were stable in 1997 11 1.2 Spreads on Brady bonds shrank, then expanded 13 1.3 Secondary market spreads on international bonds decline and rise 14 1.4 Countries in all regions were affected by rising secondary market spreads on sovereign bond issues, October-December 1997 15 1.5 Capitali7ation of emerging equity markets keeps growing, 1986-97 17 1.6 Both the amount and the share of global FDI to developing countries are rising, 1991-97 19 1.7 FDI flows remained strong in most regions in 1997 21 1.8 During the 1990s privatization projects have attracted substantial foreign investment 21 1.9 Service's share of U.S. and Japanese FDI to developing countries is large and growing, 1985 and 1995 22 1.10 FDI inflows are positively correlated with GDP growth 23 2.1 East Asia's debt-export ratios were lower than Latin America's in 1996 33 2.2 East Asia's real effective exchange rates have appreciated in recent years 35 2.3 As dollar prices fell, East Asia's exports slowed 35 3.1 Net ODA from DAC countries has slipped to its lowest level since 1970... 50 3.2 ... .and net ODA from industrial countries has shown a steep decline since the mid-1980s 51 3.3 Export credit agencies' new commitments to developing countries have shown strong growth in the 1990s 58 3.4 Export credit agencies' exposure in developing economies remained concentrated in 1996 59 3.5 The financial performance of export credit agencies strengthens, bringing a surplus in 1996 59 3.6 Growth in investment insurance has paralleled the surge in foreign direct investment in developing countries 60 Boxes 1.1 Measuring resource flows 10 1.2 What are derivatives? 16 1.3 Multilateral Agreement on Investment and developing countries 24 lA.1 Why is development of the bond market desirable? 26 1A.2 Considerations in foreign sovereign borrowing 27 2.1 Warning signs of a currency crisis 32 2.2 The quality of investment in crisis countries 34 2.3 Capital controls in Chile and Colombia 42 2.4 Initiatives of the Group of Ten and the Basle Committee on Banking Supervision 45 3.1 What are sustainable debt levels? 55 3.2 World Bank Group guarantees in support of private sector growth 61 Preface Global Development Finance was formerly pub- between debt stock and its components; the com- lished as WorldDebt Tables. The new name reflects putation of net flows, aggregate net resource flows, the report's expanded scope and greater coverage and aggregate net transfers; and the relation of private financial flows. between net resource flows and the balance of pay- GlobalDevelopmentFinance consists of two vol- ments (a discussion of the relation between net umes: Analysis and Summary Tables and Country resource flows and balance of payments is also pro- Tables. Analysis and Suimmary Tables contains vided in box 1.1 in Analysis and Summary Tables). analysis and commentary on recent developments Exact definitions of these and other terms used in in international finance for developing countries, Global Development Finance are found in the with particular focus on the East Asian financial Sources and Definitions section. crisis. Summary statistical tables are included for The economic aggregates presented in the selected regional and analytical groups comprising tables are prepared for the convenience of users; 150 developing countries. their inclusion is not an endorsement of their value Country Tables contains statistical tables on the for economic analysis. Although debt indictors external debt of the 138 countries that report pub- can give useful information about developments lic and publicly guaranteed debt under the Debtor in debt-servicing capacity, conclusions drawn Reporting System (DRS). Also included are tables from them will not be valid unless accompanied of selected debt and resource flow statistics for by careful economic evaluation. The macroeco- individual reporting countries as well as summary nomic information provided is from standard tables for regional and income groups. Reported sources, but many of them are subject to consid- for the first time are the external debt obligations erable margins of error, and the usual care must be of Eritrea and South Africa. taken in interpreting the indicators. This is partic- For the convenience of readers, charts on pages ularly true for the most recent year or two, when xi to xiii summarize graphically the relation figures are preliminary and subject to revision. This reportwas pieared by a team led by WiDla Shaw,and FredKiby under the direction of Ur- Dadush. Team mernbers iludAbayOmiAwde Mlan Bamalt; ipk Dasgupta% Ashok Dhareshwar, oPesa Economa, aon Lee F uan,r Kmikoli; Ke Iakura, Jalal Jall, Ronad Johanns, Kwang Jn. Himmat Kai, un Ju Kim: :L NgaLi Robert Lynn, : Carmini Michelira Musrapha Nab Mara Pokt Michael Pomerleano, cRiordan, Shatalov, TG. Sni&asan,Abd S i Manuel Truco, and Chiaki Yamamoto. Tih team ben- efited fiom comments from Joe ph Stigit Masood Ahmed, and Amatendra Bhattacharya, com- ments from John Williamon and Jed Shilling,; discussants at the Bankwide review; a sunmmary of -the IPC Debt Inittive by Marc Stepes andJefiy Katz, assistance from the regional chief econ- onists :and input from numerou a saff The chapter on East Asia also draws on the forth-_ comuiigAsian:Developm=ent Bank-World Bank volume"Managing Finanial Inr Asia -Th teamwas: assisted by Sa Cbrow Chi Ezenwa 'Sheilab King-Watson, Margarita Ctz, and Jac-eyn Queen. a quereaumoent and Paul Hoitz were the principal editors. The volume was laid outsby Garrett Cruce and Glenn McGrAth. Dtaphe Levitas assisted in project management. The work was arde ou er h o management of MasoodAmed and Joseph Stiglitz ix Acronyms and g abbreviations ASEAN-4 Countries most affected by the currency crisis (Indonesia, Malaysia, Philippines, Thailand) BIS Bank for International Settlements CFA Communaute financiere africaine (franc zone) CRS Creditor Reporting System (of the OECD) DAC Development Assistance Committee (of the OECD) DDSR Debt and debt service reduction DRE Debt reduction equivalent DRS Debtor Reporting System (of the World Bank) Ecu European currency unit EDT Total external debt, including shorr-term and use of IMF credit ESAF Enhanced Structural Adjustment Facility (of the International Monetary Fund) ERM European Exchange Rate Mechanism FDI Foreign direct investment GATS General Agreement on Trade in Services GDP Gross domestic product GNP Gross national product IBRD International Bank for Reconstruction and Development (the World Bank Group) IDA International Development Association (of the World Bank Group) IFC International Finance Corporation (of the World Bank Group) IMF International Monetary Fund INT Total interest payments on long-term and short-term debt, including IMF charges HIPC Heavily indebted poor countries LDOD Total long-term debt outstanding and disbursed LIBOR London interbank offered rate LILIC Less indebted low-income country LIMIC Less indebted middle-income country Mercosur Southern Cone Common Market (Argentina, Brazil, Paraguay, Uruguay) MIGA Multilateral Investment Guarantee Agency (of the World Bank Group) MGS Imports of goods and services MILIC Moderately indebted low-income country MIMIC Moderately indebted middle-income country MYRA Multiyear rescheduling agreement NGO Nongovernmental organization ODA Official development assistance OECD Organisation for Economic Co-operation and Development PV Present value RES International reserves RXD Revised External Debt (reporting system of the World Bank) SDR Special drawing right (of the International Monetary Fund) SILIC Severely indebted low-income country SIMIC Severely indebted middle-income country TDS Total debt service on long-term debt and short-term (interest only), including IMF credits TRIMs Trade-related investment measures TRIPs Trade-related aspects of intellectual property rights XGS Exports of goods and services Dollars are current U.S. dollars unless otherwise specified. x xi Debt stock and its components Total external debt (EDT) Short-term debt ong-term debt Use of IMF by debtor PrivatePulcadpbiy nonguaranteed debt guaranteed debt by creditor Official creditors Private creditors I I__ __ _ _ Multilateral Bilateral Commercial Bonds Other banks xii Aggregate net resource flows and net transfers (long-term) to developing countries Loan disbursements minus I I equals Foreign direct Debt service Net resource flows plus- investment (FDI), equals- Aggregate net (LTDS) on debt portfolio equity resource flows flows, and official grants minus minus Loan interest Interest payments and FDI profits equals equals Net transfers on Aggregate net debt transfers Note. Includes only loans with an original maturity of more than one year (long-term loans). Excludes IMF transactions. xiii Aggregate net resource flows (long-term) and the balance of payments Credits Debits * Exports of goods and services * Imports of goods and services * Income received * Income paid Current account * Current transfers * Current transfers Including workers' remittances and private grants Official unrequited transfirs (by foreign Official unrequited transfers (by national governmentts) government) Official unrequited transfers (by foreign - Official unrequited transfers (by national governments)- government) Foreign dirc investment (by nonresidents) * Foreign direct investment (by residents) (di si t h as negative) (disinvestment shown as negative) 111on..- . 11.,T. . * Portfolio investment (abroad by Capital and ... s1lwu g - . residents) (amortizations shown as financial accountneai) §i~ws + 4. ' in European developing countries through 400 September 1997 (figure 1.3). Spreads declined on 200 primary issues as well, though not by as much, January January January December 1995 thaergesred1996 1997 1997 with the average spread on new international bond Source: Blnomberg and World Bank. 14 Figure 1.3 Secondary market spreads transactions in 1993 to 5.3 million transactions in on international bonds decline and rise 1996 (Emerging Markets Traders Association). Seleded East Asinn countries It is difficult to measure just how much each of Basis points these factors contributed to the fall in spreads. One 500 study attributes much of the decline in spreads from 1995 to early 1997 to a shift in market sen- 400 r g timent toward a greater acceptance of developing country debt, rather than to improvements in 300 tP creditworthiness or increases in international liq- Philippines~ ee Iuidity (as evidenced by falling rates of U.S. 200 < ,} Treasuries; Eichengreen and Mody 1998). Indonesia Another study of the decline in spreads from the 100 <.+.>ta .@-Xhollnd second quarter of 1995 to the third quarter of 1997 found more than half of it to be attributable 0 to the increased global supply of capital rather than January July January July December to improved economic fundamentals in borrowing 1996 1996 1997 1991 1997 omrv cnm nu etasbrw g countries (Cline and Barnes 1997). Selected European and (entral Asian countries and Lebanon The rise in spreads Basis points 500 Secondary market spreads increased dramatically 400 in late 1997, as the average spread on Brady bonds 400 jumped from 350 basis points in September to Lebanon more than 600 basis points in early November (see 300 exw6>, 9., figure 1.2). Countries in all regions were affected. 200 'Fy0S} Yt= ! countries. For example, spreads in Thailand, which Hungary=.g . } had risen from 88 basis points in June to 187 basis 100 9 points in mid-October, leaped past 400 basis Poland points by mid-November (figure 1.4). Spreads for 0 - major Latin American borrowers also increased. January July January July December 1996 1996 1997 1997 1997 Secondary market spreads on eurobonds issued by Brazil (which was also a target of speculative pres- sures against the exchange rate) jumped dramati- Besis points cally, from 151 basis points in mid-October to 541 1,500 Mexico basis points in mid-November. By the end of December, however, there were signs that investors 1,200 were differentiating more carefully among devel- oping country borrowers. Spreads for Argentina 900 | Argenfina and Brazil, which had reached the highest levels in Latin America, fell by 100 and 250 basis points 600 Brazil from their November peaks, and spreads for Mexico, which had risen less, eased slightly. By con- 300e olombia 9 trast, spreads for Southeast Asia registered further 300 Colombia increases, with Indonesia's spread up more than 150 basis points from November. January January January Deaember The increase in secondary market spreads led to 1995 1996 1997 1997 a pronounced contraction in bond market activity. Developing country borrowers were reluctant to Souroe: Bloomberg and World Bank staff eslimates. issue new bonds at these higher rates, preferring to 15 Figure 1.4 Countries in all regions were affected by uncertainty) increases. This cycle of declining rising secondary market spreads on sovereign bond spreads and rising borrowing followed by sharp issues, October-December 1997 reversals may be accentuated by acute competition Basis points among fund managers, as they strive to increase Octeobene November yields when spreads decline and stage a general December withdrawal to cover find redemptions when spreads increase. Increasing financial integration of emerging eolombia markets Mexico g ] Bond markets Hungary Despite the extreme volatility in debt and equity markets late in the year, 1997 also saw developing Poland countries deepen their participation in interna- tional financial markets. There were market debuts by sovereign borrowers (Croatia, Moldova), sub- sovereign borrowers (Buenos Aires, Moscow), and China corporate borrowers. Corporate debt issue, which had risen from $2.1 billion in 1991 to $23 billion Indonesia in 1996, jumped to $36 billion in 1997. And there was increased diversification and P _ __ippines expanded use of more sophisticated finance instru- ments. International bonds issued by developing countries had greater currency diversification, Thoilond 9 including a Mexican issue denominated in euros 0 100 200 300 400 500 600 (in advance of the introduction of the European Souwce:13lonmberg- single currency) and domestic currency issues on international markets by Argentina and a South defer borrowing until market conditions became African utility. The growth in bond issuance was more settled. Argentina issued the only sovereign aided by market innovations, including the recent eurobond during the last two months of the year. development of collateralized bond and loan Some borrowers were able to tap the loan market obligations. Under this structure, unrated devel- late in the year, however, some through securitized oping country bonds or loans are packaged togeth- loans. er and used as collateral for notes. The notes can The risk reappraisal that set off the widespread be enhanced by issuing notes of lesser dollar value increase in spreads may reflect both a reevaluation than the debt instruments used as collateral and by of economic fundamentals in some countries and using guarantees (such as standby letters of cred- a correction to what has come to be viewed as mar- it). The notes also qualify for a higher credit rat- ket overshooting earlier in the year. Just as asset ing than individual bonds because they are more prices may exhibit excess volatility (volatility diversified. This structure has enabled institution- beyond that justified by fundamentals; see Shiller al investors (such as insurance companies) that are 1989), there may be excess volatility in spreads. restricted to highly rated debt securities to lend to Low spreads will lead borrowers to increase their developing countries. exposure and encourage lenders to be too opti- mistic in their assessments of risk, thus leading to Derivatives high volumes of flows. This overshooting of mar- kets, inducing excessive borrowing, implies a The growing use of derivative instruments also reversal of flows when market volatility (and thus demonstrates the greater depth and sophistication 16 the~~~~~~~~~~~~~r co~i ofacmoiyo laGa ntueta otrc s eemMndintaigo uue pfem rgicgmre.ts (buye and2sellerivroaptives far abexchangetroversyhasoverrteounter fturepssil aond dimprove thbe effi vered or markes Derivataseified opibtions coftat aere agveed dorectlabtwvee presnur- timea iuthre fst kndotowns aslo theommercia usersA pagaindt txhaneiraters depnd ton the shapdecif nder-i iof markes to hdgtheirriskanofafinatmntilos ures squtand iesduingseaedb the prs Aiancipals. cStrrcsur . thatic canfers, therbuye lorwhslerin the rot ainght(u dcnrcs Foreg example aditreet yersate weivt-ve a buiness? (orbexample, buy improving the g asseity arese theabl mntroduition taof cutoemiergopions warkth toattorac detp oanpid pouinty fundsthFutures markers nnthandare equiyderivatives, (sunch as barider parodutso when~th~ey workwel apria kprin aiscoverymrech- Gl ineth ~ors, wthinrs an priee range). atciat abnimfer the c~li o ~lleton andisneminatiohn tof Atog a ctualose deiliey to f thren underlingees orat- iellrmato on ~expectied prices. Therioativsca ano t mdctuytors fiAnccardinstruen canrakel' p olsacei de alsou imroe thnie teffiimenyo markets byiiniichothe Meraeotiof te conturact, many contrd-agets heia-e agingr icrieasedprct1icipation.b spetulvan tions, wre provisos forkti cahe worldin at umaturit (fornstrancs can addlaiquidity n elpprce f to converdge witd tiadex, futures). npractie, mrostducturs a countedfratore undelingidualsnd,amntalsket tfhandsdoe o6r periort tof deits try oe nh ashe of rimerging marketsv(boxe1.2).aDerivatives cany ablercontroversy halste surrou anderdtepsibled orin- improveth effiuciencyl of markets.nmi Deerivtivs. eribuhing ofntrstrt derivatives toseuativeapressures suhe cas b fut res oplx and opin tllwcmerciale userios agatedinst he exan eratsind to rkthe sardelinesg in ofumarkens aotoeg thei riansk ofe financia lossde eurnrcsdrn the nmeofLtnA eriast masiantcurrepancycisis lmtopied changes,o therkebyan loerigulthcorsts facing- Forteig exodctorhange adinterouest rthem dherivaive tolatracti debtadequpity foundris). FTures markltaets thnaof qut derivatives,tae sineplce aide ovaritet cofne whren thveywrkawell arpo piesicoeytehiivetrsnraes and other marketf participantsprodcts loisse for theviua collec utionsan disecmina xtio offsare exosedral toe cuntrrencyivand insteresturaed infodrmvation ornsexpctedn prces. rduerivtives crans fluctuations. Asccsordiong , to qrail's Boswaps ade palsoc ipove the effciency of mopoarket byd finncurl Mercadorliasked Futurs,sszbeioeo the thrlaretdriv- aingttton,tu increasedpricipations byr spveultors, who ties emarketing maktsThe wrdinumber of coantaractsd canmddliqutidit bandin huelpvprices to d convergerwit taed,uinty erestvaties proucs aouintraed for mren underlying f : ' .CE Q undamental,.ta 602S peren QQ} SofQ al, I I1 .()Q... it'is trove01 l(rill'§ inQI.lijs.1'1 J |9111S1B:............21 th past three Deriv>at,2ive als havQQQj QQQ}eQ the tQ6QQ ir pa dr awbtacks, 2 and the0QQy y ite S.,ar s S S' E'Qtt. Acivt in, lited (o sta ndadied foreign000t 0 0X ma not prodc al of2 th'^>e sAdye y 2{ Eto , 2 1 n omi bnfts. exchan SSl0Ci0;g;e and0g l0 int0 eret rat deiv SW 0a ; SE tie appears to b0W0 | e questonsaot tedmnshey lae ote umero Lati AmBerica markets ar plannin limitedu carpact of marktsan reglator,patic- lstepoucts or hav inrduce thm hebl inraers,smtieeutingii in spctcua Ancoa l evidecesugetstat onshore and lossb i,ndividual intttons. The cmleit offhoe ve th cone acivt in stru ctured patnc S orf the acout of corporte and fiania euiylne noes is siable in some of th Iag institutin,a tu ,< increasingrsk or nves(tors an er em U ergSing mrkets.The use 0 ofstndrdze complicating bankciengyo sup rvi tsio. ADconidaier- equibtyoo deriaivesihas alo inceuasied inrecsuent 17 years. An investor or speculator wishing to take a funds earned a compound annual return of 19 per- position in East Asian stock markets could pur- cent from 1992 through mid-1997; emerging chase derivatives contracts on five indexes (cover- market equity mutual funds earned 12 percent ing Hong Kong, China; the Republic of Korea; over that period. The volatility of returns was Kuala Lumpur; Singapore; and Taiwan, China, ) greater for the more highly leveraged activities of and match the Morgan Stanley Combined Far the hedge funds, however. East Index (excluding Japan) with only a 4 percent By and large, hedge funds dedicated to emerg- error. Among European countries, the Czech ing markets are specialized boutiques. The average Republic, Hungary, Poland, the Slovak Republic, emerging market hedge fund has about $130 mil- and Russia introduced standardized derivative lion in assets, and only six of these funds manage products in 1997. A combined index is listed on more than $250 million in assets. Many hedge the Austrian Stock Exchange to track the Czech, funds have chosen specialized market sectors, suich Hungarian, Polish, and Russian markets. as distressed debt, Latin American bonds, or Russian stocks. Most of the funds specialize in Hedgefunds stocks and bonds plus one or more other instru- ments. Twenty-five of the fifty-seven emerging Another important institutional development has market hedge funds, with $2.3 billion of the total been the growth of hedge funds. Under U.S. law $7.1 billion in assets, are involved in the transition hedge funds are private investment partnerships economies of Europe and Central Asia, possibly with wide flexibility to invest in securities or finan- reflecting the large opportunities for profit that cial futures and few regulatory requirements. have opened up during their transformation to Hedge funds domiciled in tax havens (such as market economies. Bermuda) have no legal limits on the number of non-U.S. investors, and some meet U.S. Securities Expansion of equity markets and Exchange Commission requirements for accepting U.S. investors.4 Hedge funds invest in a Developing countries' equity markets became wide variety of investment vehicles, including increasingly integrated with global markets in equities, bonds, currencies, and derivative prod- 1997. Emerging markets' international equity ucts. Typically, they supplement their equity with issues rose an estimated 44 percent in 1997 to $18 a high degree of leverage. billion, or from 15 percent of global issues in 1996 The number of dedicated emerging market to about 25 percent. The surge in portfolio equi- hedge funds increased from 5 (with assets of $682 ty flows to developing countries in the 1990s has million) in 1992 to 57 (with assets of $7.1 billion) contributed to a substantial rise in equity market in 1997 (table 1.6).5 Dedicated emerging market hedge funds have only a small share of worldwide Figure 1.5 Capitalization of emerging equity markets hedge fund assets, which are estimated at $300 bil- keeps growing, 1986-97 lion (Rodgers 1996). Emerging market hedge Billions of U.S. dollars Percent Table 1.6 The rise of emerging market hedge funds, 2,500 Share of world market 1 S 1992-97 2,000 copitaln zi 12 Asset- Assets eighted Number (billions of return 1,500 9 Yea r offunds U.S. dollars) (percent) 1992 5 0.7 7.0 1,000 1993 9 1.0 69.6 Cpllzto 1994 16 3.2 3.1 500 3 1995 29 3.4 8.8 1996 38 4.6 30.9 0 1997' 57 7.1 64.0 1986 1989 1992 1995 1991 a. As of July. a. As of October Source: Managed Account Reports, Inc. Source: IFE Emerging Morkets Global Composite Index. 18 capitalization, which rose from $200 billion in Despite the lower returns and higher risk of 1986 to $2.1 trillion in 1997 for the 18 develop- emerging markets over the past five years and the ing countries in the IFC Emerging Markets Global dramatic fall in equity prices in some East Asian Composite Index (figure 1.5). The events of late markets during the second half of 1997, there are October, when stock market declines in develop- good reasons to anticipate that foreign investors ing countries set off a period of volatility in indus- will continue to be attracted to emerging markets. trial country markets, demonstrated the growing To begin with, the comparison of U.S. and emerg- importance of emerging equity markets in the ing markets during the short sample time period global economy, as well as the risks inherent in (made necessary by the data limitations in emerg- increasingly integrated international markets. ing markets) may not reflect long-term trends. Have the returns on investments in emerging Returns on the S&P 500 in the United States dur- markets relative to those in industrial countries ing this period (19.9 percent) exceeded the histor- justified the huge increase in foreign investment? ical average (11.1 percent) by a wide margin. And Over the past five years U.S. mutual funds invest- despite relatively poor returns for emerging mar- ing in developing country stock markets have kets over the past five years, there is good reason to earned average annual returns of 8 percent while anticipate that returns in those markets may funds investing in the United States and Europe exceed returns in industrial country markets over have earned 19 percent.6 Risks were also relatively the medium term: output in developing countries high for mutual finds investing in emerging mar- is expected to rise significantly faster than in indus- kets. The standard deviation of the return on trial countries, and differences in patterns of pop- funds over the past five years was 22 percent for ulation growth are likely to widen the differential emerging markets and 13 percent for U.S. and in expected rates of return on capital between European funds (table 1.7). The risk-adjusted industrial and developing countries (World Bank return as measured by the Sharpe ratio (which 1997b). Also, one way to reduce risk is to increase shows excess return over the risk-free rate per unit the diversification of a portfolio, by holding assets of risk7) was 0.2 for funds investing in emerging whose returns are not highly correlated with those markets, much lower than the 1.1 ratio for funds on other assets, such as emerging market invest- investing in the United States. In short, for this ments and industrial country investments. Thus sample of emerging market mutual funds over the investments in emerging markets can provide limited time period under consideration, returns opportunities for improving return-risk tradeoffs were lower and variability was higher than for U.S. in the portfolios of investors from industrial coun- and European markets. Note that the results tries, even if expected returns are lower and risks would not change materially if the time period are higher in emerging markets. were extended past October 1997 to the end of the Another possible contributor to the lower year or cut off at June (before the onset of the East returns in emerging markets over the past five years Asian currency crisis). is the generally higher transactions costs (registries, Table 1.7 Performance of stock market indexes and mutual funds, 1992-97 Standard Sharpe Load-adjusted deviation for ratio for One-year five-year five-year five-year Market/index return' return6 return ' return' United States (mutual funds) 27.79 18.87 13.45 1.07 S&P 500 Index 32.10 19.85 12.51 1.19 Europe (mutual funds) 24.80 18.58 13.08 1.04 MSCI Europe ND (including reinvested net dividends) 25.98 18.56 13.19 1.03 Emerging markets (mutual funds) -3.68 8.08 21.60 0.17 IFC Emerging Markets Global Composite Index -10.01 7.49 18.74 0.13 a. November 1996 to October 1997. b. Load-adjusted returns are adjusted to count for front or back end sales charges. Index returns are not load adjusted. c. November 1992 to October 1997. Source. World Bank and Morningstar. 19 custodial arrangements, taxes, brokerage fees, countries' liberalization of their economies, par- administrative costs related to failed trades, delays ticularly privatization and the lifting of restrictions in dividend repayments) in emerging markets than on FDI; strong growth in the GDP and trade of in industrial countries. For example, total expens- the major developing country recipients of FDI es (including management fees and brokerage flows; and the falling cost and rising quality of costs) of the average U.S. stock fund are about 1.6 communication and transportation services, percent of assets, while expenses of the average which encouraged private corporations to shift emerging market fund absorb 2.9 percent of assets. toward more integrated global investment and Similar differences show up in the bid-ask spread, production. Like trade, FDI provides an impor- which reflects the costs of dealer operations tant channel for global integration and technolo- (finance, inventories, operational costs, and some gy transfer. of the costs noted above). Average bid-ask spreads Multinational corporations took advantage of are 67 basis points in the United Kingdom but 130 the liberalization of the rules governing FDI in to 170 basis points in even the more advanced developing countries and the privatization ofstate- emerging markets, such as Brazil, Indonesia, owned enterprises to access new markets and Mexico, the Philippines, and Thailand. resources. The largest FDI source countries These higher transactions costs reflect the increased their FDI in developing countries, both undeveloped infrastructure in emerging equity in absolute amounts and as a share of total stocks. markets. Efforts to improve trading systems, pro- Notably, the share of developing countries in the vide for safe and efficient depositories, and stock of the outward FDI of the five main FDI strengthen clearance and settlement systems have sources-the United States, Japan, Germany, the great potential for reducing transactions costs and United Kingdom, and France-increased during thus improving the performance of developing the 1990s in all but Japan. And in Japan the share country stock markets. Further, developing coun- declined only marginally in the 1990s, mainly tries need to implement regulatory reforms that because of its large new investments in North improve the credibility of equity markets in order America. Still, more than 30 percent of Japan's to attract both foreign and domestic investors and outward FDI stock remains in developing to realize the benefits of financial integration. countries. Reforms would include more attention to trans- parency through improved disclosure, promotion FDI concentration and low-income countries of self-regulatory organizations, and credible pro- tection of minority shareholder rights. Not all developing countries shared in the rapid growth of FDI during the 1990s. Almost three- Foreign direct investment has surged in the 1 990s Figure 1.6 Both the amount and the share of global FDI to developing countries are rising, 1991-97 Foreign direct investment (FDI) in developing Billions of US. dollars Percent countries was $120 billion in 1997, slightly above 150 FBI to developing countries 40 the $119 billion of 1996 and five times the level of as shore of global FDI 1990. The ratio of FDI to GDP in developing 120 countries jumped from 0.8 percent in 1991 to 2.0 / China percent in 1997. Developing countries have also 90 increased their share of global FDI flows, from 21 20 e All developing mornies percent in 1991 to an estimated 36 percent in 60 except China 1997 (figure 1.6). 01 0 Why the growth of FDI inflows? 1991 1992 1993 1994 1995 1996 19970 The surge in net FDI flows during the 1990s has o. Prelim inary. been spurred mainly by three factors: developing Source: World Bank doto. 20 Table 1.8 FDI flows to the top ten recipient developing countries, 1991, 1994, and 1997 (billions of U.S. dollars) Country 1991 Country 1994 Country 1997, Mexico 4.7 China 33.8 China 37.0 China 4.3 Mexico 11.0 Brazil 15.8 Malaysia 4.0 Malaysia 4.3 Mexico 8.1 Argentina 2.4 Peru 3.1 Indonesia 5.8 Thailand 2.0 Brazil 3.1 Poland 4.5 Venezuela 1.9 Argentina 3.1 Malaysia 4.1 Indonesia 1.5 Indonesia 2.1 Argentina 3.8 Hungary 1.5 Nigeria 1.9 Chile 3.5 Brazil 1.1 Poland 1.9 India 3.1 Turkey 0.8 Chile 1.8 Venezuela 2.9 Top ten share in FDI to all developing countries (percent) 74.2 76.1 72.3 a. Preliminary. Source: World Bank data and staff estimates. quarters of net FDI flows to developing countries because of an unfavorable business environment has gone to just 10 countries (table 1.8). The con- (poor policies, unstable government, civil strife, centration rate remains high, but it has eased weak infrastructure, poorly trained workforce) or somewhat in recent years, reflecting growing a small domestic market. access to FDI. Most of the top recipients of FDI are middle- Moderatinggrowth in 1997 income countries, reflecting their large markets and rapid growth in recent years. The only low- After growing strongly for several years, FDI lev- income countries in the top 10 in 1997 were eled off in 1997. Preliminary estimates indicate China and India. The ratio of FDI flows to GDP that net FDI flows to developing countries has remained smaller in low-income countries remained high in 1997, at $120 billion, but only (excluding China) than in middle-income coun- slightly above the $119 billion in 1996. Still, net tries throughout the 1990s (table 1.9). Recently, FDI flows continued to be the largest component however, some low-income countries besides of external resource flows to developing countries. China and India have begun to receive large The largest reversal in the upward trend occurred amounts of FDI. FDI flows to Vietnam have in East Asia and the Pacific, where flows declined increased significantly, from about $380 million a 9 percent in 1997, to $53 billion (figure 1.7). FDI year in 1991-93 to $1.3 billion a year in 1995-97. flows to Indonesia fell 27 percent, and flows to Some countries, such as Nigeria, have long been China fell an estimated 10 percent (in part simply large recipients of FDI in natural resource sectors. a reflection of the surge in FDI in 1996 as investors But many small, poor countries (especially in sought to take advantage of preferential tax treat- Africa) receive little foreign investment, often ment that expired that year). By contrast, flows to Table 1.9 FDI flows to developing countries by income group, 1990-97 (percentage of GDP) Income group 1990 1991 1992 1993 1994 1995 1996 1997' All developing countries 0.6 0.8 1.1 1.6 2.0 2.1 2.0 2.0 Low-income 0.5 0.8 1.5 3.4 3.7 3.5 3.0 2.8 Excluding China 0.2 0.6 0.6 0.9 1.0 1.3 1.2 1.2 China 1.0 1.2 2.7 6.4 6.5 5.4 4.7 4.2 Middle-income 0.6 0.8 0.9 1.0 1.4 1.6 1.6 1.8 Lower-middle-income 0.4 0.5 0.6 0.7 1.2 1.4 1.4 1.3 Upper-middle-income 0.8 1.1 1.3 1.3 1.6 1.8 1.9 2.2 a. Preliminary. Source: World Bank data. 21 Figure 1.7 FDI flows remained strong in most regions Figure 1.8 During the 1 990s privatization projects in 1997 have attracted substantial foreign investment Billions of U.S dollars Billions of U.S. dollors 60 EastAsia and 25 50 * 20 Portfolio 40 ..i- Pnovestmens Latin America 1 5 30 and thre Caribbean 20 Mirddle East and Central Asia anld North Africa 10 Asb-oSamon 5 Afrca South Asia 0 _ _ 0 1992 1993 1994 1995 1996 1997 1991 1992 1993 1994 1995 1996 1997° Source: World Bank dato and sraff estimates. o: Preliminry. Source: World Bank Privunzoaion Dlatabase and stoff estimates. Latin America and the Caribbean rose 10 percent, rienced large increases in FDI through privatiza- to $42 billion. Large privatization projects in tion transactions in 1997, while Hungary, Peru, infrastructure and other services (notably in and others saw large declines in FDI in 1997 as pri- Brazil), better economic performance (Mexico), vatization programs slowed or ended. ongoing liberalization measures (particularly in financial services), and strong flows into Mercosur Growth of FDI in service sectors and among its members all contributed to Latin America's FDI performance in 1997. FDI also flowed heavily into services during the During the 1990s active privatization programs 1990s. Several countries attracted especially large in many developing countries generated a steady shares of FDI to their service sectors (table 1.10). stream of FDI inflows, attracting some $60 bil- Japan and the United States (which account for lion-$15 billion in 1997 alone (figure 1.8). nearly half of the stock of FDI in developing coun- Brazil, Venezuela, and some other countries expe- tries) both showed large shares of services in their FDI to developing countries. Services account for Table 1.10 FDI in selected developing countries, 58 percent of the total stock of FDI in developing by sector countries that originates from Japanese firms, (percentage of total stock) while the share of services in the stock of U.S. FDI Country Year Primary Secondary Tertiary in developing countries has nearly doubled in the past decade (figure 1.9). A similar trend is observed Argentina 1990 16.3 35.3 48.4 1994 12.7 34.8 52.5 in most major FDI source countries, except for the Brazil 1990 16.1 70.4 13.6 United Kingdom. 1995 12.6 68.1 19.4 Several factors have contributed to the growing Czech Republic 1992 1.4 84.4 14.2 importance of services in FDI flows. Developing 1996 5.4 50.8 43.7 countries are lucrative markets for multinational Hungary 1992 2.3 52.9 44.8 service providers, and many services firms need a 1996 2.5 40.3 57.2 physical presence in the market to compete effec- Indonesia 1990 16.1 70.4 13.6 tively (UNCTAD 1996). Services have also become 1995 12.6 68.1 19.4 more attractive now that advances in communica- Peru 1990 37.9 34.4 27.7 tions technology allow separated-service providers 1996 19.6 17.2 63.2 (data processing, softwaredevelopment) to reap the Thailand 1990 6.1 37.8 56.1 cost advantages of developing countries (Allen and 1996 3.9 33.4 62.7 Morton 1994). Liberalization of FDI regimes and So.rce: World Bank staff calculations based on national sources. privatizations of service enterprises have also accel- 22 Figure 1.9 Services' share of U.S. and Japanese FDI to developing countries is large and growing, 1985 and 1995 (shofe of total stock) 1985 Japan 1995 Plrimar/y Primaoroy Tertiary Secondary 55.3%335 Secondary 29.2% United Slates Primary 3 7 5 ° / °ySecondary Secondary ~ ~ Seonar Note. Developing countries as classified by each data source and not necessarily corresponding to the Woeld Benks country classifiaotons. Source: U.S. Department of Commerce; Japan Ministry of Finance. erated the rush of FDI to services. Overall, some 40 FDI outflows from developing countries increased percent of the revenues generated by developing by more than three and a half times to $11.1 bil- countries through privatization programs during lion, indicating the growing integration of devel- the 1990s have been directed to services, and this oping economies in world markets (table 1.11). share has increased rapidly in the past few years. In China is not only the largest recipient among Brazil, the largest FDI recipient through privatiza- developing countries, but also the largest investor, tion operations in 1997, roughly 70 percent of the followed by Malaysia, Chile, Brazil, Thailand, and revenues were raised in service sectors such as South Africa. telecommunications, power, transport, and finance. Multinationals from these economies are invest- The rapid growth of FDI in services reflects a ing predominantly in other countries in the same fundamental change in the role of the private sec- region. In Latin America and the Caribbean the tor in providing services. Technological changes intraregional increase in FDI is driven by increas- and regulatory innovations have made it possible ing liberalization, new investment opportunities, to open these services to market forces, thus proximity, and trade integration. In Argentina the encouraging competition, which can bring sub- stock of FDI originating from other Latin stantial benefits for consumers. Greater private American countries accounted for 15 percent of its investment in services can also free scarce public FDI stock in 1994, up from 11 percent in 1992. resources for pressing social needs. To realize these opportunities, governments will need to ensure Prospects frr FDIflows that price and regulatory rules enable service providers to operate on a sound commercial basis. What impact the financial crisis in East Asia will have on FDI flows to developing countries in 1998 FDI outflows during the 1990s is uncertain (see chapter 2). Nevertheless, medium- term prospects remain good. FDI flows are likely Between 1991 and 1996 (the latest year for which to be supported by strong growth in developing reliable numbers are available) the dollar value of countries' output and exports, greater economic 23 Table 1.11 FDI outflows from developing countries,1991-96 (billions of U.S. dollars) Coenty 1991 1992 1993 1994 1995 1996 All developing countries 3.0 7.3 9.4 10.1 11.3 11.1 China 0.9 4.0 4.4 2.0 2.0 2.2 Malaysia 0.4 0.5 1.3 1.8 2.6 1.9 Chile 0.1 0.4 0.4 0.9 0.7 1.1 Brazil 1.0 0.1 0.5 1.0 1.4 1.0 Thailand 0.2 0.1 0.2 0.5 0.9 0.9 South Africa -0.2 0.8 0.3 0.3 0.6 0.7 Mexico 0.2 0.7 0.0 1 0.6 0.6 Source: IMF balance of payments data. integration and globalization of production, and (Markusen 1983; Helpman 1984; Pearce 1993; continued liberalization of investment rules. Pfaffermayr 1996). Analysis at both the firm and Studies have consistently found a strong, posi- the country level shows a strong positive link tive correlation between rapid GDP growth and between foreign investment and trade flows FDI inflows (figure 1.10), although the direction (Eaton and Tamura 1994; Kawai and Urata 1995). of causality is difficult to determine and may go in World trade is projected to grow by some 7 per- both directions (Lizondo 1990; Hein 1992). cent annually in the coming years, a further indi- Developing countries' average GDP growth rate is cation of the bright prospects for FDI flows projected to exceed 5 percent over the next 10 (World Bank 1997b). years, or about twice the rate in industrial coun- tries, indicating that developing countries are like- Regional economic integration may induce higher ly to continue to attract a substantial share of foreign investmentflows. Regional integration agree- global FDI (World Bank 1997b). ments can promote FDI flows by expanding mar- kets. FDI flows tend to rise in the run-up to a Trade expansion and export orientation promote regional agreement, as shown by the increase in FDI flows. The complementary relationship FDI flows to Spain and Portugal in the years imme- between FDI and trade flows is well established diately preceding their entry into the European Figure 1.10 FDI inflows are positively correlated with GDP growth Ratio of FDI to GDP Percent 1 2 Hungary Vietnam 10 5 Trinidad n Malaysia and Tobago 8 a XPeru Dominitan Republic Chino Mexico Costa Ritca t zech Rep. Ecadr Chil 4 Egypt C P7 i b i Greee \ PMorlpne mrPapua New Guinea 2 AreP ahiclippnesi * rThailand Rusine.r P°itn7ieR.n tBrazillnoei Russi0 n Fed. Venezuelo Turkey Romnia India -9-6 -3 0 3 6 9 l 2 l S Reol GOP growth (percent) Note: olculotions are boued on 1993-96 overogesfor the top 30 redpients o FDI in 1996. Source: World Bonk Debtor Reporting System ond Worid gonk datn. 24 Union, EU experiences prior to the 1992 single Anrex, inaue Aoth sapital flows and domestic market, and the rise in FDI to Mexico before bond markets NAFTA. Automobile multinationags are gearing their production to the entire Mercosur market, Though domestic bond markets in developing even though barriers to automobile trade have not countries have lagged behind equity markets, they yet been removed. There is one caveat, however. If have grown rapidly in recent yearss a result of regional integration results in increased distortions improving economic fundamentals (particularly (high external trade barriers, market-sharing agree- lower inflation), finat cial liberalization, and pri- ments between ologopolistic producers), the result- vatization programs. oehe share of foreign partici- ing inefficiencies could outweigh the potential pation in these markets appears to be rising. The benefits of FDI. importance of domesthc bond markets varies wide- Continuing liberalization of invJestmnent rules ly among countries, from 4 percent of GDP in willeurtherencouragepFDI. Morethan 50rcountries Indonesia to 51 percent in South Africa (table enacted more open laws on foreign investment IA. 1). during the 1990s, and the number of bilateral The volumesof flows to developingrcountriestfor investment treaties on fhe lamtohn the purchase of domestic fixed-income securities is protection of foreign investment increased three- unknown, although anecdotal evidence suggests fold to 1,100 (UNCTAD 1997). There have also that foreign participation is significant in some been advances in multilateral instruments with domestic bond markets. For example, it is report- provisions on cross-border investment. The ed that a minimum of $10 billion of Brazil's total Multilateral Agreement on Investment, under domestic sovereign bond market ($163 billion) is negotiation among OECD countries since May held by foreigners. Pefindo, the Indonesian credit 1995, would encompass the three cornerstones of rating agency, reports that foreign investors FDI rule-making: investment protection, invest- account for 30 percend: of the market. Other incti- ment liberalization, and dispute settlement (box cations of increased international interest in devel- 1.3). Recent agreement in the negotiations over oping countries' domestic bond markets are the trade in financial services should also encourage growing use of international services for clearance, increased FDI inflows in the future. settlement, and custody and J.P. Morganr s develop- 25 Table 1A.1 Domestic bond markets in developing countries (millions of U.S. dollars) Percentage of GDP Domestic _Total Domestic Region or country External' Total Sovereign Corporate bonds bonds East Asia and the Pacific 46,878 139,100 102,719 20,600 13.5 10.1 Europe and Central Asia 72,167 170,289 166,789 3,500 20.4 14.3 Latin America and the Caribbean 256,284 309,280 275,492 33,788 35.3 19.2 Middle East and North Africa 14,494 35,910 35,760 150 24.8 17.7 South Asia 4,927 79,999 43,177 - 19.1 - Sub-Saharan Africa 15,988 78,236 60,756 17,480 44.7 37.1 Largest domestic bond markets Brazil 80,933 186,823 185,972 851 39.2 27.3 SouthAfrica 3,602 77,600 60,340 17,260 53.8 51.4 India 4,382 73,530 36,708 - 21.3 - Russia 37,365 60,711 60,691 20 19.9 12.3 China 11,901 45,400 41,400 4,000 6.7 5.3 Malaysia 6,227 41,600 29,100 - 50.6 - Poland 6,673 41,080 41,020 60 35.5 30.5 Mexico 59,620 38,304 21,258 17,046 15.7 14.2 Turkey 13,438 23,214 22,604 610 17.2 10.9 Chile 7,142 31,042 19,849 11,193 57.1 46.4 Not available. Note: Data represent either 1995 or 1996. Data are compiled from different sources and for different years and are not necessarily consistent or comprehensive. a. Includes tradable loans. Source: Merrill Lynch, Salomon Brothers, and World Bank. ment of a domestic debt market benchmark (the convertible bonds issuance,8 while the imposition Emerging Local Markets Index), which tracks total of merit rather than disclosure regulations discour- returns for local currency-denominated money ages bond issuance in some Asian markets.9 Many market instruments in 10 emerging markets. countries lack the large pool of domestic fixed-rate Some developing countries still maintain investors needed to support a large bond market restrictions on foreign participation in domestic (ADB and World Bank forthcoming). Recently, a bond markets, to help manage short-term capital number of governments have taken steps to reduce flows. Countries impose minimum holding peri- impediments to bond market development by set- ods for bonds (Chile), restrict the range of issues ting up settlement and clearance services, estab- that foreigners may purchase (Czech Republic, lishing market benchmarks to facilitate corporate Hungary, South Africa), or require that foreigners bond issuance, and encouraging the development operate through domestic intermediaries (Brazil, of pension and insurance companies that would Chile). In general, the rules governing foreign par- increase the pool of domestic savers. ticipation in domestic bond markets are less restrictive than those for equity markets. Notes Considerable potential exists for increased flows to developing countries' domestic bond markets. 1. Canada, France, Germany, Italy, Japan, the (For a discussion ofwhy bond market development United Kingdom, and the United States. 2. Spreads on established borrowers declined some- is important, see box 1A. 1.) However, a number of what, but spreads were relatively higher on new and factors impair the efficiency of markets and their riskier borrowers that entered the market during 1997. attractiveness to foreign investors. Poor settlement 3. The high-yield index measures the yields on procedures and a lack of depositories are common fixed-income securities issued by U.S. corporations that shortcomings. In many countries the tax regime are below investment grade (for example, BB rated). discriminates against bond issuance (income from 4. The U.S. Securities and Exchange Commission limits hedge funds to 99 investors, a portion of whom bank deposits is often tax free while interest pay- must have a net worth of at least $1 million. ments on bonds are taxed). Protection of minority 5. The following overview data come from shareholders through preemptive rights hinders Managed Account Reports' MAR/Hedge, which is 26 based on information supplied by investment managers. 7. The Sharpe ratio is calculated by subtracting the This discussion does not include the activities of global average monthly return of the 90-day U.S. Treasury bill hedge funds that invest in both industrial and develop- from the fund's average monthly return, giving the fund's ing markets. excess return. This excess return is then annualized and 6. Based on Morningstar's Principia Plus database, divided by the fund's annualized standard deviation. which provides a sample of mutual funds that includes 8. Preemptive rights are requirements that existing 2,130 U.S.-based mutual funds (with assets of $1.3 tril- equity owners have the right of first refusal on new equi- lion) investing in industrial country markets and 245 ty issues on a pari passu basis. U.S.-based funds (with assets of $40 billion) investing 9. Merit regulation refers to detailed guidelines in emerging markets. regarding issuance, rather than disclosure. In most developing countries th e financial system. is * Mobilization of savings and nnovason. There 27 [Box IA2 Corsideratimns in foreign sovereign borowing Access by developing country governments to inter- domestic capital market, and the conduct of mone- national bond markets has increased markedly in tary policy. recent years. The government debt of 51 developing * Cost ofborrowing Inmost developing countries countries is now rated by the main international cred- the ex ante return to capital is likely to be higher than it agencies (just 10 were fated in 1990), and 25 devel- the cost of foreign borrowing, reflecting both capital oping countries achieved investment grade status on scarcity and high volatility in developing markets. their sovereign debt (prior to the East Asian crisis). However, international lenders may lack information Publicly guaranteed debt from private sources in on domestic private sector borrowers, and there may these countries ranges froma 0.5 percent of GDP in be a lack of creditworthy companies able to borrow Malta to 50.7 percent in Panama, with an average of on the required scale. In this case it is cheaper at the 6 percent. Many governments in developing coun- margin for the government to borrow rather than for tries are now able to choose between borrowing in individual private firms to do so, relecting the gov- local currency in domestic markets and borrowing in ernments greater access to the market. At the same foreign crrency in international markets. Some gov- time, private sector borrowers can rely on domestic ernments have issued debt linked to foreign curren- banks, which generally face lower cosrs than do inter- cies in domestic markets, such as the Mexican natioual lenders inrevaluating the riskiness of domes- tesobonos (Guidotti and Kurnar 1991). tic firmns By borrowing abroad the government In deciding where to borrow, governments in avoids crowding out firms on the domestic market developing countries typically face a highly segment- and reduces the cost of capital for the whole econo- ed marke, with policy and institutional impedinents my. Ofcourse, whether this trwns out to be a net ben- "to capital flows asymmetric information, and differ- efit depends on whether the use of ifnds by the public ing levels of risk all contributing to the segmentation secor is as efficient as in the private sector. Also, gov- between domestic and foreign markes Governments eminent borrowing can improve the faniliarity of that have access to international private sources of international lenders with the country and establish finance in effectface two distinctsiupply schedules for a benchmark that can be used to price loans to pri- finds. Both the domestic and foreign supply sched- vate cmpanies, so that initial borrowing by the gov- ules are likely tv be upward sloping, reflecting (in ernment may reuce the ultimate cost of borrowing both domestic dnd international arkters) rising risk by the private sector. perceptonswithincreasedborrowi-ngand (in domes- * Frign kebts amd crrditworthines. Defaults by tic manrkts): a limited supply of savings, (See, ifr private sector borrowers can affect a country's credit exampie, Baron 1989; Cohen 1991; Eaton, standing in the eyes of foreign investors. If relatively Gersovitz, and Stiglitz 1986; Tanzi and Blejer 1988; risky borrowers can gain access to international capi- Stiglitz 1988; and Calve and Guidotti 1993.) tal markets (rexample, during a period of uniusual- Thereis, therefore, a rationaleo rboitrowing-to be ly high liquidity when investors may- accept higher distributed beween domestic ahd foreign markid so leve-ls of risktO ronmatnmin yields), the government may as to equalize marginalcost-H in decidin decide to borrow rnore abroad rather than from whether tO borro abroad or atiome, governments domestic markets, thus lessening pressures on inter- (which are typically by for the largest borrower) go estrates dincreasiincentives for firms to borrow beyond direct cost and maturity comparisons toconu- domestically. side the implications for the econ . govern- . - C. y Ane Since foreign rowing is met's choicean affect the cost offoreign borwng .usualnibrimfign curenrcy it iiplies an increase in fio. fot the eonomw as a whole, the countrys credito - eIgxcageec Tn and Velrsco thSiness, t4w freign curreny expsuri of the goern- 1996). (This is not alw4rue, however. For example ntand theprate sector, theidevelopment of the oticb niarkers itn Brazil and Indonesia # i: 0 f - - V; 0 - , ' S -, ' ' ' - ' -, 'ox ' - nu on ne'qir1ge) 28 Bo X ° IA.2 Consideaion in oripsoveregn borroing (coninued)-t U d 'ttt 8 ;00ii0 Afic hav£e iss0ured local curenc bonds. Hoevr deeningo g#i jts: g r stailzat ion ojective s, irely in affeacted bysifts in teechnege rate (ieoome domsti aivt slackeons. (On th thoretical 0 aspect reliant onprXirnar com j, ajmodiis), bovrjrowing i fr ojggf themixberw&n Sfo:re)Sgnand10i§ doesic¢ bonds see 4 eign~ ~ ~ ~i currenc may reuc the jj conr' ntexoSsure i Bohn 19ifi90. Oe(n machroeconomi management2;i42 Q toe tcurrecyrsk. For eamle, a gvenent t that ii isi iiVaspeci ts, seei Anan40 ;Sggj and va Wi jnbre 198 i I andgwQ dependn n dollr- enoniate d y a t e p s 0 export fo it rev- Cudnro 97 enuejs can jbo>rirow abroadi ini dollas.i An ilossess (or00fi9 The 'pssibl adatae enmrae abov ofgi0i0i X Int stitu0tioa eeo iet oenent partlici - eipsrs an stabilization) willten to diin ish0 as I p tio m y eth s reta td atstw yo f evlo - apta m rk t be om more in te rtd ,i ncoms ing te domstic ond mrker,thus ptnin tip ew an capial inensit r ise,ecnomc otiliy? dim;in-X;; a n d m o r v e r s til fi n a c i n i n t u e t f o r.. . d o m e s - is h e s w i. h i n r e s e ..d i v c r. ii. .. ?f i r o of . . . . W . t h e c o n o m y ,S CHAPTER 2 East Asias financial crisis: causes, evolution, and prospects The financial crisis that began in Thailand, and led the second half of 1997 equity prices in the region to the 20 percent devaluation of the baht on July fell 50-75 percent (in dollar terms). 2, quickly spread to other countries in Southeast The crisis began to affect other economies in Asia and eventually to the Republic of Korea, and October, when speculative pressures intensified had repercussions around the globe. Between June against the Hong Kong dollar, the Korean won, and 1997 and the end of the year the median curren- the Taiwan (China) dollar, accompanied by sharp cy devaluation in 12 of the largest emerging mar- falls in stock markets in these economies. A severe kets was 39 percent,' and in the 5 East Asian plunge in Hong Kong's equity market precipitated a countries hardest hit by the crisis-Indonesia, global drop in equity prices (affecting Eastern Korea, Malaysia, the Philippines, and Thailand- Europe, Latin America, Japan, Europe, and the it was 80 percent. The International Finance United States) and increased pressures on the cur- Corporation's (IFC) emerging stock market index rencies ofdeveloping countries. In November Korea, dropped 25 percent between June and December, the world's eleventh largest economy, became the and its Asian index fell 53 percent. Among the object of the largest ever international rescue pack- group of 12 countries the median rise in short- age. Most countries in Latin America and Eastern term interest rates (from July 1 to each country's Europe were able to maintain exchange rate policies peak rate) was more than 600 basis points. or stabilize rates after the initial devaluation. But the The crisis began in mid-1 997 with intense pres- monetary (and in some cases fiscal) tightening sures on the Thai baht. Efforts to prop up the cur- required to restore confidence has undermined rency-including intervention, increases in growth prospects in several emerging markets. interest rates (to 18 percent inJune 1997 compared This chapter draws the following conclusions with 12 percent in January), and restrictions on for- in analyzing the causes and implications of the eign speculators-ultimately failed because by still-unfolding crisis. mid-1997 domestic companies sought to protect themselves from foreign exchange risks to their bal- * The East Asian crisis differs from previous ance sheets by repaying foreign debt and hedging developing country crises-such as the Mexicanpeso their foreign exchange exposure. After spending devaluation of 1994 or the debt crisis of the $8.7 billion in reserves to defend the currency (and 1980s-in that private sector financial decisions undertaking $23 billion in forward contracts were the main source of difficulties. Public sector maturing over the next 12 months), Thailand's borrowing played only a small role. Furthermore, central bank let the exchange rate float on July 2. East Asias crisis occurred despite a benign interna- By the end of the year the baht had depreciated 93 tional environment, with low international inter- percent and the stock market had fallen 34 percent est rates and solid global growth in output and (in dollar terms) relative to June 1997. The deval- exports. uation of the baht triggered a collapse of market confidence in Indonesia, Malaysia, and the * Despite impressive macroeconomic perfor- Philippines, causing their exchange rates to plum- mance and prudent fiscal policies, East Asian met in July and August 1997. Further declines in economies have become increasingly vulnerable currencies and equity markets followed. During during the 1990s. The mostpressingproblems are 29 30 in the financial sector, where distorted incentives, private decisions were the main source of difficul- lax regulatory standards, poorly managed finan- ties, public borrowing played a limited role, and cial liberalization, and inadequate disclosure and inflation was low relative to most other developing supervision have encouraged excessive risk taking, countries (and had declined in the 18 months pre- particularly in terms of maturity and currency ceding the crisis). Unlike Latin American coun- mismatches. Financial sector weaknesses led to a tries in the 1980s, most East Asian countries had misallocation ofinvestmentanda buildup ofnon- been running a fiscal surplus prior to the crisis performing loans. Large capitalinflows amplified (table 2.1). Unlike Mexico before the 1994-95 the problems of the financial sector and fueled peso crisis, public debt was not a relevant factor; domestic demand which, coupled with the depre- the short-term maturities of private, not public, ciation of the yen relative to the dollar (to which debt were the main source of financial vulnerabil- these countries were closely tied), caused real ity. And excessive consumption by the private sec- exchange rates to appreciate. tor was not an issue in East Asia, which has high savings rates. The increase in vulnerabilities does notfully The crisis also differs from previous ones in that accountfor the spread and depth of the crisis, how- it occurred against a benign international back- ever. The severity of currency and stock market ground, with high rates of growth in world trade, declines and thefew warningsfrom marketpartic- low international interest rates (both nominal and ipants indicate that a self-frulfiling loss of market real), and declining spreads on international bor- confidence played an important role. With the loss rowing (see chapter 1), which implied low debt of confidence and substantial uncovered foreign servicing costs and a favorable environment for exchange exposure, currency depreciation became additional borrowing and equity flows. Still, exter- self-perpetuating: the rise in the localcurrency value nal factors did play a role, in the form of the siz- ofliabilities impaired balance sheets, lowered stock able appreciation of the U.S. dollar during prices, and increased demandforforeign exchange 1996-97. Southeast Asian currencies' close links to cover open positions. Increased demandfor for- to the dollar contributed to exchange rate appreci- eign exchange led tojfurther currency depreciation, ation, which was an important factor in the lower andso on. Both healthy and insolventfirms suffered export growth rates relative to the extraordinary because of the lack of transparency (investors, growth in the first half of the 1990s. And although unable to distinguish amongfirms, withdrewfrom Korea's real effective exchange rate was stable, a all of them), the effect of the currency depreciation weaker market for com.puter chips in 1996 hurt its on dollar-denominateddebt, the increases in inter- economy. est rates to defend the currency, the contraction in A number of factors made the East Asian credit resulting from the rapid drop in equity of economies especially vulnerable to a sudden crisis highly leveragedfinancial institutions (due both to of confidence. As noted, these problems (includ- their own losses and to the insolvency of their bor- ing misallocation of investment and maturity and rowers), and increased uncertainty and the eco- currency mismatches) were rooted in private sec- nomic downturn. tor financial decisions. But policies-especially * The fallout from the East Asian crisis trig- Table 2.1 Fiscal balances and gross national savings gered a steep decline in portfolio flows to develop- in East Asian countries, 1993-96 ing countries toward the end of 1997, and little (percentage of GDP) r . v n ~~~~~~ ~ ~~~Fiscal Gross national recovery is expected in the near term. Private fows bal,ancessavings to developing countries will likely decline moder- ately in 1998, and downside risks are significant. Country 1993-95 19961993-95 1995 Indonesia 1.2 0.9 31.8 28.7 The crisis aind its caiuses Korea, Rep. of 0.4 0.3 35.2 34.3 Malaysia 2.3 1.1 31.8 37.8 Philippines 0.0 -0.4 20.1 19.6 The East Asian crisis differs from previous finan- Thailand 2.3 2.3 34.7 34.3 cial crises in developing countries in several ways: Source, World Bank and IMF data. 31 mismanaged exchange rates, insufficient financial Philippines (table 2.2). Only in Thailand did net regulation, and implicit or explicit government private capital flows as a share of GDP remain sta- guarantees-also played an important role in cre- ble, though averaging more than 15 percent over ating the incentives that led to the particular size the three years. The surge in flows reflected these and character of external financing and internal countries' strong economic performance, includ- resource allocation. Of course, every loan involves ing rapid growth, sustained improvements in a decision to lend as well as to borrow. Thus macroeconomic balances (public sector balances, lenders, many of them international banks, share inflation), and structural changes that have fos- with borrowers responsibility for the consequences tered a market-led, outward orientation since the of bad loans. Similarly, responsibility for resolving late 1980s. The cyclical downturn in internation- these problems lies not just with bank supervisors al interest rates in the early 1990s provided the ini- in the borrowing countries but also with bank tial impetus for the surge in flows (particularly supervisors in the lending countries-particularly portfolio flows); continued increases reflected if the international community believes that there structural changes that have increased the respon- is sufficient systemic risk to the global economy to siveness of capital to cross-border investment warrant intervention. opportunities (see World Bank 1997c). The buildup of short-term, unhedged debt left Absorbing large inflows of private capital and East Asian economies vulnerable to a sudden col- deploying them productively presented challenges lapse of confidence. The loss of confidence led to to economic management. Despite the good capital outflows, and thus to depreciating currencies macroeconomic track record of the East Asian and falling asset prices, which further strained pri- countries, three sources of vulnerability emerged: vate balance sheets and so proved self-fulfilling. And large current account deficits and misallocated once started, the scale of capital outflows may have investment, bad loans and currency and maturity had littde to do with currency mismatches or even mismatches, and misaligned real exchange rates financial sector weaknesses, except to the extent that and lost competitiveness. they contributed to perceptions ofweaknesses in the economy. East Asia's close integration with interna- Large current account deficits and misallocated tional financial markets makes it easy to export cap- investment. Although East Asia has high savings ital (although in a major confidence crisis even rates (averaging more than 30 percent of GDP in countries with relatively closed capital accounts can Indonesia, Korea, Malaysia, and Thailand during experience substantial capital flight). In fact, there 1993-96), foreign capital was required to finance is evidence that domestic investors were major par- its even higher investment rates (which rose to ticipants in the outflow of capital-an experience about 40 percent of GDP in Korea, Malaysia, and similar to that in Mexico during 1994-95. The Thailand and 34 percent in Indonesia during downward spiral has accelerated as financial prob- 1993-96). Still, financial crises result more from a lems have led to restricted credit, undermining the loss of confidence in borrowers' ability to repay real sector and further contributing to financial than simply from the existence of large current fragility. If economic problems fuel social and polit- account deficits. Indeed, the statistical link ical tensions, the crisis will deepen further. between current account deficits and currency crises appears weak (box 2.1). Manifestations of vulnerability Table 2.2 Net private capital flows to East Asia, Large amounts of private capital have flowed to 1994-96 East Asia in recent years, although the scale and (petage of GDP) phasing have differed significantly among coun- Country 1994 1995 1996 tries. Between 1994 and 1996 net private capital Indonesia 0.3 3.5 6.1 inflows as a share of GDP increased considerably Korea, Rep. of 1.2 2.0 4.9 several Eas Asian counries-for by Malaysia 1.2 6.2 8.4 in several East Asian countries-for example, by 7 Philippines 7.9 8.4 12.7 percentage points in Malaysia, 6 percentage points Thailand 14.3 17.3 14.5 in Indonesia, and 5 percentage points in the Sourre: World Bank data. 32 Racent tteconomic literstatur hlei0nrp70lo0redteatiosn ofl predictor of curec csl:$*risesfabankigcrisi mayth shipnoenweesi vareious econonn t is ¢ndXicnatoXrls aind th u cnidermie confdec pein thte ecounomy,sS Strigein of |ieiho of an exchange rare ptcries0;is The preudtciv curenc cfh risis(am0iniskyadRinhart 1996). Man:y poe f hs mds-speinally iifou«p0iten of rwdsample-is b9an-k 0ingcrises hiavebeen receded b ftlending beos; 9,00 o curec Crie col _ made, dOethe ciseswould iflatows' an financial liberaliation Atourgssona thex predp tifor woeaast Asa'sde b norm broing ths Thusedf cure c urisis asofen heraldedm utreaie tn Genra wase luhitheratr finds99 that eal ena aset ino iterationaveseorvs. Winthaen anb gexanger rate apurchalowr(igurei do1)esOnly ordl ec vner craedi is e loed, a o decline uin it ree squetz- Itn banksings crdest ano thepo eve rtof22 inercnatioa- tl leadse t pclaiat on the maurrencymach resegve hare bencusefu foreu concheprob hsthie Th utiability of an rga 99.Svrlstdeharge fourrnd accoin- ofchane ratecrissal i ceb rrenots raouwsnot defcit- a denficaits reartlydpndsh bewen hwthe le fuindarerna-d thariscly bideingcontareinot o utebroig tIfr reservesianbroeshd terbainvetyeo ah currency Theiarioe toabe ifmsortan-itpedicting thed collape ofi dinerkyfand, Reinhaprout199t) Onevespntionts tor vat necom,hwee,inalrechane drthe regime ' ( vam ne1988 thficis reatonshi ithate beserne musayideclie. prio intoa ab plet eovidncethrat an subtarntal apprcistatio cofl the h todfen the currel ncy w(adte.oncdn Mnexicneipvestorcsis,forfiexample,Evherealsexchange policies. The dosticinavietyfsuchnt) adefiipt dea d tos yearsbefor the risis and ost o the ountres ny invetmentand wetherthe eterna iaebfiltes IAppeitiong wascept asignircant facto Ain h aseo hls de tetrmbrmoine the suseadaito ofrcas currngter relaive won,D hwever.) ihe i 99ta acoutdeficit asetuis gehanverally have been unable tob LtnAexpcading debom musatihve an aexp uatea findtial systematcy relatoshipretwee the rfsize of crdthes enredit to uNchve longwer(iod.gute rapi growt (Onl bankilg sysem, inedictlies thatt macroeconomiy policesz Idomestic credit can exceeds thatiwarante byercon- mybeaunsusoftaiabe, mandripersistndeitsmayterod omich funavebentcalse andohu canc signal Tunhie Tesustain-ofdne andpeipitatef larg currenrs.Stude acofn ablte maerocnomi detepolicis. Satuies have foundcea whethrficiscparly deficits are useow preditr ofuncurrencyed sarlignificnt reatonshrip t beoen frapid bored wipng. rIse have hsadnie torrow srs howeiverte ( hedwards 1989 Tihn lande thaeprbilt of ahr-tr devautione b (tahsri dichenrsfeend o, h anprductpvity in995)nt, pehap vtesorne, n hoeveasic ese teego9s)ler befciuse warge hisale desaen ofsteinarelaed Bto o ntherard ablt stuyof 25era bankietengl crises anda7 courrnc michr of curenc crpises (suchw (asnapd texpanincidnt relatmive tonvPewsto mchnhighner ivn 1996e thant external asset(sothat investorsnt didar not havesb 33 Figure 2.1 East Asia's debt-export ratios ment in GDP rising from 30 percent in 1983-89 were lower than Latin America's in 1996 to 38 percent in 1995-96 while GDP growth fell Percent from 10 percent to 8 percent. The increase in the Argentinae capital intensity of output partly reflected a shift Brnza E.~ iE iEEiiEiEEsfEfEE away from labor-intensive sectors (such as gar- ments, footwear, and light assembly) toward what Mexico T were considered strategic sectors for Korean indus- trialization (including automobiles and electron- Indonesi ics). High capital spending led to overcapacity in many sectors, as the country's chaebol (industrial Korea, Rep. of conglomerates) pursued investments hoping to build market share both domestically and overseas. Moloysia To some degree, however, the decline in GDP growth relative to investment could also reflect Philippines changcs that lowered the return on the capital ThailInd stock-such as shifts in international demand for computer chips-rather than a decline in the pro- 0 50 100 l50 200 250 300 ductivity of new investment. Source: World Rank Debtor Reporting System. The low productivity of investment and the growing vulnerability of the corporate sector can been directed into risky, low-productivity invest- also be seen in the drop in Korean companies' ments. For example, a large portion of domestic return on assets (to 1 percent in 1996) and the credit flowed into nontradables, especially real spate of bankruptcies in 1997 (prior to the spread estate. Market sources report that real estate loans of the crisis to Korea). Similarly, the return on accounted for about 25 percent of outstanding equity in Indonesia, Malaysia, and Thailand bank loans in Malaysia and the Philippines2 and declined between 1992 and 1996 to below money 20 percent in Thailand. The surge in the region's market interest rates, indicating that there was no real estate investments rapidly outpaced demand: compensation for the risk of investing in these by 1996 vacancy rates had reached nearly 15 per- economies. cent in Bangkok and Jakarta. The poor quality of real estate investment in Thailand was reflected in Bad loans and currency and maturity mismatches. the stock market, where the index for building and Low-productivity investments were reflected in furnishing companies collapsed from a peak ofjust banking systems' high nonperforming loans, which below 8,200 in late 1994 to 1,100 in late 1997. according to market sources reached 19 percent of The concentration of loans in the highly cyclical loans in Indonesia, 16 percent in Malaysia, 13 per- property market led to asset price bubbles and cent in the Philippines, and 17 percent in Thailand made the banking system vulnerable to a down- in the immediate aftermath of the devaluation.3 turn in domestic demand. And the concentration Conservative estimates suggest that by the end of of bank portfolios in nontradables-along with 1997 the equivalent of 6.5 percent of GDP was currency mismatches-made them more vulnera- required to restore the capital adequacy of Korea's ble to a depreciation in the currency. merchant, commercial, and specialized banks. More generally, an increase in incremental cap- Some East Asian banking systems were made more ital-output ratios in East Asia may reflect the poor vulnerable by the fact that a lot of external borrow- quality of a large portion of new investment dur- ing (especially in the past few years) was short term ing the 1990s (box 2.2). Although in recent years and was used to fund local currency-denominated the share of investment in GDP rose across East assets, particularly in real estate and other nontrad- Asia, GDP growth rates remained roughly con- ables. Even when banks avoided currency mis- stant and even fell in some countries. For example, matches, their clients often had significant foreign the efficiency of Korean investment declined in the exchange exposure, so the banks remained vulnera- years preceding the crisis, with the share of invest- ble to a fall in the exchange rate. 34 domesi inextment in threnc yarsprcdin mtheuritmiss. fieryeat peiod endoeing 199y Juo as9 heaigh as 25he Butnwaies threqaity ofxschaingeexpstmenta failing? Onetiedastbe1el 996gertedspt)2 u easre of Theiprodu of invstmntn is t e rA driing is ten indei te p torer yqusa inem'nefoeintalicapitliotput rasetio (2C0 ) which netent (fpirt esptecillefr neerinetmnt)n ompaesti cumultiv in vest,cmpaent with thperchgeint geTher losinancialvulerailityes to aSlowdowns (sifath gin do0Bymesicdoutpu sovt-erma gien t perid. riCsen icnv mesn ias rfinacted bybroinga sbthantwill neewdtow ca00lculae s fieya oig averages toj reduce be $re:paid out offuure oupt. VButt othr0 fators tylia pefects, havel drisen s yinKreaaadnd6 structuralchaes. cpita d9eepening, Sthe Ausiness iThailand ( omparbout in thefive-yearperiod end- cycle e is dgurngy anslowewndf fI during tig s990etoe w re haigin Indo ny 16 binased sg aelowegry-sowt rafe oth valu thade aCR Crises The i thernd cut riesc and mayteris and t h t afs f erced Etoe sianl and dot r o ries decinvohe mathilpies (tabled conideabl preiptong EatAin 1889 (vrg;figure 2.32 IThiand tbexp4) ortev- ecoomis. oregn xchngeexpsur wa patic mantanues ros jstablpercentli txhane yapreedesitegau27 declid significal in C a.es i n l apa aompe 1997s here roprd iy an mshoar rise in Crhre s durnstie Dures ing re 199ma capiarealoutput ratis the3firsd in lf The l eos o o ay andshave ilan dpa- aprctinTatead, by pae wit1 percent in the Pipins13 iesxportfievenues grewrbydanraverageson to believe preThaind Maayia 9owercn in Thalaysia and 5the at volumengrowth felland debta prclbes adelne Phlppns (al .) rcptul fgr .) nTaln xotrv 1900nes 1urin the 98 1 1990s 1992 1994ng rates 19he first 1af 18f 19he 1990s 1may 1994 giencopa Thpeextent of curnyadmauiyms percent in Indoilppnes,1 nisadfacia(b Juner edi1997 relative to theliv uparcey highaasi,9pecn in Thailand, where theaakn y- pret drop inhisermsd dofltardebinth three years- 35 Table 2.3 Ratios of short-term debt to total debt Table 2.4 Real effective exchange rate indexes, and to reserves in East Asia and Latin America, December 1 996-December 1997 mid-I 997 (1988-92 average= 100) (percent) December June December Short-term Short-term Country 1996 1997 1997 Country debt/total debt debt/reserves Indonesia 104 105 59 Indonesia 24 160 Korea, Rep. of 89 87 59 Korea, Rep. of 67 300 Malaysia 109 113 82 Malaysia 39 55 Philippines 114 118 85 Philippines 19 66 Thailand 108 109 74 Thailand 46 107 Source: J.P. Morgan data. Argentina 23 108 Brazil 23 69 because the crisis (and its severity) partly arose from Chile 25 44 the interaction among these factors. Furthermore, Colombia 1 9 57 Mexic 1]6 126 these problems were neither necessary nor suffi- cient for a crisis to occur. For example, financial Sorce, Bank for International Settlements data; IMF Interational c Financial Statistics; and World Bank data. sector weaknesses clearly contributed to the crisis, but many countries have weak financial sectors and quately serviced. The export slowdown, however, do not experience a crisis. Moreover, stronger limited their ability to do so. financial regulation may not have averted the crisis in East Asia, because incentives for unhedged cor- Causes of vulnerability porate borrowing would still have existed. And the small size of these economies meant that they were The root causes of the mounting vulnerability of not widely diversified-and thus were potentially East Asian economies (the extent of which differed more volatile-providing incentives during a peri- significantly among countries) were financial sec- od of uncertainty for capital flight to international tor weaknesses and poor corporate governance, currencies. Thus what follows should be seen as an macroeconomic conditions and the policy initial presentation of the main reasons these response, and a lack of due diligence by external economies developed the vulnerabilities that made creditors. But the precise relationship between a crisis possible. Further work is needed to analyze these problems and the crisis is complicated, the contributions of each problem. Figure 2.2 East Asia's real effective exchange rates Figure 2.3 As dollar prices fell, East Asia's exports have appredated in recent years slowed Index: 1988-92 overage 100 Year-syear pertentage thGnge Yen/dollar 120 35 140 1hihppins 1 30 j Korean exports /s1\/* w \ ~~~~~YendollurJrX 110 Mlys Philippines 25 't rate , 20 ASEAN' ' / 120 \ 'h,,,""\ l ' j - " Thailand exports, 1 00 15idnoi 100 X ' ~~~~~~~~~Indonesia - .F7X........ 10 1 90~~~~~~ V , A Sgr Ia.. ~ ~ Korea, Rep. of O 80 -5 80 Januory Januory January January Jonuary June Jonuory January January Jonucry January January Sept 1993 1994 1995 1996 1997 1992 1993 1994 1995 1996 1997 Note: Calculated as 12-month moving average in current dollars. Source: J.P. Morgan dota. Source: World Bonk and IMF dots. 36 Financial sector weaknesses and poor corporate money-losing units with favorable transfer pricing governance. Distorted incentives, inadequate dis- and implicit and explicit debt guarantees. While the closure and supervision, poorly managed financial extensive system of cross-guarantees was developed liberalization, and lax regulatory standards result- to serve a market need and worked well for many ed in weak financial sectors and poor corporate years, it made it difficult for outsiders (like minority governance in many parts of the region. In many shareholders or the market in general) to ensure countries financial systems allow insider lending appropriate accounting for and allocation of losses. and close connections between lenders and bor- Poor corporate governance was also evident in the rowers. Implicit or explicit government guarantees lack of impartial audit committees and of indepen- have encouraged banks to take excessive risks (as dent directors in many companies. Although East the banks would receive high profits if projects Asian equity markets grew quickly, limited trans- were successful, and the government would absorb parency made it hard for them to impose discipline the losses if projects failed), further eroding incen- on corporate behavior. The largest corporates (for tives to evaluate credit and contributing to inflat- example, some of the chaebol) may also have been ed prices of financial and physical assets (see viewed as "too big to fail," thus encouraging exces- Krugman 1998). For example, Korea's chaebol sive risk taking. may have been able to obtain loans from banks in Financial sector supervision in East Asia gener- part because of their direct shareholdings in and ally has been weak, and regulations relatively lax. support for associated firms (and, until recently, Countries lacked the institutional capacity to cope through their influence on the government, which with the rapid expansion of domestic credit during directed a portion of bank lending). Thus, despite the 1990s. Reporting and provisioning require- their high debt to equity ratios-which for the 25 ments for nonperforming loans were inadequate biggest averaged 4:1 in 1996-the chaebol in several countries. (In Indonesia, for example, a obtained sufficient credit to finance continued loan can be nonperforming for 24 months or more expansion. And when the financial position of sev- before it is deemed uncollectable.) In several East eral chaebol deteriorated in 1997, the government Asian countries capital adequacy requirements are provided funds to cover some of these losses. more lenient than those suggested by the Bank for The lack of transparent and timely balance International Settlements (BIS) even though these sheet and other information in most East Asian economies face higher risks than the industrial economies led many banks to base credit decisions countries that follow BIS standards. Finally, many on the availability of collateral rather than on an countries lacked effective exit mechanisms for analysis of cash flows. This practice was particu- failed institutions, so insolvent banks (which have larly problematic in Korea, where nontransparent little incentive to behave prudently, since they systems of interrelated ownership within a chaebol have no equity to lose) were allowed to continue made it difficult to judge the value of collateral lending. offered as security for loans. Overreliance on col- These problems were exacerbated by the rapid lateral skewed lending toward real estate and con- liberalization of financial markets without a com- struction (where the property or building can serve mensurate strengthening of supervision and regu- as collateral), and the use of equities for collateral lation. In the late 1980s and early 1990s the Thai increased the vulnerability of banking portfolios to government reduced reserve requirements, eased a downturn in the stock market. the rules governing nonbank financial institutions, Distorted incentives and low transparency affect- and expanded the scope of permissible capital ed corporates as well. Many major companies in East market activities (allowing banks to finance equi- Asia are closely held, and despite recent improve- ty purchases on the margin, for example). ments in disclosure and other requirements there is Deregulation of the Korean financial sector, intro- inadequate protection for minority shareholders. duced in 1993 (partly under pressure from (For example, a recent scandal in Malaysia demon- Western governments), eliminated many interest strated that shareholders lack sufficient recourse in rate controls, removed restrictions on corporate the event of malfeasance.) In Korea and Indonesia debt financing and cross-border flows, and allowed profitable units of industrial groups subsidized increased competition in financial services. 37 Liberalization in the absence of adequate regu- rency regime was adopted) and June 1997. This lation or supervision greatly increased the vulner- stability inspired confidence that the exchange rate ability of financial systems. Opening banking would continue to hold in the future. When com- systems to new entrants lowered the franchise bined with liquid capital markets and low inter- value of existing banks. Lax enforcement, implic- national interest rates, exchange rate regimes it government guarantees, and inadequate mini- created strong incentives for unhedged borrowing mum capital requirements encouraged risk taking. from abroad. And increased access to offshore funding made it During 1994-96 Indonesia, Malaysia, and easier for banks to take on excessive foreign Thailand saw a widening of their current account exchange risk. The liberalization of financial mar- deficits (driven primarily by an increase in imports) kets at a time of easy global monetary conditions and a counterpart surge of inflows on their capital encouraged a surge in borrowing, and domestic accounts, accompanied by an increase in asset credit jumped in Southeast Asia (table 2.5). For prices in some sectors. By early 1997 these pressures example, in 1996 domestic credit rose to 130 per- had abated. Inflation increased slightly in cent of GDP in Thailand (compared with ratios of Indonesia in 1995, from 8.5 percent to 9.3 percent, 22 to 73 percent in Latin America), further but by 1996 it was down to 8.0 percent (table 2.6). increasing the economy's vulnerability to shocks to In Malaysia inflation jumped from 3.5 percent in the banking system. 1994 to 6.0 percent in 1995, but by 1996 it was back to 3.6 percent. Although inflation in Macroeconomicconditionsand thepolicy response Thailand increased to 5.9 percent in 1996 (from during 1994-96. Regional vulnerability also 5.0 percent in 1995), it remained moderate by increased as a result of the incentives for borrow- international standards (although above the U.S. ing created by macroeconomic policies-particu- rate when the baht was closely tied to the dollar). larly exchange rate regimes. This inclination to In all three countries capital inflows generated borrow was reinforced by liquid international strong upward pressures on domestic demand and markets. In most East Asian countries exchange the exchange rate. Given the size of the capital rates were tied to a basket of currencies in which inflows, the healthy state of public finances, and the the dollar had a large weight. In Thailand, for small size of governments, it was impractical to instance, the exchange rate ranged from 25 to 27 entirely offset the rise in demand through a fiscal baht to the dollar between 1984 (when the cur- contraction, although these countries did run fiscal Table 2.5 Credit growth in various countries, 1990-96 (percent) Annual growth of Annual growth Loan growth! Net domestic credit/GDP Country nominal GDP of loans GDP growth 1990 1996 Indonesia 17 20a 122 45 55 Korea, Rep. of 14 17 123 68 79 Malaysia 13b I8a b 134b 80 136 Philippines 13 33 264 26 72 Thailand 14b 24b 176b 84 130' Argentina 28 23 82 32 26 Brazil 540 447 83 88 34 Chile 21 20 93 78 73 Colombia 28 34 120 36 46 Mexico 24 14 60 37 22 Germany 6 9 138 123 141 Japan 3 2 80 162 157 United States 6 8 140 109 123 Note: Loans include nonbank financial intermediaries. a. Does not include nonbank financial intermediaries in 1990 and 1996 for Indonesia and in 1995 for Malaysia. b. Data are for 1990-95. c. 1995. Source: World Bank 1 997c; IMF International Financial Statistics, and Goldman Sachs. 38 Table 2.6 Macroeconomic indicators in Southeast Asia, 1993-96 (percent) Indonesia Mwlaysia Indicator 1993 1994 1995 1996 1993 1994 1995 1996 Current account balance (percentage of GDP) -1.3 -1.6 -3.5 -3.4 -4.5 -5.9 -8.5 -5.2 Export growth in U.S. dollars (percentage change) 8.4 8.8 13.4 9.7 15.7 24.7 26.0 9.3 Import growth in U.S. dollars (percentage change) 3.8 12.9 27.9 6.2 14.5 30.4 30.5 0.7 Inflation 9.3 8.5 9.3 8.0 3.7 3.5 6.0 3.6 Source: World Bank data. surpluses in 1996 (of 2.3 percent of GDP in some East Asian economies. The extent to which Thailand). Thus, as current account deficits rose, the end-1994 unification of the Chinese yuan the authorities turned to monetary policy to restrain affected this process is open to question, as about the increase in demand. Indonesia and Thailand 80 percent of foreign exchange transactions were offset a significant proportion of the increase in cap- already being carried out at the swap market rate. ital inflows through a contraction in domestic cred- There does not appear to have been any marked it-that is, they relied on sterilized intervention. change in the trend improvement in China's share Although sterilization reduced demand below of trade after the unification. the level that would otherwise have occurred, it proved to be a double-edged sword. For example, Lack of due diligence by external creditors and in 1996 short-term money market interest rates in poor external intermediation. East Asia's problems Thailand were 400 basis points higher than were not limited to or entirely caused by inade- comparable U.S. rates, encouraging short-term, quate financial regulation or bad choices by unhedged borrowing. By preventing interest rates borrowers. Lenders also bear much of the respon- from falling in response to capital inflows, sibility. These loans came from countries with sterilization also encouraged creditors to continue seemingly well-regulated and transparent financial lending. Finally, sterilization was reflected in large institutions operated by sophisticated managers accumulations of reserves. Between 1994 and without government intervention. Yet foreign 1996 international reserves in Indonesia, lenders and investors were not restrained by inad- Malaysia, the Philippines, and Thailand increased equate financial statements, high short-term debt, by about $30 billion (about one-fifth of net pri- or the unhedged foreign exchange exposure pre- vate capital flows). Holding such large reserves sent in the financing structure of East Asian banks imposes substantial costs on the economy, espe- and firms. In Thailand international banks were cially in developing countries, because the return willing to funnel short-term loans denominated in on liquid reserves is typically far lower than the foreign currency through offshore banking facili- cost of attracting capital from abroad. ties in return for the opportunity to establish The major swing in the dollar-yen exchange domestic operations. rate contributed to the appreciation of East Asian Foreign lending continued despite the improve- currencies. A shift in fiscal policy in Japan-a ments in the collection and dissemination of infor- major trading partner for East Asian developing mation on emerging markets that followed the countries-equal to about 2 percent of GDP, cou- Mexican peso crisis. Indicators such as fiscal deficits pled with monetary easing, led to a decline in yen and capital flows were being reported regularly. As long-term interest rates (which reached 1.1 per- in Mexico, there seems to have been sufficient pub- cent by October 1997) and a 50 percent appreci- licly available data in Thailand to allow observers to ation of the dollar against the yen between April foresee problems at least a year before the devalua- 1995 and July 1997. tion of the baht. Yet few appreciated the depth of In addition, structural reforms in China have the structural weaknesses in East Asian economies. improved its international competitiveness in In any event, rating agencies and international insti- recent years, resulting in large increases in Chinds tutions failed to adequately assess the region's eco- share of export markets, perhaps at the expense of nomic vulnerabilities. 39 Philippines Thailand 1993 1994 1995 1996 1993 1994 1995 1996 -5.5 -4.6 -2.7 -4.3 -5.1 -5.7 -8.1 -7.9 13.8 19.9 31.6 26.9 13.2 22.7 25.1 1.8 22.1 19.9 25.7 21.3 13.5 17.9 30.0 3.1 7.8 9.4 7.9 8.4 3.6 5.3 5.0 5.9 Loss of investor confidence: triggeringfactors and estate bubble dropped by nearly two-thirds). the emergence of crisis and contagion Confidence was further eroded by the eurobond default of Somprasong Land (a Thai property Although increased vulnerability did not necessar- company), liquidity support to failing financial ily mean that a crisis would erupt, it did make a institutions (some $16 billion had been channeled loss in investor confidence more likely. That mar- to nonbank financial institutions by July 1997), kets and market observers failed to anticipate the and the provision of guarantees to the depositors scope and severity of the crisis is strong evidence and creditors of banks and finance companies in that a self-fulfilling loss of confidence played an the face of continued pressures on the baht. important role. The foreign currency debt of most The rapid spread of the crisis to Indonesia, East Asian countries had the same investment Malaysia, and the Philippines was due to aspects grade ratings in June 1997 as a year earlier, and rat- of their economies that were similar to those of ing downgrades occurred only once the crisis was Thailand's. While intraregional trade among the in full swing (table 2.7). Just a few months before four countries accounts for only 6 percent of the crisis the government of Thailand was able to their total trade, they are close competitors in borrow in the eurobond market at a spread of only world markets (and also compete intensely for 90 basis points over U.S. Treasury bills. In consid- foreign investment). Thus currency devaluation ering the effects investor confidence had on the cri- in Thailand created expectations of a fall in the sis, it is usefuil to distinguish between early events others. Market participants (including domestic in Southeast Asia and the spread of the crisis to corporates) reassessed the sustainability of other regions in October and November. exchange rate policies once it had been demon- strated that continued close ties to the dollar were The crisis in Southeast Asia. The crisis was trig- not guaranteed. And concerns about the sustain- gered by events in Thailand during the first half of ability of exchange rates dramatically increased 1997, when exports remained flat (in dollar the perceived riskiness of the currency and matu- terms), capital flows slowed (bond issues and syn- rity mismatches built up during the 1992-96 dicated loan commitments fell to $4.0 billion surge in capital inflows. Foreign investors also from almost $6.5 billion in the first half of 1996), may have perceived that loans to banks carried an and the stock market dropped 34 percent (as the implicit government or central bank guarantee. price of companies caught up in the burst real When this proved not to be the case in Thailand, Table 2.7 Credit ratings for East Asian countries, 1996-97 Stanaard e- Poor's Moodys Country June 1996 June 1997 December 1997 June 1996 June 1996 December 1997 Indonesia BBB BBB BBB- Baa3 Baa3 B2 Korea, Rep. of AA- AA- B+ Al Al Bal Malaysia A+ A+ A+ Al Al A2 Philippines BB BB+ BB+ Ba2 Bal Bal Thailand A A BBB A2 A2 Bal Source: Standard & Poors and Moody's. 40 risks and asset prices were reassessed. Despite government commitment to financial sector their differences (for example, the Philippines restructuring appears to have impeded recovery. In had a smaller burden of short-term debt and less some countries political transition created uncer- overbuilding in real estate than Thailand), the tainties over likely government responses to the cri- market treated these economies similarly-indi- sis. Potential increases in unemployment also raised cating that the market failed to discriminate ade- the specter of social unrest in some countries, con- quately among them or that their role as tributing to the downvard spiral. competitors was crucial to evaluating the sus- tainability of exchange rate regimes. The spread of the crisis. Confidence factors also The severity of the crisis, which far exceeded may help explain the spread of the crisis in effects resulting from the misalignment of October, as there was little new information to jus- exchange rates, developed because the weaknesses tify global declines in stock markets. The drop in of corporates and financial intermediaries encour- stocks reflected a flight to quality, as evidenced by aged a self-fulfilling loss of confidence. For exam- the 35 basis point drop in long-term U.S. Treasury ple, the large volume of short-term debt failing due bill yields between October 22 and 27. Large in Indonesia, Korea, and Thailand increased pres- financial losses reduced investor appetite for risk, sures on their currencies, as the turmoil in finan- resulting in an exaggerated response to the sudden cial markets made creditors reluctant to roll over drop in Hong Kong's stock market. And the spread credit lines. The resulting currency depreciation of the crisis began to lower expectations of global increased the local currency value of uncovered demand, reducing earnings projections and stock dollar liabilities of banks, finance companies, and market valuations. corporates, impairing balance sheets, lowering Given the large gaps in information about and stock prices, and increasing demand for foreign familiarity with emerging markets, financial exchange to cover open positions. Increased shocks can spread extremely rapidly among bor- demand for foreign exchange led to further cur- rowers of varying quality, resulting in severe over- rency depreciation, and so on. shooting. The advent of speculative attacks against The economic contraction further eroded con- the currencies of Singapore and Taiwan, China, fidence. The increases in interest rates made to which have strong economic fundamentals, mas- defend the currency and the rapid fall in the equi- sive reserves ($80 billion and $90 billion in mid- ty of highly leveraged financial institutions (due to 1997), and current account surpluses (estimated at their and their borrowers' losses from currency $19 billion and $10 billion in 1997), also suggest mismatches and the decline in real estate and share the importance of confidence factors in the spread prices) led to a credit contraction that impaired the of the crisis. position of otherwise healthy firms and increased Market concerns about emerging markets, foreign lenders' reluctance to roll over short-term combined with economic vulnerabilities, caused debt. And because investors had limited informa- investors to lose confidence in Korea. Short-term tion about these economies, market perceptions debt had increased from $18 billion in 1993 to were likely to change rapidly once the crisis began, more than $100 billion in September 1997. The exacerbating the decline in prices. current account deficit had increased from 1 per- The crisis was also exacerbated by a faltering pol- cent of GDP in 1991-95 to 5 percent in 1996. icy response and more generally by political uncer- The terms of trade h ad dropped 15 percent in tainties. Thailand's government did not move fast 1996 with the falling prices of computer chips and enough to deal with insolvent financial institutions, the appreciation of the dollar against the yen. and market confidence also suffered when it was GDP growth had slowed from 9 percent a year in revealed that losses on forward foreign exchange 1994-95 to 7 percent in 1996, and debt servicing transactions to defend the baht were much higher by the highly leveraged chaebol depended on rapid than initially believed. In Malaysia government growth. Starting with the January 1997 collapse of statements caused concern about the possible impo- Hanbo Iron and Steel, Korea's fourteenth largest sition of capital controls, encouraging further capi- conglomerate, 8 of the 30 largest chacbol received tal outflows. In Indonesia uncertainty about court protection from creditors. Market confi- 41 dence was further impaired by the national assem- will improve transparency. In addition, the ceiling bly's failure to provide government guarantees for on foreign ownership of stocks was eliminated, the debts of the chaebol.4 and the limit on individual ownership of banks The onset of the economic crisis in Korea in will be lifted for all investors. November appears to have been triggered by the general collapse in confidence in emerging mar- Ramifications of the crisis beyond EastAsia. Most kets in the context of mounting short-term debt. Latin American economies were not greatly affect- Korea's macroeconomic environment had actually ed by spillover effects from the devaluation of improved somewhat during the first three quarters Southeast Asian currencies in July and August, of 1997. Export growth had started to pick up, although the Brazilian market dropped 20 percent with export revenues averaging about $1 billion a in July. The stock market turbulence in October month more than the average for 1996. GDP did more damage: between October 22 and 27 growth was 6 percent, inflation was 4 percent, the stock markets in Argentina, Brazil, and Mexico fell current account deficit had narrowed, and the 20-24 percent, and some countries experienced budget was in surplus. exchange rate pressures (the Mexican peso depreci- Mounting bankruptcies, exchange rate pres- ated 9 percent by October 27). By mid-November sures, and concerns about the liquidity support three-month interest rates had increased by 2,300 that could be required for corporates and banks led basis points in Brazil, 600 basis points in Mexico, Standard & Poor's to downgrade Korea's long- and 500 basis points in Argentina. In addition, sec- term foreign currency debt rating in late ondary market spreads on the debt of Brazil and November. Creditor uncertainty about the extent Argentina rose by some 200 basis points and of of financial sector problems made it difficult to roll Mexico by 100 basis points. By the end of the year, over short-term credit lines. The stock market fell however, Argentina and Brazil had maintained 19 percent in November (and by the end of their exchange rate regimes and the Mexican peso December was down 51 percent since July), the had recovered to within 4 percent of its October 22 won depreciated 20 percent in November (down level. Although stock markets remained below 80 percent from July to December), and interna- October levels, for the year as a whole stock prices tional reserves dropped below $10 billion. These in the seven largest Latin American economies rose difficulties led the Korean authorities to ask the by an average of 21 percent (in dollar terms). International Monetary Fund (IMF) for financial A number of factors explain the ability of Latin assistance to help meet foreign exchange obliga- American countries to weather the storm. In recent tions and restore investor confidence, and to ask years a few countries in the region have come to commercial banks for a negotiated refinancing of depend more than the most affected countries in a sizable portion of short-term debt obligations. East Asia on foreign direct investment to finance Korea's crisis worsened in December because of current account deficits, and foreign direct invest- concerns over the policy response. The sharp jump ment tends to be more stable than debt flows and in estimated short-term debt following the portfolio equity. Though much remains to be announcement of the IMF program increased done, countries in Latin America have been market worries about the severity of the refinanc- strengthening their financial systems, particularly ing problem. Uncertainty over the economic poli- since the 1994-95 Mexican peso crisis. Efforts have cies that would follow the December presidential been made to improve prudential regulation and election also undermined confidence. In early banking supervision. Weak institutions have been January, however, the exchange rate stabilized, the closed or absorbed by stronger institutions. Public incoming administration reiterated its support for banks have been privatized. And restrictions on for- the IMF agreement, and the government moved eign participation in the financial sector have been quickly to implement economic reforms. The eased. In recent years Latin American countries national assembly passed legislation that gives the have not experienced lending booms comparable Bank of Korea more autonomy over monetary pol- to those in East Asia, and domestic credit accounts icy, unifies regulation of the financial sector, and for a smaller share of GDP than in East Asia (see mandates consolidated financial reporting that table 2.5). Moreover, currency and maturity mis- 42 matches are probably less of a problem in Latin currencies and increasing interest rates. Policies America, where a history of high inflation and large discouraging short-term external borrowing also exchange rate fluctuations has made lenders and may have helped Chile and Colombia avoid many borrowers familiar with and adverse to the risks of of the problems plaguing East Asia (box 2.3). unhedged exposure. Still, many Latin American Although Latin America managed to avoid economies continue to harbor significant financial sharp currency devaluations and experienced a sector weaknesses. recovery in stock prices and secondary market Another factor in Latin America's favor was the spreads, the region has not been untouched by the swift policy response by the authorities, particu- turmoil in Asia's financial markets. Secondary larly in Brazil. Brazil was perceived as vulnerable market spreads on external debt instruments because of its high fiscal deficit (rather than high remain high in several countries. For example, private capital inflows, as in East Asia), growing spreads on sovereign bond issues in Argentina and external account deficit, and inflexible exchange Brazil exceeded 300 basis points at the end of rate regime. To protect the economy, the govern- December 1997, more than 100 basis points above ment doubled overnight nominal interest rates to September levels. And prospects for near-term more than 40 percent and introduced emergency growth in the region have been revised downward fiscal measures projected to yield about 2.5 percent in view of the increased uncertainty, fiscal con- of GDP. The measures, with some amendments traction, higher interest rates, more difficult access but the same aggregate impact, were approved by to foreign financing, deterioration in terms of Congress in mid-December. trade, and increased competitiveness of East Asian Argentina also was considered vulnerable economies. The World Bank anticipates that GDP because of its exchange rate system (a currency in Latin America will rise by 2.7 percent in 1998; board) and its close commercial ties to Brazil. At before the crisis growth was expected to reach 3.7 first, interest rates increased markedly. But unlike percent. The greatest drop will occur in Brazil, after the Mexico crisis, there was no significant which has had the strongest policy response to the outflow of international reserves, bank deposits crisis. Brazil's GDP is expected to increase only continued to grow (although with increased signs marginally in 1998, compared with a mid-1997 of dollarization), and interest rates subsequently forecast of 3 percent. At the time of writing, the declined. Argentina regained access to the dollar impact of contractionary policies on output in bond market by early December. Latin American Brazil were just beginning to be seen in the data. countries with more flexible exchange rate The industrial federation reported a fall in sales of regimes-Chile, Colombia, Mexico, Peru- 9.4 percent in November, about half of which may absorbed the external shocks by depreciating their have been due to seasonal factors J.P. Morgan). Pa 2 Bor 23apaS cntrol in - - is and Csulonbk thog neueAXe reserve requirlements o:n hc01rave ngi thn0e voue of flows, asa hnge int flos th atnvariesivrseflywihX an mauit. Inid 199 i reserv i0t taxta dc si+gnif00icarndy; increse the Ointerest di; er financial investments, including issues of American~~~~~~~~~~fi ineetrts cnmticsitudie tha use thisi 43 Markets in Europe and Central Asia also suf- percent between late October and mid-December, fered from the turmoil that roiled global stock mar- and the stock market fell 15 percent. By the end of kets in October 1997. While exchange rates in the the year, however, the exchange rate and stock mar- region emerged relatively unscathed (except for the ket had recovered somewhat. Otherwise, the conta- depreciation in Turkey later in the year), Russia's gion effects of the crisis were limited in South Asia, stock market dropped 12 percent between October possibly because current account deficits are smaller 22 and 27, Turkey's fell 9 percent, and Poland's lost and capital accounts are more heavily regulated than 7 percent. By the end of the year Russia's interest in East Asia (although financial systems in the region rates had reached 42 percent and Ukraine's 45 per- share many of the weaknesses that affect East Asia). cent. Unlike in Latin America, the decline in many The crisis erupted at a time when foreign exchange of the region's stock markets continued, and by the reserves were relatively comfortable (except in end of December the market in Russia had fallen Pakistan), particularly in India (twice the stock of by 26 percent, in Poland by 18 percent, and in short-term debt and the stock of portfolio invest- Turkey by 12 percent relative to October. Still, ment). South Asian currencies were experiencing stock markets remained well above levels at the only a modest real appreciation and, except for beginning of the year (on 31 December Russia's Pakistan, only small current account deficits. Private market was 125 percent higher than on 1 January). capital flows to India had remained moderate- The effects the crisis had on Eastern Europe about 1.5 percent of GDP during 1994-96. may reflect the emergence of vulnerabilities simi- Although reforms are under way to liberalize lar to those in East Asia. Large current account and strengthen overregulated banking systems in deficits have emerged-notably in the Baltics, India (and a banking sector reform program is Romania, and the Czech and Slovak republics- under way in Pakistan), liberalization is being and are being financed largely by private capital accompanied by a tightening of prudential regula- inflows. Short-term borrowing has increased tions and higher real interest rates. As a result it has rapidly in some countries, although the stock of caused only a modest expansion in credit. Capital short-term debt remains low (accounting for just controls are being relaxed in India (in particular, 26 percent of reserves and less than 10 percent of Indian authorities have raised ceilings on medium- debt in Russia, for example). The share of private and long-term external commercial borrowing), capital inflows being intermediated by the finan- but restrictions on banks' and corporations' short- cial sector is also increasing rapidly. And some term borrowing have been maintained. As a result countries in the region share weaknesses with East Indian banks' short-term external debt has Asia-in particular, fragile financial systems, low remained small, as has the private sector's uncov- transparency, and poor corporate governance. ered foreign exchange position. In general, South Moreover, in contrast to East Asia, public sec- Asia has not experienced the rapid credit expan- tor deficits are large in Russia, Turkey, and sion and currency and maturity mismatches that Ukraine, and all three countries rely on private for- contributed to the crisis in East Asia. eign funding at short maturities. Although real Most countries in the Middle East and North exchange rates appreciated 34 percent in Russia Africa and Sub-Saharan Africa were untouched by and 17 percent in Poland during 1994-96, these the crisis. The main reasons were the limited pri- rates were significantly undervalued in the early vate capital flowing to these regions (the $22 bil- years of the transition from socialism. Exports lion in net long-term private flows to the two from the region have been growing strongly (part- regions in 1997 was only 9 percent of the devel- ly because of increases in the prices of commodity oping country total and 2 percent of their GDP) exports), and an asset price boom in nontradables and, with some exceptions, their relatively unde- similar to East Asia's in real estate has generally veloped stock markets. been absent. Thus there is little evidence that exchange rates are misaligned. The response of the international community The East Asian crisis had a relatively modest impact on South Asia. After a briefspeculative attack The international community is supporting the on the Indian rupee, the currency depreciated 10 efforts of countries affected by the crisis to adjust 44 economic policies. The rescue package for Thailand pens, investors may take on more risk (including may total $17 billion, including support from bilat- more leveraged positions) than is warranted and eral donors, the IMF, the Asian Development Bank, asset prices may become overvalued, increasing and the World Bank. (The World Bank's contribu- systemic risk. Note that this problem influences tion is envisaged to be a $1.5 billion structural private sector decisions, not public sector ones. adjustment loan and a technical assistance project The possibility of international support in the to address the immediate needs of the financial sec- event of a crisis is unlikely to be viewed as suffi- tor.) The IMF has agreed to extend the Philippine cient compensation for the difficulties that gov- $1.05 billion Extended Fund Facility, which had ernments confront in the event of a crisis been set to expire in July 1997. The emergency aid -witness the recent experience of governments in package for Indonesia, including funds from the East Asia. IMF, the World Bank, the Asian Development Moral hazard can be mitigated if financing is Bank, and bilateral donors, totals $33 billion. (The provided in the context of strong policy reforms, World Bank is preparing $2 billion in adjustment including measures to ensure greater prudence in lending and a $15-20 million technical assistance future borrowing. Moral hazard will also be project to support the banking sector, in addition to reduced to the extent that investors are made to the $2.5 billion that was already planned to be dis- absorb some of the losses. There can be a tradeoff bursed over the next three years.) Most recently, the between establishing the right incentives for the largest ever international financial rescue package future and responding to immediate problems. was organized for Korea, with bilateral and multi- Future incentives are improved if owners of insol- lateral contributions expected to total $57 billion, vent domestic banks bear the first losses (to the induding a three-year, $21 billion standby arrange- extent of their equity). But in the absence of ade- ment from the IMF and up to $10 billion from the quate regulation, falling net worth could encour- World Bank. The Bank approved an initial $3 bil- age further risky behavior, and failing banks could lion economic reconstruction loan for Korea in late further contract the economy. December. Combined, international rescue pack- In the current crisis, portfolio equity investors ages for Indonesia, Korea, and Thailand could total in the countries affected have generally lost large some $107 billion, with the World Bank con- amounts as a result of'steep price declines and cur- tributing up to $14 billion. rency depreciations. Bondholders have experi- International financial institutions have an enced defaults, and the market value of their important-and difficult-role to play in respond- holdings has declined precipitously. By contrast, ing to the crisis. They must strike a balance short-term lending by foreign commercial banks between responding promptly to requests for has generally been protected, and in some coun- financial support (which may be required to reduce tries governments have provided implicit or systemic risk) and ensuring that governments com- explicit guarantees of short-term debt incurred by mit to strong corrective actions to address funda- domestic corporations and financial institutions. mental weaknesses. Moreover, these actions must While such actions may be needed to maintain be carefully designed to restore confidence and financial stability, they adversely affect the com- address the problems that caused the crisis. A pri- position of flows to emerging markets and the level ority in many countries is to strengthen financial of risk that market participants are willing to take. institutions and corporate governance. Socializing private sector losses that otherwise At the same time, the ready availability of funds would have been borne by foreign lenders could to support countries suffering financial crises rais- also be perceived as unfair to taxpayers, especially es concerns about moral hazard, and the extent if workers bear the costs through higher unem- and conditions of financing offered during eco- ployment and lower real wages. Ensuring that nomic crises have been the subject of much debate. commercial banks roll over debts as a condition for Official institutions can provide crucial support international assistance would facilitate adjust- for resolving a crisis, but they can also undermine ment and reduce moral hazard by having them market discipline if support for other countries is shoulder some of the risk. For example, during the anticipated in times of difficulty. When that hap- debt crisis of the 1980s commercial banks were 45 required to absorb losses through the restructuring other countries with sufficient financial capacity of countries' debt in conjunction with the provi- will provide resources to forestall or cope with sion of international assistance under the Brady impairments of the international monetary system initiative. or to respond to a situation that threatens the sta- Recent financial crises have prompted interna- bility of the system. These arrangements will go tional initiatives to help prevent future crises. For into force when participants (including the five example, emerging market economies and coun- largest) with credit arrangements totaling SDR tries in the Group of Ten (G-10) are working 28.9 billion ($41 billion) have agreed. The IMF together to strengthen financial systems in devel- also has approved the Supplemental Reserve oping countries (box 2.4). The IMF is helping Facility, under which financing will be available to governments improve transparency by establish- member countries experiencing exceptional bal- ing voluntary standards that countries can use ance of payments difficulties due to a large short- when disclosing economic and financial data to term financing need resulting from a sudden and the public; 43 countries (including 19 developing disruptive loss of market confidence. countries) have subscribed to the Special Data A cooperative regional financing mechanism is Dissemination Standards. The IMF also has for- being prepared in East Asia to enhance prospects for malized procedures for activation of the regional stability. This framework will establish a Emergency Financing Mechanism, which helps mechanism for regional surveillance to complement countries that are experiencing large and sudden IMF activities, enhance economic and technical capital outflows. cooperation among member states (particularly for In addition, the IMF has adopted New financial sector regulation and supervision), con- Arrangements to Borrow, under which G- 10 and sider measures to strengthen the IMF's ability to Box 2.4 Initiatives of the Group of len ond the Basle Committee on Baking Supervision The East Asian crisis has highlighted the severe weak- corporate governance, and other key elements of a nees affecting fitancial systems in several developing robust financial system. countries and the serious implications that these weak- * Promotion by the IMF, the World Bank, and nesses can have for macroeconomic stability. To that regional development banks of sountd principles and end, emerging market economies and the Group of practices. Ten (C I 0) are developing a strategy to strengthen the Principles and practices are being established for financialsystems of developing countries. Thework is accounting, payments and settlements, securities gided by seveal principles. First, national authorities market supervision, insurance supervision, and bank- are ultimately responsible for policies designed to ing supervision, To support this initiative, the Basle strengthen financial systems. Second, financial stabili- Committee on Banking Supervision has released 25 l ty is possible only if countries confoirm Vith interna- core principles for effective banking supervision that tional prudential standards and allow markets tO cover licensing and structure, prudential regulations I operate in a compeiitive. professional, and transparent and requirements, methods of ongoing banking fashion. FinaLly, sound macroeconornic and structural supervision, information requirements, formal pow- policies are prerequisites for financial system stability. ers of supervisors, and cross-border banking. The strategy for strengthening financial systems Supervisory authorities from around the world has several compotnents: endorsed the core principles at the annual meetings * Development of international consensus on the key of the IMF and the World Bank in October 1997. elements of a sound financial and regulatory system. The principles are designed to serve as a basic refer- * Fornulation of principles and practices by interna- ence for the minimum requirements for effective tional authorities with relevant expertise and experi- banking supervision and are to be applied in the ence, such as the Basle Cormmittee on Banking supervision of banks in each country's jurisdiction. Supervision, the International Association of insurance The principles are designed to be verifiable by both Supervisors, and the International Organization of supervisors and financial markets at large. Working Securities Commissions, with the World Bank and the IMF, the Basle * Use of market discipline to provide incentives for Committee will monitor countries' progress in imple- the adoption of sound supervisory systems, improved menting the principles. 46 respond to financial crises, and establish a coopera- to countries with stronger fundamentals will be tive financing arrangement to supplement IMF less affected. For the countries most affected by the resources. crisis, access to capital markets will depend on the policy response. In Indonesia, Korea, and Prospects for private capital flows in 1998 Thailand measures to close insolvent institutions should be followed up over the medium term with The collapse in investor confidence and the determined policies that establish more robust uncertainty about prospects for recovery in East financial systems, which are essential to the effi- Asia, combined with the possible implications for cient mobilization of domestic savings and exter- other regions, are likely to reduce net long-term nal finance, and that channel these resources to private flows to developing countries in 1998 rel- productive investments. ative to 1997. Countries that rely on these flows for new financing will likely face considerable dif- Composition offlows ficulties in the near term. If the crisis in East Asia is bottoming out, however, the extent of this The crisis will have different effects on different decline may be moderated by the continued types of flows, so sharp downturns in some flows favorable external environment, the return to may be offset by increases in others. In the short markets of countries with strong economic fun- term portfolio equity flows (13 percent of net damentals, and changes in the composition of long-term private flows to developing countries capital inflows that may have made them more in 1997; table 2.8) will likely be undermined by resilient. the severe volatility in many emerging stock markets, as well as by uncertainties about corpo- External and domesticfactors rate earnings in the countries most affected by the crisis. The recovery of bond finance (21 per- World Bank projections prepared in mid-January cent of net long-term private flows in 1997) may 1998 anticipate that the international economic also be slow, as riskier borrowers may be shut out environment will remain broadly favorable to from bond markets for some time (the loss of developing countries for the year. International investment-grade ratings will close off some interest rates, which have a significant influence on markets) and a number of creditworthy borrow- flows-especially portfolio flows (Taylor and ers may postpone borrowing to avoid issuing Sarno 1997)-will likely remain low in 1998, with bonds at a higher spread than they have obtained the London interbank offered rate (LIBOR) aver- in the past. Over time, however, inflows should aging about 5.7 percent (in dollar terms). World recover to take advantage of asset price declines output growth rates should decline to 2.6 percent, that are deeper than is justified by market fun- from 3.2 percent in 1997. World trade volumes are damentals. expected to rise 6.4 percent in 1998. Foreign direct investment (47 percent of net Once the initial impact of the crisis has been long-term private flows in 1997), which has fully absorbed, sharp downturns in flows may longer-term objectives, may remain stable or even occur in only a limited number of countries. Flows increase in 1998 in response to the low asset prices Table 2.8 Composition of private flows, 1990-97 (billions of U.S. dollars andpercentage ofshares) 1990 1995 1996 1997 Type offlow Amount Share Amount Share Amount Share Amount Share Portfolio equity 3.2 7.6 32.5 17.2 45.8 18.5 32.5 12.7 Commercial bank loans 14.9 35.6 31.3 16.6 36.5 14.8 49.4 19.3 Bonds 0.1 0.2 23.8 12.6 45.7 18.5 53.8 21.0 Foreign direct investment 23.7 56.6 101.5 53.6 119.0 48.2 120.4 47.0 Total 41.9 100.0 189.1 100.0 246.9 100.0 256.0 100.0 Source: World Bank data. 47 and production costs resulting from currency retire short-term debt, implying the need for an devaluations-particularly in sectors producing increase in net long-term flows in 1998. But cap- tradable goods. Foreign direct investors that are ital outflows (which are not reflected in the data closely linked to local markets may postpone new on net private flows; see box 1.1) could decline in projects, however, if they believe that the econom- 1998 relative to 1997, and official financing may ic downturn is likely to become protracted. rise (as envisioned under international rescue packages for Indonesia, Korea, and Thailand). Outlook for 1998 Moreover, the baseline projection of a moderate decline in private capital flows in 1998 is subject Prospects for flows are uncertain and will depend to several risks: largely on policy responses in the countries most * Continued devaluations in the region would affected by the crisis. There is, however, reason to lower the competitiveness of countries whose cur- believe that the slowdown in flows will not last too rencies have maintained their parities (China, for long. Although it took Latin American countries example, has exports equal to about 70 percent of seven to eight years to regain access to interna- the value of exports from the five East Asian coun- tional markets after the debt crisis of the early tries most affected by the crisis). 1980s, flows to Argentina and Mexico picked up * Intensification of the crisis could cause a sig- within six months of the 1994-95 peso crisis (table nificant drop in international stock markets. 2.9). Furthermore, although flows of foreign direct Exports to the five most affected East Asian coun- investment in Mexico dropped somewhat after its tries represent only 0.6 percent of European GDP recent crisis (following a temporary surge in and 0.7 percent of U.S. GDP, so any decline is 1994), they remained well above the levels of the unlikely to substantially affect valuations of firms early 1990s (thanks in part to the North American quoted on European and U.S. exchanges (even Free Trade Agreement). after accounting for earnings of foreign sub- The financing requirements of the five East sidiaries in the region). But since price-earnings Asian countries most affected by the crisis are like- ratios are at high historical levels (especially in the ly to be about the same in 1998 as in 1997, reflect- United States), the markets are vulnerable to news ing a shift from a current account deficit to a that projected earnings growth might not materi- surplus ($41 billion), no change in amortization alize. A sharp drop in international stock markets payments on long-term debt, and a $42 billion would likely intensify the flight to quality that has change in reserves (table 2.10). Some buildup of already been observed, further reducing net flows reserves will likely be required after the $31 billion to emerging markets. drop in 1997. * International banks with exposure in East Still, the future of net long-term private flows Asia could be placed at risk. The assets of G-7 to countries affected by the crisis remains uncer- banks in the five East Asian countries hardest hit tain. This projection of external financing require- by the crisis accounted for about 27 percent of ments does not take into account the short-term total capital in 1995. At 43 percent of capital, debt ($150 billion in mid-1997) that may need to Japanese banks are the most exposed. Still, bank be rolled over. Some countries plan to borrow exposure to the five East Asian countries is a small- long-term funds on the international markets to er share of capital than it was to the countries most Table 2.9 Private flows to Argentina and Mexico, 1994-96 (billions of U.S. dollars) Bond issues and loan commitments Foreign direct investment First half Second half Country 1994 1995 1995 1995 1996 1994 1995 1996 All developing countries 128.1 56.1 107.6 163.7 184.8 86.9 101.5 119.0 Argentina 8.2 3.7 6.1 9.8 24.0 3.1 4.2 4.3 Mexico 13.3 3.1 12.0 15.1 28.6 11.0 9.5 7.6 Source: Euromoney and World Bank. 48 Table 2.10 Current account deficits, changes in reserves, and long-term amortization among major developing country borrowers, 1997-98 (billions of U.S. dollars) 1997 1998 Current Change Current Change account in account in Type of borrower deficit reserves Amortization Total deficit reserves Amortization Total Most affected East Asian countriesa 34 -31 38 41 -7 11 38 41 Other major borrowersb 51 42 106 199 66 30 102 197 Allmajorborrowers 85 11 144 241 58 41 139 239 Note: Data on net flows to these countries, which are partly based on data from international capital markets, are considerably larger than the esti- mates of financing requirements presented here. The difference reflects net flows on short-term debt, capital outflows, lags in reporting, and other statistical discrepancies (see box 1.1). a. Indonesia, the Republic of Korea, Malaysia, the Philippines, and Thailand b. Argentina, Btazil, Chile, China, Colombia, the Czech Republic, Hungary, India, Mexico, Peru, Poland, Russia, Turkey, and Venezuela. Source: World Bank, JP Morgan, and Consensus Forecasts. affected by the debt crisis in the 1980s (more than ing countries that attract private capital flows can 60 percent of capital in 1982).5 service their obligations. * Intensification of the crisis could heighten pro- tectionist sentiment. Recent World Bank projec- Notes tions indicate that the United States will account for about half of the $100 billion increase in the current 1. The 12 emerging markets are Argentina, Brazil, account surplus of the countries most affected by the the Czech Republic, India, Indonesia, the Republic of .Given that the public already Korea, Malaysia, Mexico, the Philippines, Poland, crisis. Glven that theU.S. publlc already believes Thailand, and Turkey. that the current account deficit is large (although it 2. This estimate differs significantly from that given is only 2.1 percent of GDP, well in line with histor- in official sources. Real estate lending in the Philippines ical averages and levels in other deficit countries), may have been less speculative than in the other coun- this additional increase could escalate trade friction. tries, however, because buyers are required to make sub- stantial downpayments. Developing countries may also feel pressures to raise 3. These data come from Jardine Fleming. Official trade barriers in response to the falling prices of sources report that nonperforming loans in the Asian exports, or as a way to reduce balance of pay- Philippines were 4 percent of bank assets. ments deficits. For example, Argentina and Mexico 4. Although implicit or explicit government guaran- raised tariffs temporarily during the peso crisis in tees encouraged excessive risk taking and contributed to 1995, and Mercosur tariffs recenty were raised 3 the crisis, the concern that public commitments to guar- .,, antees might not be supported by the legislature added to percentage points. Efforts to increase protection Will uncertainties over economic policy just before the crisis hit. most likely prove self-defeating, as happened in the 5. Those countries included Argentina, Brazil, 1930s, and could weaken confidence that develop- Chile, Mexico, the Philippines, and Venezuela. CHAPTER 3 The changing role of official finance This chapter covers trends in net official finance even nominal) terms since the early 1990s. This and aid allocation, efforts to help heavily indebt- decline reflects constraints on the availability of ed poor countries achieve sustainable levels of concessional resources, the decision by many gov- external debt, and the increasing volume and scope ernments to reduce direct lending as private flows of guarantee activities. to developing countries have increased, and a shift by some developing countries away from official Shifting focus of aid sources to international capital markets. Net flows also have been reduced by a sharp increase in The events of the past two decades have greatly amortization payments by middle-income coun- changed the political context of official finance to tries as the debt deferred by Paris Club creditors developing countries. The collapse of centrally during the 1980s has come due. planned economies in Eastern Europe and the The decline in net nonconcessional flows was poor performance of heavily distorted economies interrupted in 1997 by the international rescue in Africa, Latin America, and the Middle East have package for Thailand. The package, which includ- underlined the importance of providing appropri- ed $9 billion of nonconcessional finance, boosted ate incentives for private sector growth. And the net nonconcessional flows to an estimated $6.9 need to contain government spending in industri- billion for the year, compared with -$5.5 billion al countries, the end of the cold war, and the evi- in 1996 (table 3.1). Exduding the Thai package dence of mismanagement of aid flows to some and the early repayment by Mexico of the remain- countries have reduced public support for official ing amounts due under its 1995 rescue package, development assistance. This section shows that: official nonconcessional flows are estimated at $5.5 billion in 1997, compared with an average of The decline in net concessionalflows to developing $11.4 billion a year in 1990-92. Net nonconces- countries in real and nominal terms during the sional lending by multilateral institutions (exclud- 1 990s (including 1997) has been accompanied by ing the IMF) continued its upward trend over the efforts to improve the effectiveness of aid by allo- past few years, a result of strong increases in net catingagreatershare to countries with betterpoli- flows from the Inter-American Development cies. Continued improvements in aid effectiveness Bank and the World Bank. will be critical to ensure benefits for developing Net concessional assistance to developing economies and will provide the most convincing countries continued its downward trend in 1997. argument for maintaining or increasing conces- The availability of concessional resources has been sional assistance in theface offurtherplanned cut- constrained by the pressing need for fiscal consol- backs by major donors. idation in most industrial countries, the declining strategic and military importance of development Net officialfinance in 1996-97 aid since the end of the cold war, and weak public support for aid in some major donor countries, Net official development finance, consisting of due in part to skepticism about its effectiveness. loans and grants from government agencies and Net inflows of official concessional finance, con- multilateral institutions, has declined in real (and sisting of official development assistance (ODA) 49 50 Table 3.1 Net official flows of development finance, 1990-97 (billions of U.S. dollars) Category 1990 1991 1992 1993 1994 1995 1996 1997' Official development finance 56.4 62.7 53.8 53.6 45.5 54.0 34.7 44.2 Concessional finance 43.9 50.7 44.1 41.7 46.3 45.4 40.1 37.3 Grants 29.2 35.1 30.5 28.4 32.7 30.6 29.2 25.1 Loans 14.7 15.6 13.6 13.3 13.7 12.8 10.9 12.3 Bilateral 8.7 9.3 7.1 6.8 5.9 5.5 2.8 4.6 Multilateral 6.0 6.4 6.5 6.5 7.8 7.2 8.1 7.7 Nonconcessional finance 12.5 12.0 9.7 11.8 -0.8 8.6 -5.5 6.9 Bilateral 2.9 4.0 4.0 3.2 -3.4 4.5 -10.0 -2.8 Multilateral 9.6 8.0 5.7 8.7 2.6 4.1 4.5 9.7 Memo items Use of IME credit 0.1 3.2 1.2 1.6 1.6 16.8 1.0 3.3 Technical cooperation grants 14.2 15.8 17.9 18.6 17.4 20.6 19.2 18.3 Note: Official concessional finance comprises inflows of official development assistance (excluding technical cooperation grants) and official aid to Eastern Europe and the former Soviet Union. Memo items are not included in preceding aggregates. a. Preliminary. Source: World Bank Debtor Reporting System and staff estimates. and other official aid to low- and middle-income health services has worsened. The number of peo- countries, fell by an estimated $2.8 billion in 1997 ple 15 and older in developing countries who are (following a $5.3 billion decrease in 1996). Dollar illiterate rose from an estimated 848 million in appreciation of roughly 10 percent in 1997 1980 to 872 million in 1995, and the number of explains almost half this decline.' people per hospital bed rose from 921 in 1980 to Thedeclinein development assistancesince the 950 in 1993 (World ]3ank 1997d). 1980s has not been accompanied by any decline in The detailed information now available for the need for aid. The number of people in low- 1996 confirms the decline in ODA flows. Net income countries increased from 2.4 billion in ODA flows (including technical cooperation 1980 to 3.2 billion in 1995 (World Bank 1997d). grants) from the OECD's Development Assistance One study showed that the number of people liv- Committee (DAC) countries fell to 0.25 percent ing on less than $1 a day in developing countries of their GNP in 1996 (figure 3.1), the lowest level rose from 1.2 billion in 1987 to 1.3 billion in 1993 recorded since the United Nations adopted a tar- (World Bank 1996). And while life expectancy has get of 0.7 percent of GNP in 1970. This decline climbed in developing countries (from 56 years to in donors' ODA effort marks a continuation of the 63 for males and from 59 years to 66 for females steep fall since the mid-1980s, when total net between 1980 and 1995), access to education and ODA from DAC members was 0.35 percent of aggregate GNP (figure 3.2). Figure 3.1 Net ODA from DAC countries has slipped DAC members provided $55 billion in official to its lowest level since 1970... development assistance to developing countries and Billions of U.S. dollors Pertent multilateral development banks in 1996, down 80 Percentage of GNP 0.35 from $59 billion in 1995, reflecting a decline of 4 _of OAf eountries 70 0 30 percent in real terms.2 Driving this decline was a $5 60 0.25 billion fall in aid from Japan, the largest donor. The 50 x 025 sharp decline in Japsnese ODA in dollar terms m United Stotes 5 o ,ngns0.20 stemmed from a 14 percent depreciation of the yen D United Kingdom 40 0.15 against the dollar and a 25 percent drop in ODA in * lJapn 30 , l . yen terms, as difficult economic conditions and I Germany 20 0.10 pressures for fiscal consolidation constrained Japan's = France 050 O.OS 1996 ODA budget to 1 trillion yen, down from 1.4 M Other 0 trillion yen in 1995. Also contributing to the ° 990 1991 1992 1993 1994 1995 1996 decline in Japanese ODAwere a delay in payments Soarre: OECD data. to the International Development Association 51 Figure 3.2 ...and net ODA from industrial countries has shown a steep decline since the mid-I 980s Percentage of GN1P 1986 Percentage of GNP 1996 Norway 55f Denmark M Netherlands 1' Norway l. 5 Denmark 0 Netherlands Sweden . . Sweden Belgium 3 France T Ccadod Belgium France Finland Australia Switzerland Finland Germany Germany Canada Italy Ireland United Kingdom Australia New Zealand Austric Switzerland United Kingdom Japan New Zealand Ireland Italy United States Japan Austria United States All DAC All DAOC 0 0.2 0.4 0.6 0,8 1.0 1.2 0 0.2 0.4 0.6 0.8 L.0 1.2 Source: OEtD drta. (IDA) and other multilateral institutions planned percent of DAC countries' GNP.3 In Europe con- for 1996 due to protracted multilateral negotiations tinued efforts to meet the Maastricht target for fis- and a rise in repayments of bilateral concessional cal deficits (3 percent of GDP) will restrict the loans to Japan. Austria, Canada, France, and scope for real increases in ODA. In Japan, the Portugal also registered significant falls in ODA. largest donor by volume, pressure to reduce the fis- But 9 of the 21 DAC member countries increased cal deficit led to a 10 percent cut in the ODA bud- aid as a share of GNP ODA flows from the United get for fiscal 1998, and further cuts are expected States rose by 21 percent in real terms in 1996, and through 2000. In the United States, which has those from Italy by 34 percent. In both cases the halved the ratio of net ODA to GNP since the flows included fiinds that had been approved but mid-1980s, opinion polls show continued public not spent in 1995. concern over the fiinds devoted to foreign aid, in The ranking of countries by their ratio of net part because of exaggerated perceptions of aid's ODA to GNP has changed little over the past 10 share in the budget (currently about 1 percent).4 years, however. Norway, the Netherlands, Denmark, Many donors are channeling an increasing and Sweden continue to have the highest ratios. share of their ODA flows through nongovern- Among the major donors, Denmark has increased mental organizations (NGOs). While in 1980 its ODA ratio (from 0.9 percent in 1986 to more official sources accounted for less than 10 percent than 1 percent in 1996), France has maintained a of NGO budgets, by the 1990s their share had constant ratio, and the others have seen declines in risen to 35 percent. But the private resources that their ODA ratios of between 15 and 50 percent. NGOs have channeled to developing countries Prospects for ODA allocations are dismal. have stagnated at around $6 billion since the mid- Preliminary information for 1997 shows that net 1990s and are estimated to have declined to $5.7 ODA flows could have declined to as low as 0.21 billion in 1997. 52 Aid effectiveness on the basis of strategic considerations than does multilateral assistance. Burnisde and Dollar found There is increasing evidence that development that in 1970-93 countries with poor policies assistance is most effective when it supports sound received as much bilateral aid relative to income as economic policies. Experience over the past 20 countries with good policies, reflecting the influ- years has shown that development projects have ence that strategic and political factors have had on little chance of success in an environment of high bilateral aid allocations. But they found that the inflation and severely distorted prices produced by allocation of multilateral aid-including assis- inadequate policies. tance from the World Bank's IDA-favored coun- In a recent study of World Bank-financed pro- tries with good policies. Killick (1991), McKinlay jects, the Bank's Operations Evaluation Depart- and Little (1978), Boone (1996), Kreuger, ment confirmed that the macroeconomic Michalopoulos, and Ruttan (1989), and Maizels environment plays a critical part in determining and Nissanke (1984) have also cited the impor- the success of borrowers' project portfolios tance of strategic considerations in bilateral aid (Piciotto 1997; World Bank 1997a). Another allocation. study by World Bank researchers provides strong Studies have also provided some support for the empirical support for the view that sound eco- view that multilateral aid has been less responsive nomic policies are vital to ensure effective use of to strategic considerations. Maizels and Nissanke aid (Burnside and Dollar 1997). It found that large (1984) found that the allocation of aid by multi- aid flows tend to be associated with faster growth lateral donors in the late 1980s (though not in the in countries with a sound policy environment- late 1960s) was based on countries' needs. defined as an open trade regime, fiscal discipline, Trumbull and Wall (1994) found that the needs of and the absence of high inflation. And it found recipients played a significant part in the allocation that aid made little or no contribution to growth of ODA by multilaterals, although Frey and in countries with poor policies. In a sample of 41 Schneider (1986) concluded that World Bank low-income developing countries in 1970-93, lending was best explained by a combination of countries with good policies that received large economic and political indicators rather than by amounts of aid achieved GDP per capita growth such considerations as need or development rates of 3.5 percent a year, while countries with potential. good policies and small amounts of aid had per But these studies mostly explain the allocation capita growth rates of 2 percent. Countries with of aid before the 1990s. The increasing recogni- poor policies achieved virtually no per capita tion of the importance of the policy environment growth, regardless of whether they received large for aid effectiveness has recently led donors to or small amounts of aid. focus more attention on efforts to increase aid to Other studies have tended to confirm the good performers. One such effort, the Special importance of sound economic policies for aid Program of Assistance for Africa, helps coordinate effectiveness (although their results are sometimes aid efforts to ensure that African countries that are ambiguous). Both Killick (1991) and Kreuger, effectively implementing economic reform pro- Michalopoulos, and Ruttan (1989) argued that grams receive the quick-disbursing assistance they the effectiveness of economic assistance is strong- need. The program identifies funding gaps in per- ly affected by the economic policy environment in forming countries and provides timely informa- recipient countries. Boone (1996) found that aid tion on the status of reform programs so that most directly assists the poor (a different focus donors can adjust the allocation of assistance from Burnside and Dollar, who were more con- accordingly. The program reflects donor countries' cerned with the impact of aid on growth) in coun- commitment to improving the effectiveness of aid tries that have liberal political regimes and through better coordination of aid delivery at the democracies.5 national and local levels. Has aid generally gone to countries with good Donor countries also are working to coordinate policies? The evidence is mixed. The literature the goals of aid. The DAC has set out a strategic finds that bilateral aid tends to be allocated more agenda of goals to be achieved in developing coun- 53 tries by 2015 with the support of official assistance. countries with access to private flows (countries The agenda includes reducing by half the propor- with bond ratings from international credit rating tion of people living in extreme poverty, ensuring agencies) was 33 percent in 1996, essentially the universal primary education in all countries, elim- same as it was in the mid-1980s. And the share of inating gender disparity in primary and secondary net ODA going to low-income countries has education, reducing by two-thirds the mortality declined in the 1990s, from 50 percent in 1991 to rates for infants and children under age five, reduc- 46 percent in 1995. This decline reflects a fall in ing by three-fourths maternal mortality, providing the share of bilateral ODA to low-income coun- access to reproductive health services for all indi- tries (from 45 percent in 1991 to 28 percent in viduals of appropriate age, and reversing the cur- 1996), due in part to the fact that in many large rent trend in the loss of environmental resources. recipients civil war or economic or political con- Has the increased attention to aid effectiveness ditions have precluded substantial official assis- and policy performance led to any improvement in tance. The share of multilateral ODA to donors' allocation of aid to good performers in low-income countries has remained unchanged recent years? There is some evidence that it has. over the decade, but it has shifted sharply toward Between 1990 and 1995 the share of net ODA going humanitarian support (which averaged close to 20 to the best-performing countries-the first and sec- percent in 1994-96). ond quintiles in a ranking of policy performance by Another way to determine whether concerns the World Bank-rose from 38 percent to 45 per- over income levels or capital market access affect cent (table 3.2). Most of the change stemmed from the data on aid allocation by policy performance is a decline in the share of the fourth quintile. to look at the allocation among low-income coun- Donors might use other allocation criteria to tries (see table 3.2). 6 The best-performing 40 per- enhance the development impact of aid that could cent of these countries received 53 percent of lead to less observable improvement in aid alloca- ODA in 1995, up from 36 percent in 1990. But tion by policy performance. For example, donors this increase in ODA to the top two quintiles came might allocate more aid to countries heavily at the expense of the third quintile; the share allo- dependent on concessional flows than to those cated to the worst performers changed little. In with access to nonconcessional flows, or more aid short, aid allocation did improve among low- to lower-income than to middle-income coun- income countries, but it is unclear whether it tries. But there does not appear to have been any improved more than among all developing coun- improvement in aid allocation by these criteria tries. And there is inevitably a problem of "adverse during the 1990s. The share of net ODA going to selection"' in the allocation of some forms of aid. For example, the significant share of aid provided Table 3.2 Recipients' share of official development primarily for emergency assistance and humani- assistance by policy performance, 1990 and 1995 tarian reasons will go disproportionately to coun- (percent) tries where political or economic conditions may All developing Low-income not be conducive to good policies. Performance countries countries' One area where progress has been made is in quintile 1990 1995 1990 1995 donor efforts to increase the concessionality of aid 1 st (best) 20 23 16 22 and to minimize the use of ODA loans (as opposed 2nd 18 22 20 3 1 to grants), particularly for countries with an unsus- 4th 34 24 18 1 5 tainable debt burden. Several donors now provide 5th (worst) 13 15 13 16 all their ODA in the form of grants, most provide Total 100 100 100 100 all their ODA to heavily indebted poor countries Note: The categorization by performance is based on World Bank rat- as grants and many have canceled the ODA oblig- ings that evaluate the effectiveness of economic policies in supporting ations of these countries. As a result several donors growth and reducing poverty. The ratings for each year are used, so (Australia, Norway, and the United Kingdom) no the composition of each group differs between 1990 and 1995. a. Excluding China. India, and Pakistan, which also borrowv from longer have any outstanding ODA daims. IBRD and other nonconcessional sources of official development Yet the terms on ODA from some donors are finance. Source: OECD data and World Bank staff estimates. hardening, owing to an increase in the share of 54 loans. Despite the fall in total net ODA, new com- The HIPC Debt Initiative was formed against mitments of concessional loans from bilateral a background of weak economic performance and donors rose by 15 percent between 1995 and high debt levels in a large number of poor coun- 1996. Because of the lag in actual disbursements, tries. Economic reform programs undertaken by however, this trend will not start to show up in net many of these countries have improved economic ODA flows until 1997. The tendency to mix performance. And debt relief efforts-through ODA and commercial loans in a single package such mechanisms as the Paris Club, the IDA Debt also appears to be increasing, reversing the trend Reduction Facility, and the World Bank's Fifth of the early 1990s. Fully a third of ODA lending Dimension Program-have provided substantial committed in 1996 had associated commercial benefits. But for many countries the debt relief funds, underlining the continued importance of available through these mechanisms could not tied aid and the influence of commercial interests ensure that debt sustainability could be achieved on aid flows. within a reasonable period even with good policy performance. Progress under the HIPC Debt Initiative Economic performance and debt of the HIPCs The Debt Initiative for Heavily Indebted Poor Countries (HIPCs) was launched in September Forty-one countries are classified as heavily indebt- 1996 out of a recognition that despite considerable ed poor countries.7 Mostly in Africa, these countries efforts by countries and creditors, many of the account for 12 percent of the debt of developing poorest developing economies still faced unsus- countries, but less than 5 percent of their exports tainable debt burdens. This section explains that: and only 3 percent of their GNP (table 3.3). Though economic and policy performance has var- TheHIPCDebtInitiativeha-splayedan important ied widely among the HIPCs, average GDP growth part in the evolution of donor policies and in the for the group was only 2.2 percent between 1985 increased effectiveness ofaid by making the achieve- and 1990. The GDP growth rate fell to 1 percent ment of debt sustainability its explicit goal and by during 1990-95, well below the population growth involvingallcreditorsin debt reliefefforts. In itsfirst rate of 2.7 percent. The volume of exports from year seven countries that had established the HIPCs rose by 2.7 percent a year in 1985-90, but required track record ofgoodeconomicperformance the increase slowed to 2.2 percent in 1990-95 were consideredfor additional debt relief under the despite more vigorous reform efforts (table 3.4). initiative, and agreements were reached to reduce The HIPCs' poor average growth can be attributed the debt offour of these countries by $1.2 billion in in large part to civil disturbances, weak governance, present value terms. The HIPC Debt Initiative is unstable macroeconomic policies, and severely dis- geared to encouraging improved economic and torted prices in many of the countries and to declin- socialpolicies, facilitating the provision of interim ing terms of trade during the late 1980s in some. finance to strongperformers, contributing to a more Large debt burdens also have played a part in productive relationship between creditors and depressing investment and growth.8 debtors, and, ultimately, enabling countries to exit The median ratio of the present value of debt from repeated debt rescheduling exercises. to exports among HIPCs is 340 percent, and some Table 3.3 Heavily indebted poor countries relative to all developing countries, 1996 (percent) Number of Share of Share of Share of Share of Region countries population" external debt GNP exports Africa 33 10.6 8.8 2.3 3.3 Asia 4 2.8 1.9 0.8 1.0 Latin America 4 0.4 0.8 0.2 0.3 Total 41 13.8 11.5 3.3 4.7 a. Data are for 1995. Source: World Bank data. 55 Table 3.4 Economic performance of the heavily half of the 1990s it was only 17 percent of GNP, indebted poor countries, 1980-95 compared with 26 percent of GNP for other devel- (average annualpercentage change orpercent) oping countries. Second, even when a country Indicator 1980-85 1985-90 1990-95 receives a positive net transfer, cumbersome nego- Export growth -0.8 2.7 2,2 tiations with creditors over the amount of debt to GDP growth 1.3 2.2 1.0 be rescheduled and the constant dependence on 1985 1990 1995 new inflows to cover debt service can impose sig- Debt/exports 383 506 465 nificant administrative costs and increase uncer- Debt service/exports 26.2 22.1 22.6 tainty about government spending on projects Net transfers/GDP' 8.2 10.4 9.1 (deterring private contractors). Third, the size of a. Net transfers from official sources. future repayments may be seen as the outcome of Source: World Bank data. bargaining between the country and its creditors, with higher economic growth resulting in higher have ratios of more than 1,000 percent (the aver- repayments. In these circumstances private agents age ratio of debt to exports for other developing may be hesitant to invest because of uncertainty countries is 130 percent). Debt burdens of this size over future tax rates (which may rise to finance raise concern that debt may not be sustainable, debt service on old loans). Similarly, the impetus meaning that the debtor is unlikely to be able to for undertakingdifficultadjustment programs will service the debt and, ultimately, to repay it with- be weakened if governments expect that foreign out resorting to further rescheduling or the accu- creditors will claim a large share of the resulting mulation of arrears (box 3.1). growth in output. Finally, unsustainable debt lev- els can preclude access to international capital The effect of unsustainable debt on growth markets, making it more difficult for the private sector to finance trade. Providing an exit from Unsustainable debt levels can affect economic per- rescheduling by reducing debt to sustainable lev- formance through several channels. First, and els can thus play an important part in removing most directly, the payment of debt service reduces constraints on growth. the resources available for investment. This effect is limited for the HIPCs, many of which continue Goals and design of the initiative to receive positive net transfers from official cred- itors (which hold the bulk of HIPC debt). The basic goal of the HIPC Debt Initiative, which Nevertheless, for a variety of reasons the invest- was endorsed by the Development Committee of ment rate in HIPCs is relatively low; in the first the World Bank in September 1996, is to reduce Box:31 Whatuwest * dkl? ;i The coct of sustainable debt levels is critial toi fnance the full amount of interest payments due (see Understadin posiinon of the HIPS an h Cuddon- 19. s condition is met only if goals of trhe PC Dbbt InitiatUve. Sustinalilhyme db .grows more slowly thn theinteres ratee- e ened as m in tt country w ale in al Defuing rhe srstnhity ofdebt-fom public likelih tom current antd fiture eternal sous di more difficult, since the conditons -obigtionsin full without resorting to rescheing under w ic ntary finae is l robe forth- in the fute accuraulao f (Clessi-s c g m offial Sources are ls cl t fr and eter 1996).. pivate sources. Orne approach is to defineAsustain- ---For-b from priatea sources,. as of olicis ability s the countr/s abity to achieve, over a (and | h resulting debt leve) is sstainable if t . ed pi, equilibrium in the balance f y- j-: ere path of key- maacoonomic targets is consis- incr, -and to reach level debt by the end of rWenlt wVifth- the: v of financing likely sobe- -theperiod that is low enough to makefuture debt -irtc on ; olunar basis. In she long run service -problemns unikiy (Claessens and others new lending (at positive red interest rat c t 199 56 HIPC debt burdens to levels that can be serviced tion point). The performance period addresses the without recourse to further rescheduling, in the potential moral hazard problem that could arise if context of a sound growth and development pro- debt relief were seen as a reward for poor econom- gram.9 The initiative is intended to provide a per- ic policy performance. The requirement of a track manent exit from debt rescheduling and thus record also provides some assurance that debt relief marks an important extension of previous debt ini- will be provided in a context in which the resources tiatives. To qualify for assistance under the initia- released as a result will be used for sound develop- tive, countries must be eligible to borrow from ment purposes. IDA, but not from the IBRD, must be eligible to Under the HIPC Debt Initiative the target borrow from the IMF's Enhanced Structural ranges for debt sustainability are defined case by Adjustment Facility (ESAF), must have established case within the range of 200-250 percent for the a track record of adjustment and reform supported ratio of debt (on a net present value basis) to by the IMF and the World Bank, and must face an exports and 20-25 percent for the ratio of debt ser- unsustainable debt situation even after the full vice to exports.10 The initiative recognizes the fis- application of existing debt relief mechanisms. cal dimension of external debt; for countries that By involving all of a country's creditors, and for at the decision point have a ratio of exports to the first time multilateral creditors, the initiative GDP of at least 40 percent and are making a sub- provides an orderly process for allocating the costs stantial fiscal effort, as reflected in a ratio of fiscal of debt relief. And by ultimately removing exces- revenue to GDP of at least 20 percent, the target sive pressures to refinance debt service, it will help ratio of debt to exports may be further reduced to to strengthen the credibility of the adjustment dia- a level that achieves a ratio of the net present value logue between donors and countries. In addition, of debt to revenues of 280 percent at the comple- an important consideration in debt relief agree- tion point. The resources to be provided under the ments under the HIPC Debt Initiative has been to HIPC Debt Initiative are based on the target val- establish monitorable programs for macroeconom- ues of these ratios at the completion point, not on ic and structural reform and for social development their current value. policies focused on basic health, primary educa- Countries for which existing debt relief mech- tion, and rural development. Debt management anisms would not achieve sustainability in the capacity is also assessed, and programs are being three years following the decision point would put in place to strengthen capacity where necessary receive enhanced relief under the initiative, ade- to help ensure that debt problems do not reemerge. quate to achieve debt sustainability by the end of The HIPC Debt Initiative incorporates lessons that period, assuming continued strong policy learned in the 1980s from such efforts as the Brady reform. For borderline cases, where there is uncer- Plan and builds on existing debt relief mecha- tainty about the robustness of their ability to reach nisms, including the Paris Club. Countries are a sustainable debt position, there are provisions for considered for eligibility after maintaining a three- further monitoring that leave open the possibility year track record of macroeconomic, structural, of receiving enhanced assistance under the initia- and social policy reforms, monitored by the World tive if needed to achieve debt sustainability. The Bank and the IMF. A country's eligibility for assis- requirement of a six-year performance period is tance is determined by the Boards of the Bank and implemented flexibly- and case by case; countries the IMF on the basis of a tripartite debt sustain- receive credit toward the first three years of per- ability analysis undertaken by the country's gov- formance for progranis already under way, and in ernment and Bank and IMF staff. Toward the end exceptional cases the second three-year stage may of the three-year performance period required for be shortened for countries with sustained records a country to be considered for a Paris Club stock- of strong performance. of-debt operation (the decision point), an analysis is made of whether this operation, together with at Progress in 1997 and outlook least comparable action by other nonmultilateral creditors, would be enough to achieve debt sus- During 1997 the eligibility of seven countries that tainability after another three years (the comple- had established track records of performance was 57 Table 3.5 Countries with a decision point in 1997 Guinea-Bissau, Mali, Mauritania, Mozambique, under the HIPC Debt Initiative Senegal, Togo, and Vietnam). Decisions on indi- (millions of U.S. dollars) _ vidual countries will be made by the Boards of the Estimated Bank and the IMF at the appropriate time. Decision Estimated nominal debt Country point debt relief' service relief b Funding Uganda April 1997 338 700 Benin July 1997 0C 0 Burkina Faso Sept. 1997 115 200 All creditors participate in the debt relief packages Bolivia Sept. 1997 448 600 under the initiative. Costs are shared broadly in Guyana Dec. 1997 253 442 proportion to each creditor's outstanding claims, Total 1,155 1,942 expressed in net present value terms at the decision a. Net present value of debt relief at the completion point. l l l - b. Nominal debt service relief refers to the cumulative amount of relief pont. Pars ClUb creditors PartiCIPate bY grantng over time. It exceeds debt relief in net present value terms, which dis- Lyon terms. Each creditor may define the mecha- counts the nominal debt service relief back to the completion point. nism through which it will participate as long as it c. Sustainable case. Source: World Bank and IMF staff estimates. achieves the agreed reduction in net present value claims by the completion point. Regular meetings reviewed by the Boards of the World Bank and the have been established among multilateral creditors IMF under the HIPC Debt Initiative. Six of these to coordinate their participation. countries were judged to qualify for relief packages The total cost of assistance under the HIPC that could amount to about $3 billion in net pre- Debt Initiative is estimated at about $7.4 billion sent value terms and debt service relief of $5 bil- in net present value terms."2 This relief is in addi- lion over time. The debt of these six countries in tion to that provided through traditional mecha- 1996 was equivalent to about 40 percent of the nisms such as Paris Club Naples terms or debt of the 19 HIPCs considered likely to qualify commercial bank buybacks funded through the over the life of the initiative, assuming that they IDA Debt Reduction Facility and other donors. establish the necessary track records. Debt relief The World Bank has established the HIPC Trust packages were agreed on for four of the six coun- Fund to facilitate participation by multilateral tries (Bolivia, Burkina Faso, Guyana, and Uganda) institutions. The Bank remains committed to by the end of 1997, totaling about $1,155 million meeting its full share of the costs out of its own in net present value terms (table 3.5) ." For all four resources, and it has transferred $750 million from countries the normal three-year interim period IBRD net income to the HIPC Trust Fund. between the decision point and the completion Together with the relief provided through IDA point was shortened in view of their strong policy grants, these funds are expected to cover the Bank's performance. share of commitments made in 1997. The HIPC The HIPC Debt Initiative is open to countries Trust Fund has also received about $170 million that begin undertaking Bank- or IMF-supported in pledges and contributions from 11 bilateral adjustment programs by September 1998, at donors to help meet the costs of other multilater- which point the initiative will be reviewed and a al development banks. The IMF has established decision made on whether it should be extended. the ESAF-HIPC Trust for financing special ESAF Although the projections are subject to much operations under the initiative, to which up to uncertainty, preliminary Bank and IMF staff SDR 250 million (equivalent to about $340 mil- analyses suggest that about 15 additional countries lion) can be transferred from the ESAF Trust are likely to receive assistance under the initiative, Reserve Account. assuming good performance. Of the countries likely to require assistance under the initiative, Growth and innovation in offidal guarantees about three-quarters could qualify with continued good performance in the first three years of the As developing countries have greatly expanded program (that is, by 2000). Nine countries could their access to private capital markets and foreign qualify for assistance in 1998, assuming continued direct investment, export credit agencies have satisfactory performance (C6te d'Ivoire, Ethiopia, responded to the rise in private flows by increasing 58 support through their traditional export guarantee large infrastructure projects in power generation, business and through the development of invest- telecommunications, and transport. Project ment insurance. This section argues that: finance, one of the most rapidly growing forms of external finance in the 1990s, typically involves a Export credit guarantee commitments have package of financing arrangements that may increased strongly during the 1990s, in part include export credit guarantees, commercial bank because of more aggressive export promotion by loans, equity, debt, and different types of contin- many countries and the improvedfinancial con- gent liabilities of the host government. ditions ofexport credit agencies. The rapidgrowth New export credit agency commitments in of investment insurance has helped developing 1996 totaled $94 billion, well above the average of countries involve theprivatesectorin theprovision the late 1980s but 5 percent lower than in 1995. of infrastructure and supported the privatization There were modest increases in several major mar- ofstate enterprises. kets (Brazil, Russia, and Turkey) and a big jump in South Africa. But these increases were more than Trends in the export credit market offset by a substantial and widespread decline in Asia (China, Indonesia, Malaysia, the Philippines, Export credit agencies' new commitments-the and Thailand), which followed a doubling of com- value of new business insured, new lending facili- mitments to this region between 1992 and 1995. ties, and guarantees for new foreign direct invest- The decline in new export credit commitments ment (but excluding trade finance with maturities in 1996 (and probably in 1997 as well) reflected of less than one year)-to developing countries some contraction in new infrastructure projects increased to an average $110 billion a year in and growing concerns over macroeconomic imbal- 1990-96, up from $83 billion in the second half ances and financial sector fragility in emerging of the 1980s (figure 3.3) i3 markets, particularly in Asia. The health of the Driving this strong expansion in export credit banking sector in borrowing countries is a partic- agency activities have been more aggressive export ularly important consideration for export credit promotion by many countries and, perhaps more agencies in deciding on new credits because the important, the changing nature of international prospects for repayment are directly affected by the financing for developing economies, which has financial situation of correspondent commercial shifted toward project finance and direct invest- banks. For some countries the downward trend in ment projects. Roughly half of new export credit new commitments in the past two years reflects agency commitments in recent years have gone to increasing competition from private insurers and support project financing activities, mainly for commercial banks and increased access to interna- tional bond issues that require no guarantees from Figure 3.3 Export credit agencies' new commitments export credit agencies. to developing countries have shown strong growth New commitments continue to be concentrat- in the 1 990s ed in a handful of economies that are viewed as rel- Billions of U.S. dollars atively low risk and that are large purchasers of 140 industrial country exports. The top 12 recipients 120 accounted for nearly 70 percent of new commit- 100 mnents in 1996, and the top 20 for more than 80 Investnent percent (figure 3.4). 80 insurance Export credit agencies' total exposure to devel- 60 Input oping countries reached an estimated $463 billion 40 ~~~~~~~~~~credit by the end of 1996, or 22 percent of the total exter- 40 ~~~~~~~~~~~~nal debt owed by developing countries (28 percent 20 of their long-term debt). Export credit agencies are 1988 the largest official creditor of developing countries, 1987 1988 1989 1990 199] 1992 1993 1994 1995 1996 accounting for 31 percent of their debt to official Soorce: Berne tnion ditr. creditors. But the agencies' share of countries' 59 Figure 3.4 Export credit agendes' exposure in developing economies remained concentrated in 1996 Billions of U.S. dollors Pereontage shore 60 100 50 Uj' ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~80 50 0 O~~~~~~~~~~~~~~~~umulative soe-8 30 | S 60 U _ | , Agencies' 40 20~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~2 Saurce: Berne Union date, IMF date, and Warld Bank staff esnimeten. external debt varies considerably. A few countries position in recent years is due largely to increased (such as Algeria, Iran, and Nigeria) owe more than recoveries from payments of moratorium interest 60 percent of their total debt to export credit agen- under Paris Club rescheduling agreements, cies as a result of large arrears or rescheduled debt. reduced claims owing to generally improved pay- ment performance by debtor countries, and F7inancial perfo smance of export credit agencies increased premium income reflecting the trend toward more market-oriented risk pricing systems. The net cash flow of the 41 export credit agencies Export credit agencies have recently made that are members of the Berne Union, measured as progress toward harmonizing their export credit the difference between receipts (of premiums and policies under the Arrangement on Guidelines for recoveraes on old claims) and payments (of new Officially Supported Export Credits, commonly laims and operating costs), totaled $1.5 billion in referred to as the OECD Consensus. The consen- 1996. This was the first surplus since before the sus was established in 1978, and its most recent debt crisis in the early 1980s, and it compares modification (intended in part to further reduce favorablywiththe $0.4 billionloss in 1995 andthe the provision of subsidized credits) became effec- recordlossof$7.1 billion int1990 (figure 3.5).aThe tive in 1994. The consensus defines aset of limits improvement in export credit agencies' financial on the terms of officially supported export credits, including the minimum cash down payment, Figure 3.5 The finandal performane of export credit maximum repayment period, minimum currency- agencies strengthens, bringing a surplus in 1996 specific interest rates, and minimum concessional- illions of avo oollars th Reioneies receined ity for mixed credits (which combine aid resources 20 Pm Pemium rereived and commercial credits). The consensus has sup-- ii Caitms paid ported much progress in rationalizing the struc- 10 r tu re of premiums by reducing interest rate Figure 3E5 Th finmndl performance of exportcredi subsidies; export credits of more than two years agencies srengthens, brnging a surplus in 196 specmaturity are now close to market rates, and this 7'. of, U l .S. dollars = Rcoveres receiveditrend is expected to continue. Although the recent -10 S i _ I _: \ t i tm j developments are not likely to produce fully bar- 0 monized premium rates, there is now agreement Net Cash _ among rthe OECD countries on minimum, or -20 floor, premium rates. And after a short transition 1990 199] 1992 1993 1994 1995 1996 period there should be much greater convergence Sourro: Reran Union data, in the premium rates charged by agencies. 60 Investment insurance andforeign direct expropriation without compensation, losses on investmentflows the investment due to war or civil unrest, and inability to convert and transfer or remit profits One of the most notable developments in the and dividends. This narrow coverage may limit the export credit market has been the rapid growth expansion of investment insurance, which remains in agencies' provision of investment insurance. much smaller in amount than export credits. But The Berne Union member agencies extended a the relative importance of investment insurance in record $15 billion of insurance against foreign export credit agencies' business is expected to direct investment projects in developing coun- grow, given the favorable outlook for sustained tries in 1996, five times more than in 1990. This foreign direct investment flows to developing growth trend has been closely associated with the countries. The latest data show that insured invest- surge in foreign direct investment flows in the ment is approaching $20 billion a year. 1990s (figure 3.6). The total investment under cover by member agencies (the outstanding Multilateralguarantees exposure, or stock) rose to $43 billion by the end of 1996, up from $17 billion in 1990. Ranked by Multilateral institutions also have expanded their the outstanding amount under cover, the Berne guarantee activities during the 1990s in response Union member agencies active in this market are to increasing private sector flows and the growing led by EID/MITI (Uapan), OPIC (United States), involvement of the private sector in infrastructure. C&L (Germany), and the Multilateral Invest- The World Bank Group, which has issued guar- ment Guarantee Agency (MIGA) of the World antees for some time (box 3.2), is further expand- Bank Group. Investment insurance by official ing its guarantee activity to meet the expected agencies has covered about 10-15 percent of the increase in demand for risk mitigation instru- foreign direct investment flows to developing ments. The Bank's Development Committee countries. agreed in principle to double MIGA's capital to $2 The strong growth in investment insurance in billion and urged swift action to finalize the capi- recent years reflects the rising demand for political tal increase by the 1998 spring meetings of the risk cover for long-duration projects, often in con- World Bank Group and the IME The IDA direc- junction with large privatization programs. Unlike tors approved a pilot program for IDA's provision export credits, investment insurance by export of partial risk guarantees to private lenders against credit agencies excludes commercial risks and is country risks. And the Bank's executive directors normally limited to coverage of nationalization or approved the use of IBRD partial risk guarantees to support private enclave projects in IDA-only Figure 3.6 Growth in investment insurance has countries. paralleled the surge in foreign direct investment Other multilateral institutions also provide in developing countries guarantees. The Inter-American Development Investment insurance FDI flows Bank approved its first private sector partial risk Billions of US. dollofs Billions of US. dollars guarantee in April 1997. The Asian Development 25 120 Bank has approved $252 million in guarantees 20 100 since 1988 for projects in which it has a stake through a direct loan, bond subscription, or equi- 15 Investment 80 ry investment. The European Bank for Recon- insuronce 60 struction and Development had approved 14 1 0 guarantee operations for a total of Ecu 401 million 40 (approximately $497.2 million) by the end of 5 / 11 1 20 1996. Smaller multilateral institutions providing guarantees include the West African Development 198 0m R m Ba-nk, the East African Development Bank, and 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 the Islamic Corporation for the Insurance of Souroe: Berne Union dota ond World Bonk Debtor Reporting System. Investment and Export Credit. 61 Box 3.2 World Bank Group gurantees In support of private sedor growth Supporting private sector growth is central to the a contract for the entire amount ofinsurance request- World Bank Group's core mission of reducing pover- ed by an investor but retains only part of the expo- ty, Its guarantee programs are intended to serve as a sure for its own account, the rest is underwritten by catalyst for private sector activities in devetoping privte insurets. countries by mitigating noncommercial risks facing IFC offers a loan syndation program under investors and lenders. Bank Group guarantees have which participating lenders enjoy ftC's lender-of- supported a growing volume of private flows; the record umbrella coverage, which ptovides a degree of amount covered increased from $1 A4 billion in fiscal currency transfer protection. IFC also provides guar- 1991 to $4.5 billion in fiscal 1997, with most of the antees to support the credit of parties engaged in coverage . financed by the International Finance derivatives transactions, attracs institutional investor Corporation (IFC\. financing for certificates backed by a pool of IFC loans Bank Group members offer different kinds of and additional IFC credit enhancement, and provides guarantees. IBRD guarantees require a sovereign a limited amount of ul risk coverage guarantees, counterguarantee, while MICA and IFC do not principally for domestic currency intermediation. MICA affers political risk guarantees, primarily I3RD offers paral credit guarantees and partial for equity and related debt investmnents, which cover risk guarantees designed to help open new areas liar expropriation, war and civil disturbance, currency project financing and other forms af funding by pri- transfer, and breach of contract provided that the vate capital. Its partil credit guarantees cover all claimant is denied apropriate judicial or arbritral events of nonpayment for a designated part of a relief In addition, under the Cooperative Under- financing (usually the later maturities), and its partial writing Programt that MIGA has developed, it issues risk guarantees cover sovereign risks. Printe hwestnent flows covered by World Baon Grua guerantees, fiscal 1991 ..d 1997 (bdwn of U S. dol&n) Percentage Typ of garante 1991 1997 change IFCsyndications 1.3 3.4 161 MIGA insurance 0.% 0.7 1,067 IBRR mainstream guaranrtee n.a 0.4 n.a. Total :-A+4 4.5 221 u.ts.Not pa$btc. a. C-raredin 1994 Soure; W1Forld Bank Group. The benefits of expanded guarantee activity for Notes developing countries 1. About 70 percent of aid is tied to the dollar, so The growth and innovation in guarantees have done appreciation of the dollar affects the dollar value of 30 percent of aid flows (about $11 billion). Thus the 10 much to help developing countries expand the role percent appreciation of the dollar in 1997 lowered the of the private sector and to increase their integration dollar value of aid flows by approximately $1.1 billion. with the global economy. Export credit guarantees 2. Average prices in donor countries expressed in help countries access a wider range of finance and dollars declined by 3 percent in 1996 as the dollar appre- obtain improved financing terms. The growth in ciated (particularly against the yen). 3. Part of the decline in the ratio of ODA to GNP in 1997, perhaps 0.2 percentage point, is due to the enterprises and increasing the private sectors removal of Israel and a few smaller countries from the involvement in the provision of infrastructure ser- DAC's list of ODA-eligible countries. vices. Investment guarantees can play a key role in 4. These perceptions were reported by a study by strengthening investor confidence when a govern- the Program on International Policy Attitudes, which included a nationwide poll, focus groups, and telephone ment is just beginning to implement economic interviews. reforms. But where reforms are more firmly in place, interviews. refor. B5. However, Levy (1988) found that aid was posi- some governments have decided that the private sec- tively and significantly correlated with investment and tor should bear the full risk of investments. economic growth for low-income countries in Sub- 62 Saharan African, without distinguishing between good 9. The Bank has taken the lead in coordinating mul- and bad performers. tilateral creditors, with contacts involving 20 multilat- 6. Excluding the "blend" countries China, India, eral institutions. The IMF has taken the lead in contacts and Pakistan, which have access either to private capital with bilateral creditors. The Paris Club has stated its markets or to official nonconcessional resources. willingness to increase the concessionality of debt 7. These countries are Angola, Benin, Bolivia, rescheduling from 67 percent to up to 80 percent (Lyon Burkina Faso, Burundi, Cameroon, Central African terms) in the context of the HIPC Debt Initiative. Republic, Chad, Democratic Republic of the Congo, 10. The experience of the 1980s shows that coun- Republic of Congo, Core d'Ivoire, Equatorial Guinea, tries typically reschedule their debt or build up arrears if Ethiopia, Ghana, Guinea, Guinea-Bissau, Guyana, the ratio of the present value of debt to exports is on the Honduras, Kenya, Lao People's Democratic Republic, order of 200-250 percent or the ratio of debt service to Liberia, Madagascar, Mali, Mauritania, Mozambique, exports is on the order of 20-25 percent. Myanmar, Nicaragua, Niger, Nigeria, Rwanda, Sao 11. The IMF and Bank Boards have held prelimi- Tome and Principe, Senegal, Sierra Leone, Somalia, nary discussions on the eligibility of Cote d'Ivoire and Sudan, Tanzania, Togo, Uganda, Vietnam, Republic of Mozambique, which are expected to reach their decision Yemen, and Zambia. points in the spring of 1998. 8. Many studies have presented empirical evidence 12. This estimate excludes several countries showing that high debt levels reduce growth and invest- (Liberia, Somalia, and Sudan) because of insufficient ment. Studies focusing on Africa, where most of the data. HIPCs are located, include Hadjimichael and others 13. Export credit commitments are excluded from 1995; Cohen 1996; Oshikoya 1994; and Savvides 1992. the data on official development finance. Part II Appendixes APPENDIX 1 Debt burden indicators and country classifications The 138 countries that report debt data to the of the present value of public and publicly guar- World Bank's Debtor Reporting System (DRS) anteed debt to exports of goods and services, are classified by degree of indebtedness, which is excluding worker remittances.) measured using debt and debt service data from * The ratio of the present value of total debt the DRS and GNP and export data from World service in 1996 to average exports in 1994, 1995, Bank files, as shown in the Country Tables volume and 1996. of Global Development Finance. Export figures are If either ratio exceeds a critical value-80 per- earnings from goods and services, including work- cent for debt service to GNP ratio and 220 percent er remittances. Data on official grants are not for the debt service to exports ratio-the country included, although they may be a stable source of is classified as severely indebted. If the critical value foreign exchange in some countries. is not exceeded but either ratio is three-fifths or more of the critical value (that is, 48 percent for Country classifications for 1997 the present value of debt service to GNP and 132 percent for the present value of debt service to Global Development Finance 1997 classified indebt- exports), the country is classified as moderately edness based on two ratios: the ratio of the present indebted. If both ratios are less than three-fifths of value of total debt service to GNP and the ratio of the critical value, the country is classified as less the present value of total debt service to exports. indebted. Countries are further classified as low These ratios cast a country's indebtedness in terms of income if 1996 GNP per capita was $785 or less two important aspects of its potential capacity to ser- and as middle income if 1996 GNP per capita was vice the debt: exports (because they provide foreign more than $785 but less than $9,636. Combining exchange to service debt) and GNP (because it is the these criteria leads to the identification of severely broadest measure of income generation in an econ- indebted low-income countries (SILICs), severely omy). This methodology was applied by calculating indebted middle-income countries (SIMICs), the average of the ratios for 1993, 1994, and 1995. moderately indebted low-income countries This approach does not, however, fully capture (MILICs), moderately indebted middle-income the debt position of reporting countries. For countries (MIMICs), less-indebted low-income instance, the resulting classification based on this countries (LILICs), and less-indebted middle- methodology does not reflect the current position of income countries (LIMICs; table Al.1). countries that benefited from debt relief in 1996-97. The use of critical values to define the bound- To reflect such instances of a permanent reduction in aries between indebtedness categories means that debt, this year's Global Development Finance makes a changes in country classifications should be inter- slight change in the dassification methodology, cal- preted with caution. If a country has an indicator culating the two indebtedness ratios as follows: that is close to the critical value, a small change in * The ratio of the present value of total debt the indicator may trigger a change in indebtedness service in 1996 to average GNP in 1994, 1995, classification even if economic fundamentals have and 1996. (In cases where a joint debt sustainabil- not changed significantly. Accordingly, the use of ity analysis has been undertaken in the context of critical values implies a greater degree of precision the Highly Indebted Poor Countries Debt in the exercise than is warranted by their capacity Initiative, countries are classified based on the ratio to signal discrete changes in country indebtedness. 65 66 Table Al.1 Income and indebtedness classification criteria Indebtedness classification PV/XGS less than 220 percent PV/XGS less than PV/XGS higher than 220 but higher than 132 percent or 132 percent percent or PVIGNP higher PV/GNP less than 80 percent and PV/GNP less Income classification than 80 percent bht higher than 48 percent than 48 percent Low-income: GNP per capita Severely indebted Moderately indebted Less-indebted less than $785 low-income countries low-income countries low-income countries Middle-income: GNP per capita Severely indebted Moderately indebted Less-indebted middle- between $786 and $9,635 middle-income countries middle-income counrries income countries Note, PV/XGS is present value of debt service to erports of goods and services. PV/GNP is present value of (iebt service to GNP Source, World Bank. Moreover, these indicators do not represent an discounted to compute its present value. For vari- exhaustive set of useful indicators of external debt. able-rate loans, for which the future debt service They may not, for example, adequately capture the payment cannot be precisely determined, debt ser- debt servicing capacity of countries in which gov- vice is calculated using the rate at the end of 1996 ernment budget constraints are key to debt service for the base specified for the loan. difficulties. Countries (such as the franc zone coun- tries in Africa) that allow the use or free conversion Classification of low-income countries of a foreign currency can face government budget difficulties that are related to servicing external Applying the present value methodology to public debt but that are not necessarily reflected in 1994-96 data, 36 countries are classified as balance of payments data. In other countries the SILICs, 13 as MILICs, and 12 as LILICs (table servicing of domestic public debt may be a source Al.2). There are five changes from last year in the of fiscal strain that is not reflected in balance of pay- indebtedness classification of low-income coun- ments data. But rising external debt may not nec- tries: Cambodia, Kenya, and Togo joined the essarily imply payment difficulties, especially if moderately indebted group because of a decrease there is a commensurate increase in the country's in the debt to GNP ratio (Kenya) and decreases in debt servicing capacity. Thus these indicators debt to export ratios (Cambodia and Togo). should be used in the broader context of a country- Burkina Faso and Haiti became severely indebted specific analysis of debt sustainability. because of increases in debt to export ratios. The discount rates used to calculate present value are interest rates charged by Organisation for Classification of middle-income countries Economic Co-operation and Development (OECD) countries for officially supported export In the middle-income group, 12 countries are clas- credits. They represent, on average, the most sified as SIMICs, 16 as MIMICs, and 47 as favorable terms for fixed-rate nonconcessional debt LIMICs. There were 13 changes in the indebted- that countries are able to contract in international ness classification since last year. Algeria and loan markets. The rates are specified for 19 curren- Indonesia became severely indebted (because of cies, induding G-7 currencies-British pounds, increases in the present value of debt to exports Canadian dollars, French francs, German marks, ratios). Five countries joined the moderately Italian lire, Japanese yen, and U.S. dollars. Inter- indebted group: Georgia (because of an increase in national Bank for Reconstruction and Development the debt to exports ratio), Malaysia, Panama, and (IBRD) currency-pool loans, International Thailand (because of increases in debt to GNP Development Association (IDA) credits, and ratios), and Mexico (because of an increase in International Monetary Fund (IMF) loans are dis- exports). Six other countries joined the less- counted at the Special Drawing Rights (SDR) lend- indebted group: Egypt (because of improvements ing rate. For debt denominated in other currencies, in both the debt to exports and debt to GNP discount rates are the average of interest rates on ratios), Poland and Russia (because of an improve- export credits charged by other OECD countries. ment in debt to exports ratio), and Papua New In present value calculations, debt service on Guinea, Samoa, and Trinidad and Tobago fixed-rate loans is determined and each payment (because of improvements in debt to GNP ratios). 67 Table A1.2 Classification of DRS economies Severely indebted Severely indebted Moderately indebted Moderately indebted Less-indebted Less-indebted low-income middle-income low-income middle-income low-income middle-income Angola Algeria' Bangladesh Chile Armenia Albania' Burkina Faso' Argentina Benin Colombia Azerbaijan Barbados Burundi Bolivia Cambodiab Georgia"' Bhutan Belarus Cameroon Brazil Chad Hungary China Belize Central African Republic Bulgaria Comoros Macedonia, FYR Eritrea Botswana Congo, Dem. Rep. Ecuador Gambia, The Malaysia' Kyrgyz Republic Cape Verde Congo, Rep. Gabon India Mexicob Lesotho' Costa Rica C6te d'lvoire Indonesia' Kenyab Morocco Moldova' Croatia Equatorial Guinea Jamaica Lao PDR Panamab Mongolia Czech Republic Ethiopia Jordan Pakistan Philippines Nepal Djibouti Ghana Peru Senegal St. Vincent and the Grenadines Sri Lanka Dominica Guinea Syrian Arab Republic Togob Thailand' Tajikistan Dominican Republic Guinea-Bissau Zimbabwe Tunisia Egypt, Arab Rep.' Guyana Turkey El Salvador Haiti' Uruguay Estonia Honduras Venezuela Fiji Liberia Grenada Madagascar Guatemala Malawi Iran, Islamic Rep. Mali Kazakhstan Mauritania Latvia Mozambique Lebanon Myanmar Lithuania Nicaragua Maldives Niger Malta Nigeria Mauritius Rwanda Oman Sao Tome and Principe Papua New Guineab Sierra Leone Paraguay Somalia Polandb Sudan Romania Tanzania Russian Federation' Uganda Samoab Vietnam Seychelles Yemen, Rep. Slovak Republic Zambia Slovenia Solomon Islands South Africa St. Kitts and Nevis St. Lucia Swaziland Tonga Trinidad and Tobagob Turkmenistan Ukraine Uzbekistan Vanuatu Note: Tables Al.2 and Al.3 classify all World Bank member economies and all other economies with populations of more than 30,000. Economies are divided arnong income groups according to 1996 GNP per capita, calculated using the WorliBankAtlas method. Income groups are low-income, $785 or less; lower-middle-income, $786-3,115; upper-middle- income, $3,116-9,635; and high-income, $9,636 or more. The tuble excludes Bosnia and Hereegovina and the Federal Republic of Yugoslavia (Serbia and Montenegro) because of lack of data. a. Countries whose indebtedness has increased. b. Countries whose indebtedness has decreased. c. Countries whose income classification has changed. Source: WXorld Bank Debtor Reporting System. 68 Table A1.3 Classification of non-DRS economies Severely indebted Severely indebted Moderately indebted Less-indebted low-income middle-income middle-income middle-income Afghanistan Cuba Gibraltar Antigua and Barbuda Namibia Iraq Bahrain Saudi Arabia Kiribati Suriname Korea, Dem. Rep. Libya Note: Tables A1.2 and Al.3 classify all World Bank member economies and all other economies with populations of more than 30,000. Economies are divided among income groups according to 1996 GNP per capita, calculated using the Workl Bank Atlas method. Income groups are low-income, $785 or less; lower-middle-income, $786-$3,115; upper-middle-income, $3,116 $9,635; :nd high-income, $9,636 or more. Source: World Bank Debtor Reporting System. Table Al.4 Major economic indicators, 1996 (millions of U.S. dollars) Country ED T PV TDS INT XGS GNP Albania 781 759 34 21 956 2,750 Algeria 33,260 29,154 4,215 2,147 15,215 43,349 Angola 10,612 8,943 694 199 5,215 3,455 Argentina 93,841 88,065 14,021 5,904 31,740 290,960 Armenia 552 387 49 12 457 1,621 Azerbaijan 435 357 10 10 758 3,595 Bangladesh 16,083 8,681 693 215 5,907 31,818 Barbados 581 565 101 40 1,349 - Belarus 1,071 970 120 56 6,047 22,165 Belize 288 233 40 13 308 613 Benin 1,594 1,056 43 17 628 2,168 Bhutan 87 42 6 2 110 271 Bolivia 5,174 3,344 413 182 1,334 6,395 Bosnia and Herzegovina - - - - - - Botswana 613 478 150 76 3,090 4,760 Brazil 179,047 167,555 25,091 10,637 61,065 731,510 Bulgaria 9,819 9,278 1,284 548 6,262 9,111 Burkina Faso 1,294 684 49 17 448 2,527 Burundi 1,127 536 31 10 56 1,123 Cambodia 2111 1,509 10 6 829 3,116 Cameroon 9,515 8,161 528 265 2,238 8,437 Cape Verde 211 128 6 3 196 564 Central African Republic 928 511 13 5 204 1,038 Chad 997 503 31 14 319 1,133 Chile 27,411 30,179 6,270 1,594 19,404 72,358 China 128,817 116,539 15,756 5,496 180,670 807,210 Colombia 28,859 30,121 5,401 2,036 15,616 81,793 Comoros 206 120 1 1 61 230 Congo, Dem. Rep. 12,826 11,636 48 12 2,009 6,050 Congo, Rep. 5,240 4,405 339 199 1,588 1,878 Costa Rica 3,454 3,196 585 215 4,140 9,003 Cote d'lvoire 19,713 14,708 1,347 487 2,521 9,795 Croatia 4,634 4,175 450 163 8,235 19,036 Czech Republic 20,094 19,217 2,590 940 31,044 54,210 Djibouti 241 137 11 3 209 - Dominica II1 75 7 3 128 221 Dominican Republic 4,310 3,854 445 198 3,888 12,770 Ectuador 14,491 12,952 1,314 669 5,821 17,661 Egvpt, Arab Rep. 31,407 20,866 2,309 1,194 19,872 67,850 El Salvador 2,894 2,381 314 137 3,316 10,374 Equatorial Guinea 282 213 5 1 180 243 69 Table A1.4 Major economic indicators, 1996 (continued) (millions of U.S. dollars) Country EDT PV TDS INT XGS GNP Eritrea 46 18 0 0 328 756 Estonia 405 381 43 19 3,284 4,354 Ethiopia 10,077 7,997 348 55 823 5,950 Fiji 217 206 49 14 1,342 1,986 Gabon 4,213 3,677 378 243 3,411 4,823 Gambia, The 452 236 29 7 226 - Georgia 1,356 1,042 13 12 - 4,471 Ghana 6,202 3,249 471 148 1,728 6,203 Grenada 120 91 9 3 - 286 Guatemala 3,785 3,317 353 160 3,212 15,586 Guinea 3,240 2,153 114 47 775 3,785 Guinea-Bissau 937 617 11 5 23 266 Guyana 1,631 1,150 105 32 - 663 Haiti 897 446 27 13 192 2,607 Honduras 4,453 3,459 564 173 1,960 4,006 Hungary 26,958 26,221 8,362 1811 20,401 43,411 India 89,827 70,611 12,669 4,457 52,605 350,320 Indonesia 129,033 124,373 21,459 6,647 58,270 216,280 Iran, Islamic Rep. 21,183 19,969 3,405 1119 - 142,580 Jamaica 4,041 3,638 682 228 3,795 4,284 Jordan 8,118 7,074 656 333 5,319 7,102 Kazakhstan 2,920 2,723 695 147 6,995 20,949 Kenya 6,893 5,130 840 277 3,048 8,966 Kyrgyz Republic 789 577 51 15 555 1,673 Lao PDR 2,263 776 29 7 466 1,857 Latvia 472 440 65 26 2,768 5,025 Lebanon 3,996 3,795 301 231 4,724 13,283 Lesotho 654 399 36 14 595 1,235 Libcria 2,107 1,974 1 1 - - Lithuania 1,286 1,164 125 42 4,265 7,688 Macedonia, FYR 1,659 1,402 53 27 1,347 2,003 Madagascar 4,175 3,191 76 22 815 3,986 Malawi 2,312 1,215 89 33 425 2,157 Malaysia 39,777 41,672 7,657 1,830 93,878 94,526 Maldives 167 97 12 4 387 284 Mali 3,020 1,265 116 62 646 2,597 Malta 953 902 50 38 3,243 - Mauritania 2,363 1,588 121 34 557 1,038 Mauritius 1,818 1,753 198 90 2,732 4,241 Mexico 157,125 148,458 40,786 11,722 115,156 321,260 Moldova 834 748 57 36 921 1,774 Mongolia 524 307 48 8 493 953 Morocco 21,767 19,730 3,174 1,313 11,445 35,609 Mozambique 5,842 5,565 175 73 479 1,543 Myanmar 5,184 4,146 158 18 - Nepal 2,413 1,155 85 31 1,111 4,521 Nicaragua 5,929 4,842 221 88 913 1,672 Niger 1,557 806 57 12 327 1,958 Nigeria 31,407 29,394 2,509 1,091 15,669 31,125 Oman 3,415 3,178 751 180 7,610 - Pakistan 29,901 23,190 3,276 1,190 11,936 64,633 Panama 6,990 6,147 954 491 8,888 8,073 Papua New Guinea 2,359 1,822 376 94 2,994 4,783 Paraguay 2,141 1,937 240 107 4,353 9,564 Peru 29,176 23,743 2,932 1,705 8,280 59,406 Philippines 41,214 38,615 5,778 2,265 42,249 87,136 Poland 40,895 35,140 2,552 1,438 39,640 134,110 Romania 8,291 7,861 1,230 459 9,736 35,107 Russian Federation 124,785 91,623 7,002 4,076 106,682 432,384 (table continues on next page) 70 Table A1.4 Major economic indicators, 1996 (continued) (millions of U.S. dollars) Country EDT PV TDS INT XGS GNP Rwanda 1,034 513 19 8 91 1,317 Samoa 167 73 5 2 124 177 Sao Tome and Principe 261 141 3 1 13 41 Senegal 3,663 2,384 274 117 1,717 5,025 Seychelles 148 122 15 5 351 514 Sierra Leone 1,167 661 59 12 112 922 SlovakRepublic 7,704 6,879 1,328 508 11117 18,919 Slovenia 4,031 3,712 951 208 10,958 18,713 Solomon Islands 145 89 8 2 231 359 Somalia 2,643 2,225 4 3 - - South Africa 23,590 22,014 3,783 1,013 34,168 123,430 Sri Lanka 7,995 5,262 427 160 5,868 13,710 St. Kitts and Nevis 58 40 6 2 139 232 St. Lucia 142 110 13 6 427 554 St. Vincent and the Grenadines 213 182 14 9 170 265 Sudan 16,972 15,640 48 12 961 Swaziland 220 146 33 12 1,138 1,037 Syrian Arab Republic 21,420 18,815 254 178 6,665 16,415 Tajikistan 707 502 1 1 772 2,030 Tanzania 7,412 5,561 258 105 1,383 5,714 Thailand 90,824 90,818 8,652 4,343 75,385 180,400 Togo 1,463 936 56 26 517 1,388 Tonga 70 40 4 1 84 184 Trinidad and Tobago 2,242 2,164 473 194 3,021 4,991 Tunisia 9,887 8,969 1,479 559 8,952 18,459 Turkey 79,789 76,358 10,940 4,619 50,473 183,990 Turkmenistan 825 775 180 39 1,691 4,346 Uganda 3,674 1,690 150 43 250 6,069 Ukraine 9,335 8,682 1,254 466 20,597 43,436 Uruguay 5,899 5,597 665 367 4,260 17,993 Uzbekistan 2,319 2,062 289 104 3,551 23,907 Vanuatu 47 24 2 1 136 227 Venezuela 35,344 33,571 4,506 2,141 26,814 65,769 Vietnam 26,764 24,234 346 196 9,867 23,340 Yemen, Rep. 6,356 5,242 84 26 3,591 5,288 Yugoslavia, FR (Serbia and Montenegro) 13,439 13,434 19 17 - - Zambia 7,113 5,078 324 122 1,319 3,294 Zimbabwe 5,005 4,240 664 240 3,126 7,227 -Not available. Note: For definition of indicators, see Sources and Definiriorr section. Numbers in italics are from debt sustainability analyses undertaken in the context of the Heavily Indebted Poor Countries Debt Initiative. Present value estimates for these countries are for public and publicly guaranteed debt only, and export figures exclude worker remittances. Source. World Bank Debtor Reporting System and staff estimates. 71 Table Al.5 Key indebtedness ratios, 1994-96 (percent) Country EDT/XGS PV/XGS EDT/GNP PV/GNVP TDS/XGS INTIXGS Albania 104 101 32 32 4 3 Algeria 260 228 81 71 33 17 Angola 260 219 368 310 17 5 Argentina 344 323 33 31 51 22 Armenia 162 114 39 27 14 4 Azerbaijan 55 45 12 10 1 1 Bangladesh 308 166 56 30 13 4 Barbados 47 46 34 33 8 3 Belarus 23 21 5 4 3 1 Belize 94 76 50 41 13 4 Benin 324 215 86 57 9 3 Bhutan 87 42 33 16 6 2 Bolivia 418 270 89 57 33 15 Bosnia and Herzegovina - - - - - - Botswana 22 17 14 11 5 3 Brazil 310 293 28 26 43 18 Bulgaria 160 151 94 89 21 9 Burkina Faso 456 241 58 31 17 6 Burundi 1,131 538 100 47 31 10 Cambodia 267 191 75 54 1 1 Cameroon 465 399 124 106 26 13 Cape Verde 121 74 40 24 3 2 Central African Republic 439 242 92 51 6 2 Chad 359 181 101 51 11 5 Chile 151 166 44 48 35 9 China 84 76 19 17 10 4 Colombia 197 206 38 40 37 14 Comoros 343 200 93 54 2 2 Congo, Dem. Rep. 764 693 140 127 3 1 Congo, Rep. 406 342 309 260 26 15 Costa Rica 90 83 40 37 15 6 Cote d'Ivoire 400 299 230 171 27 10 Croatia 62 56 27 24 6 2 Czech Republic 73 70 44 42 9 3 Djibouti 110 63 49 28 5 1 Dominica 93 63 52 35 6 3 Dominican Republic 121 108 37 33 12 6 Ecuador 275 246 88 78 25 13 Egypt, Arab Rep. 176 117 53 35 13 7 El Salvador 95 78 31 26 10 5 Equatorial Guinea 249 188 168 126 4 1 Eritrea 15 6 7 3 0 0 Estonia 15 14 10 9 2 1 Ethiopia 1,377 1,093 187 149 47 8 Fiji 18 17 12 11 4 1 Gabon 141 123 98 86 13 8 Gambia, The 216 113 122 64 14 3 Georgia 272 209 33 26 3 2 Ghana 397 208 107 56 30 9 Grenada 78 59 45 34 6 2 Guatemala 126 110 26 23 12 5 Guinea 449 298 92 61 16 7 Guinea-Bissau 3,509 2,312 376 248 42 19 Guyana 255 180 294 207 16 5 Haiti 597 297 41 20 18 9 Honduras 257 200 118 92 33 10 Hungary 162 158 64 62 50 11 India 194 152 28 22 27 10 Indonesia 245 236 67 64 41 13 (table continues on next page) 72 Table A1.5 Key indebtedness ratios, 1994-96 (continued) (percent) Country EDT/XGS PV/XGS EDT/GNP PV/GNP TDS/XGS INT/XGS Iran, Islamic Rep. 105 99 16 15 17 6 Jamaica 112 101 102 92 19 6 Jordan 170 148 126 110 14 7 Kazakhstan 51 48 15 14 12 3 Kenya 238 177 86 64 29 10 Kyrgyz Republic 178 130 50 37 12 3 Lao PDR 517 177 132 45 7 2 Latvia 21 20 9 9 3 1 Lebanon 94 90 35 33 7 5 Lesotho 114 69 55 33 6 2 Liberia 414 388 130 121 0 0 Lithuania 39 35 17 16 4 1 Maccdonia, FYR 125 106 87 74 4 2 Madagascar 557 426 127 97 10 3 Malawi 560 294 145 76 21 8 Malaysia 48 50 49 52 9 2 Maldives 51 29 66 38 3 1 Mali 624 261 133 56 24 13 Malta 30 29 30 29 2 1 Mauritania 473 318 234 157 24 7 Mauritius 76 73 47 45 8 4 Mexico 163 154 47 44 42 12 Moldova 102 92 44 39 7 4 Mongolia 111 65 62 36 10 2 Morocco 204 185 68 61 30 12 Mozambique 1,411 1,345 432 411 42 18 Myanmar 370 296 43 34 11 1 Nepal 214 102 55 26 8 3 Nicaragua 934 763 395 322 35 14 Niger 548 284 88 45 20 4 Nigeria 256 240 121 114 20 9 Oman 51 48 33 31 11 3 Pakistan 265 206 50 39 29 11 Panama 78 69 91 80 11 6 Papua New Guinea 79 61 48 37 13 3 Paraguay 52 47 25 22 6 3 Peru 391 318 53 43 39 23 Philippines 124 116 54 51 17 7 Poland 119 102 36 31 7 4 Romania 94 89 25 23 14 5 Russian Federation 132 97 34 25 7 4 Rwanda 1,374 682 95 47 25 11 Samoa 157 69 92 40 5 2 Sao Tom6 and Principe 2,132 1,150 647 349 25 8 Senegal 231 150 82 53 17 7 Seychelles 50 41 29 24 5 2 Sierra Leone 909 515 138 78 46 9 Slovak Republic 73 66 46 41 13 5 Slovenia 39 36 23 21 9 2 Solomon Islands 68 42 43 26 4 1 Somalia 3,671 3,090 406 341 5 4 South Africa 72 67 19 18 12 3 Sri Lanka 147 97 63 41 8 3 St. Kitts and Nevis 43 29 26 18 4 1 St. Lucia 36 28 27 21 3 2 St. Vincent and the Grenadines 136 116 86 74 9 6 Sudan 2,131 1,964 282 260 6 2 Swaziland 20 13 21 14 3 1 Syrian Arab Republic 342 301 136 120 4 3 73 Table Al.5 Key indebtedness ratios, 1994-96 (continued) (percent) Country EDT/XGS PV/XGS EDT/GNP PV/GNP TDS/XGS INT/XGS Tajikistan 97 69 33 24 0 0 Tanzania 666 499 152 114 23 9 Thailand 131 131 56 56 12 6 Togo 298 191 125 80 11 5 Tonga 100 57 41 23 6 1 Trinidad and Tobago 83 80 48 46 17 7 Tunisia 117 106 59 53 18 7 Turkey 192 184 49 47 26 11 Turkmenistan 42 39 19 18 9 2 Uganda 639 294 70 32 26 7 Ukraine 51 48 19 18 7 3 Uruguay 151 143 34 33 17 9 Uzbekistan 63 56 10 9 8 3 Vanuatu 37 19 22 12 1 1 Venezuela 154 147 54 51 20 9 Vietnam 355 322 136 123 5 3 Yemen, Rep. 194 160 107 88 3 1 Yugoslavia, FR (Serbia and Montenegro) - - - _ _ _ Zambia 545 389 225 161 25 9 Zimbabwe 182 154 79 67 24 9 - Not available. iVote. For definition of indicators, see Sources and Definitioro section. Numbers in italics are from debt sustainability analyses undertaken in the context of the Heavily Indebted Poor Countries Debt Initiative. Present value estimates for these countries are for public and publicly guaranteed debt only, and export figures exclude worker remittances. Source- World Bank Debtor Reporting System and staff estimates. 74 Table A1.6 Classification of economies by income and region, 1998 Europe and Middle East Sub-Saharan Africa Asia CentralAsia and North Africa Eastern Income East and East Asia Europe and Rest of Middle North group Subgroup Southern Africa West Africa and Pacific South Asia CentralAsia Europe East Afi-ica Americas Loxv- Angola Benin Cambodia Afghanistan Armenia Yemen, Rep. Guvana income Burundi Burkina Faso China Bangladesh Azerbaijan Haiti Comoros Cameroon Lao PDR Bhutan Bosnia and Honduras Congo, Dem. Central Afri- Mongolia India Herzegovina Nicaragua Repa can Republic Myanmar Nepal Kyrgyz Republic Eritrea Chad Vietnam Pakistan Moldova Ethiopia Congo, Rep. Sri Lanka Tajikistan Kenya Cote d'lvoire Lesotho Equatorial Guinea Madagascar Gambia, The Malawi Ghana Mozambique Guinea Rwanda Guinea-Bissau Somalia Liberia Sudan Mali Tanzania Mauritania Uganda Niget Zambia Nigeria Zimbabwe Sao Tome and Principe Senegal Sierra Leone Togo Middle- Lower Botswana Cape Verde Fiji Maldives Albania Turkey Iran, Islamic Algeria Belize income Djibouti Indonesia Belarus Rep. Egypt, Arab Bolivia Namibia Kiribati Bulgaria Iraq Rep. Colombia Swaziland Korea, Dem. Estonia Jordan Morocco Costa Rica Rep. Georgia Lebanon Tunisia Cuba Marshall Islands Kazalkhstan Syrian Arab Dominica Micronesia, Latvia Republic Dominican Fed. Sts. Lithuania West Bank Republic Papua New Macedonia, and Gaza Ecuador Guinea FYRb El Salvador Philippines Romania Guatemala Samoa Russian Jamaica Solomon Islands Federation Panama Thailand Turkmenistan Paraguay Tonga Ukraine Peru Vanuatu Uzbekistan Sr. Vincent Yugoslavia, FR' and the Grenadines Suriname Venezuela Upper Mauritius Gabon American Croatia Isle of Man Bahrain Libya Antigua and Mayotte Samoa Czech Malta Oman Barbuda Seychelles Malaysia Republic Saudi Arabia Argenitinia South Africa Palau Hungary Barbados Slovak Republic Brazil Slovenia Chile Grenada Guadeloupe Mexico 75 Table A1.6 Classification of economies by income and region, 1998 (continued) Europe and Middle East Sub-Saharan Africa Asia CentralAsia and North Africa East and Eastern Income Southern East Asia Europe and Rest of Middle North group Subgroup Africa WestAfrica and Pacific SouthAsia CentralAsia Europe East Africa Americas Puerto Rico St. Kitts and Nevis St. Lucia Trinidad and Tobago Uruguay Subtotal 158 26 23 22 8 27 3 10 5 34 High- OECD Australia Austria Canada income Japan Belgium United States Korea, Rep. Denmark New Zealand Finland France Germany Greece Iceland Ireland Italy Luxembourg Netherlands Norway Portugal Spain Sweden Switzerland United Kingdom Non-OECD Reunion Brunei Andorra Israel Aruba French Channel Kuwait Bahamas, The Polynesia Islands Qatar Bermuda Guam Cyprus United Arab Cayman Hong Kongd Faeroe Islands Emirates Islands Macao Greenland French Guiana New Liechtenstein Martinique Caledonia Monaco Netherlands N. Mariana Antilles Islands Virgin Singapore Islands (U.S.) OAE' Total 211 27 23 35 8 27 28 14 5 44 Note: This table classifies all World Bank member economies and all other economies with populations of more than 30,000. For operational and analytical purposes, the World Bank's main criterion for classifying economies is gross national product (GNP) per capita. Every economy is classified as low- income, middle-income (subdivided into lower-middle and upper-middle), or high-income. Other analytical groups, based on region and external debt, are also used. Classification by income does not necessarily reflect development status, although low-income and middle-income economies are sometimes referred to as deveLoping economies. The use of that term is convenient; it is not intended to imply that all economies in the group are experiencing similar development or that other economies have reached a preferred or final stage of develop- ment. Economies are divided among income groups, according to 1996 GNP per capita, calculated using the WorldBankAtla method. Income groups are low-income, $785 or less; lower- middle-income, $786-$3,115; upper-middle-income, $3,116 -$9,635; and high-income, $9,636 or more. a. Formerly Zaire. b. Former Yugoslav Republic of Macedonia. c. Federal Republic of Yugoslavia (Serbia and Montenegro). d. As of July 1, 1997, Hong Kong is part of China. e. Other Asian economies (Taiwan, China). Soarce: World Bank data. 76 Table A1.7 Classification of economies by income group and indebtedness category, 1997 Not classified by Income group Subgroup Severely indebted Moderately indebted Less indebted indebtedness Low- Afghanistan Bangladesh Armenia income Angola Benin Azerbaijan Bosnia and Herzegovina Cambodia Bhutan Burkina Faso Chad China Burundi Comoros Eritrea Cameroon Gambia, The Kyrgyz Republic Central African Republic India Lesotho Congo, Dem. Rep.a Kenya Moldova Congo, Rep. Lao PDR Mongolia C6te d'Ivoire Pakistan Nepal Equatorial Guinea Senegal Sri Lanka Ethiopia Togo Tajikistan Ghana Zimbabwe Guinea Guinea-Bissau Guyana Haiti Honduras Liberia Madagascar Malawi Mali Mauritania Mozambique Myanmar Nicaragua Niger Nigeria Rwanda Sio Tome and Principe Sierra Leone Somalia Sudan Tanzania Uganda Vietnam Yemen, Rep. Zambia Middle- Lowver Algeria Colombia Albania Marshall Islands income Bolivia Georgia Belarus Micronesia, Fed. Sts. Bulgaria Macedonia, FYRb Belize West Bank and Gaza Cuba Morocco Borswana Ecuador Panama Cape Verde Indonesia Philippines Costa Rica Iraq St. Vincent and Djibouti Jamaica the Grenadines Dominica Jordan Thailand Dominican Republic Peru Tunisia Egypt, Arab Rep. Syrian Arab Republic Turkey El Salvador Venezuela Estonia Fiji Guatemala Iran, Islamic Rep. Kazakhstan Kiribati Korea, Dem. Rep. Larvia Lebanon Lithuania 77 Table A1.7 Classification of economies by income group and indebtedness category, 1997 Not classified by In comegroup Subgroup Severely indebted Moderately indebted Less indebted indebtedness Middle- Lower Maldives income Namibia Papua New Guinea Paraguay Romania Russian Federation Samoa Solomon Islands Suriname Swaziland Tonga Turkmenistan Ukraine Uzbekistan Vanuatu Yugoslavia, FR' Upper Argentina Chile Antigua and Barbuda American Samoa Brazil Hungary Bahrain Guadeloupe Gabon Malaysia Barbados Isle of Man Mexico Croatia Mayotte Uruguay Czech Republic Palau Grenada Puerto Rico Libya Malta Mauritius Oman Poland Saudi Arabia Seychelles Slovak Republic Slovenia South Africa St. Kitts and Nevis St. Lucia Trinidad and Tobago High-income OECD Australia Austria Belgium Canada Denmark Finland France Germany Greece Iceland Ireland Italy Japan Korea, Rep. Luxembourg Netherlands New Zealand Norway Portugal Spain Sweden Switzerland United Kingdom United States (table continues on next page) 78 Table A1.7 Cassification of economies by income group and indebtedness category, 1997 Not classified by Income group Subgroup Severely indebted Moderately indebted Less indebted indebtedness Non-OECD Andora Aruba Bahamas, The Bermuda Brunei Cayman Islands Channel Islands Cyprus Faroe Islands French Guiana French Polynesia Guam Greenland Hong Kongd Israel Kuwait Liechtenstein Macao Martinique Monaco Netherlands Antilles New Caledonia North Marina Islands Qatar Reunion Singapore United Arab Emirates Virgin Islands (U.S.) OAE' Total 211 52 29 68 62 Note: This table classifies all World Bank member economies and all other economies with populations of more than 30,000. Incomegro up: Economies are divided according to 1996 GNP per capita, calculated using the World BankArlas method. Income groups are losv-income, $785 or less; lower-middle- income, $786-$3,115; upper-middle-income, $3,116-$9,635; and high-income, $9,636 or more. Indebtedness: Standard World Bank definitions of severe and moderate indebtedness are used to classify economies in this table. Severely indebted means either that the present value of debt service to GNP exceeds 80 percent or that the present value of debt service to exports exceeds 220 percent. Moderately indebted means that either of the two key ratios exceeds 60 percent of, but does not reach, the critical levels. Present value calculation is not possible for economies that do not report detailed debt statistics to the World Bank Debtor Reporting System (DRS). Instead, the following methodology is used to classify non-DRS economies. Severely indebted means that three of four key ratios (averaged over 1994-96) are above criti- cal levels: debt to GNP (50 percent), debt to exports (275 percent), debt service to exports (30 percent), and interest to exports (20 percent). Moderately indebted means that three of the four key ratios exceed 60 percent of, but do not reach, the critical levels. All other classified low- and middle-income economies are listed as less indebted a. Formerly Zaire. b. Former Yugoslav Republic of Macedonia. c. Federal Republic of Yugoslavia (Serbia and Montenegro). d. As of July 1, 1997, Hong Kong is part of China. e. Other Asian economies (Taiwan, China). Sourre: World Bank data. APPENDIX 2 Official debt restructuring Most agreements to restructure official source debt itor governments forgave some or all of the devel- have been reached under the aegis of the Paris opment aid debts owed by the poorest countries Club. Paris Club debts are mostly export credit and increasingly provided new development aid loans-commercial debts at market intercst rates on grant rather than loan terms. that become official debt upon default by the In response to the increasingly difficult situa- debtors, and after export credit agencies have paid tion of the poorest countries, G-7 leaders agreed out claims to the private policyholders who had at the 1991 London summit to increase the degree insured the commercial loans against nonpay- of concessionality in reschedulings for the poorest ment. Paris Club agreements cover government- countries. In December 1991 Paris Club creditors guaranteed debt and intergovernmental loans, adopted London terms, under which the maxi- including aid loans. mum available debt relief on eligible pre-cutoff date debt was increased from one-third to one-half Evolution of terms for low-income countries in present value terms. Following the G-7 summit in Naples in 1994, Paris Club creditors agreed Until the late 1980s official creditors addressed the that, where necessary, concessionality could be repayment difficulties of heavily indebted poor increased to two-thirds on debt eligible for restruc- countries by helping them overcome liquidity prob- turing-arrangements known as Naples terms lems.' The debt relief granted was a rescheduling or (box A2.1). refinancing of some portion of the payments and Creditors also agreed in principle to go beyond arrears falling due during an adjustment program repeated flow rescheduling and to restructure the supported by the International Monetary Fund full stock of eligible debt outstanding for poor (IMF)-in most cases a period lasting between one countries that established a track record of good and three years. But repeated rescheduling and the macroeconomic adjustment policies and debt ser- refinancing of interest falling due on pre-cutoff date vice payments. The aim of such stock-of-debt debt did not solve the debt problem. Instead, it con- operations is to allow a debtor to exit the resched- tributed to a steady increase in the stock of out- uling process and manage its future debt service standing debt for many of the poorest countries. obligations without additional relief Although By the late 1980s it had become evident that in payments requested from the debtor are not nec- some of the poorest countries the problem was not essarily higher under a stock operation than under liquidity but solvency and that some debts were a flow rescheduling, a stock-of-debt deal may unlikely to be repaid in full. In October 1988, fol- require an increase in the actual payments being lowing an agreement at the 1988 Toronto summit made to creditors, as debtors forgo the opportuni- of the Group of Seven (G-7) industrial countries, ty to negotiate future comprehensive flow resched- Paris Club creditors started to provide debt relief ulings of payments falling due. An important on concessional terms to the poorest rescheduling motivation for a debtor to negotiate such a deal is countries by forgiving up to a third of debt pay- that the reduction in the stock of outstanding debt ments rescheduled or by reducing the interest rate to levels judged to be sustainable reduces uncer- charged on rescheduled amounts to below the pre- tainty about future external debt service and so vailing market rate. At the same time, many cred- reduces the concern that foreign creditors might 79 80 Box A Pan'ds mb Napliesi ad Liyonte-ins Ke elements ofNaples terms indude * Flow resrevings provide for the rescheduling Eligibility Determined creditors on a case- 4ofdebt ser on eligiblledebtfalling due during the by-case basis based on a countrys income and indebt- consolidation period (generally in line with the peri- edness. Countries that have previously received od of the IMF arrangement). concessional rechdulings (on Toronto or London i Stock bt operations, under which the stock terms) can be considered for Naples terms. of eligible pre-cutoff date debt is rescheduled conces- Concesionaiy Moiiist countries have received a sionally, are reserved for countries with a satisfactory 2 reduion in eligible non-ODA debt of 67 percent in track record for a minimum of three years with respect net presentvalue terms, Some countries with per capi- to both payments under rescheduling agreements and ta incomes of more than $500 and a ratio of debt to performance under IMF arrangements. Creditors exports ofless thai 350 percentin present valueterms must be confident that the country will be able to may receive a 50 percent net present value reduction, respect the debt agreement as an exit rescheduling to be determined on a case-by-ase basis. because in most casesno further rescheduling can take Coverage covrage (incusion in die rescheduling place. (There is an exception for countries that have agreement) of non-ODA pre-cutoff date debt is deter- been granted a stock-of-debt treatment under Naples mined on a case-by-case basis in light of balance of terms and are eligible under the HIPC Debt Initiative payments need, Debt previously rescheduled on con- for a topping up of debt relief on their stock of debt cessional terms may be subject to further rescheduling of up to 80 percent in net present value terms.) to top up the amount of concessionaiitygranted. Under Key elements of Lyon terms include: such topping up the net present value reduction is Elgibility. Countries that are IDA-only, increased from the original level given under Toronto or Enhanced Structural Adjustment Facility-eligible, London terms to the level agreed under Naples terms. low-income countries, that are in full compliance Choice foptions. Creditors can chose one oftwo with IMF and Paris Club agreements and have sus- concessional options for achieving a 67 (or 50) per- tainable macroeconomic programs, and that need to cent net present value reduction:1 the debt reduction achieve a sustainable debt situation (that is, a net pre- option, under which repayment is made over 23 years sent value of debt to exports of 200-250 percent and with a 6-year grace period, or the debt service reduc- debt service to exports of 20-25 percent) in the medi- tion option, under which the net present value reduc- um term. tion is achieved through concessional interest rates Concessionaity. Qualifying countries will receive (with repayment in less than 23 years).2 There is also debt relief on eligible debt of up to 80 percent in net a commercial, or long maturities, option that provides present value terms on the basis of the sustainability for no net present value reduction and repayment over target agreed on by creditors of the country concerned. 40 years with a 20-year grace period. Creditors choos- X Coverage. Various categories of debt. ing this option undertake best efforts to change to a Choice of options: The debt service reduction concessional option at a later date when feasible. option repayment period will be lengthened to 40 - ODA credits. Pre-cutoff date credits are resched- years, including 8 years of grace. The repayment and uled at interest rates at least as concessional as the orig- grace periods for the debt reduction option would inal interest rates over 40 years with a 16-year grace remain the same as under Naples terms. Under the period (30-year matuity with a 12-year grace period capitalization of moratorium interest option, half of for the 50 percent net present value reduction). the moratorium interest during the grace period Creditors can also choose an option reducing the net would be capitalized and repaid in semiannual equal present value of ODA debt by 67 (or 50) percent. installments during the maturity period. 1. For a 50 percent net present value reduction, the debt service reduction option provides for repayrnent over 23 years with a 6-year grace period; the long maturities option provides for repayment over 25 years with a 16-year grace period. 2. For flow reschedulings there is no grace period; for stock-of-debt operations the grace period is three years. There is also a capitalizadon of moratorium interest option, which achieves the net present value reduccion through a lower interest rate over the same repayment and grace periods as the debt service reduction option. appropriate some of the benefits accruing from Creditors have implemented stock-of-debt stronger future growth. Through the reduction of operations on Naples terms since 1995. There uncertainty-and by making an eventual return to have been six such operations-Uganda and creditworthiness a real possibility-stock-of-debt Bolivia in 1995 and Mali, Guyana, Burkina Faso, deals improve a country's ability to implement and Benin in 1996-covering more than $2 bil- economic adjustment. lion. In addition, 25 flow reschedulings worth 81 more than $11 billion have been concluded on Depending on the financing needs of a resched- Naples terms-6 of them in 1997 (table A2. 1). uling country, Paris Club reschedulings on Naples Although Naples terms provide for a net pre- terms typically involve a comprehensive restruc- sent value reduction of 50 or 67 percent on eligi- turing of pre-cutoff date debt including, where ble debt, the effective net present value reduction necessary, arrears and debts outstanding resulting on total debt to Paris Club creditors is typically from earlier rescheduling agreements with Paris lower, for two reasons. First, post-cutoff date and Club creditors (known as previously rescheduled short-term debt are normally excluded from eligi- debt). Some previously rescheduled debt may ble debt (although in some cases creditors may already have been treated concessionally, under provide exceptional treatment on these debt cate- either Toronto terms (up to 33 percent net present gories, principally by deferring or reprofiling such value reduction) or London terms (up to 50 per- debts without cancellations). Second, official cent net present value reduction). Naples terms development assistance (ODA) debt is resched- agreements often grant an additional reduction in uled over 40 years, with 16 years' grace, at an order to raise the total net present value reduction interest rate at least as concessional as the original granted on previously rescheduled debt (including loan rate. The net present value reduction on nonconcessional previously rescheduled debt) to rescheduled ODA debt depends on the interest 67 percent. The coverage of the agreement and rate on the rescheduled debt compared with the extent of topping up is negotiated on a case-by- interest rate on the original loan and the current case basis. market interest rate, as well as the remaining In the context of the Debt Initiative for Heavily maturity of the loan. A number of creditors, how- Indebted Poor Countries (HIPC), Paris Club ever, have systematically forgiven all ODA claims, creditors agreed to support the efforts of the inter- resulting in substantial debt relief for poor coun- national financial community to ensure that the tries and more debt relief than is required under debt problems of the world's poorest countries are Naples terms. dealt with in a comprehensive way so as to reduce Table A2.1 Paris Club agreements on Naples terms, 1996-97 Earliest Eligible monthfor Amount concessionality level Consolidation consideration Signature Cutoff rescheduled (percentage of period Length of stock of debt Country date date (USS millions) present value) start date (months) rescheduling Benin 25 Oct 96 31 Mar 89 209 67 Debt stock rescheduling n.a. n.a. Burkina Faso 20 Jun 96 1 Jan 91 64 67 Debt stock rescheduling n.a. n.a. Cameroon 24 Oct 97 31 Dec 88 1,270 50 1 Oct 97 35 Aug 98 Chad 14Jun96 30Jun89 12 67 1Jan96 32 Congo 16 Jul 96 1 Jan 86 1,758 67 11 Jul 96 36 Jul 99 Ethiopia 24Jan97 31 Dec89 184 67 1Jan97 34 Jan00 Guinea 26 Feb 97 1 Jan 86 123 50 1 Jan 97 36 Dec 99 Guyana 23 May 96 31 Dec 88 793 67 Debt stock rescheduling n.a. n.a. Honduras 1 Mar 96 1 Jun 90 112 50 1 Jan 95 12 Jan 99 Madagascar 26 Mar 97 1 Jul 83 1,247 67 1 Jan 97 35 Nov 99 Mali 20 May 96 1 Jan 88 33 67 Debt stock rescheduling n.a. n.a. Mozambique 21 Nov 96 1 Feb 84 664 67 1 Nov 96 32 Nov 99 Niger 19 Dec 96 1 Jul83 128 67 1 Dec 96 31 Jun 99 Sierra Leone 28 Mar 96 1 Jul 83 39 67 1 Jan 95 24 Jan 98 Tanzania 21 Jan 97 30 Jun 96 1,608 67 1 Dec 96 36 Nov 99 Yemen, Rep. 24 Sep 96 1 Jan 93 113 67 1 Sep 96 10 b Yemen, Rep. 20 Nov 97 1 Jan 93 1,444 67 1 Nov 97 36 Oct 00 Zambia 28 Feb 96 1 Jan 83 566 67 1 Jan 96 36 Feb 99 n.a. Not applicable. a. In accordance with normal Paris Club practice to base reschedulings on agreed terms of reference when a small number of creditors is involved, the reschedulings for Chad wvere not based on full-fledged agreed minutes. Chad obtained Naples terms, but no date for stock-of-debt operation was specified in the terms of reference. b. The goodwill clauses in the Yemen agreement provided for continuation of debt rescheduling if certain conditions were met. Sourre: World Bank Debtor Reporting System and Paris Club data. 82 their debt to sustainable levels. Following the Lyon claims that were previously being dealt with on a summit of 1997, creditors agreed to augment the bilateral basis. Thus Russia will be able to con- support provided to the poorest countries by tribute to the efforts of Paris Club creditors to help implementing concessional debt relief (especially heavily indebted countries. Naples terms) and bilateral debt relief programs for eligible heavily indebted poor countries. Lyon Rescheduling agreements in 1997 terms increase the concessionality of debt relief on flow reschedulings and stock-of-debt reschedul- In 1997 six agreements were signed with heavily ings to up to 80 percent in net present value terms. indebted poor countries under Naples terms (see Countries that qualify for support under the table A2. 1). Four countries (Ethiopia, HIPC Debt Initiative-that is, countries that have Madagascar, Tanzania, and the Republic of demonstrated a sound and continuing record of Yemen) received a 67 percent reduction (in present economic performance, and for which traditional value terms) on eligible pre-cutoff date debt, while debt relief mechanisms would not be sufficient to two (Cameroon and Guinea) received a 50 percent reach a sustainable external debt position in the reduction (in present value terms). Four of the medium term-will be considered for Lyon terms. agreements (Cameroon, Guinea, Madagascar, and In September 1997 the Russian Federation Tanzania) also included topping-up arrangements, became a full-fledged member of the Paris Club as whereby arrears and debt outstanding from earlier a participating creditor country. To make Russian Paris Club reschedulings (that is, previously claims comparable with the claims of other Paris rescheduled debt) on Toronto or London terms Club members, Russian claims inherited from the would be eligible for a further reduction in order former Soviet Union are reduced by an upfront to achieve a total net present value reduction of 67 discount that will be implemented when these percent. Creditors also agreed to consider the eli- claims are first treated by Russia within the Paris gibility of these countries for a stock reduction Club framework. The discount takes into account after a three-year period. Only one agreement was a debtor's financial and economic situation. Thus signed with a middle-income country (Uordan) in a higher discount is applied to the poorest coun- 1997. tries, given the need to ensure their financial sus- rainability. In addition, Russia will provide debt Note reduction or debt relief on the discounted value of these claims in accordance with the terms of the 1. The Paris Club also restructures the debt of applicable Paris Club agreement for that,counr middle-income countries but not on concessional applicable Paris Club agreement for that country. terms. Middle-income countries, however, have largely Russia's participation in the Paris Club paves the graduated from official rescheduling, and in 1997 there way for the resolution of a substantial amount of was only one nonconcessional rescheduling, for Jordan. APPENDIX 3 Commercial debt restructuring Since 1989 the restructuring of developing coun- cash payments and for principal and interest col- try debt to commercial banks has occurred largely lateral needed to guarantee the debt exchanges- through buybacks supported by the International have totaled $23.6 billion. Financing, net of the Development Association's (IDA) Debt Reduction $3.7 billion of concerted new money provided by Facility for low-income countries' and through commercial banks, came in almost equal shares officially supported debt and debt service reduction from debtor countries and official lenders. The programs (Brady operations) for middle-income World Bank's participation amounted to $4.7 bil- countries.2 These programs have helped resolve lion, or about 37 percent of foreign financing long-standing concerns of debtors and commercial requirements net of concerted commercial bank bank creditors and have improved these countries' lending. creditworthiness, in some cases contributing to the restoration of market access. Some middle-income Developments in 1997 countries have recently come full circle, entering the market to retire collateralized Brady bonds Nine debt reduction agreements between debtor through exchanges for uncollateralized instru- countries and their commercial bank creditors were ments and through debt buybacks. concluded in 1997, restructuring $19.1 billion in Officially supported programs and associated debt and reducing outstanding debt by $6.9 billion market swap operations reduced developing coun- (see table A3. 1). Among low-income countries, tries' debt to commercial banks by $53.2 billion Togo bought back $46.1 million at an average price between 1989 and December 1997 (table A3.1). of 12.5 cents per dollar in a deal supported by the This reduction, equivalent to 23 percent of the Debt Reduction Facility (table A3.2) and C6te $231.2 billion of eligible commercial bank debt d'Ivoire and Vietnam restructured $7.3 billion (induding interest arrears), was effected through under the Brady initiative. Bosnia and Herzegovina buybacks, cash payments, and writeoffs. Since conduded an agreement with commercial bank 1989, 33 countries have completed 41 debt and creditors to restructure $1.3 billion, including $0.7 debt service reduction operations under the aegis billion in interest arrears. Among middle-income of the Debt Reduction Facility, the Brady Plan, countries, Argentina, Brazil, Ecuador, Panama, and and, most recently, voluntary swap operations by Venezuela retired $10.4 billion of collateralized major Latin American countries. Eighteen low- Brady bonds through debt buybacks and discount- income countries have extinguished $12.6 billion ed swaps for unsecured bonds. The Russian of the $18.2 billion of eligible principal and inter- Federation concluded an agreement, initiated in est arrears due to commercial banks under the 1995, to restructure $33.0 billion of debt due to Debt Reduction Facility and, more recently, under commercial banks. debt and debt service reduction operations. Fifteen middle-income countries have eliminated nearly Debt and debt service reduction operations 20 percent of their $213.0 billion in commercial in low-income countries bank debt. Financing costs of officially supported opera- Cdte d'Ivoire. On 6 May 1997 C6te d'Ivoire tions-funds expended for buybacks and other completed a debt and debt service reduction agree- 83 84 Table A3.1 Debt and debt service reduction operations, 1989-97 (billions of US. dollars) Gross amount Face value of Face value of Type of op eration Country Closing date restructured' debt reduction restructured debt Low-income countries Buybacks funded by the Cumulative, 1996 4.2 4.1 0.1 Debt Reduction Facility Togo 12 Dec 97 0.1 0.1 0.0 Cumulative, 1997 4.3 4.2 0.1 Debt and debt service Cumulative, 1996 5.3 3.3 2.0 reduction agreements C6te d'lvoire 6 May 97 6.5 4.1 2.4 Vietnam 16 Dec 97 0.8 0.2 0.6 Cumulative, 1997 12.6 7.6 5.0 Other Bosnia and Herzegovina 30 Dec 97 1.3 0.9 0.4 Middle-income countries Buybacks funded by the Cumulative, 1996 0.3 0.3 0.0 Debt Reduction Facility Cumulative, 1997 0.3 0.3 0.0 Debt and debt service Cumulative, 1996 197.9 37.8 160.1 reduction agreements Cumulative, 1997 197.9 37.8 160.1 Brady bond exchanges Cumulative, 1996 4.4 0.8 3.6 Argentina 12 Sep 97 2.4 0.6 1.8 Brazil 4 Jun 97 2.7 0.4 2.2 Ecuador 18Apr 97 0.2 0.1 0.2 Panama 19 Sep 97 0.7 0.1 0.6 Venezuela 11 Sep 97 4.4 0.4 4.0 Cumulative, 1997 14.8 2.5 12.3 Total Cumulative, 1996 212.1 46.2 165.8 Operations in 1997 19.1 6.9 12.2 Cumulative, 1997 231.2 53.2 178.0 Note: Components may not add up to totals because of rounding. a. Includes past-due interest. b. Includes buybacks, discounts, down payment on past-due interest, and forgiveness. Source: World Bank Debtor Reporting System and Euromoney Bondware. ment to restructure $6.5 billion of debt owed to guaranteed. Neither principal nor interest securi- commercial banks (table A3.3). Of the $2,271.5 tization was required for the past-due interest million of eligible principal, C6te d'Ivoire bought bonds. back $681.5 million at 24 cents per dollar and The debt and debt service reduction agreement exchanged $159.0 million for 50 percent discount allowed C6te d'Ivoire to reduce its debt to com- bonds and $1,431.0 million for front-loaded mercial banks by $4.1 billion in nominal terms, or interest-reduction bonds. The $4,190.3 million in close to 63 percent of the total amount restruc- past-due interest was restructured as follows: tured. The operation, including a cash payment, $30.0 million was paid in cash at closing, $867.3 buyback, and principal and interest collaterals, million was exchanged for past-due interest bonds, cost $226 million, of which $19 million was pro- and $3,293.0 million was written off. vided by C6te d'Ivoire and $207 million was fund- The principal component of the discount ed by foreign loans and grants. Key lenders bonds was collateralized with 30-year U.S. included the IMF ($70 million through the Treasury zero-coupon bonds delivered at closing. Enhanced Structural Adjustment Facility), France Both the discount and the front-loaded interest ($52 million concessional loan), and IDA ($50 reduction bonds required a six-month rolling million credit). In addition, the Institutional interest guarantee, calculated at a fixed rate of Development Fund provided a $35 million grant, about 2.5 percent and secured by cash or permit- of which $20 million came from the IBRD, $10 ted investments. The principal component of the million from Switzerland, and $5 million from the front-loaded interest reduction bonds was not Netherlands. 85 Table A3.2 Completed operations financed by the IDA Debt Reduction Facility as of December 1997 (millions of U S. dollars) Percentage Price of eligiblI Principal (cents per principal Total IBRD Country Date completed extinguished dollar) extinguished resourcesb resourcesb Niger March 1991 107.0 18.0 99.0 19.4 8.4 Mozambique December 1991 123.8 10.0 64.0 13.4 5.9 Guyana November 1992 69.2 14.0 100.0 10.2 10.0 Uganda February 1993 153.0 12.0 89.0 22.6 10.2 Bolivia May 1993 170.0 16.0 94.0 27.3 9.8 SaoTome and Principe August 1994 10.1 10.0 87.0 1.3 1.3 Zambia September 1994 199.7 11.0 78.0 24.9 11.7 Albania July 1995 371.3 26.0 99.0 97.4 26.0 Sierra Leone September 1995 234.7 13.0 73.0 31.3 21.0 Nicaragua December 1995 1,100.0 8.0 81.0 88.8 40.0 Ethiopia January 1996 226.0 8.0 80.0 18.83 6.2 Mauritania August 1996 53.0 10.0 98.0 5.90 3.3 Senegal December 1996 71.0 20.0' 95.0 15.4 6.6 C6te d'Ivoired May 1997 681.5 24.0' 30.0 185.4 20.0 Vietnam' June 1997 1.0 1.0 Togo December 1997 46.1 12.5 90.0 6.4 5.4 Total 3615.7 14.7' 83.9' 570.1 187.6 a. Of original face value of principal. b. Includes technical assistance grants. c. At 16 cents for cash buyback and 20 cents for long-term bonds. d. Only the buyback portion of the debt and debt service reduction operation was funded by the Debt Reduction Facility. e. Except for a $1.0 million advance for consultant services filnded by the Facility, the operation vwas financed by an IDA c.edit and by the government. f. Weighted average. Source: World Bank staff estimates. Vietnam. On 16 December 1997 Vietnam These operations reduced Vietnam's debt to signed a debt and debt service reduction agree- commercial banks by $237.6 million in nominal ment to restructure $797.1 million of debt owed terms. Taking into account interest service savings to commercial banks, including $486.2 million of and cash payments resulting from bond collateral, past-due interest (table A3.4). Eligible principal of the debt reduction value of the debt and debt ser- $310.9 million was rescheduled according to a vice reduction (excluding additional official lend- menu of choices that included $51.6 million of ing) was equivalent to 30 percent of the nominal discount bonds (at a 50 percent discount) and amount restructured. The upfront costs of the $238.9 million of par bonds. In addition, $20.4 operation (including cash payments and bond col- million was bought back at 44 cents per dollar. lateral) totaled $54 million, of which $19 million Past-due interest of $486.2 million was discharged was funded by Vietnam and $35 million by an as follows: $15.0 million was paid at closing, IDA credit. $294.8 million was exchanged for past-due inter- est bonds, $21.8 million was tendered for a buy- Togo. On 12 December 1997 Togo concluded back, and $154.6 million was forgiven following an agreement (sponsored by the Debt Reduction the recalculation of interest at a lower spread and Facility) to restructure $75.0 million due to com- the waiving of penalty interest. mercial banks, including thewriteoffof $28.9 mil- The principal component of the discount and lion of past-due interest and a buyback of $46.1 par bonds was collateralized with 30-year U.S. million at 12.5 cents per dollar. Almost 70 percent Treasury zero-coupon bonds delivered at closing. of the eligible principal stemmed from 1980 and But while 100 percent of the discount bonds is 1983 rescheduling agreements, some 25 percent guaranteed, only 50 percent of the par bonds is originated from debt owed by public enterprises, collateralized. Payment of six months of interest is and the remaining 5 percent resulted from promis- guaranteed on a rolling basis by cash or permitted sory notes issued by the government to foreign investments only on the discount bonds. The past- contractors. The operation cost $6.4 million, of due interest bonds do not carry a principal or which $5.4 million came from the IBRD and $1.0 interest guarantee. million came from France. 86 Table A3.3 Terms of Cote d'lvoire's debt and debt service reduction agreement, May 6, 1997 Debt retired (millions Share Maturity/grace Interest rate Instrument of U.S. dollars) (percent) (years) (percent) Comments Principal subject to the menu approach 2,271.5 100.0 Discount bonds (50 percent discount) 159.0 7.0 30/bullet Year 1-2 Principal collateralized with 30- 3-6 b year U.S. Treasury zero-coupon 7-8 bonds or French treasury 9-10 zero-coupon bonds. No interest 11-30 ' collateral. Six-month rolling interest guarantee secured by cash or permitted investments. Front-loaded interest reduction bonds 1,431.0 63.0 20/10 Year 1-7 2.0 No principal collateral. Graduated 8-9 2.5 ansoritization schedule rising from 10 3.0 2.0 percent in year 11 to 8 percent 11-12 4.0 in year 20. Six-month rolling 12-13 5.0S interest guarantee through year 13 14-20 secured by permitted investments. Buyback (24 cents per dollar) 681.5 30.0 Past-due interest 4,190.3 Down payment at closing 30.0 Past-due interest par bonds 867.3 20/0.5 Year 1-10 h No principal or interest collateral. 11-15 Graduated amortization schedule. 16-20 c rising from 2.5 percent during the first five years to 5.0 percent in year 20. Forgiveness 3,293.0 Total amount restructured 6,461.8 a. 2.5 on dollar debt; 3.0 on French franc deht. b. 3.0 on dollar debt; 3.5 on French franc debt. c. 3.5 on dollar debt; 4.0 on French franc debt. d. 4.0 on dollar debt; 4.5 on French franc debt. e. U.S. LIBOR + '/,I on dollar debt; French franc LIBOR + on French franc debt. f. Through the first half of year 12. gs From the second half of year 12. h. 2.0 on dollar debt; 1.9 on French franc debt. i. 3.0 on dollar debt; 2.9 on French franc debt. Source. Republic of C6te d'lvoire, 1996 Financing Plan." Bosnia and Herzegovina. On 30 December tization schedule that rises from 1 percent in years 1997 Bosnia and Herzegovina finalized an agree- 1-2 to 7 percent in years 11-20. ment to restructure $1.3 billion, including $0.7 Bosnia and Herzegovina is not required to billion in past-due interest owed to commercial make debt service payments on the remaining bank creditors under the aegis of the London tranche of $250 million for at least ten years. In Club. Past-due interest, including penalty interest, addition, no debt servicing will be required there- was forgiven. Eligible principal of $600 million after until per capita income exceeds $2,800 for was exchanged for $400 million of uncollateral- two consecutive years.3 The agreement initially ized discount bonds. An innovative feature of the reduces nominal debt by 69.2 percent and may agreement links scheduled repayments to the reach 88.5 percent throughout the life of the country's economic performance. bonds if the per capita target is not exceeded. Servicing of interest due on $150 million of the new bonds begins in mid-1998, and the repay- Swaps in middle-income countries ment of principal is subject to a stepped-up amor- tization schedule. This tranche of the bonds has a In 1996 Mexico and the Philippines swapped $4.4 grace period of seven years, concessional fixed billion of Brady bonds for uncollateralized long- interest rates of 2.0-3.5 percent for the first 10 term bonds. This important development contin- years and LIBOR + 13/16 thereafter, and an amor- ued in 1997, when such swaps more than doubled 87 Table A3.4 Terms of Vietnam's debt and debt service reduction agreement, December 16, 1997 Debt retired (millions Share Maturity/grace Interest rate Instrument of US. dollars) (percent) (years) (percent) Comments Principal subject to the menu approach 310.9 100.0 Discount bonds (50 percent discount) 51.6 16.6 30/bullet Principal collateralized with 30- year U.S. Treasury zero-coupon bonds. Six-month rolling interest guarantee secured by cash or permitted investments. Par bonds 238.9 76.8 30/15 Year 1-15 3.00 Last principal payment equal to 16-25 4.00 50 percent of principal collateralized 26-30 5.50 by U.S. Treasury zero-coupon bond. Buyback (44 cents per dotlar) 20.4 6.6 Past-due interest 486.2 Down payment at closing 15.0 Past-due interest bonds 294.8 20/10 Year 1 3.00 No principal or interest collateral. 2 3.33 Graduated amortization schedule. 3 4.67 4-7 4.00 8-18 Buyback (44 cents per dollar) 21.8 Forgiveness 154.6 Total amount restructured 797.1 a. LIBOR t 13/t 6. Source: Republic of Vietnam, "1997 Financing Plan." in value. These voluntary deals show the renewed bonds plus the freed collateral and the replacement confidence of foreign investors in these countries' bond were estimated at $242 million. prospects-particularly significant in Mexico given the uncertainties of recent years. Such swaps offer BraziL On 4 June 1997 Brazil completed a $3.0 two benefits to debtor countries. First, the collater- billion, 30-year bond offering involving a $0.8 bil- al associated with Brady bonds, including interest lion cash sale and a $2.3 billion exchange for $2.7 earned on escrow accounts, is released on a pro rata billion of Brady bonds. The new issue carries an basis and can be used to meet other obligations. For interest rate of 395 basis point above the U.S. example, in 1997 Ecuador used the freed collater- Treasury rate. Nominal savings of $1.0 billion al to clear debt service arrears with Paris Club cred- resulted from the differential between Brady itors. Second, because the swap is effected at a bonds' par and market values ($0.4 billion) and discount based on secondary market prices, debt from the pro rata release of the collateral of the outstanding is commensurately reduced. The Brady bonds ($0.6 billion). Net present value sav- advantage to the original bondholder lies in higher ings were estimated at $186 million. yields on the uncollateralized bonds. Ecuador. On 18 April 1997 Ecuador issued a Argentina. On 12 September 1997 Argentina $150 million eurobond to buy $214 million of retired $2.4 billion of Brady bonds for $1.8 billion Brady bonds. The principal amount is due at of uncollateralized 30-year bonds at an interest rate maturity in April 2004 and carries an interest rate of 305 basis points above the U.S. Treasury rate. The of 475 basis points above the U.S. Treasury rate. offering allowed for direct exchange and cash sales of Nominal savings of $114 million resulted from the Brady bonds. Nominal savings of$ 1.1 billion result- differential between Brady bonds' par and market ed from the differential between the par and market values ($164 million) and from the pro rata release values of these securities ($0.6 billion) and from the of the collateral of the Brady bonds ($50 million). pro rata release of the collateral of the Brady bonds The $50 million saved from the collateral was ($0.5 billion). Net present value savings stemming applied toward clearance of debt service arrears from the cash-flow differential between the Brady with Paris Club creditors. 88 Panama. On 19 September 1997 Panama com- ing $7.5 billion of interest arrears to commercial pleted a $600 million offering of 30-year uncol- banks. Eligible principal will be repaid over 25 lateralized bonds for $713 million of Brady bonds years, including a 7-year grace period. A graduat- that had been issued on 3 April 1996. The new ed amortization plan will result in annual pay- offering carries an interest rate of 250 basis points ments of about 2 percent a year for the first two to above the U.S. Treasury rate. Nominal savings of three years following expiration of the grace peri- about $1,320 million resulted from the differen- od, peaking at 15 percent in years 9 and 10 and tial between Brady bonds' par and market values declining thereafter. Interest due will be calculat- ($112 million) and from the pro rata release of the ed at LIBOR + 13/16 , with actual payments rising collateral of the Brady bonds ($20 million). from about 25 percent of the amount due in 1996 to full payment beginning in 2002. The shortfall Venezuela. On 11 September 1997 Venezuela in interest payments will be covered by issuance of retired $4.4 billion of Brady bonds in exchange for interest notes with a 14-year payment profile. $4.0 billion of uncollateralized 30-year bonds at About $2 billion of the past-due interest arrears an interest rate of 325 basis points above the U.S. will be paid up front into escrow accounts. Treasury rate. The operation-the largest to date Implementation of the agreement occurred in by a single country-resulted in nominal savings phases. In September 1996 the government of about $1.8 billion, of which $0.4 billion announced that the London Club had agreed to stemmed from the differential between Brady complete the rescheduling of $20 billion. The bonds' par and market values and $1.4 billion remaining $13 billion were rescheduled in the accrued from the pro rata release of the collateral fourth quarter of 1997. of the Brady bonds. Net present value savings stemming from the cash-flow differential between Debt conversion programs the Brady bonds plus the freed collateral and the replacement bond were estimated at $0.4 billion. The number of countries participating in debt conversions and the face value of debt restructured Future Brady swaps. Whether the market for increased rapidly after May 1985, when Chile these transactions continues to grow at the pace established the first institutionalized debt-for- observed in 1997 will depend, in the near term, equity swap program. Since 1985 debt-equity con- more on demand than on supply factors. About $67 versions have totaled $38.6 billion (table A3.5). billion in collateralized par and discount bonds are Debt conversion activities declined during potentially eligible for exchange. Latin American 1992-96, however. Debt-for-equity swaps totaled countries account for about 84 percent of these just $100 million in 1996, and debt-for-develop- issues. Expansion of the market will depend, among ment swaps totaled less than $ 100 million. Several other things, on investors' continued confidence, factors contributed to the drop in debt conver- secondary market prices, spreads on the new instru- sions. Investor interest in debt conversion pro- ments, and accounting rules. Accounting rules lim- grams declined as rising secondary market prices ited demand for the Mexican issue by domestic of several countries' commercial bank debt banks, which in accordance with accounting guide- reduced the discount that could be captured. A sig- lines list Brady bonds as assets at face value rather nificant amount of debt conversion activity was than as tradable instruments. As a result the books linked to specific privatization programs that have of any Mexican banks participating in the swap turned to other instruments or have been winding showed a loss equal to the difference between the down. And Brady operations, which have enabled face value and market value of the transaction. debtor countries to regularize relations with com- mercial bank creditors, have permitted more flex- Other restructuring in middle-income countries ibility in debt management and reduced some governments' interest in conversion programs. Russian Federation. On 6 October 1997 the After a general decline during 1992-96, debt- Russian Federation concluded an agreement that for-development swaps rebounded in 1997, reach- began in 1995 to reschedule $33.0 billion, includ- ing $108 million. Led by Mexico and Nigeria, this 89 increase is expected to be sustained in the next few Two small debt-equity operations in FYR years. Low-income countries are expected to Macedonia accounted for $ 0.1 million. contribute to the expansion, which not only con- tributes to development but also suggests evidence Debt for development of proactive debt management by countries in complementing action plans developed under the In debt-for-development swaps an international Heavily Indebted Poor Countries Debt Initiative. organization (usually a nongovernmental organi- Debt conversions are permitted under both the zation, or NGO) purchases sovereign debt in the Paris Club minutes and within the Enterprise for secondary market at a deep discount and then the Americas Initiative. exchanges the debt at a redemption price negoti- ated with the country. The funds are then used for Debtfor equity a development project approved by the country and managed by the NGO. Debt-for-equity swaps generally involve the pur- chase of debt by the investor at a discount in the Debtfor nature. Debt-for-nature operations are secondary market and the sale of the debt to the used to reduce developing countries' debt and allo- central bank for funds that are used to acquire cate funds to the protection of nature preserves. public assets or invest in private equity. Debt-for- Since 1987, when Conservation International and equity conversions can be a useful tool for acceler- Bolivia signed the first debt-for-nature agreement, ating a country's privatization program, as has 15 countries have retired $152.7 million in face been done in Argentina, Mexico, and the value of debt though such programs, at an average Philippines. discount of 69.8 percent (table A3.6). Since 1994 Debt-for-equity swaps were negligible in 1995, Mexico has converted $3.7 million in face value of however. The only two sizable operations were in debt through nine debt-for-nature operations. Latin America, both in Peru. The privatization of The magnitude of debt-for-nature swaps has EDEGEL (the electric utility company of Lima) been declining over time, however. After reaching generated $524 million, of which $100 million $43.9 million in 1989, debt-for-nature swaps was debt-for-equity. The privatization of Banco declined to $576,000 in 1997-two operations Continental delivered $255 million to the effected by Mexico. Peruvian Treasury, of which $60 million was con- verted debt. Other debt-for-developmentswaps. Three orga- Debt-for-equity swaps were also used on a nizations-Finance for Development, New York small scale in Mexico as part of the debt relief and Bay, and the United Nations Children Fund recapitalization measures adopted by the govern- (UNICEF)-have been the main participants in ment in 1995-96 to contain the banking crisis. debt-for-development swaps that provide local Through these swaps, banks acquired major shares currency funds for projects other than nature of several companies, including Mexicana and preserves. Finance for Development and New Aeromexico (airlines) and Grupo Gigante (retail). York Bay have swapped $566.7 million since Table A3.5 Types of debt conversion instruments, 1985-96 (millions of U.S. dollars) 1985-96 Share Instrutment 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Total (percent) Debt-equiry swaps 550 882 3,577 7,457 5,114 8,901 2,171 4,806 4,662 212 200 100 38,632 30 Debt buybacks or exchanges 0 0 0 1,415 945 12,170 960 13,643 6,708 19,953 649 800 57,243 44 Localcurrencyconversions 156 438 796 1,503 1,384 1,521 1,459 1,224 134 0 8 0 8,623 7 Local currency payments 0 63 87 3,580 2,269 5,242 800 342 0 0 0 0 12,383 10 Private sector restructuring 89 279 3,454 4,341 3,113 337 788 371 293 0 0 0 13,065 10 Total 795 1,662 7,914 18,296 12,825 28,171 6,178 20,386 11,797 20,165 857 900 129,946 100 Source: Institute for International Finance and World Bank Debtor Reporting System. 90 1992, of which $107.8 million was swapped in Role of the Debt Reduction Facility in debt con- 1997 (table A3.7). The funds have been invest- versions. Through the IDA Debt Reduction ed in various sectors, including health, popula- Facility, the World Bank has expanded the menu tion, agriculture, ecotourism, and low-income of debt reduction options to include provision for housing. debt-for-development swaps. Under this provision By 1995 UNICEF, a pioneer in debt-for-devel- commercial banks can choose to donate or tender opment swaps, had completed 21 transactions, debt to be repurchased by NGOs (at the same generating $52.9 million in local currency while price as the debt buyback option). NGOs can then helping participating countries reduce their exter- convert the debt into local currency to finance nal debt stock by $199.3 million (table A3.8). development projects. Two countries, Bolivia (Although UNICEF is planning additional debt (1993) and Zambia (1994), have implemented conversion operations in developing countries, no such options. No debt-for-development swap has operations have been carried out since 1995.) A been concluded through the Debt Reduction wide range of entities has been involved in these Facility since 1994. transactions, including an IDA Debt Reduction Facility operation for Zambia. The funds used Notes come from national committees for UNICEF in industrial countries. These funds help finance pro- 1. For derails, see page 27, WorldDebt Tables 1990-91. gramsfor rimay edcatio, woen i devlop- 2. For details, see box A2.2, World Debt Tables grams for primary education, women indlop 199495 ment, children in especially difficult circumstances, 3. In 1997 per capita income was estimated at primary health, and water supply and sanitation. $1,079. 91 Table A3.6 Debt-for-nature swaps, 1987-97 (thousands of U.S. dollars) Face value Cost to Conservation Year Country Buyer of debt' donor' funds' 1997 Mexico Ci 310 238 299 1997 Mexico CI 266 186 243 1996 Mexico Ci 671 440 561 1996 Mexico CI 496 327 443 1996 Mexico Cl 391 192 254 1995 Mexico CI 488 246 337 1994 Mexico CI 290 248 290 1994 Mexico CI 480 399 480 1994 Mexico CI 280 236 280 1994 Madagascar" CI/WWF 200 50 160 1993 Madagascar CI/WWF 3,200 1,600 3,200 1993 Mexico CI 252 208 252 1993 Bolivia CI n.a. n.a. 397 1992 Ecuador Japan n.a. n.a. 1,000 1992 Brazil TNC 2,200 746 2,200 1992 Boliviad TNCIVW F/JP Morgan 11,500 n.a. 2,800 1992 Guatemala CI/USAID 1,300 1,200 1,300 1992 Panama TNC n.a. n.a. 30,000 1992 Ecuador WWF 1,000 n.a. n.a. 1992 Philippines' NWWF 9,900 5,000 8,800 1992 Mexico CI/USAID 441 355 441 1991 Ghanaf DDC/CIlSI 1,000 250 1,000 1991 Jamaica TNC/USAID/CTPR 437 300 437 1991 Guatemalag TNC 100 75 90 1991 Mexico" CI 250 n.a. 250 1991 Nigeria NCF 149 65 93 1991 Philippines USAID/WWF n.a. n.a. 8,000 1991 Mexicoh,i CI 250 183 250 1991 Costa Ricaj g RA 600 360 540 1991 Madagascar" CI/UNDP 119 59 119 1990 Madagascar WWF 919 446 919 1990 Philippines NWWF 900 439 900 1990 Costa Rica \W9WF/TNC/Sweden 10,800 1,900 9,600 1990 Dominican Rep. TNC/CTPR 582 116 582 1990 Poland WWF n.a. n.a. 50 1989 Zambia VWWF 2,300 454 2,300 1989 Madagascar WWF 2,100 950 2,100 1989 Ecuador /WWF/TNC/MBG 9,000 1,100 9,000 1989 Costa Rica Sweden 24,500 3,500 17,100 1989 Costa Rica TNC 5,00 784 1,700 1989 Philippines' WWF 390 200 390 1988 Costa Ricam Netherlands 33,000 5,000 9,900 1988 Costa Rica CI/WWF 5,400 918 5,400 1987 Ecuador NWWF 1,000 354 1,000 1987 Bolivia CI 650 100 250 n.a. Not applicable. Cl = Conservation International, CTPR = Conservation Trust of Puerto Rico, DDC = Debt for Development Coalition, MBG = Missouri Botanical Gardens, NCF = Nigerian Conservation Foundation, RA= Rainforest Alliance, SI = Smithsonian Institute, TNC - The Nature Conservancy, USAID = U.S. Agency for International Development, WWF = World Wildlife Fund. a. Expenditures by environmental agency to acquire sovereign debt. b. Value in dollar equivalent to the local currency part of the swap (either face value of the environmental bond or local currency equivalent). Overhead fees charged by government are not deducted. c. Conservation finds as a percentage of face value. d. Debt donated by JP Morgan. e. Face value of debt includes $200,000 debt donation by Bank of Tokyo. f. Involves buying blocked local currency funds from multinational organizations; includes Midwest universities, Consortium for International Activities, and U.S. Committee of the International Council on Monuments and Sites. g. Purchase of Central American Bank for Economic Integration debt. h. Total amount of program is $4 million. i. Debt donated by Bank of America. j. W`WF contributed $1.5 million on top of the swap. k. Total amount of program is $5 million. 1. Total amount of agreement is $3 million. m. Includes $250,000 donated by Fleet National Bank of Rhode Island. Saurce. World Wdidlife Fund, Nature Conservancy, Conservation International, and World Bank. 92 Table A3.7 Face value of debt conversions originated by Finance for Development and New York Bay, 1992-97 (millions of US. dollars) Country 1991 1992 1993 1994 1995 1996 1997 Total Mexico 7.8 4.3 15.5 92.2 37.6a 49.2 56.0 262.5 Nigeria 0.0 0.0 1.7 1.6 6.9 33.3 51.8 95.3 Zambia 0.0 0.0 0.2 87.0 0.0 0.0 0.0 87.2 Philippines 9.0 7.0 9.1 13.0 3.0 0.0 0.0 41.1 Bolivia 0.0 0.0 0.0 0.0 32.8 0.0 0.0 32.8 South Africa 0.0 0.0 7.4 18.9 0.0 0.0 0.0 26.3 Tanzania 0.0 0.0 4.5 0.0 9.6 0.0 0.0 14.1 Ecuador 5.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0 Kenya 0.0 1.5 0.0 0.0 0.0 0.0 0.0 1.5 Ghana 0.0 1.0 0.0 0.0 0.0 0.0 0.0 1.0 Total 21.8 13.8 38.4 212.7 89.9 82.5 107.8 566.7 a. Includes special conversion transactions. Source: New York Bay and Finance for Development. Table A3.8 UNICEF debt-for-child-development swaps, 1989-95 (thousands of U.S. dollars) Face Development Purchase price Country Year Sector valuie Cost findsa (percent) Mexico 1995 UNICEF programs 6,400 3,647 4,935 57.0 Madagascar 1994 Emergency 2,000 1,000 2,000 50.0 Madagascar 1994 Water/education/health 1,200 576 950 48.0 Perub 1994 UNICEF programs 10,880 0 2,720 0.0 Zambiac 1994 UNICEF programs 66,614 7,328 10,990 11.0 Mexico 1994 Health/education 1,870 1,015 1,902 54.3 Philippinesd 1993 Education 250 0 180 0.0 Philippines 1993 Education 1,226 864 1,000 70.5 Bolivia 1993 Education/institutions 15,000 2,400 3,600 47.0 Madagascar 1993 Water/education/health 2,000 940 2,000 16.0 Senegal 1993 Water/education/health 24,000 6,000 11,000 25.0 Jamaica 1992 Health/street children' 4,000 2,877 4,000 71.9 Madagascar 1992 Health/education/nutritionr 4,000 2,000 4,000 50.0 Philippines 1992 Children in armed conflict 486 245 329 50.4 Sudan 1992 Water/sanitation/healthg 38,068 0 1,200 0.0 Sudan 1991 Water/sanitation/healthg 5,000 0 460 0.0 Sudan 1991 Water/sanitation/healthg 3,000 0 276 0.0 Sudan 1990 Water/sanitation/healthg 7,023 0 801 0.0 Sudan 1989 Water/sanitation/healthg 2,732 0 244 0.0 Sudan 1989 Water/sanitation/healthg 2,732 0 225 0.0 Sudan 1989 Water/sanitation/healthg 800 0 80 0.0 Total or average 199,281 28,892 52,892 14.5 a. Value in dollar equivalent to the local currency part of the swap (either face value of the bond or local currency equivalent). For bonds, figures do not include interest earned over the life of the bond. b. Debt donated. c. IDA Debt Reduction Facility operation. d. Brady bonds donated. e. Primary health care, women in development, and children in especially difficult circumstances. f. Healtn, nutrition, education, social mobilization, and area-based UNICEF programs. g. UNICEF water, sanitation, and health education programs in rural areas. Source. UNICEF and World Bank. 93 Table A3.9 Multilateral debt relief agreements with commercial banks, January 1980-December 1997 Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount rescructured (US$ millions) New Short-term Maturity Grace Country and Length ---- long-term credit (years! (years! Interest date ofagreement Start date (months) Deferment' Reschedulingb money' maintenance" months) months) (marginj' Albania July 1995 Debt buyback (see notes) Algeria February 1992 See notes 1,500 5-8/0 3/0 1 1/2/1 31/ June 1995 March 1994 3,200 12/6-16 6/6 13/16 Argentina January 1983 1 January 1983 12 1,300 1/2 0/7 1 August 1983 500 4/6 3/0 2 /j4 August 1985 1 January 1982 48 9,777 3,593 3,100 10/0 3/0 I 318 August 1987 See notes 24,286 1,253 3,500 19/0 7/0 13/l6 April 1993 DDSR agreement (see notes) Bolivia December 1980 1 August 1980 8 200 1/0 1/0 1 314 April 1981 1 April 1981 24 411 6/0 3/0 2 1/4 July 1988 Buyback arrangement (see notes) July 1992 DDSR agreement (see notes) May 1993 Debt buyback (see notes) Bosnia and Herzegovina December 1997 Debt rescheduling (see notes) Brazil February 1983 1 January 1983 12 4,800 4,195 15,675 8/0 2/6 2 '/4 January 1984 I January 1984 12 5,900 6,510 15,100 9/0 5/0 2 July 1986 1 January 1985 12 9,600 6,552 14,750 6/3 4/3 1 /4 November 1988 1 January 1987 84 61,482 5,200 14,833 20/0 8/0 3/16 July 1992 Interest arrears end-1990 (see notes) April 1994 DDSR agreement (see notes) Bulgaria July1994 DDSR agreement (see notes) Chile July 1983 1 January 1983 24 2,151 1,294 1,700 8/0 4/0 2 / January 1984 Short-term debt only 1,204 8/0 4/0 2 '/4 June 1984 785 9/0 5/0 1 3/4 November 1984 1,700 0/6 0/6 November 1985 1 January 1985 36 3,891 1,037 1,700 12/0 6/0 1 3/8 June 1987 1 January 1988 48 9,732 1,700 15/6 5/0 1 August 1988 Modification of terms (see notes) 13/I6 December 1990 1 January 1991 48 4,173 320 8-12/0 4/0 13/,6 Colombia December 1985 1,000 8/6 3/0 1 1/2 June 1989 1,640 11/0 5/6 April 1991 12/6 7/6 1 Congo, Rep. October 1986* See notes (table continues on nex- page) 94 Table A3.9 Multilateral debt relief agreements with commercial banks, January 1980-December 1997 (confinued) Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount restructured (US$ millions) New Short-term Maturity Grace Country and Length long-term credit (years! (years/ Interest date of agreement Start date (months) Deferment' Rescheduling' money' maintenance4 months) months) (margin)' Costa Rica September 1983 1 January 1983 24 706 202 202 8/0 4/0 2 /4 May 1985 1 January 1985 24 470 75 10/0 3/0 1 5/8 May 1990 DDSR agreement (see notes) 1,457 Cote d'lvoire March 1985 1 December 1983 25 485 104 8/0 3/0 1 7/a November 1986 1 January 1986 48 851 9/0 3/0 I 5/8 April 1988* See notes May 1997 DDSR agreement (see notes) Cuba December 1983 1 September 1982 28 130 490 5/6 2/0 2 1/14 December 1984 1 January 1984 12 103 490 7/0 2/6 1 7/8 July 1985 1 January 1985 12 90 490 10/0 6/0 1 '/2 Dominican Republic December 1983 1 December 1982 13 500 5/0 1/0 2 i/4 February 1986 1 January 1985 60 787 13/0 3/0 1 318 August 1994 DDSR agreement (see notes) Ecuador October 1983 1 November, 1982 14 2,770 433 700 7/0 1/0 2 /4 December 1985 1 January 1985 60 4,219 200 700 L2/0 3/0 1 318 November 1987* See notes February 1995* DDSR agreement (see notes) Ethiopia January 1996 Debt buyback (see notes) Gabon December 1987 1 September 1986 16 27 10/0 4/6 13/, December 1991 1 January 1989 36 75 13/0 3/0 7/, May 1994 10 July 1994 6 187 10/0 2/6 7/8 Gambia, The February 1988 Balance as of 18 December 1986 19 8/0 3/6 1 /4 Guinea April 1988 Short-term debt only 28 3/0 0/6 13/4 Guyana August 1982 11 March 1982 13 14 2'12 June 1983 1 July 1983 7 12 2'12 July 1984 1 August 1984 12 11 2 /2 July 1985 1 August 1985 18 15 2 /2 July 1988 8 November 1992 Debt buyback (see notes) Honduras June 1987* 1 April 1987 33 248 8/0 6/0 1 /8 August 1989 See notes 101 Iran, Islamic Rep. March 1993 Balance as of March 1993 2,800 1/1 1/0 131/6 December 1994 Balance as of December 1994 10,900 6/0 2/0 13/16 95 Table A3.9 Multilateral debt relief agreements with commercial banks, January 1 980-December 1997 (continued) Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount restructured (US$ millions) New Short-term Maturity Grace Country and Length long-term credit (years! (years! Interest date of agreement Start date (months) Deferment" Rescheduling6 moneyc maintenanced months) months) (margin)' Jamaica April 1981 1 April 1979 24 126 5/0 2/0 2 June 1981 1 July 1981 21 89 89 5/0 2/0 2 June 1984 1 July 1983 21 164 5/0 2/0 2 /2 September 1985 1 April 1985 24 359 10/0 3/0 17/8 May 1987 1 January 1987 39 366 12/6 9/0 1 1/4 June 1990 1 January 1990 24 315 14/0 0/6 13/,6 Jordan September 1989* 1 January 1989 30 580 11/0 5/0 13/l6 November 1989* 1 January 1989 18 50 50 10/6 4/0 13/16 December 1993 See notes Liberia December 1982 1 July 1981 24 29 6/0 2/9 1 3/4 June 1983 Sce notes 26 Madagascar November 1981 Arrears only 155 3/6 0/0 I I/2 October 1984 See notes 379 8/0 2/6 2 June 1987 See notes 9/0 0/0 1 5/8 May 1990* 1 April 1990 69 49 12/0 0/2 7/8 Malawi March 1983 1 September 1982 24 59 6/6 3/0 17/S October 1988 Balance as of 21 August 1987 36 8/0 4/0 1 /4 Mauritania August 1996 Debt buyback (see notes) Mexico August 1983 23 April 1982 28 23,280 5,007 8/0 4/0 178 April 1984 3,873 10/0 5/6 1 /2 March 1985 1 January 1987 48 28,000 14/0 0/0 1 1/4 August 1985 1 January 1985 72 20,256 14/0 1/0 I 1/4 October 1985 950 March 1987 44,216 7,439 20/0 7/0 13/,6 August 1987 1 January 1988 48 9,700 20/0 7/0 13/16 March 1988 Debt exchange (see notes) March 1990 DDSR agreement (sce notes) 48,231 1,091 May/September 1996 Voluntary debt swap (see notes) Morocco February 1986 9 September 1983 16 531 610 7/0 3/0 1 /4 September 1987 1 January 1985 48 2,415 11/0 4/0 1 316 September 1990 Balance as of 31 December 1989 3,200 18/4 8/10 13116 Mozambique May 1987 Entire stock of debt 253 15/0 8/0 1 1/4 December 1991 Debt buyback (see notes) (table continues on next page) 96 Table A3.9 Multilateral debt relief agreements with commercial banks, January 1980-December 1997 (continued) Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount restructured (US$ millions) NVewv Short-term Maturity Grace Country and Length long-term credit (years! (years/ Interest date of agreement Start date (months) Deferment' Reschedulingb money' maintenance d months) months) (margin) Nicaragua December 1980 Arrears 582 12/0 5/0 3/4 December 1981 See notes 192 12/0 5/0 3/4 March 1982 See notes 100 12/0 5/0 3/4 February 1984 1 July 1983 12 145 8/0 0/0 1I/4 December 1995 Debt buyback (see notes) Niger March 1984 1 October 1983 29 29 7/6 3/6 2 April 1986 1 October 1985 39 36 8/6 4/0 2 March 1991 Debt buyback (see notes) 107 Nigeria November 1987 1 April 1986 21 4,714 9/0 3/0 11 /4 March 1989 Short-term debt only 5,671 20/0 3/0 7/, January 1992 DDSR agreement (see notes) 5,436 Panama September 1983 278 217 6/0 3/0 2 l/4 October 1985 1 January 1985 24 578 60 190 12/0 3/6 1 8 April 1996 DDSR agreement (see notes) Peru January 1980 1 January 1980 12 364 5/0 2/0 1 I/4 July 1983 7 March 1983 12 432 650 2,000 8/0 3/0 2 1/4 November 1996 DDSR agreement (see notes) Philippines January 1986 17 October 1983 38 5,885 925 2,974 10/0 5/0 1 5/8 December 1987 1 January 1987 72 9,010 2,965 17/0 7/6 71, January 1990 DDSR agreement (see notes) 1,337 715 December 1992 DDSR agreement (see notes) 135 17/0 5/0 3/16 September 1996 Voluntary debt swap (see notes) Poland April 1982 26 March 1981 9 1,956 7/0 4/0 1 3/4 November 1982 1 January 1982 12 2,225 7/6 4/0 1 3/4 November 1983 1 January 1983 12 1,254 10/0 4/6 1 7/8 July 1984 1 January 1984 48 1,480 335 10/0 5/0 1 3/4 September 1986 1 January 1986 24 1,940 5/0 5/0 1 3/4 July 1988 1 January 1988 72 8,310 1,000 1]5/0 0/0 15/16 June 1989* 1 May 1989 20 206 October 1994 DDSR agreement (see notes) 206 138 Romania December 1982 1 January 1982 12 1,598 6/5 3/0 1 3/4 June 1983 1 January 1983 12 567 6/5 3/6 1 3/4 September 1986 1 January 1986 24 800 5/6 4/0 1 3/8 September 1987* 1 January 1986 24 800 5/6 4/0 1316 Russian Federation December 1991 See notes July 1993 See notes November 1995 Balance as ofl 5 November 1995 32,500 25/0 7/0 13/,6 97 Table A3.9 Multilateral debt relief agreements with commerdal banks, January 1980-December 1997 (continued) Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount restructured (US$ millions) New Short-term Maturity Grace Country and Length long-term credit (yearsl (years! Interest date of agreement Start date (months) Deferment' Rescheduling' money maintenancel months) months) (margin) Sao Tome and Principe August 1994 Debt buyback (see notes) Senegal February 1984 1 May 1981 38 96 6/0 3/0 2 May 1985 1 July 1984 24 20 7/0 3/0 2 January 1989 37 9/0 0/0 7/8 December 1996 Debt buyback (see notes) Sierra Leone January 1984 Principal arrears 25 7/0 2/0 114 August 1995 Debt buyback (see notes) South Africa September 1985 28 August 1985 7 13,628 March 1986 28 August 1985 22 650 1/3 Bullet/variable March 1987 1 July 1987 36 4,500 3/0 Bullet/variable October 1989 1 July 1990 42 7,500 September 1993 See notes 5,000 8/0 0/6 I l/8 Sudan November 1981 1 January 1980 28 593 7/0 3/0 1 3/, March 1982 Interest arrears only 3 0/9 0/5 1 3/4 April 1983 See notes 702 6/0 2/0 1 3/4 October 1985 See notes 1,037 8/0 3/0 1/ December 1997 Debt buyback (see notes) Togo March 1980 See notes 69 3/6 1/0 October 1983 See notes 84 7/3 0/0 2 May 1988 See notes 48 8/0 4/0 I 1/8 December 1997 Debt buyback (see notes) Trinidad and Tobago December 1989 1 September 1988 48 473 12/6 4/6 1S/16 Turkey March 1982 See notes 2,269 10/0 5/0 1 314 Uganda February 1993 Debt buyback (see notes) Uruguay July 1983 1 January 1983 24 555 240 6/0 2/0 2 1/4 July 1986 1 January 1985 60 1,720 12/0 3/0 I March 1988 1 January 1990 24 1,512 17/0 3/0 February 1991 DDSR agreement (see notes) 1,284 89 Venezuela February 1986 1 January 1983 72 21,089 12/6 0/0 1 November 1987 See notes 100 14/0 1/0 September 1988 See notes 20,388 13/0 0/0 December 1990 DDSR agreement (see notes) 19,598 1,212 (table continues on next page) 98 Table A3.9 Multilateral debt relief agreements with commercial banks, January 1980-December 1997 (continued) Other assistance Repayment terms (US$ millions) (consolidation portion only) Consolidation period Amount restructured (US$ millions) Nvew Short-term Miaturity Grace Country and Length long-term credit (years! (years! Interest date ofagreement Start date (months) Defermenta Rescheduling' money' rnaintenance' months) months) (margin)' Vietnam December 1997 DDSR agreement (see notes) Yugoslavia, FR (Serbia and Montenegro) October 1983 1 January 1983 12 1,300 600 800 6/0 3/0 1 May 1984 1 January 1984 24 1,330 7/0 4/0 1/8 December 1985 1 January 1985 48 4,004 10/6 4/0 1 '/4 September 1988 1 January 1988 24 7,000 300 18/0 6/0 131,6 Zaire April 1980 See notes 402 10/0 5/0 1 7/8 January 1983 1 January 1983 12 58 10/0 0/0 2 June 1984 1 January 1984 16 64 10/0 0/0 2 May 1985 1 May 1985 12 61 10/0 0/0 2 May 1986 1 May 1986 12 65 10/0 0/0 2 May 1987 1 May 1987 12 61 10/0 0/0 2 June 1989 See notes 61 Zambia December 1984 1 January 1985 74 September 1994 Debt buyback (see notes) * Agreement in principle. DDSR = Officially supported debt and debt service reduction agreement (Brady initiative); DRF IDA Debt Reducrion Facility; FLIRB Front-loaded interest reduction bonds; MYRA = Multiyear rescheduling agreement; PDI = Past-due interest. a. Short-term rollover of current maturities. b. Consolidation of debt into new long-term obligations; may include arrears as well as future maturities; interest and short-term debt included only if indicated in country notes. For DDSR agreements, figures include face value of buybacks and all debt exchanges. c. Loans arranged for budgetary or balance of payments support in conjunction with debt rescheduling, usually in proportion to each creditor bank's exposure; sometimes referred to as concerted lending. d. Understanding by banks to maintain the sime of existing trade or other short-term credit facilities, arranged in conjunction with debt rescheduling. e. Percentage points above LIBOR. Sources. World Bank Debtor Reporting System, Institute of International Finance, and IMF Country Notes past due interest $0.7 billion was paid in cash at closing, $400 million was written off, and the remainder was A/geria exchanged for bonds (17-year maturiry, 7 years' grace), Feb 92 1991-93 Financing Facility designed to refinance maturities repayable in rising installments and yielding LIBOR +'3/1 falling due between October 1991 and March 1993. Tranche percent. A covers debts with a maturity of 2 years or more and is repayable in 8 years, including 3 years' grace bearing interest Bolivia at LIBOR + ]'/2 percent. Tranche B covers debts with a matu- Dec 80 Includes short-term debt. rity of more than 360 days and less than 2 years, and is Apr 81 Includes debt deferred it, August 1980. repayable in 5 years, including 3 years' grace. Jun 88 Ongoing program in which commercial bank debt retired through a buyback ($272 million) and a local currency bond A/bania exchange ($72 million). Applies only to previously deferred Jun 95 Restructuring of $501 million due to commercial banks. loans. $371 million bought back for $96.5 million funded by grants Jul 92 DDSR term sheet. Cash buyback at 84 percent discount; col- from IDA DRF and other donor countries; $130 million lateralized interest-free 30-year bullet-maturity par bonds; converted into long-term bonds. short-term discount bonds (84 percent) convertible on matu- rity into local currency assets at a 1:1.5 ratio, exchangeable Argentina into investments for special projects. Past-due interest can- Jan 83 Bridge loan. celed under all options. Value recovery clause based on price Aug 83 New money initially $1.5 billion. of tin. Aug 85 Agreed in principle December 1984. May 93 Buyback of $170 million commercial bank debt, funded by Aug 87 Agreement extended maturity and lowered spreads on 1983 grants from IDA DRF and other donor countries. and 1985 agreements. Also includes noncollateralized debt exchange with interest reduction ($15 million). Bosnia and Herzegovina Apr 93 DDSR agreement in which outstanding stock of $19.3 bil- Dec 97 Agreement to restructure $1,300 million of principal and lion exchanged for 30-year honds yielding a market interest past-due interest owed to commercial banks under the aegis rate (LIBOR +13/l percent) at a 35 percent discount or 30- of the London Club. Past-due interest of $700 million was year par FLIRBs. First-year interest rate 4 percent, rising to 6 wvritten off. Eligible principal of $600 million was exchanged percent in year 7 and remaining there until maturity. Both for $400 million of uncollateralized discount bonds. The bonds collateralized for principal and contain rolling 12- tenor of 37.5 percent of the new bonds is 20 years' maturity, month guarantee. Agreement also included $9.3 billion of including 7 years' grace, and stepped-up interest rates rising 99 from 2.0 percent in years 1-4 to LIBOR + 13/16 in years Congo 11-20. Servicing on 62.5 percent of the new bonds is linked Oct 86 Agreement in principle, never concluded, to restructure to economic performance. The country is not required to 1986-88 maturities, repayable in 9 years, including 3 years' make principal or interest payments for the first ten years. grace, bearing interest at LIBOR +2 7/B percent. Approximately After that the country is required to make debt service pay- $200 million of debt would have been restructured. In addi- ments if per capita income income exceeds $2,800 for two tion, there was a new money provision of $60 million. consecutive years. Per capital income in 1997 was estimated at $1,079. Costa Rica Sep 83 Includes principle arrears. Brazil May 85 Includes deferment of revolving credit ($2 million). Jul 86 Includes deferment of 1986 maturities. May 90 DDSR agreement including cash buyback at 84 percent dis- Nov 88 Includes broad package of creditor options. count ($992 million), debt-for-bond-exchange ($579 mil- Jul 92 Interest arrears: December 31, 1990. Cash payment during lion), and write-offof$29 million of PDI. 1992: $863 million. Balance converted into 10-year bonds (3 years' grace), bearing market interest rates. Cote d'Ivoire Apr 94 DDSR agreement under which four components of debt, Nov 86 MYRA. totaling $48 billion, were restructured (a) debt to foreign Apr 88 Agreement designed to replace MYRA. Includes new money banks under the 1988 multiyear deposit faciliry agreement to refinance interest. Interest on the new money portion was ($32.5), (b) debt to Brazilian banks under the MDFA, (c) LIBOR +1 /2 percent. Agreement was not put into effect debt resulting from the 1988 new money facilities ($8.1 bil- because interest arrears were not cleared, and current interest lion), and (d) interest arrears accruing from 1991-94 ($6.0 payments were suspended in April 1988. billion). The first category of debt was restructured following May 97 DDSR agreement restructuring $6.5 billion of principal and a six-choicc menu: (a) discount bonds, 35 percent discount, past-due interest. For eligible principal of $2,2751.5 million, 30-year bullet marurity yielding LIBOR -"/,,percent with creditors agreed to exchange $159 million for discount bonds principal collateral and a 12-month rolling interest guarantee (50 percent discount)subject to stepped-up interest rising ($11.2 billion); (b) par bonds with a reduced fixed-rate inter- from 2.5 percent in years 1-2 to LIBOR + 13/16 in years est (yielding 4 percent in the first year, gradually rising to 6. 11-30: exchange $1,431 million for front-loaded interest percent in year 7), 30-year bullet maturity, with principal col- reduction bonds (FLIRBs) with a maturiry of 20 years, lateral and a 12-month rolling interest guarantee. ($10.5 bil- induding 10 years' grace, and stepped-up interest rising from lion); (c) FLIRBs ($1.7 billion), with interest rising from a 2.0 percent in years 1-7 to LIBOR + 13116 in years 14-20; fixed rate of 4 percent in year I to 6 percent in years 5 and 6 and buy back $681.5 million at 24 cents per dollar. Principal and then reverting to LIBOR +13/,- percent from year 7 to is collateralized with 30-year U.S. Treasury zero-coupon maturity, 15 years maturity including 9 years' grace, 12- bonds for the discount bonds but not for the FLIRBs. A six- month rolling interest guarantee; (d) C-bonds, par reduced month rolling interest guarantee is required for the FLIRBs interest rate bonds with capitalization of interest ($7.1), with but not for the discount bonds. For past-due interest of repayment terms of 20 years maturity, including 10 years' $4,190.3 million, $30 million was settled in cash at closing, grace, interest beginning at 4 percent and the applicable rates $867 million was exchanged for bonds with a 20-year matu- in the first 6 years being capitalized, no collateral; (e) conver- rity (half a year of grace period) repayable on a graduated sion bonds ($1.9 billion) combined with new money bonds amortization schedule, and $3,293 million was written off. in a 1:5.5 ratio, interest at LIBOR +7/8 percent, 18-year manuriry, including 10 years' grace, for the conversion bonds Dominican Republic and 15 years, including 7 years' grace, for the new money Dec 83 Includes short-term debt. bonds, no collateral; (f) interest reduction loan with capital- Feb 86 MYRA. Includes arrears as of December 31, 1984. ization, maturiry of20 years including 10 years' grace, inter- Aug 94 DDSR agreement covering principal and PDI ($1.2 billion). est rising from 4 percent in year I to 5 percent in year 6 to The agreement has a menu consisting of (a) buybacks ($.4 LIBOR +±3/ ,from year 7 to maturiry. billion); (b) discount exchange bonds ($.5 billion), 35 per- cent discount, 30-year bullet maturity, interest rate of LIBOR Bulgaria +131"6percent; and (c) PDI bonds ($171 million) bearing Jul 94 DDSR agreement under which creditors agreed to restruc- interest at LIBOR +13/,6 percent, with 3 years' grace and 15 turing of $8.3 billion in public external debt, including years maturity. The accord also included a write-off of$112 about $2.1 billion in PDI. The menu for the original debt million of PDI, and $52 million paid in cash at closing. includes (a) buyback at 0.25 cent per dollar ($.8 billion); (b) discount bond 50 percent discount on face value (30-year Ecuador bullet maturity, market rate, $3.7 billion), with the discount Dec 85 MYRA. bonds collateralized for principal; and (c) FLIRBs, 18 years Nov 87 Replaces MYRA. maturiry, including 8 years' grace, interest beginning at 2 Feb 95 DDSR agreement restructuring $7.8 billion of principal and percent, rising to 3 percent in the 7 year and thereafter PDI. For principal, creditors agreed to exchange $2.6 billion for LIBOR +t/,; ($1.7 billion). FLIRBs have 1 year's interest discount bonds (45 percent discount) yielding LIBOR +13/,6 rolling interest guarantee. PDI includes cash payment of percent and $1.9 bilhon for par reduced-interest rate bonds. about 3 percent, a buyback ($.2 billion), write-off of $0.2 Both bonds have a 30-year bullet maturity, are collateralized for billion, and PDI par bonds ($1.6 billion) having a 17-year principal, and have a 12-month rmlling interest guarantee. The maturity, including 7 years' grace, and a yield of LIBOR interest rate on the par bonds is 3 percent for the first year, ris- +:sl percent. ing to 5 percent in year 11. For PDI, $75 billion is to be settled in cash at dosing, $2.3 billion was exchanged for bonds with a Chile 20-year maturiry (no grace period) repayable on a graduated Jan 84 Short-term debt consolidated. amortization schedule, $191 million was exchanged for interest Nov 84 Short-term debt rolled over to June 30, 1985. equalization bonds, and $582 million was written off Nov 85 Short-term trade credit rolled over to 1990. Aug 88 Interest spread reduced to 3/,, percent. Also cash buybacks Ethiopia ($439 million). Jan 96 Debt buyback at 8 cents per dollar of $226.0 million owed to Dec 90 New money bonds not tied to existing banks' exposure. commercial banks. Funding for the operation provided by Rescheduling includes previously rescheduled debt. the IDA DRF. Colombia Gabon Dec 85 New money without restructuring. May 94 Reschedulled principal due throstgh 1994 on debt contracted Jun 89 New money without restrmcturing. prior to September 20, 1986 (debt covered by the 1991 Apr 91 New money without restructuring. agreement, which had not been implemented). Ten-year (table continues on next page) 100 maturity including 2%/, years' grace, interest at LIBOR +7'/ May! On May 7 Mexico swapped $2.4 billion of Brady bonds for a percent. Arrears of interest and arrears of post cut-off maturi- Sep 96 $1.8 billion, 30-year utcollateri-rd bond at an interest rate of ties as of July 1, 1994, are to be repaid betveen 1994 and 11.5 percent. On September24 Mexico bought back $1.2 1996. billion of Brady bonds at a cost of8l cents per dollar This operation was funded by a $1.0 billion, 20-year bond at an Guyana interest rate of 445 basis points above U.S. Treasury rates. Aug 82 One-year deferment. Jun 83 Extension of 1982 deferment. Morocco Jul 84 Extension of previous deferment. Feb 86 Agreement in principle initiated August 1983. Jul 85 Extension of previous deferment. Jun 90 Phase one of this agreement restructures debt; phase rwo is a Nov 92 Buyback of $69 million under the IDA DRF at 14 cents per DDSR arrangement that will take effect if Morocco signs an dollar. EFF agreement with the IMF by December 31, 1991. Honduras Mozambiqrue Jun 87 Two agreements, in 1983 and 1984, were not implemented; May 87 Outstanding balance consolidated, including interest arrears. this agreement incorporated 1981-85 maturities but was not Dec91 Buybackof$124 million of outstandingcommercial bank signed. debt at a 90 percent discount, funded by grants from the Aug89 Bilateral rescheduling of debt to two commercial banks. The IDA DRF and from France, the Netherlands, Switzerland, agreement includes interest arrears. Grace period varied from and Sweden. 7 to 10 years. Interest rates were fixed, ranging from 4 to 6'/, percent. Nicaragua Dec 80 Covers government debt, all maturities, including arrears. Jamaica Dec 81 Covers nationalized bank debts, all maturities, including May 87 Includes reduced spreads on earlier agreements. arrears. Jun 90 Agreement also includes a reduction of spreads on earlier Mar 82 Covers debts of nonfinancial enterprises, all maturities, agreements to '3/,, percent. including arrears. Feb 84 Deferment of service tin rescheduled debt. Jordan Dec 95 Buyback of$ 1.1 billion of outstanding commercial bank Dec93 DDSR agreement restructuring $736 million of principal debt at 8 cents per dollar and $153 million of PDI. For restructured principal, a small amount was repurchased at 39 cents per dollar, $243 million Niger was exchanged for discount bonds (35 percent discount), and Mar 91 Buyback of all commercial bank debt at 82 percent discount $493 million was exchanged for par fixed interest bonds. ($107 million). Resources provided by grants from the DRF Both bonds have a 30-year bullet maturity with principal col- ($ 10 million), Switzerland ($3 million), and France ($ 10 mil- lateral and a 6-month rolling interest guarantee. The lion). discount bonds yield LIBOR +13I. percent interest; the yields on par bonds begin at 4 percent in the first year, rising Nigeria to 6 percent in year 7. At closing $29 million in PDI was Nov 87 Includes short-term debr. paid; $91 million was exchanged for noncollateralized bonds Mar 89 Includes line of credit arrears. with a 12-year maturity, including 3 years' grace, yielding Jan 92 DDSR agreement providing for a cash-back at 60 percent LIBOR +13/, percent; and $33 million was written off Up- discount on $3.3 billion and debt exchanges on $2 billion for front costs totaled $ 147 million, all of which was provided collateralized 30-year bullet maturity par bonds with reduced from Jordan's own resources. interest rates of 5.5 percent for the first 3 years, 6.25 percent thereafter Creditor selections: 62 percent for the buyback, 38 Liberia percent for the debt-reduction bond. A third option, new 1983 Consolidation of oil facility debt. money combined with conversion bonds, *vas not selected by participating creditor banks. Madagascar Oct 81 Restructuring of entire stock of debt, including arrears. Panama Nov 81 Arrears on overdrafts consolidated into long-term debt. Apr96 DDSR agreement under which creditors agreed to restruceur- Jun 87 Modified terms of October 1984 agreement. ing of $3.9 billion in public external debt, including $2.0 billion in PDI. The menu for the principal includes (a) dis- Malawi count bonds at a 45 percent discount of face value (30 years Oct 88 Rescheduled balances as of August 21, 1987. bullet maturity, market rare, $87.8 million); (b) par bonds with reduced interest rates and a 30-year bullet repayment Mauritania ($268.0 million); and (c) FLIRBs for $1,612.2 million with Aug 96 Debt buyback of $53.0 million, at a 90 percent discount. 18 years maturity, including 5-year grace period. The Funding for operation prouided by the IDA DRE discount and the par bonds are collateralized with respect to the principal by U.S. Treasury zero-coupon bonds and with Mexico respect to interest in the form of a nine month rolling inter- Mar 85 MYRA covering previously rescheduled debt. est rate guarantee in the first year rising to 12 mouths in years Aug 85 MYRA covering debt not previously rescheduled. 2-3. The FLIRBs do nor require guarantee for the capital, Oct 85 Deferment of first payment under March 1985 agreement. but include a 6 month rolling interest guarantee. PDI settle- Mar 87 Modification of terms of earlier agreements. ment includes progress payments of $30.0 million, a Aug 87 Private sector debt restructured. payment at closing of$ 100.0 million, a write-off of $590.4 Mar 88 Exchange of debt for 20-year zero-coupon collateralized million arising from she recalculation of penalry interest at a bonds ($556 million). lower interest rate, and PDI par bonds of S1,247.6 million Mar 99 DDSR agreement. In addition to new money of $1 billion, with 20-year maturity, including 7 years' grace, and an inter- the agreement provides for the exchange of $20.5 billion of est rate of LIBOR st5l16percent. Neither principal nor inter- debt for bonds at a 35 percent discount, an exchange of est is guaranteed, and Panama may capitalize for the first 6 $22.4 billion of debt at par for reduced interest rate bonds, years any positive differences between LIBOR +tS/16and 4.0 and conversion bonds totaling $5.3 billion. Conversion percent. bonds are not collateralized and have a 15-year maturity, May 96 DDSR agreement: Creditors agreed to restructure $2.0 bil- including 7 years' grace, and an interest rate of LIBOR + 5/ lion in past-due interest (PDI). The menu for the principal The total base also includes $693 million nor committed to includes: discount bonds at a 45 percent discount of face any option, value (30 years bullet maturity, market rate, $87.8 million); 101 par bonds with reduced interest rates and a 30-year bullet debt covered by the 1988 restructuring agreement ($8.9 bil- repayment ($268.0 million); and front-loaded interest reduc- lion), (b) debt due under the Revolving Short-Term tton bonds (FLIRBs) with a tenor of 18 years maturity, Arrangement ($1.2 billion), and (c) PDI not othersvise including 5 years' grace ($ 1,612,2 million). The discount and restructured ($4.3 billion). The first category was subject to a the par bonds are collateralized with respect to the principal menu approach, with $2.1 billion of long-term debt repur- by U.S. Treasury tero-coupon bonds, and with respect to chased at 41 cents per dollar and $0.3 billion of RSTA debt interest in the form of a 9-month rolling interest rate guaran- repurchased at 38 cents per dollar. For the remaining long- tee in the first year rising to 12 months in years 2-3. The term debt, creditors chose between (a) discount bonds at 45 FLIRBs do not require guarantee for the capital, but include percent discount ($5.4 billion), (b) par reduced fixed interest a six-month rolling initerest guarantee. PDI settlement bonds ($0.9 billion), and (c) conversion bonds combined includes progress payments of $30.0 million, a payment at with new money bonds equal to 35 percent of the amount closing of$ 100.0 million, a writeoff of $590.4 million aris- converted ($0.4 bilhon). The discount bonds and par bonds ing from the recalculation of penalty interest at a lower inter- have 30-year bullet maturirirs and feature collarcralization of est rate, and PDI par bonds of $1,247.6 million with 20 principal only. Interest on the discount bonds is LIBOR years' maturity, including 7 years' grace, and interest rate of +±/1 ,percent. Interest on the par bonds is 2.75 percent for LIBOR + 13/16 percent. Neither principal nor interest is the first year, rising to 5 percent in year 21. The conversion guaranteed. bonds have a 25-year maturity, including 20 years' grace peri- od. Yield in year I is 4.5 percent, rising to 7.5 percent in year Peru 11. New money bonds have a 15-year maturity, including Nov 96 DDSR agreement under which creditors agreed to restructure 10-year grace, and yield LIBOR 15/, percent. New money $8.0 billion in public external debt, including $3.8 billion in and conversion bonds are not collateralized. RSTA debt not PDI. The menu for the principal includes (a) discount bonds repurchased ($0.9 billion) is exchanged for 30-year bullet at a 45 percent discount of face value (30 year bullet maturi- maturity fixed interest bonds, with similar (but slightly differ- ty, market rate, $947.0 million); (b) par bonds with reduced ent) step-down/step-up arrangements as the par bonds, start- interest rates and a 30 year bullet repayment ($189 million); ing at 2.75 percent in year I and gradually rising to 5 percent (c) FLIRBs for $1,779.0 million with a 20-year maturity, in year 21. For PDI $.8 billion was repurchased wyith related including 8 years' grace period; and (d) a buyback of long-term and RSTA principal. A portion is to be settled $1,266.0 million at 38 cents per dollar The discount and the with cash payments at closing ($63 million), a portion was par bonds are collateralized with respect to the principal by written off ($0.8 billion). and the remainder ($2.7 billion), U.S. Treasury zero-coupon bonds and with respect to interest was converted into fixed-interest rate bonds yielding 3.25 in the form of a 6-month tolling interest rate guarantee percent in year R. rising to 7 percent in year 9. Maturity is 20 secuted by cash or permitted investments. The FLIRBs do years, including 7 years' grace. Amortization is graduated. not require guarantee for the capital, but include a 6-month rolling interest guarantee. PDI settlement includes progress Romania payments of $83.0 million, a payment at closing of $225.0 Sep 86 Covers previously rescheduled debt onlv. million, a buyback of $1,217.0 million at 38 cents per dollar, and PDI par bonds of $2,284.0 million with a 20-year matu- Russ/an Feder/sion rity. including 10 years' grace, and an interest rate of LIBOR Dec 91 Deferment of principal due between December 1991 and ,3I/G percent. Neither principal nor interest is guaranteed, March 1992 on pre-1991 debt. Deferment was extended for and Peru may capitalize for the first 6 years, the difference each consecutive quarter until the end of 1993. between LIBOR +/ and 4.0 percent a year. Jul 93 Agreement in principle to reschedule the stock of debt of the former Soviet Union to commercial banks contracted prior to Philippites January 1, 1991 ($24 billion), to be repaid with 15-year Jan 90 DDSR agreement provided for $1,337 million of buybacks at maturity, including 5 years' grace. In the fourth quarter of 50 percent discount. 1993, $500 million was to be paid on interest accruing dur- Dec 92 DDSR agreement: Following implementation of a cash buy- ing 1993. At the end of 1993, all remaining unpaid interest back of $1.3 billion on May 14, 1992, banks selected debt (estimated at $3 billion) wvould then be consolidated and exchanges from three options: (a) FLIRBs; yielding LIBOR repaid with 10-year maturity including 5-year grace. The +13/" percent from year 7 to maturity (15 years for series A 1993 interest payments were not made, and the agreement and 1512 year for series B, both including 7 years' grace); (b) was not implemented, mainly because of Russian refusal to collateralized step-down/step-up interest reduction bonds accept the bankers' requirement that sovereign immunity be yielding 6.5 percent from year 6 to maturity (25-year bullet waived. However, an understanding was reached on 5-Oct-94 maturity for series A and 25112 year for series B); and (c) new that the banks would drop their insistence on the waiver of money combined with conversion bonds in a 1:4 ratio, with sovereign immunity and that the Vneshekonombank (or both bonds attaining 17'/2 (series A) or 17-year (series B) another public sector enrity) svould guarantee the maturity, including 5 years' grace and yielding LIBOR +13/,t restructured debts. Signing and payment of the $500 million percent. Interest payments on both interest-reduction bonds expected by end-1994. covered by a rolling 14-month guarantee. Creditor choices Nov 95 Agreement in principle. Heads of terms were signed for a ($4.4 billion, 96 percent total eligible debt): buybacks, $1.3 comprehensive rescheduling of the debt of the former Soviet billion (27.5 percent): option (a) $0.8 billion (46.3 percent). Union in the amount of $25.5 billion in principal outstand- option (b) $1.9 billion (41.1 percent), option (c) $0.5 bil- ing and $7.5 billion in accrued interest due. This total of $33 lion, (11.7 percent). billion was to be repaid over 25 years, with 7 years' grace, Sep 96 Philippines issued $0.7 billion eurobond in exchange for beginning December 15, 1995, in 37 semi-annual payments Brady bonds originally issued to replace commercial bank on a graduated schedule at LIBOR 5t3/1, percent a year. It debt in 1989. The Eurobond was issued in the form of a 20- was further agreed that an inreiest note would be issued with year note at an annual interest rate of 8.75 percent. a 20-year maturity and 7 years' grace from December 15, 1995, that would bear the same interest rare, listed on the Poland Luxembourg Stock Exchange. The remaining $1.5 billion in Jul 84 Includes some short-term trade credits. interest arrears was paid over 1995-96. By September 1996 Sep 86 Covers debt rescheduled in 1982. the minimum subscribership by individual commercial banks Jul 88 MYRA. Also improved terms of earlier agreements. of $20 billion in outstanding principal was reached, which Jun 89 Principal due May 1989-December 1990 deferred until triggered the Russian agreement to the rescheduling package. December 1991; in October the $145 million in interest due in the fourth quarter of 1989 was deferred until the second Sao Toma and Princrpe quarter of 1990. Aug 94 Buyback under the IDA DRF at 10 cents per dollar Oct 94 DDSR agreement under vhich creditors restructured $14.4 extinguished $10.1 million of principal (87 percent of eligi- billion. Three categories of debt were affected: (a) long-term ble debt). (ratle conrinues on nextpage) 102 Senegal Dec 88 Exchange of debt for bonds outside the framework of the Dec 96 Debt buyback under the IDA DRF at 8 cents per dollar of main negotiations. $80.0 million owed to commercial banks. Dec 90 DDSR agreement featuring buybacks in the form of 91-day collateralized short-term notes ($1,411 million), exchange for Sierra Leone bonds at 30 percent discount ($1,810 million), exchange at Jan84 Covers arrears asof Deember 31, 1983. par for reduced fied-r ate inerest bonds ($7,457 million), Aug 95 Buyback at 13 cents on average per dollar of $235 million exchange for bonds at par with temporary step-down interest due to commercial banks. Funded by grants from IDA DRF rates ($3,027 million), and nesv money combined with debt and other donor countries. conversion bonds ($6,022 million). Sudan Vietnam Nov 81 Includes arrears of principal and some short-term debt. Dec 97 DDSR agreement restructuring $310.9 million of principal Mar 82 Covers arrears of interest and modifies 1981 agreement. and $486.2 million of past-due interest (PDI). For restruc- Apr 83 Modification of 1981 agreement. tured principal, $20.4 million was repurchased at 44 cents Oct 85 Covers arrears of interest. per dollar, $51.6 million was exchanged for discount bonds (50 percent discount), and $238.9 million was exchanged for Togo par fixed interest bonds. Both bonds have a 30-year maturity Mar 80 Balance of debts to French banks, including arrears of princi- but the discount bond is subject to an interest rate of LIBOR pal. Interest rates vary by currency. + 13/16 while the par bond is subject to step-up interest rates Oct 83 Covers all commercial bank debt, including debt previously rising from 3.0 percent in years 1-2 to 5.5 percent in years rescheduled. 21-30. One hundred percent of the discount bonds and fifty May 88 Restructuring of 1983 agreement. percent of the par bonds are guaranreed by U.S. Treasury Dec 97 Debt buyback at 12.5 cents per dollar of $46.1 million owed zero-coupon bonds, atld the discount bonds have a 6-month to commercial banks. Funding for the operation was provid- rolling interest guarantee. Regarding PDI, $1 5.0 million was ed by the IDA Debt Reduction Facility. paid at closing, $294.8 million was exchanged for noncollar- eralized bonds with an 18-year maturity (including 7 years' Turkey grace and step-up interest rates), $21.8 million was Mar 82 Improved the terms of the August 1979 agreement. repurchased at 44 cents per dollar, and $154.6 million was written off. Uganda Feb 93 Buyback of $153 million commercial bank debt, funded by Yugoslavia, FR (Serbia andMontenegro) grants from IDA DRF and orher donor countries. Oct 83 Includes a 1-year rollover of short-term bonds. Dec 85 MYRA. Uruguay Jul 86 MYRA. Zaire Mar 88 Improved terms of July 1986 agreement. Apr 80 Covered stock ofdebt as of end-1979, including arrears. Feb 91 DDSR agreement provided for cash buyback at a 44 percent Jan 83 Rescheduled principal due under April 1980 agreement. discount ($628 million), collareralized debt reduction bonds Jun 84 Rescheduled principal due under April 1980 agreemenr. ($535 million), and new money ($89 million) combined May 85 Rescheduled principal due under April 1980 agreement. with debt conversion notes ($447 million). Repayment terms May 86 Rescheduled principal due under April 1980 agreement. are 30-year bullet maturiry and 6.75 percent fixed interest for May 87 Rescheduled principal due under April 1980 agreement. the interest reduction bonds, 16-year maturity; including 7 Jan 89 Financed monthly payments on ourstanding claims, mainly years' grace, with LIBOR +7/8 percent interest for the conver- interest on arrears. sion notes, and 15-year maturity including seven years' grace with LIBOR + I percent interest for the new money notes. Zambia Dec 84 Includes arrears as of February 28, 1983. Venezuela Sep 94 Buyback under the IDA DRF at 11 cents per dollar Feb 86 MYRA. Agreed in principle in September 1984. extinguished $200 million of principal (75 percent of eligible Nov 87 Reduced spread and extended maturities of 1986 agreement. debt), using $10.5 million of IDA resources and $22.3 mil- Sep 88 Reduced spread on February 1986 agreement. lion from other donors. APPENDIX 4 Progress on privatization Privatization is the transfer of productive assets Shanghai exchange and Hong Kong dollars on the from the state to private investors through such Shenzhen exchange. H shares are Chinese corpo- methods as auctions, stock offers, stock distribu- rate shares issued on the Hong Kong market. tions, negotiated sales, management-employee Buoyed by a surge of activity on the Hong Kong buyouts, and voucher or coupon exchanges. exchange, Chinese state-owned companies issuing Other methods include leasing, joint ventures, H shares fared well in 1996, raising more than management contracts, and concessions, includ- $900 million. This compares favorably with 1995, ing build-own-operate (BOO), build-operate- when new issues were constrained by poor market transfer (BOT), and build-own-operate-transfer conditions and only three new H share issues were (BOOT) arrangements. completed. The Guangshen Railway share offering In 1996 global proceeds from privatization in 1996 raised nearly $544 million, making it one exceeded $25 billion, mainly from sales of large- of China's largest deals and accounting for about 15 scale infrastructure (figure A4. 1). Brazil was the percent of the H share market. Also of interest was most active privatizer, accounting for more than $5 the Anhui Expressway offer, which raised more billion in sales, largely from the sale of energy util- than $112 million and was the first state-owned toll ities and railway concessions. Europe and Central road to be listed outside China. To raise the quali- Asia saw privatization receipts fall in 1996, follow- ty of listings on the H share market, the China ing a major selloff program by Hungary in 1995. Securities Regulatory Commission established new Privatization activity also slowed in South Asia and criteria in 1996 for selecting companies, including East Asia and the Pacific, in part because of a lull a minimum issue size, a minimum annual profit, in primary issuing activity. The Middle East and and a three-year earnings record. North Africa saw sales revenues jump because of an Indonesia, which has played a sizable role in the active program in Egypt, while Sub-Saharan Africa region's privatization activities in recent years, got continued to privatize on a smaller scale, off to a late start in 1996. Primary issues, the coun- Eaist Asial and the Pacific Figure A4.1 Privatization revenues in developing With the exception of China, privatization activi- countries, 1990-96 ty in East Asia and the Pacific slowed considerably 30 during 1996, with total revenues estimated at nearly $2.7 billion, about half the level in 1995 25 (table A4.1). The mixed performance of the 20 . Sub-Saharan Africa region's stock markets was partly responsible for 15 U i s d Nr AfSouth Asia the absence of privatization-related equity deals.X 0 00 Middie and teCibAn Although China does not have an official priva- 1 uoe n etrlAi tization program, it sells stakes in state-owned 5 Europeand entalAsia enterprises in the form of B and H shares. B shares 0 U East Asia and the Pacifc are issued to foreigners through the domestic mar- 1990 1991 1992 1993 1994 1995 1996 ket and are denominated in U.S. dollars on the Source:World Bank Privotizaon Database. 103 104 Table A4.1 Privatization revenues in East Asia and the Pacific, 1990-96 (millions of U.S. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total China - 11 1,262 2,849 2,226 649 919 7,916 Indonesia - 190 14 31 1,748 2,031 1,008 5,022 Malaysia 375 387 2,883 2,148 798 2,519 214 9,326 Philippines - 244 754 1,638 494 207 22 3,359 Thailand - - 238 471 242 - 291 1,241 Other 1 2 10 18 - 4 226 261 Total 376 835 5,161 7,155 5,507 5,411 2,679 27,123 - Not available. Source: World Bank Privatization Database. try's main mode of privatization sales, slowed con- Thailands stalled privatization program received siderably. Only two sales were concluded in 1996, a much-needed boost when the Electricity and one was a secondary offering, for PT Telekom. Generating Authority sold off the Khanom The company, which was partly privatized through Electricity Generating Company. The transaction an initial offering in 1995, raised nearly $611 mil- included a public offering for $240 million. And lion for a 4.2 percent stake in its second offering. Papua New Guinea tnarked its first privatization Combined proceeds from both deals were close to public offer with a 49 percent sale of Orogen the originally estimated sales price of $3 billion. Minerals for nearly $224 million. The second sale in 1996, of Bank Negara Indonesia, was the first offer of a state-owned bank Latin America and the Caribbean and was judged a huge success based on investor response to the domestic tranche, which raised Latin America raised more than $14 billion from nearly half of the total sales price of $400 million. sales of state enterprises in 1996 (table A4.2). Sales The Philippine privatization program, which by federal governments continued to wind down had made a significant contribution to the gov- in many of the region's larger economies, while ernment's budgetary surplus in recent years, sales by state and provincial authorities began to slowed considerably in 1996. State selloffs raised a pick up speed. fraction of the revenue generated in previous years, Brazil raised more than $5.7 billion in 1996, due in part to delays in the completion of some mainly from the sale of Rio de Janeiro-based elec- deals and the postponement of others. Bidding for tricity distribution comnpany Light, which was sold the Manila Waterworks and Sewerage System for $2.4 billion. Proceeds from the sale of Light commenced in 1996 but was not completed until accounted more than half of the targeted privati- 1997 and sparked controversy over the selection of zation revenue for 1996. Relative to Argentina and two concessionaires. Chile, Brazil has only recently embarked on full- Table A4.2 Privatization revenues in Latin America and the Caribbean, 1990-96 (millions of U.S. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total Argentina 7,532 2,841 5,741 4,670 894 1,208 642 23,527 Bolivia - - 9 13 - 789 34 844 Brazil 44 1,633 2,401 2,621 2,104 992 5,770 15,564 Chile 98 364 8 106 128 13 187 904 Colombia - 168 5 391 170 - 2,075 2,809 Mexico 3,160 11,289 6,924 2,131 766 167 1,526 25,964 Peru 3 212 127 2,840 1,276 1,751 6,209 Venezuela 10 2,278 140 36 8 39 2,017 4,527 Other 71 147 120 393 1,289 132 140 2,292 Total 10,915 18,723 15,560 10,487 8,198 4,615 14,143 82,641 -Not available. Source: World Bank Privatization Database. 105 scale privatization of its state and federal utilities, international tranche of American depository which are estimated to be worth several billion shares (ADSs), making it the largest offering (in dollars. The first state telephone company selloff terms of size and investor response) in Peru and the occurred in 1996 with the highly successful sale of most successful equity deal in Latin America since a $660 million minority stake (35 percent) in the Mexican peso crisis in December 1994. The CRT, a utility operated by the state of Rio Grande sale was also a showcase for the government's citi- do Sul. Restructuring of the Brazilian railway sys- zen participation program, whereby stakes in state- tem was nearly completed in 1996, with conces- owned enterprises are offered to foreign as well as sions worth more than $1 billion awarded for five domestic institutional investors and domestic retail of the six networks. The sale of mining giant investors. EGENOR, one of three units of Companhia Vale Rio Doce was delayed numerous Electroperu, was sold for $228 million to a strate- times, but the company was eventually sold in gic foreign invcstor. The sale of Petroperu, post- 1997. The sale raised more than $3 billion, mak- poned several times, was initiated in 1996. The ing it the largest deal in Latin America to date. government decided to split the company into Chile resumed its privatization program in units to be sold separately. La Pampilla refinery was 1996 after a pause of several years with the sale of the first unit to be sold, raising $180 million. a railway company, Ferronor, and an electricity generating company, Tocopilla-Codelco. The gov- Europe and Central Asia ernment plans to unload remaining shares in 2 other energy firms and 14 water companies over Countries in Europe and Central Asia raised more the next few years. than $5 billion from privatization in 1996 (table Privatization programs in Colombia and A4.3). Although privatization programs in some Venezuela, which had been moving slowly in countries in the region (the Czech Republic, recent years, were boosted by several noteworthy Hungary) wound down, the process continued sales. As part of an overall restructuring of its ener- apace in others. gy sector, Colombia offered several key electric Privatization proceeds in Hungary dropped sig- power plants for sale and raised more than $1 bil- nificantly following a record year of sales in 1995. lion through the selloff of two large hydropower Hungary has nearly completed its bank privatiza- plants. Foreign participation in Colombia's econ- tion program, begun in 1995, with the sale of five omy, including sales of public utilities, has been of its six largest banks, two of which were sold in facilitated by changes in the foreign investment 1996. The government also sold its largest petro- law that were introduced in 1996. Prior autho- chemical firm, TVK, raising more than $162 mil- rization is no longer required for foreign invest- lion from a placement of global depository receipts ment in sectors such as energy, transport, and (GDRs). One of the remaining power plants, communications. In Venezuela the government's Tiszai Power Co., also found a buyer in 1996, rais- 49 percent stake in television station CANTV was ing $149 million. sold for more than $1 billion. The government While Hungary and the Czech Republic were also reprivatized several banks that had been taken nearly finished with their privatization programs, over during the banking crisis of 1994. But other other countries in the region, including Russia, deals, including the sales of aluminum and steel were just beginning to sell off state shares in large companies, were postponed. enterprises. Following the controversial loans-for- Mexico continued to promote private partici- shares program initiated in 1995-which enabled pation in infrastructure. In 1996 the first railway the government to secure loans from Russian concession was awarded for $1.4 billion. Mexico banks by entrusting shares in large companies to also continued to privatize its ports and storage the banks-the banks were allowed to sell these facilities. shares if the government had not repaid the loans Peru's government sold its remaining 26.6 per- by September 1996. Several of these stakes were cent stake in Telefonica del Peru through a place- put up for sale in 1996 and were bought by affili- ment of domestic and international shares. The ates of the banks. In 1996 Lukoil became the first deal raised $1.2 billion, 74 percent of it through an Russian corporate to issue American depository 106 Table A4.3 Privatization revenues in Europe and Central Asia, 1990-96 (millions of U.S. dollars) Country 1990 199] 1992 1993 1994 1995 1996 Total Bulgaria - - - 45 147 111 48 350 Czech Republic - - - 645 7 1,645 - 2,297 Hungary 483 798 779 1,685 1,507 3,988 945 10,185 Poland 62 338 240 733 641 980 605 3,599 Russia - 35 88 110 - 1,002 1,192 2,426 Slovakia - - - 63 415 1,004 486 1,968 Turkey 437 212 780 483 354 572 297 3,135 Other 280 1,168 1,739 224 886 440 1,893 6,630 Total 1,262 2,551 3,626 3,988 3,956 9,741 5,466 30,591 - Not available. Source. World Bank Privatization Database. receipts (ADRs), giving international investors progress on privatization, Bulgaria has one of the access to Russian stocks for the first time. The deal lowest levels of foreign investment in the region. raised nearly $131 million, equivalent to a 3.3 per- Kazakhstan continued to actively pursue priva- cent stake in the company. In another offering tization during 1996. As in Hungary, the program Gazprom, Russia's largest corporate and the is targeted at strategic investors, who receive a world's biggest producer of natural gas, raised majority stake in the privatized company in $429 million through an ADR issue. The sale, rep- exchange for transferring management and tech- resenting 1.2 percent of the company, was the nological know-how. largest international equity offering by a Russian The first internation-al listing from Croatia was company at the time. a success by any measure. The GDR issue by phar- Polands privatization program slowed in 1996, maceutical group Pliva raised $161 million, of held up in part by the delay in the listing of nation- which about $90 million came from foreign al investment funds. The funds, through which investors. The Pliva deal also marked the first the government plans to sell 513 small manufac- London Stock Exchange listing for an Eastern turing, construction, and trading companies, rep- European industrial corporate. resent Poland's version of mass privatization. Polish citizens willing to pay the equivalent of $7 Middle East and North Africa per voucher were awarded privatization certificates that are exchangeable into shares in the funds once Proceeds from privatization in the Middle East they are listed on the Warsaw Stock Exchange. and North Africa was estimated at nearly $1.5 bil- Bulgarias mass privatization got under way in lion in 1996, largely due to a successful selloff pro- 1996, with voucher books distributed to citizens in gram in Egypt (table A.4.4). January. The government also announced an accel- The government of Egypt accelerated its priva- erated program of cash sales to meet budgetary tar- tization program by aggressively selling off a long gets. The only significant sale in 1996 was the list of public enterprises that generated more than Sheraton Sofia, which was sold for $22.3 million to $1 billion in 1996. In 1993-95, by contrast, pro- a foreign investor. Partly because of its slow ceeds totaled just $773 million. One significant Table A4.4 Privatization revenues in the Middle East and North Africa, 1990-96 (millions of U.S. dollars) Counny 1990 1991 1992 1993 1994 1995 1996 Total Egypt - - - 118 393 262 1,150 1,923 Morocco - - 273 347 240 271 1,130 Tunisia 2 17 60 - - 32 36 147 Other - - 9 26 42 212 21 310 Total 2 17 70 417 782 746 1,477 3,510 - Not available. Source, World Bank Privatization Database. 107 development in 1996 was the resumption in the sales, for steelmaker SONASID and fertilizer com- sale of majority stakes in state enterprises. Except pany FERTIMA, were concluded in 1996. for three companies sold outright in 1994, only In Tunisia Societe des Produits Chimiques minority stakes (10-20 percent) had been offered Detergents became the first company to sell shares to investors in recent years. The government to foreigners after the introduction of new legisla- offered equity in 12 companies through stock tion allowing foreign investors to purchase up to offers and direct sales to investors, several of which 10 percent of a Tunisian company. were for majority ownership. Among the first to be sold under the new policy was a 75 percent stake in South Asia Medinet Nasr Housing and Construction, which raised $172 million through a stock flotation. In As in East Asia and the Pacific, privatization activ- 1996, to introduce competition into the domestic ity in South Asia slowed considerably during banking system, the government allowed foreign 1996. In the absence of major sales, revenues minority shareholders to increase their sharehold- totaled less than $900 million (table A4.5). ings above 49 percent. The government also In India, where privatization has focused on the instructed all state-owned banks to sell their minor- sale of minority equity shares in public companies, ity stakes in joint venture banks by the end of 1996. there were two quasi-divestments during 1996. Only one such sale took place in 1996, however. The first, a $125 million GDR issue by the Steel The National Bank of Egypt, one of the "big four" Authority of India, the largest listed company on public banks, reduced its stake in Commercial the Bombay Stock Exchange, met with a muted International Bank, which it owns jointly with a response from investors. It was the first time a U.S. bank. Commercial International Bank state-owned Indian enterprise had been listed on became the first Egyptian issuer of GDRs with a the London Stock Exchange. The government still deal worth $120 million, reducing the National controls about 85 percent of the company's equi- Bank of Egypt's stake to 22.62 percent. ty. The second, the State Bank of India's $370 mil- The pace of privatization in Morocco, which lion deal through a rule 144A placement, was the had slowed in 1995, recovered in 1996. Of inter- largest equity offering ever from India and the first est was the largest privatization sale in Morocco, a Indian listing on the New York Stock Exchange. public offering of shares in petroleum refinery The sale of VSNL, the state-owned telecommuni- Societe Marocaine des Industries du Raffinage cations company, had been postponed several (SAMIR), which raised nearly $173 million for a times but finally came to the market in the first 30 percent stake. It was the country's first deal quarter of 1997. The issue raised $527 million, involving privatization bonds. In 1996 the gov- making it India's largest deal to date. ernment issued about $308 million of these bonds Pakistan saw many large deals postponed or to raise revenue for the treasury in advance of sales cancelled, including the sale of its telephone com- of state-owned enterprises. To encourage partici- pany, Pakistan Telecom, and United Bank, the sec- pation in the issue, bondholders were given the ond largest commercial bank. The sale of the Kot option of converting the bonds into privatization- Addu Power Station was concluded, however, related stock offerings. Two previously postponed bringing $215 million from a foreign investor. Table A4.5 Privatization revenues in South Asia, 1990-96 (millions of U.S. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total Bangladesh - - - 43 12 5 0 59 India - 931 1,098 861 1,505 810 495 5,700 Pakistan 11 63 343 17 1,106 36 317 1,893 SriLanka 18 2 105 52 42 65 77 363 Other - - 11 1 1 - 0 13 Total 29 996 1,557 974 2,666 917 889 8,028 - Not available. Source: World Bank Privatization Database. 108 Sub-Saharan Africa (table A4.7). Latin American countries, particu- larly Brazil, were especially active in power and African countries continued to pursue privatiza- transport. Power accounted for more than 40 per- tion, raising $745 million in 1996 (table A4.6). cent of infrastructure sales and nearly 25 percent Ghana led the way with revenues totaling more of proceeds from privatization in 1996. Selloffs of than $180 million. The government sold a 5 per- electricity distribution assets were particularly cent stake in Ashanthi Goldfields worth $112 mil- popular with investors, as returns are typically lion through an international placement of shares. higher from distribution than from generation. In South Africa the government sold 6 of its 23 Telecommunications generated more than a quar- radio stations as part of its plan to increase com- ter of the proceeds from infrastructure sales in petition in the news media. Local investors paid 1996, mostly due to vwo large deals from Peru and $122 million for the stations. The highlight of Venezuela that raised more than $1 billion apiece. Kenyas privatization program was the sale of a 26 In transport, sales of railway networks accounted percent stake in Kenya Airways to a strategic for more than 13 percent of privatization revenues investor, KLM, for $26 million. Another 54 per- in 1996, concentrated mainly in Latin American cent of shares were sold through the stock market countries such as Brazil and Mexico. for $46.3 million, making it the largest domestic Another important source of revenue was the offer ever. financial sector, as governments in all regions Zambia continued to make progress on sales of sought to shed their holdings in banking institu- state-owned enterprises. But the sale of Zambia tions. Hungary moved closer to its objective of Consolidated Copper Mines, which accounts for privatizing its six largest banks, which account for nearly 90 percent of the country's foreign exchange nearly 60 percent of the domestic market. After earnings, was delayed once again. The government selling three in 1995, it sold two more to foreign had prequalified companies and consortiums for investors in 1996, together with several smaller bidding on the 10 separate units through which banks. In Egypt state-owned banks began divest- the company would be sold. ing shares in joint venture banks as part of the government's strategy to withdraw from the bank- Sectoral distribution ing sector. Venezuela continued to reprivatize banks taken over by the government in the wake Infrastructure-related selloffs continued to domi- of the 1994 banking crisis. In Slovenia an amend- nate the world's privatization programs in 1996, ment to the law on privatization of strategic enter- capturing 60 percent of proceeds as governments prises was passed in 1996 that precluded the continued to unload their stakes in power, direct privatization of the country's four biggest telecommunications, and transport companies banks. Table A4.6 Privatization revenues in Sub-Saharan Africa, 1990-96 (millions of US. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total CBte d'Ivoire - 2 10 5 19 74 103 213 Ghana 10 3 15 28 476 87 186 804 Kenya 12 1 12 10 19 13 137 203 Mozambique 3 5 9 6 2 26 38 89 Nigeria 16 35 114 541 24 - - 730 South Africa - 1,073 - - - - 122 1,195 Tanzania - - 3 27 5 77 13 125 Uganda - - 12 19 24 47 30 132 Zambia - - - 3 14 69 30 115 Zimbabwe - - - - 13 75 - 88 Other 33 2 32 2 9 5 86 168 Total 74 1,121 207 640 602 472 745 3,861 - Not available. Source: World Bank Privatization Database. 109 Table A4.7 Privatization revenues by sector, 1990-96 (millions of U.S. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total Infrastructure 9,704 6,863 9,715 5,360 9,399 9,240 15,063 65,344 Telecommunications 7,643 5,981 3,007 1,083 6,069 3,691 3,814 31,288 Power 59 359 4,892 1,741 2,180 4,523 6,156 19,911 Manufacturing 1,402 5,558 7,188 7,491 6,091 5,787 3,546 37,063 Steel 185 2,145 1,614 2,900 1,219 135 193 8,391 Chemicals 156 466 315 415 1,285 291 488 3,417 Construction 196 484 732 491 790 592 745 4,030 Other manufacturing 864 2,462 4,528 3,685 2,798 4,769 2,120 21,224 Primary sector 1,367 3,608 3,394 6,215 4,068 4,336 2,787 25,776 Petroleum 568 2,085 2,760 5,162 2,115 2,781 1,687 17,158 Mining 485 235 382 187 1,220 618 468 3,595 Financial services 47 7,793 5,263 3,411 1,065 1,933 2,895 22,406 Banking 47 7,505 5,099 2,464 779 1,853 2,646 20,392 Other services 138 420 621 1,184 1,088 606 1,108 5,165 Total 12,658 24,242 26,181 23,661 21,712 21,901 25,399 155,753 Source: World Bank Privatization Database. Method of sale relatively developed, has been public offers. In Latin America and Europe and Central Asia direct Direct sales continue to be the preferred method sales to strategic investors have been widely used to of divesting state-owned shares, accounting for attract management and technological know-how. more than 60 percent of privatization revenues in In the Middle East and particularly Egypt, the 1996. Public offers, which tend to be large and flood of liquidity resulting from the large number highly liquid investments, were the second most of privatization offers has been the primary reason preferred method, raising more than $8.1 billion for increased activity on the region's stock markets. in 1996. International public offers in the form of In Africa, even though direct sales have been pop- GDRs were also popular in 1996, with three coun- ular in the past, maturing stock markets have tries (Croatia, Egypt, Morocco) becoming first- helped boost the number of privatization-related time issuers in order to sell stakes in state-owned equity offers. Leases and concessions are common- companies. GDRs are popular because they offer ly used for infrastructure-related projects, particu- a wider investor base, particularly if there are larly in transport. domestic ceilings on foreign ownership of shares. Several Russian companies established level 1 Foreign participation depository receipt programs in 1996 in prepara- tion for selling stakes to international investors.' Foreign investors provided nearly 44 percent of the Sales methods have varied across regions for var- proceeds from privatization sales in 1996. Of the ious reasons, including privatization strategy and estimated $11.3 million raised in foreign exchange market conditions. The most preferred method in (table A4.8), equity investors accounted for $5.6 East Asia and the Pacific, where stock markets are million-twice as much as in 1995, when equity Table A4.8 Foreign exchange raised through privatization in developing regions, 1990-96 (millions of US. dollars) Country 1990 1991 1992 1993 1994 1995 1996 Total EastAsiaandthePacific 1 102 1,556 4,156 4,036 2,026 1,990 13,865 Europe and Central Asia 586 1,892 3,069 2,932 1,588 4,778 1,880 16,726 LatinAmericaandtheCaribbean 6,358 7,384 4,037 3,765 5,058 2,206 6,448 35,257 Middle East and North Africa 0 3 19 183 246 16 126 594 South Asia 11 4 44 16 997 38 528 1,638 Sub-Saharan Africa 38 5 66 566 453 275 299 1,702 All developing countries 6,994 9,390 8,791 11,619 12,378 9,338 11,271 69,782 Source, World Bank Privatization Database and staff estimates. 110 Table A4.9 Portfolio investment and foreign direct investment in privatization, 1l990-96 (millions of US. dollars) Type 1990 1991 1992 1993 1994 1995 1996 Total Foreign direct investment 6,888 5,517 6,039 6,429 6,414 6,380 5,644 43,310 Portfolio investment 106 3,873 2,752 5,190 5,965 2,959 5,627 26,472 Total 6,994 9,390 8,791 11,619 12,378 9,338 11,271 69,782 Source: World Bank Privatization Database. issues plunged in the wake of the Mexican peso cri- or are converted from local currencies into U.S. sis (table A4.9). Foreign direct investment account- dollars at the annual average exchange rate. For ed for another $5.6 million, including foreign direct sales, if a buyer's identity is unknown, it is concessionaires, who provided $686 million. Latin assumed that the buyer is domestic. For public American countries were the main recipients of for- offers, the foreign exchange component is not esti- eign direct investment, attracting more than $4 bil- mated if the number of shares purchased by for- lion. East Asia led in portfolio investment, with eign investors is unknown. nearly $2 billion, followed closely by Latin America, with almost $1.9 billion. Central and Note Eastern European countries attracted more than $1 billion in equity investment in 1996. 1. A level I depository program offers new issuers an entry into the U.S. market by allowing over the World Dank's counter trading through pinksheets of existing shares. Sources of data for the World Bank's Level I programs are popular with non-U.S. companies Privatization Database that are unable to meet the more stringent requirements of level 2 or level 3 programs. A level 2 program involves The data contained in the World Bank's the listing of a company's shares on the New York Stock Privatization Database is drawn from various Exchange or other major exchange, which requires com- pliance with U.S. stock exchange requirements, includ- sources, including reports from official privatiza- ing Securities and Exchange Commission (SEC) filing tion agencies and other internal World Bank requirements and generally accepted accounting princi- Group databases, and is supplemented with data pals. With a level 3 program, a company is allowed to contained in publications such as Privatization issue new shares in a public offering after meeting level Internationa4 International Financing Review, 2 requirements and additional requirements involving underwriters and the like. Level 3 programs can also be Latin Finance, Middle East Economic Digest, and established as private placements under the SEC's rule Euromoney. All data are in U.S. dollars as reported 144A. APPENDIX 5 Trends in flows from international capital markets Gross flows from international capital markets, Japanese yen fell to 8 percent and that of deutsche including portfolio debt flows (bond issues, cer- marks (DM) to 5 percent, in part because of the tificates of deposit, and commercial paper), com- strong performance of the U.S. dollar against mercial bank loan syndications, and portfolio other major currencies. equity flows, totaled $267 billion in 1997, up from The past year saw some diversification to other $225 billion in 1996. Gross flows increased currencies, especially by Latin American borrow- strongly for the first three quarters of 1997 and ers who issued in Italian lira, British pounds, then declined in the fourth quarter as the East Argentine pesos, Canadian dollars, French francs, Asian crisis spread and reduced developing coun- Swiss francs, Austrian schillings, and Spanish pese- tries' access to the capital markets. tas. Latin America accounted for 85 percent of the value of issues in currencies other than the U.S. Portfolio debt flows dollar, deutsche mark, or yen. East Asian countries borrowed primarily in U.S. dollars or yen, with the International bond issues (over 90 percent of port- exception of small issues by China in deutsche folio debt flows) totaled $94 billion from January marks and Thailand in Austrian schillings (fig- through October 1997 and then fell to only a lit- ure A5.2). U.S. dollar funding by Europe and tde over $1 billion for the rest of the year as the East Central Asia shot up to 67 percent in 1997 from Asian financial crisis spread. Certificates of deposit 40 percent in 1996, largely owing to Russia's heavy (CDs) fell from $4 billion in 1996 to $3 billion in borrowing in dollars. Most of the remaining issues 1997, and commercial paper issues were stable at from the region were denominated in deutsche $3 billion. marks, with Turkey and Russia the major borrow- ers. South Asian issues were almost entirely in U.S. Borrower composition Figure A5.1 Borrower composition of international Sovereign borrowers retained the largest share (48 bond issuance by region, 1996-97 percent) of developing countries' bond issues in Bilins of U.S. dollars 1997, but almost 60 percent of the $21 billion rise 1997 in issues came from increased participation by the 50 996 private sector (figure A5.1). Greater familiarity 40 aPnovaute with developing country borrowers and expanded 30 Nnbeic use of collateralized transactions enabled more Government corporate borrowers to participate, and at general- 20 ly improved terms (at least until October). 10 C Currency composition Sub-Sahoran EastAsia South Europe and Lahin Middle East Africa and the Asia Central Asia America and North Pactifc and the Africa Three-fourths of the bond volume issued by devel- Caribbean oping countries was denominated in U.S. dollars, Nate: Data tar 1997 ire preliminary. compared with 67 percent in 1996. The share of Source: Euromoney Bondware ind World Bank. 111 112 Figure A5.2 Currency composition of international Figure A5.3 International bond volumes and issues bond issues by region, 1997 from the ASEAN-4, 1997 East Asia and the Pacific Millions of U.S. dollars Number of issues Yen 1,0006 2; 800 Bond 8 400 3% 200 2 Europe and Central Asia20 Yen 3l I 0 January- July August September October November December June overage Note: Bond volume exclusive of offshore issuonce; the ASEAsN-4 are Indonesio, Moloysio, JSS tise Philippines, and Thailand. r3h°/e' t 67% Source: Euromoney Bondwore ond World Boank. lar through a currency board arrangement); and Latin America and the Caribbean the century bond from a Chilean public sector Lier electricity company. Bond volumes fell sharply Y4e%n 80i N r with the rise in secondary market spreads in late October, to $0.7 billion in the last two months of the year (1 percent of the total for 1997). /% 05 5 _ US$ Argentina issued the only sovereign bond among 74% developing countries in November or December. Total portfolio debt flows to East Asia and the Pacific increased marginally from $19.2 billion in 1996 to $20 billion in 1997. The region contin- Sued to account for almost all of the $3.9 billion issued by developing countries through the CD dollars, with the exception of a British pound issue market in 1997. Private sector borrowers took up from India. South Africa issued primarily in the about 65 percent of total issues. Bond issues were Japanese Samurai market and in its local currency, strong in the first half of the year ($12 billion, or and Middle East and North African issues were 60 percent higher than the amount issued in the principally in U.S. dollars. first half of 1996). However, the region's access to the bond markets declined in the second half of Regional composition the year with the East Asian financial crisis. Issues dropped to $7 billion (some 60 percent of the vol- Latin America and the Caribbean accounted for 60 ume in the second half of 1996), and secondary percent of developing country issues in 1997. Key market spreads skyrocketed. Issues following the market developments prior to the spread of the devaluation of the baht in July were primarily by East Asian crisis included the exchange by established borrowers, offshore registered corpora- Argentina, Brazil, Ecuador, Panama, and tions (which may have had greater access to for- Venezuela of Brady bonds for $12.3 billion in eign exchange), and a few borrowers willing to pay unsecured global bonds (see appendix 3), follow- very high rates (a Thai private sector company ing the precedent set by Mexico in 1996; paid a spread of 911 basis points over U.S. Argentina's issue of a local currency-denominated Treasuries). Of the major East Asian borrowers, eurobond (the country's currency is tied to the dol- only China was relatively unaffected by the crisis, 113 as bond issues rose from $3.8 billion in 1996 to an Bond volumes for the Middle East and North estimated $5 billion in 1997, $1.5 billion of it in Africamorethan doubled,with sovereign borrowing the second half of the year. from Lebanon, Oman, and Tunisia accounting for Bond volumes from South Asia doubled in half the regional volumes, Lebanon issued a World 1997, although there were no bond issues in the Bank-backed bond that enabled it to secure last quarter of the year. Bond issues from India, improved terms and attract a wider range of which accounted for over 80 percent of the region's investors. South Africa was the only issuer fromn Sub- total, increased as the government allowed blue- Saharan Africa, with most of the borrowing from the chip companies to access the international bond public sector. South Africa also issued a local markets at long-dated maturities. A private sector currency-denominated zero-coupon eurobond. company issued a century bond. The government of Pakistan issued bonds worth $450 million, and Terms the Sri Lankan government, returning to interna- tional capital markets for the first time since 1982, The average maturity of developing countries' issued $50 million. bond issues has approximately doubled since Bond volumes from Europe and Central Asia 1992, reaching 10 years in 1997. (The decline in rose from $7.5 billion in 1996 to $15 billion in secondary market spreads through September 1997. Issuance out of Russia grew sixfold, 1997 and their subsequent rise with the East Asian accounting for 70 percent of the total regional financial crisis is covered in some detail in chapter increase (figure A5.4). Russian credit has become 1.) Favorable market conditions in the first half of increasingly established in the markets, with the year led several borrowers to extend maturities. issues from the government, the cities of Moscow Most of the borrowing beyond 10 years was from and St. Petersburg, and the region of Nizhiny East Asia (before the crisis) and Latin America. Novgorod and a strong increase from private sec- Latin America accounted for the greater part of the tot banks and finance companies (paying much longer-term borrowing through the exchange of higher rates and for shorter durations than the Brady bonds for 30-year global bonds. Maturities public sector). Volumes for Poland and Turkey were lower in Europe and Central Asia, with 40 also surged ahead, with Poland's rising almost percent of the total volume at less than 5 years and threefold and Turkey's 45 percent (despite its most of the remaining at less than 10 years. The credit downgrade earlier in the year). Some coun- share of fixed-rate issues in the total volume of tries reportedly issued bonds to establish bench- bond issues rose to more than 85 percent from marks rather than to meet any pressing financing about 75 percent in 1996, reflecting the increased requirement. Moldova and Lithuania issued their demand for fixed-return instruments as spreads first public eurobonds. declined (until October). The use of floating-rate notes almost halved, with East Asia accounting for Figure A5.4 Surge in international bond volumes more than half the total floating-rate issues. from Russia, 1996-97 The average spread paid by developing coun- Billions of U.S. dollars tries in the international bond markets declined 8 from about 313 basis points in 1996 to about 260 7 basis points in 1997 (figure A5.5), reflecting a 6 modest decline in the average public sector spread 5 and a sharp fall (from 345 to 260 basis points) in 2, ~ spreads on private sector issues.' 4 0 @ Among the major borrowing regions, primary 3 t @ spreads fell most in Latin America (from 367 basis 2 points in 1996 to 285 basis points in 1997; figure I @mM A5.6). Spreads fell from close to 200 basis points O in East Asia and the Pacific to 180 basis points 1996 1997 prior to the onset of the financial crisis. By con- Sourre: Euromoney Bondwre and World Bank. trast, average spreads for Europe and Central Asia 114 Figure A5.5 Spreads on bonds Primary market spreads have dropped much more for developing countries Spreads on Brady bonds have also dropped sharply since 1996 than for industrial countries, 1994-97 Basis points over risk-free rote Basis points 400 50 1,000 < ~~~~~~Industrirl countri'esA t 6 basis-point dedine 45 800 350 -(__ight axis) Developingcountries ~~ ~> 40 600 / 300 (leftoxis) 35 400 250 30 200 1994 1995 1996 1997 January July January July December 1996 1996 1997 1997 1997 Source: Bloomberg. increased from 194 basis points in 1996 to 255 kets, in part because of the greater use of securi- basis points in 1997, as new and relatively less tized lending. Loan commitments totaled $10 bil- established borrowers with higher borrowing costs lion during the last two months of the year, down gained access to international markets. The largest more than 22 percent from the monthly rate reduction in spreads has been for borrowers with through October. ratings of below-investment grade or the lowest investment grade rating (figure A5.7). Short-term borrowing Commerdal bank loan syndications Short-term borrowing increased from about 15 percent of the total volume in 1996 to about 20 Loan commitments increased to $133 billion in percent in 1997. The increase was due largely to 1997 (figure A5.8). While commitments slowed new entrants from Europe and Central Asia, many in the last two months of the year owing to the of which could contract funds only for short dura- spread of the East Asian financial crisis, the crisis tions, and to a shift to short-term borrowing from had less impact on developing countries' access to East Asia and the Pacific as the crisis hit in the sec- the loan market than on their access to bond mar- ond half of the year. Figure A5.6 Changes in primary market spreads in private and public sectors, by region Private sector Nonsovereign public sector Basis points over risk-free rote Bosis points over risk-free rate 500 500 1 994 400 400 1994 200 94 197 997t19720 I 1996 ~~~~~~~~~~~~1 997 1 1997 1997 1995 1 994 1 996 ~ , 994 199 100 100 n 0 East Asio South Europe and Middle Eost L[an Americu East Asia South 1Europe and Middle East Lohn America and the Asia Centnal Asia and North and the ond the Asia Central Asia and North and the Pacific Africa Caribbean Pacific Africa Caribbean Note: Arrows indicate direchon of chonge. Spreuds are cululated on U.S. dollar-denominated securines. Datr for 1997 ore based on informrfion available as of mid-September. Source: Euromoney Bondwore and Woald Bank. 115 Figure A5.7 Changes in spreads by risk category, Figure A5.8 Loan syndication for developing 1994-97 countries, by region, 1990-97 Bosis points over risk-free rate Billions of U.S. dollars 500 Below-investment grade 150 Middle East v Y and North Africa 400 120 300 90 S-aarnand Cenftrl Ai 200 1991 i 60 ~~~~~~ ~ ~~~~~~~~~Africa South 100 199410AS' 1 99A Source: Euramnnes aed woiii ~~~~30~ Al A3 Baal Boo3 Bat Bo2 Ba3 Bt 1990 1991 1992 1993 1994 1 995 1996 1997 Note: Calculoaions bosed on U.S. dollor-denominated issues and Moody's rotings. Source: Euromoney Loanwroe and World BEank. Source: Euramoney Bondware and World Bank. Regional composition accounted for more than 70 percent of the region- al syndications, with most of the rest going to the Syndicated loan commitments to East Asia and the nonsovereign public sector in China (figure A5. 10). Pacific fell from $41 billion in 1996 to $39 billion Though short-term borrowing increased in the sec- in 1997, as lending to Thailand fell in the first half ond half of the year, the average maturity for the of the year and lending to Indonesia, the year was still about 4.5 years. Philippines, and Malaysia plummeted with the Loan commitments to South Asia fell to $6 bil- onset of the crisis (figure A5.9). While borrowers in lion in 1997, about 20 percent lower than the year the region could still access the loan market in the before, due primarily to a drop in syndications for third quarter (when volumes were about equal to Pakistan. The volume for India increased marginal- the second-quarter level), the last quarter saw a ly. Sub-SaharanAfrica registered marginal growth in sharp fall in loan commitments. Strong lending to its syndications. South Africa accounted for three- Indonesia in the first half of the year made it the fourths of the total, with an average maturity of leading borrower from the region for the year as a 2.5 years. Small quantities of short-term financing whole, followed by China. Private sector borrowers Figure A5.1 0 Borrower composition for loan Figure A5.9 Bond and loan volumes for ASEAN-4 syndications, 1997 countries, 1997 Percent 30 Billions of U.S. dollars ... Nonsovereign 12 25 public sector Lonos Sovereign ta 20 6 10 4 5 2Bands0 199701 199702 1997Q3 199704 4. 4 > Note: Volumes are exclusive of offlhore issuance; the ASEAN-4 countries are Indonesia, Malaysia, the Philippines and Thailand. Note: Data are preliminary esimotes. Source: Eurmoney Bondwore and Launwore and Wodd Bank. Source: Euromoney Loanware and World Bonk. 116 were put together for C6te d'Ivoire, Ghana, loans, and banks are typically less constrained in Namibia, Senegal, Zambia, and Zimbabwe. lending to unrated or speculatively rated borrowers Syndicated loan commitments to Europe and than are many bond investors. Central Asia rose to $28 billion in 1997, 85 per- In recent years, however, this gap has been nar- cent higher than in 1996, owing largely to rowed by the rise in bond issues, as more devel- increased lending to Croatia, Poland, Russia, and oping countries earned investment-grade ratings Slovakia. In contrast to other regions activity (20 had done so before the onset of the East Asian picked up as the year came to a close, with $8 bil- crisis) and market familiarity with developing lion raised in the last quarter. About 60 percent of countries increased. Thus in 1996-97 bond issues the funds were raised by the nonsovereign public averaged $85 billion, while loan volumes averaged sector. Volumes fell 40 percent for Hungary and $115 billion. The East Asian crisis has probably increased marginally for Turkey. The average interrupted this treiid, since bond market access maturity of loans was a little less than 4 years. has been substantially reduced (and several coun- Syndicated loan commitments to Latin America tries' ratings have been downgraded), while many and the Caribbean rose to $42 billion, 70 percent countries can still tap the loan market. If there is greater than in 1996, as commitments to Brazil a rapid recovery from the crisis, however, the rel- more than tripled to pass $12 billion. The $10 bil- ative increase in bond issues may continue owing lion in commitments to Mexico was the region's to the potentially larger investor base in the bond second largest after Brazil. Commitments for market and the greater flexibility of bond instru- Argentina declined but were still around $7 billion. ments in managing developing country exposure Over 95 percent of the capital went to the private (for example, through the use of put and call and nonsovereign public sector, which borrowed options). On the other hand, syndicated financ- funds with an average maturity of about 4.5 years. ing can more easily accommodate individual deals Government borrowing from the region was that are heavily tailored to the requirements of the restricted to Argentina, Colombia, and Panama. borrower. Syndications for the Middle East and North Africa surged past $9 billion in 1996 and then past Portfolio equity flows $11 billion in 1997, due primarily to Saudi Arabia and Oman, which accounted for almost three- International placements of equity issues rose from fourths of the regional volume. Egypt and $12.5 billion in 1996 to $18 billion in 1997, while Morocco returned to the loan market after being foreign investments in domestic stock markets fell absent the year before. Most of the funds were from $33 billion to $14 billion (figure A5.12 and arranged for the private sector. table A5.1). Over 85 percent of international placements were brought to the market before the Comparing the bond and syndicated loan markets Figure A5.1 1 Gross international bond issuance and loan syndications, 1990-97 Developing countries have typically relied more on Billions of U.S. dollars loan syndications than on bond issues. For exam- 250 ple, bond issues averaged $38 billion from 1993 to 1995, while loan volumes averaged $75 billion (fig- 200 ure A5.1 1). The main reasons were that bond investors lacked information about many develop- 151 ing countty borrowers and few developing coun- tries had a high enough credit rating to enable l00 institutional investors to purchase their bonds. By 50 contrast, many commercial banks have established 5 relationships with developing countty borrowers Ii (for example, in the course of financing short-term 1990 1991 1992 1993 1994 1995 1996 1997 trade facilities) that enable them to expand their Source: Euromoney Bondwrre ond Lonwire ond Wold Bink. 117 Figure A5.12 Percentage change in portfolio equity oping countries have increased, the flows remain flows from 1996 to 1997, by region concentrated in a small group of countries. Percenrt International equity issues from EastAsia and the 100 Pacific jumped to $8.5 billion in 1997, from $4.7 50 C internolionol billion in 1996, owing largely to $7.5 billion in isuesisusfo issues from China. Volumes from other East Asian C E g E Ss t ' S L countries fell, accompanied by reports of postpone- -50 ments and cancellations of highly rated issues due to the deterioration in market conditions. Issues -100 from Indonesia, Malaysia, the Philippines, and -150 Stock market Thailand almost disappeared after July. Direct -nvestment investments in regional stock markets fell to -$7 bil- c>i200 smp &SS+'* @S@M4sQ etsp @ lion (from $10 billion in 1996), with all of the s 'go . major countries except China believed to have ar @s a recorded outflows in the second half of the year. Source: Eufomoney loodwore and World Bonk. International equity placements from SouthAsia fell by half, as issues from India dropped from $1.3 spread of the East Asian crisis in late October. billion in 1996 to $0.6 billion in 1997. Investments Developing countries' share in global internation- in the region's stock markets are estimated to have al equity issues rose from about 15 percent in 1996 declined due to political uncertainties in India. to about 25 percent in 1997. Equity placements Stock markets in Pakistan and Sri Lanka rose by through privatizations accounted for about 10 more than 20 percent for the year as a whole. percent, and much of the remainder flowed Portfolio flows to Europe and CentralAsia rose through initial public offerings targeted to inter- marginally to $9 billion in 1997. The increase in national investors. Though developing country international equity issues was offset by a fall in participation in international equity markets and direct investments in local stock markets. Over 90 foreign investment in domestic markets in devel- percent of the primary market issues were in the Table A5.1 Composition of portfolio equity flows to developing countries by region, 1990-97 (millions of U.S. dollars) Region 1990 1991 1992 1993 1994 1995 1996 19970 All developed countries 3.2 7.2 11.0 45.0 32.7 32.5 46.0 32.5 International issues 0.1 4.7 5.7 11.0 18.5 8.7 12.6 Direct investment 3.1 2.5 5.3 34.0 14.2 23.8 33.4 Sub-Saharan Afiica 0.0 0.0 0.1 0.2 0.9 49 2.0 2.1 International issues 0.0 0.0 0.1 0.1 0.6 0.4 0.8 Direct investment 0.0 0.0 0.0 0.0 0.2 4.5 1.3 EastAsia and Pacific 1.7 0.7 2.1 14.6 10.1 14.7 14.4 1.5 International issues 0.0 0.0 1.3 2.6 6.8 6.3 4.7 Direct investment 1.7 0.7 0.8 12.0 3.3 8.4 9.7 SouthAsia 0.1 0.0 0.3 2.0 62 2.3 5.2 3.0 International issues 0.0 0.0 0.2 0.3 4.3 0.3 1.3 Direct investment 0.1 0.0 0.1 1.7 2.0 2.1 3.9 Eastern Europe and Central Asia 0.2 0.0 0.1 1.0 2.3 2.8 8.9 8.9 International issues 0.0 0.0 0.0 0.2 0.7 0.9 1.6 Direct investment 0.2 0.0 0.1 0.8 1.5 1.9 7.3 Latin America and Caribbean 1.1 6.2 &3 27.2 13.2 76 13.9 15.5 International issues 0.1 4.7 4.0 7.7 6.0 0.8 3.7 Direct investment 1.0 1.5 4.3 19.5 7.1 6.8 10.2 Middle East and NorthAfrica 0.0 0.0 0.0 0.0 0.1 0.2 1.6 1.5 International issues 0.0 0.0 0.0 0.0 0.0 0.0 0.4 Direct investment 0.0 0.0 0.0 0.0 0.1 0.2 1.2 a. Estimated. Source, World Bank, based on information ftom Euromoney Bondware, U.S. Treasury, central banks, Micropal, stock exchanges, and various marker reports. 118 Figure A5.13 Share of depository shares in total 1997, up from $0.6 billion in 1996. Investments in equity issuance, 1997 regional stock markets as a whole are estimated to Giobal depository have declined by 13 percent in 1997. International _ , shyres equity issues from the Middle East and North Africa equaled $0.4 billion, about the same as in 1996, all of it accounted for by global depository shares from Egypt and Lebanon. lIowever, investments in local / Americon depository stock markets declined to $1 billion in 1997, from shores $1.2 billion in 1996. 28% .Globl depository Recent initiatives in emerging equity markets shores ji l g g 32% There have been major efforts over the past two years to strengthen emerging equity markets by broadening the range of securities in primary and cAmeron depository secondary markets, improving clearance and set- shores tlement, and increasing conformance to interna- ucEoedrrWlu11% tional regulatory practices in taxation, securities lending norms, and other rules and regulations form of global depository shares and American governing securities markets. The following are depository shares from Hungary, Poland, Turkey, brief illustrative summaries of some of the initia- and Russia. Hungary ($700 million) and Poland tives undertaken in these areas: ($600 million) had the largest equity issues from In Argentina an independent clearinghouse, the region, with major activity related to issuance Argenclear, is being established that will benefit set- of global depository receipts. Other major offer- dements and securities administration. Delivery ings came from Egypt, Kazakhstan, and Slovenia. against payment will apply for all trades executed Portfolio equity flows to Latin America and the on the national securities market. Caribbean rose to $16 billion in 1997 from $14 bil- Trading in Central European securities will ben- lion in 1996, as international equity issuance efit from the establishment by the OTOB-the increased by almost 50 percent, to an estimated $5.5 derivatives section of the Vienna Stock Exchange in billion. International equity issues out of Argentina Austria-of a new derivative product line based on were four times higher in 1997 than in 1996, while the indexes of Central and Eastern European mar- those from Brazil were six times higher. Brazil kets (the CECE indexes derivatives). accounted for more than 45 percent of total issues, The Brazilian Central Bank and the Brazilian with the largest ($1.2 billion) regional equity place- Monetary Council will allow registered foreign ment coming from a private sector Brazilian bank. investors to hedge their exposure through derivatives The IFC's index of Latin American stock markets trading on the Brazilian stock exchanges, futures increased by 38 percent during the first six months exchanges, and registered over the counter markets. of the year, when investments in local stock markets Hungary is improving the infrastructure for set- surged. However, market conditions deteriorated tlement and securities administration, including with the spillover of the East Asian financial crisis the establishment of an integrated on-line system late in the year. From September to December the for futures and options clearing, settlement, and IFC index fell 12 percent, international equity issues risk management. fell more than 60 percent, and some markets (par- The Securities an.d Exchange Board of India ticularly Brazil) saw outflows of foreign investments. (SEBI) has set up a committee to prescribe a reg- Equity flows to Sub-Saharan Africa and the ulatory format for derivatives trading, and the Middle East and North Afiica remained at approxi- National Stock Exchange began trading in dema- mately the same levels as in 1996. South Africa, the terialized shares in December 1996. The Bombay only Sub-Saharan African country placing equities Stock Exchange has proposed the establishment of in international markets, issued almost $1 billion in a depository, and SEBI has allowed shares to be 119 borrowed to facilitate settlement. SEBI also is ties of securitization companies, or corporations. considering measures to simplify procedures and Also, a new securities market law is designed to to improve registration and investment by foreign improve settlements and securities administration institutional investors. The Reserve Bank (for example, by separating the clearing and set- announced guidelines enabling foreign institu- tlement services from the book-entry registry). tional investors to purchase central and state gov- In Thailand the trading of Thai Trust Fund ernment bonds, except for Treasury bills. A shares began inJune 1997. The fund acts as a nom- self-regulatory organization has been formed-the inee for foreign investors, which enables them to Indian Association of Securities Intermediaries- avoid the 49 percent limit on foreign ownership of to regulate securities market intermediaries. Thai companies. In April 1997 the Indonesian Capital Market The Istanbul Stock Exchange in Turke estab- Supervisory Board issued a decree setting require- lished an international market to provide an ments for margin trading services by stock exchange investment environment similar to that in indus- members. Another decree clarifies the permissible trial economies. Intermediary institutions will be corporate actions for custodians of securities. closely controlled, and public disclosure require- Notable progress has been achieved in the clear- ments will be fully implemented. ance, settlement, and depository facilities in International clearing organizations are also Malaysia. All listed equity issues on the Kuala active in the development of emerging markets. In Lumpur Stock Exchange have been immobilized the euromarket, the Philippine peso became the in the Central Depository System. At present 38th settlement currency at Cedel Bank. Cedel delivery is within five days of the trade and pay- Bank's Global Credit Support Service is a book- ment within seven. The Securities Commission entry real-time collateral management service authorized 15 brokerage houses to engage in secu- launched in September 1996. rities borrowing and lending activities. Foreign International organizations continue to provide companies can now apply to have their shares list- guidance to market participants on best practices. ed on the Kuala Lumpur Exchange provided that The Committee on Payment and Settlement they meet the guidelines for such listing. Systems of the central banks of the G-10 countries Mexicos achievements in pursuing effective and the International Organization of Securities securities administration are well known. Delivery Commissions have focused on mitigating the sys- against payment settlement is carried out by apply- temic risks associated with settlement systems. The ing gross settlement of securities and cash netting; recently released report "Disdosure Framework for it includes financial safeguards supported by col- Securities Settlement Systems" addressed the issue lateralized lines of credit that eliminate the princi- through enhanced transparency of clearance and pal risk in the settlement process and reduce the settlement systems. Other published documents need to use cash and credit. S.D. Indeval central- address real-time gross-settdement-systems and ized the custody and settlement of all instruments clearing arrangements for exchange-traded deriva- negotiated in the Mexican market (banking, gov- tives. The central securities depositories of 13 E.U. ernment, corporate debt, and equities) in April countries have formed the European Central 1997. A real-time connection has been established Securities Depositories Association. Its objectives between this system and the central bank payment are to develop and improve links between system. In January 1997 S.D. Indeval launched an European national central depositories, to identify electronic securities lending system. and study projects of mutual interest, and to make Peru is implementing a valuation system for recommendations for the benefit of all European fixed-income instruments and revisions to the national central depositories. Electronic Trading System (ELEX) to provide greater flexibility and ease of trading and to offer Note real-time access to information on trades. A December 1996 amendment to the income tax 1. The decline in secondary market spreads through September 1997 and their subsequent rise with the East law provides tax exemptions on income from debt Asian financial crisis is covered in some detail in chap- obligations issued by investment funds, trust equi- ter 1. APPENDIX 6 Regional trends in debt and flows Keyindicators REGIONAL TRENDS IN DEBT AND FLOWS Billions of U.S. dollors 1986 1996 1997' Total lang-term debt outstanding 1,055 1,650 1,728 G overvie w World BonkAlGA 116.6 18.6 184.4 Gloa overview Coneessienal share M%) 22.0 25.9 24. Net resource flows 73.5 281.6 300.3 Net tronsfers 8.7 174.2 187.9 Debt service/exports l 20.4 17.2 16.7 Aggregate net resource flows to developing a. Pretfminry. countries rose to an estimated $300 billion in 1997, up from $282 billion in 1996. Long-term debt outstanding, 1981-97 Net flows from private sources were up $9 Composition of private flows Billions of U.S. dollars billion, and net official development 1992 S90 billion 1997 5256 billion 2,000 finance increased by $10 billion. Official Porifolio Portfolio Regional total flows were driven by an increase in net e2ui 13°/ 1,soo nonconcessional lending by multilateral Bod institutions, from $5 billion in 1996 to I c l,000 Official $10 billion in 1997. Concessional finance ~ j 1 Loans Priv0te _ _ _continued the decline seen over the 1990s, 29L / Lo 500 to an estimated $37 billion in 1997 from 20'. $40 billion in 1996. 1987 1989 1991 1993 1995 1997 Net long-term private capital flows Foreign diredinvestment, 1 "2-97 rose slightly for the year, from $247 bil- Swources of long-term finance, 1991 lion in 1996 to $256 billion in 1997. Billions of U.S. dollars $1,728 billion Private debt flows increased strongly 150 All developing through the first three quarters of 1997, 120 then plummeted with the worldwide decline in equity prices in late October. Portfolio equity flows fell from $46 bil- 60 - - -\ 31a / : / .....lion in 1996 to $32 billion in 1997, as several countries experienced outflows t during the last few months of the year. 1992 1993 1994 1995 1996 1997 Debt indicators, 1992 and 1997 Foreign direct investment was roughly Pertent stable at $120 billion. FDI by cwntry, 1997 200 Developing countries' debt stock Billions of U.S. dollars Debtio increased only marginally in 1997, to 0 5 10 15 20 25 30 35 40 150 SO export ratio $2.2 trillion from to $2.1 trillion in 1996. Moloysia : Export receipts rose 6 percent, and the Poland 100 average ratio of debt to exports fell from Indonesia 137 percent in 1996 to 134 percent in 50 Debt 1 997. Further progress was made on debt Mexico 0 s restructuring. Nine debt reduction agree- Brozill 1992 1997 ments were concluded with commercial China . Aggregate flaws~, 1992-97 banks in 1997 reducing debt outstanding Net private flow,1997 Billions of U.S. dollars by $6.9 billion. Six low-income countries