Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Policy,Planning,and Reearch WORKING PAPERS DebtandInternallonalFinance InternationalEconomicsDepartment The WorldBank August1989 WPS254 Future Financing Needs of the HighlyIndebted Countries Ishrat Husain and Saumya Mitra What amount of external resources would be required to reverse recent investment trends and bring about modest growth in per capita incomes? Between $18 and $20 billion of net new disbursements annually. But consider the alternatives.... The Policy, Planning, and Research Complex distributes PPR Working Papers to disseminate the findings of work in progress and to enoourage the exchange of ideas among Bank staff and aU others interested in development issues. These papers carry the names of the authors, reflect only their views, and should be used and cited accordingly. The findings, iterpmetatias. and conclusions are the authors' own. They should not be attributedto the World Bank, its Board of Directors,its managemet, or any of its manber counuies. Plc,Plannlng,and Research Debtand International Finance Under base scenario assumptions, the authors estimate that the Baker 17 countries will require about $18 to $20 billion of net new disbursements annually to reverse recent investment trends and bring about modest growth in per capita incomes. The most significant shortfall is in commercial bank lending. Without an adequate burdensharing arrangement, i: is unlikely that official creditors - particularly the multilateral institutions - would be prepared to assume a disproportionately large exposure risk in these countries. Husain and Mitra conclude that with sound adjustment policies in the debtor countries, a combination of concerted new Iending, debt reduction, reflows of flight capital, and intermittant accumulation of interest arrears will be the principal means of financing. Some countries - such as Mexico, Venezuela, Nigeria, and Ecuador - need extemal financing to offset their worsening terms of trade. Others need it to restore productive investment to reasonable levels. At least some countries should be able to work their way out of the debt crisis, as their needs are feasible, so commercial bank creditors wili probably respond favorably. Others will have to balance the finance from official and private creditors. Others will be unable - even under the most stringent conditions and policies - to grow out of their present difficulties without some reduction in the stock or servicing of their debt. The lumping together of good and bad debtors is repulsing efforts of countries that should have access to voluntai, lending. The main actors should begin by abandoning the concept of a homogeneous group of 15 or 17 highly indebted countries (HICs). The contagion effect should not deter the creditors from differentiating between countries on the same continent that have managed their economies well and are close to creditworthiness from those whose economic policies and management arof questionable quality. The next logical step is to develop a cooperative framework that channels money through equitable burden sharing and promotes credit enhancemcnt, debt reduction, and other innovative financing techniques in support of growthoriented adjustment programs. The altemative is continued stagnation in the highly indebted countries - which could mean political and social unrest, a greater reluctance to maintain orderly debtor-creditor relationships, and a disruption of debt servicing even in countries that have carried out their obligations unfailingly. This paper, prepared for the conference "Dealing with the Debt Crisis," is a product of the Debt and International Finance Division, IntcrnationalEconomics Department Copies arc available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Sheilah King-Watson, room S8-029, extension 33730 (34 pages with tables). The PPR Working Paper Series disseminates the findings of work undcr way in the Bank's Policy. Planning, and Research Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The ftndings, interpretations, and conclusions in these papers do not necessarily represent official policy of the Bank. Produced at the PPR Dissemination Center i Future Financing Needs of the Highly Indebted Countries by Ishrat Husain and Saumya Mitra Table of Contents The Record 2 External Financing Needs 6 Prospects for Financing 14 Constraints on Action 18 Scope for Action 25 References 33 The external financing prospects of 17 highly indebted countries for 1989-95 fall considerably short of the requirements if a steady rate of growth is to be achieved. lending. The most significant shortfall is in commercial bank Without adequate burden sharing, it is unlikely that official creditors, particularly the multilateral institutions, will assume a disproportionately large exposure in these countries. With sound adjustment policies in the debtor countries, a combination of concerted lending, debt reduction, reflows of flight capital, and some intermittant accumulation of interest arrears will be the principal means cf financing. Some countries--such as Mexico, Venezuela, Nigeria, and Ecuador--need external finance to offset their worsening terms of trade; others need it to restore productive investment at reasonable levels. At least some countries should be able to work their way out of the debt crisis--because their resource requirements are feasible and commercial bank creditors will probably respond favorably. Others will have to balance the finance from official and private creditors. Others still will be unable, even under the most stringent conditions and most sensible policies, to grow out of their difficulties without some reduction in the stock or servicing of their debt. The lumping together of good and bad debtors is repulsing efforts of countries that should have access to voluntary lending. The main actors should begin by abandoning the concept of a homogeneous group of 15 or 17 -2- highly indebtedcountries. The contagioneffect should not deter the creditorsfrom differentiating betweencountriesin the same continentthat have managed their economieswell and are close to creditworthiness from those whose economicpoliciesand managementare of questionablequality. The next logical step is to developa cooperativeframeworkthat channelsmoney through equitableburden sharingand promotescredit enhancement,debt reduction,and other innovativefinancingtechniquesin supportof growth-oriented adjustment programs. The alternativeis continuedstagnationin the highly indebted countries--which could mean po.'itical and social unrest,a greater reluctance to maintainorderlydebtor-creditor relations,and a disruptionof debt servicingeven in countriesthat have carriedout their obligations unfailingly. The Record The net flows to highly indebted countries declined precipitiously from an annual average of $41 billionduring 1980-82to about $8 billiona year during 1986-88(table 1). 1/ In the earlieryears, privatelending provided three-fourths of resources to these countries. In the recent years, official lending and transfers account for the bulk of the resources. The picture becomes bleaker if net transfers are examined. From an aggregate net annual transfer of $16 billion in 1980-82, net transfers started to turn negative in 1982 and accelerated to an average outflow of about $29 billion a 1/ Aggregate net flows consist of net disbursements of official transfers, direct and equity investment, short-, medium-, and longterm loans from all sources to the private and public sectors in a country. Net flows, as defined in the World Debt Tables and elsewhere, consist only of net disbursements of short-, medium-, and long-term loans and exclude other components of resource flows. - 3- year in the last three years. 2/ Table 1: Highly Indebted Countries AggregateNet Resource Flows 1980-88 ($ billion) 1986-88 1983-85 1980-82 (annual average) (annualaverage) (annual average) Official transfers Private Direct Investment Official lending (net) Bilateral Multilateral IMF Private lending (net) 0.4 4.0 6.7 2.2 3.1 1.4 30.1 0.8 3.5 8.2 1.6 3.8 3.8 7.3 1.4 2.5 4.4 1.6 3.9 -1.1 -0.5 Aggregate net flows 41.2 19.8 7.8 Memo Item Aggregate Net Transfers 15.9 -20.8 -29.0 Sources: World Bank, OECD and IMF. Because commercialbanks accounted for most of the external financing to highly indebtedcountries,the dramaticdecline in their net flows from $24 billion in 1982 to -$1.0 billion in 1988 is quite revealing (table 2). In 1982 net lending by these banks financed almost 100 percent of interest payments due to them. By 1988 internallygeneratedresourcesof highly indebtedcountries financed100 percent of interestpaymentsand even a small fractionof amortization. During the same period, the domestic savings rate 2/ Aggregatenet transfersare defined as aggregatenet flows minus interestpayments. This concept differs from net transfersthat is used in the World Debt Tables and elsewhere in so far as it covers both debt-creatingas well as nondebt-creatingflows. - 4 - of these 17 countries remained almost unchanged at around 21 percent of GDP, shoving that domestic savings were financing both domestic investment and external outflows. Table 2: Highly Indebted Countries Net Lending by Commercial Banks _ ($ billion) 1982 Disbursements Principal Repayments Net lending Interest payments Net lending as Z of Interest payments 48.8 21.0 27.8 27.5 101.1% 1985 13.5 12.3 1.2 27.9 4.3% 1986 10.5 12.6 -2.1 25.6 -8.2Z 1987 10.6 10.0 0.6 21.3 2.8% 1988 11.7 12.6 -0.9 28.9 -3.1% 1/ Public and Publicly Guaranteed (PPG) Financial Markets plus Non-Guaranteed (PNG). Source: World Debt Tables, 1988, The World Bank. Private Nultilateral institutions are still the major source of positive net flows, but the volume has shrunk considerably due to repayments of large borrowings from the IMF during 1982-85 (table 3) as the Fund's resources are revolving and for temporary support only. The World Bank has stepped up its adjustment lending, and in 1985-88 it was about the only identifiable source of net new funds to these countries, with average net disbursements of $2.9 billion annually. Net lending by multilateral institutions accounted for about 29 percent of interest payments in 1988, much lower than the 326 percent in 1982. -5- Table 3: Highly Indebted Countries Net Lending by Multilateral Institutions /a ($ billiori) 1982 1985 1986 1987 1988 jisbursements 7.7 8.9 10.7 10.2 10.8 Principal Repayments 1.5 3.3 5.9 8.2 9.2 Net lending 6.2 5.6 4.8 2.0 1.6 Interest payments 1.9 3.7 4.8 5.6 5.5 Net lending as X of Interest payments 326% 151% 100% 35.7% 29.1% /a PPC Multilateral debt including Use of IMF Credit. Source: World Debt Tables, 1988, The World Bank. As the debt indicators show, most highly indebted countries are no better placed than when the debt crisis erupted six years ago--signaling the need to depart from the present approach. The stock of outstanding debt grew onethird during this period, and the debt-GNP ratio and debt-export ratio have almost doubled (table 4). Total debt servicing, despite repeated reschedulings, accounted for 43 percent of their exports in 1988, up from 37 percent in 1982. The interest to export ratio has eased only marginally and only because of rising arrears. The ratio of interest payments to exports in 1988 was still as high as 26 percent, despite the highly indebted countries' expansion of export volumes by almost 3 percent a year for the last six years. So, despite net resource transfers of 3 to 4 percent of their GDP to creditors--by compressing imports and generating a trade surplus--the highly indebted countries could not reduce their debt ratios, and they have paid a heavy price in forgone economic growth. - 6 - Table 4: Highly Indebted Countries Debt Indicators 1980 1. 2. 3. 4. 5. /a 1982 1985 1986 1987 1988 Total External Debt (USS b) 289 391 454 482 527 512 Debt-GNP ratio 33 45 59 62 63 61 Debt-export ratio 171 259 296 353 357 321 Debt-service ratio /a 26 37 39 36 36 43 Interest-export ratio /a 12 20 26 27 19 26 These ratios differ from those reported in the World Debt Tables as they include private non-guaranteed debt, short-term debt and IMP charges. Source: Author calculations. To sum up, the last six years for the highly indebted countries have not been favorable. Per capita incomes and real wages have declined, inflationary pressures have intensified, net investment rates ase abysmally low, the debt burden has ricen, and external financing flows have turned negative. This setting augurs poorly for a spe-dy, sustained recovery by these countries in the near future. The rest of this paper examines the external financing needs of these 17 countries for the next six years. How much do they need to resume modest growth in their per capita incomes? And what are the prospects for, and constraints on, this financing? External Financing Needs Two scenarios underlie the estimates here of the highly indebted countries' external financing requirements through 1995. The scenarios differ in the assumptions about the CDP growth of the highly indebted countries' trading partners and about developments in interest rates and the terms of trade. The effect of a slowdown in industrial countries' growth in 1989-90 and higher real interest rates on the external financing requirement is -7- discussed,but no effort is made to trace the effects of a lesser adjustment effort by the highly indebtedcountries.(The underlyingassumptionis that the countriesthat are not preparedto undertakestrong adjustmentefforts are not eligible for external financing support). This scenariomay seem to err on the pessimistic side, but it is intendedto provide a possiblyhigher bound measure of the external financingneeded to support growth it.s "stagflationary"environmentsimilar to that of 1979-82. The GDP growth rate of the highly indebtedcountries is projectedat 4.5 percent for 1988-95-ranging from 5.5 percent for Brazil and the Philippinesto 2 percent for Yugoslavia. The Base Scenario The projectionsunder the base scenariohave three assumptions: (1) continuedannual growth in the GDP of the highly indebtedcountries'trading partnersof 2.5 to 3 percent to 1995, enablingannual increasesin the highly indebtedcountries'export volumes of about 5 percent; (2) some recovery in the terms of trade from the historic lows of 1986 and 1987; and (3) nominal LIBORs of 8.5 percent until 1991, then gradually falling to 7 percent in 1995. The aggregateprojectionsare comprisedof individualcountries (given the common externalenvironmentfor growth, prices, and interestrates) tuat utilize the IBRD's Revised Minimum StandardModel (RMSM). This model contains a close relationshipbetween the rate of investmentwithin an economy and the rate of growth of output. This simple specificationpermits the use of this approachacross countries. It is widely familiarto individualcountry economistsand is the central tool in formulatingprojectionsin the World Bank. It has the disadvantageof failing to capture other (important) elements that influencegrowth, but it is judged that for medium-termanalysis the robustnessis sufficientlygreat to project financingrequirements required to support growth. Once the investmentrequirementsare identified (throughincrementalcapital-outputratios) and the domestic savings performanceare specitied,the externalfinancingrequirementis obtained as a residual. In most cases an improvementin the domesticmarginal savings effort has been postulated(presumedto occur because of policy improvements as adjustmentprogramsproceed); this helps to avoid a possible overstatement of the external financingrequit went. The country projectionsalso hinge on a fiscal policy correctionto realize high nmarginalsavings rates for domestic growth. Implicit in the projectionsis that externalflows are requiredto finance the investmentand importsneeded for a successfulexport-ledgrowth strategy. In all of the above cases, the projectionsare based on the successfulimplementationof strong adjustmentprograms. The improvementsin export performanceare striking,with the debtto-exportratio falling substantiallyby 1995. The dangers to this strategy are twofold. First, the domestic adjustmentmay be insufficientto redirect the requiredresources to the export and investmentsectors. Second, the external financingmay fall short of what is required. Later in this paper the likelihoodof the external financingrequireme.-is assessedat some length. The requirednet externallong-termflows (net disbursements)are projected to rise sharply from $1 billion in 1988 to a range between $14 and - 9 - 17 billion a year during 1989-95 for the highly indebtedcountries (table 5). Table 5: HICs EXTERNAL FINANCINC,1982-95 - Base Scenario 1982 1983-86 1987 1988 1989-91 1992-95 <--- ----(Annualaverages in $ billion)--------> Current account deficit Addition to reserves /a Financingrequirement Non debt creating flows /b Net LT flows IBRD/IDA Other multilateral Bilateral Total Official Private /c Short term flows Net IMF 51.3 -21.3 30.0 8.7 7.9 8.2 5.4 9.7 -5.5 14.1 17.6 2.7 16.0 4.0 20.0 17.0 6.0 23.0 7.5 4.5 5.6 5.3 6.0 7.0 36.9 1.9 4.1 2.0 8.0 28.9 13.5 2.9 4.3 1.3 8.5 5.1 6.2 2.3 1.1 1.9 5.3 0.9 1.0 1.5 1.6 2.1 5.2 -4.2 14.5 3.0 1.0 1.5 5.5 9.1 17,0 1.5 1.0 1.5 4.0 12.0 -6.7 7.1 -3.7 -0.5 -1.0 - - 58 35 46 37 -16.6 2.2 2.8 -1.3 0.1 Memo Item: Net flows as percent of interestpaymentsto creditor Official 245 166 51 62 Private 103 19 4 2 /a IncludesStatisticalDiscrepancy. /b Principallyforeign investmentand official transfers. /c Includesfinancingvia arrears. Source: Authors own calculations. Given the identifiedsources of financingfrom the official sector,the expectedcontributionfrom commercialbanks required to cover the external financingrequirement,which amounted to a net outflow of $4 billion in 1988, - 10 - amounts to $9 billion a year in 1989-91 and $12 billion a year in 1992-95. Lending on such a scale from commercialsources is fraught with uncertainty. It implies a significantincreasein exposure and in the implicitcapitalizationof interest.3/ The implicitinterestcapitalization was only 19 percent in 1983-86;and net flows were only marginallypositive in 1987. 4/ The projectionsimply interestcapitalizationsof 35 percent for 1989-91 and 37 percent for 1992-95,clearly large. The externalpayments positions,includingreserve holdings, of these countriesdiffer considerably,as does the degree of macroeconomicor structuraladjustmentof their economies. For some countries(Chile, Uruguay) a return to voluntaryaccess to private markets is within grasp if adjustment policies are sustained. For others the financingrequirementsfrom the commericalbanks remain high, even under tough adjustmentprograms. For another group of countriesexceptionallyhigh net flows are requiredfrom both commercialbanks and officialcreditors. And a few need heavy support from official creditors. An examinationof the projectionsfor individualcountries (that underlie the aggregate projectionsin table 5) show that the 17 highly /3 /4 A measure of net transfer,implicit interestcapitalizationis the ratio of net lending by a creditor (say, private banks) to the interestpaymentsdue to it in any period. The aggregate figure masks large differences. Through concerted lending packages,generally supportedby Fund arrangements,certain countries (Argentina,Brazil, Chile) were able to obtain increases in commercialbank exposure, but others suffereda withdrawalof commercialbank lending. - 11 - indebtedcountriescan be divided roughly into three categories. Four countries that account for 10 percent of total highly indebtedcountry debt require external financingover the period to 1995 that is modest in absolute terms or in light of recent history; for these few countriesa strategyof growth-orientedadjustmentbased mainly on largelyvoluntary new-money packages is unlikely to be constrainedby the external financing availability. Five more countries account for 35 percent of total highly indebtedcountry debt that is large in absoluteterms or in light of recent history and require externalfinancial flows from officialor private sources. The final categoryhas holders of over half of total highly indebtedcountry debt. They face a heavy financingrequirement,in all cases except one from the private sector; the exception is the Philippines,which has the option of an officiallysponsored"MarshallAid." 15. For severalhighly indebtedcountriec,especiallyoknesthat are not major debtors, the limits of the debt strategyprevailingin 1988 were clear to the market, making it extremely difficultto arrange financing packages. When financingpackageshave been negotiated,they have requiredgreatly increasedparticipationby officialcreditors. For several of these countries prolongedinterestarrears have become a large source of financing from commercialbanks. The forced burden sharing by banks in new lending has meant a shift toward a unilateralapproach--hindering other (trade-related)flows and rising costs and uncertainties. - 12 - The Low Scenario Growth in the industrialworld in 1988 and thus far in 1989 has exceeded initial forecasts,with the United States, Germany, and Japan in particularexperiencinga strong rise in domestic demand. But there is some risk of a marked slowdownin 1990 and 1991, to less than the 2.75 percent of GDP growth assumption. Such a temporary slowdownwould, besides lowering the demand for imports from the highly indebtedcountries,mean less improvement in the highly indebtedcountries terms of trade than that projected in the base scenario. This slowdownis triggeredby an officiallyinduced U:se in nominal and real interestrates of two percentagepoints compared to base case projectionson the assumptionthat, as in early 1989, authoritiesreact to acceleratedinflation. This rise in interestrates would not only raise the cost of the highly indebtedcountries'floatingrate debt. It would also-through reduced inventoriesand a fall-offin domesticdemand in industrial countries--dampenthe highly indebtedcountries'growth and tend to soften their terms of trade. In the simulationsin table 6 it is assumed that growth in highly indebtedcountries'export volumes is dampened by two percentage points in 1989 and one percentagepoint in 1990 and that their terms of trade worsen by two percentagepoints in 1989-90. These adverse shocks are temporary,recoveringto the parametersin the base scenarioin 1992. - 13 - Table 6: Deteriorationin highly indebtedcountriesexternal financing1989-95 caused by a weaker externalenvironment 1989-91 1992-95 (Annual averages in $ billions) Change in current account balance (- =weakening)-12 Change in net flows (+ = increase) Official Private 8 2 6 Memo Items: Total net flows as percent of interestpayments to creditora Official 75 Private 55 -8 6 2 4 66 66 Source: Author's own calculations. This scenarioshows how much the financingrequirementsof the highly indebtedcountrieswould increase. The net annual flaws requiredrise to $30 billion for 1989-91 and $24 billion for 1992-95 (tables 5 and 6). Even on the assumptionsthat the multilateraland officiallenders step up their lending, that the IBRD lends amounts equivalentto the maximum exposure permitted,and that other official lenders greatly raise the amounts extendedover the base case, the contributionsof the private sector would have to be greatly enlarged. The requirednet flows between 1989 and 1995 would amount to high rates of interestcapitalizationby banks. In light of both recent history and the current lending attitudesof the commercialbanks, such financingare large. None of the 17 highly indebtedcountriesis likely to obtain the - 14 - additionalexternal financingrequired if there is a 1989-90 recession. Brazil'snew money requirementswould increaseby $3.5 billion a year during 1989-91,with interestcapitalizationof more than 50 percent overall and about 80 percent from private sources. Chile would require commercial-bank interestcapitalizationof about 70 percent. Mexico is perhaps a little more robust but its commercial-bankinterestcapitalizationwould still be about 40 percent. Morocco and Nigeria would need extremelyheavy "new money" support in 1989-91. PROSPECTSFOR FINANCING Private Direct Investment Private direct investmentflows to the highly indebtedcountrieshave been slowing significantlysince the beginningof the decade. But because of higher growth, better po;lcies,an improvedinvestmentclimate,and the sometimesstrong incentivesof debt-equityconversionprograms,this trend may be reversed. These flows could recover stronglyif countries implement programsthat improve incentivesto the private sector, liberalizetrade and investmentregimes, privatizepublic activitiesand reform and develop financialmarkets. Such reforms are central to the objectivesof World Bank assistance,and the expansionof this assistancebeyond levels currently planned can have favorableeffects on the flow of private equity capital as well. At a minimum, it would seem reasonableto expect net direct investment flows of $4-5 billion a year (includingdebt-equityconversions)over the next few years. The same policy measures--withhigher growth rates and specific - 15 - also facilitatea return of steps to encourage capital repatriation--should residents'hcldingsabroad, or at least reduce the capital outflow from these countries,now about $9 billion a year. Net Official Transfers Net oiZicial transfers (includingbilateral aid) to the highly indebtedcountrieshave averaged $1-2 billion a year and seem likely to be sustainedat this level. Except for Bolivia (and perhaps Nigeria, Jamaica, Philippinesand Costa Rica) there is little possibilitythat official aid flows to the 17 highly indebtedcountriesas a group will be stepped up. The continuingbudget deficits in industrialcountries,the growing attention to Sub-SaharanAfrican countries,and the relativelyhigh level of per capita income in HICs make it difficult to expect any significantchange in the attitudeof official donors. Bilateral lending Bilateral net lending to the highly indebtedcountries,including export credits and mixed credits,has also averagedabout $2 billion a year. Although there will be some variation in the availabilityof these credits, depending on the country situationand the politicaland strategic attractivenessto creditorgovernments,the aggregate sums are unlikelyto be substantiallyhigher than before. The net flows from export credit agencies have been disappointingin recent years--reflectingtheir difficult financial situationsand the decline in borrowingcountries'demand for importedcapital - 16 - goods and other imports. Net new lending from these agencies has picked up somewhat,but more net flows from them in the next few years will be needed, as well as further financialrelief through reschedulingand refinancingof interest.5/ The exception to this pattern is Japan, which has been playing a prominentrole. The recent actions by Japan's EXIM Bank to make large sums of untied funds available suggest the way in which the lending policiesof export credit agencies should be adapted to suit the highly indebtedcountries' special circumstances. A combinationof some increase in Official Development Assistance(ODA) and untied loans by the EXIM Bank of Japan may boost Japan's net lending to the highly indebtedcountriesas much as $1 billion a year in the near term. But the reallocationof concessionallending to Sub-Saharan Africa and export credits to more creditworthycountries in Asia by other bilateral creditorsmay offset some of these gains. It is therefore safe to assume an annual net inflow of $1.5 billion from bilateral sources. MultilateralLending During 1986-88,the share of the World Bank and the other multilateraldevelopmentbanks (MDBs) was 50 percent of total medium- and long-termflows, up from roughly 12 percent in 1980-82. If adjustment programsand economic performancedevelop as assumed,net flows from the IBRD /5 Some $61 billion in debt relief was providedby Paris Club members to developingcountriesduring 1983-87,comparedwith only $19 billion in 1976-82. - 17 - to the highly indebtedcountriescan be expectedto be about $3 billion a year over the next five years. A further $1 billion in net flows is likely to be available to these countriesfrom the other MultilateralDevelopmentB=nks (MDBs) at a minimum. CommercialBanks Commercialbanks continue to lend to a handful of creditworthy borrowers (mainly in Asia), and some concerted loans are still arranged for such highly indebtedcountriesas Brazil and Mexico. But loan charge-offs, debt conversions,swaps, buybacks,and sales have all contributedto the decliningbank claims on highly indebtedcountries. An estimated $12.3 billion of commercialbank debt was taken off the books in the first half of 1988 alone, and the process is gaining momentum. This outcome is a major ground for pessimismabout securingadequate net flows from commercial banks. It is not likely that the commercialbanks' attitudes toward the indebtedcountrieswill be reversed soon. True, there will be reschedulings of principal,reduction in spreads and fees, and some concertednew-money packages,mainly to finance retroactiveinterestarrears.6/ But no support are likely before 1995. significantnew flows for balance-of-payments /6 Since the inceptionof the debt crisis, commercialbanks have restructuredabout $300 billion in outstandingprincipal;they have assembled $40 billion of new-money packagesand have arranged special short-termcredit lines on the order of $36 billion. The figures for new financing from commercialsources in individualcountry situationsthrough new-moneypackages do not take account of the outflows to commercialcreditors in some countries. Hence, the total new money that must be raised in the market could be somewhat larger than the aggregatenet financingrequirements. - 18 - Constraintson Action Debtor AdjustmentEfforts The highly indebtedcountries and their creditors disagreeabout the nature, extent, and outcome of adjustmentefforts. Most creditorsbelieve that the economicplight of the highly indebtedcountrieshas much to do with their weak commitmentto reforms and with their policy slippages. The highly indebtedcountrieshave found it easier to cut investmentthan to tax or reduce the consumptionof powerfulhigh-incomegroups. The stop-go cycle of economic policieshas also eroded the credibilityof adjustmentpolicies. But politicalleaders of the debtor countries,particularlyin the new democracies,argue that a decade of continuousdecline in per capita income and consumptionhas generatedenormous social and political tensionsand stretchedthe feasibilityof these reforms to the maximum. Further reforms to pay off their debts to foreigners--reforms that impose even greater sacrifice on the population--willpose real constraints. Even in the best of circumstances,structuralreforms meet formidablepoliticalobstacles across the board. The political problem is that the costs of adjustmentare immediate while the diffuse benefitsmaterializeonly gradually,far beyond the horizon of politicalleaders. George Schultz,when U.S. Secretaryof State, observed: "If the immediatevisible impact of changes in economic policy is hardshipat home to keep up service on the debt, then that debt service can - 19 - have the effect of a marginal tax on economicreform. Any effort t xed at 100 percent,or at only excessivelyhigh rates will be diocouragedand become Departmentof State (19881). politicallydifficult to sustain (u.s. Is it feasibleto achieve some minimal acceptablegrowth in output and consumptionand simultaneouslyimprove creditworthiness?Most of the highly indebtedcountriescannot restore creditworthinessor gain access to capital markets even with strong adjustmentpolicies. Moreover, the sharp cuts in investment,maintenanceexpenditure,and imported inputs disrupt the supply responsesto better relativeprices and make output gains--normally assumed as a result of successfuladjustmentprogram--difficult.Another problem is that, in assuming private-sectordebt, governmentshave had to increasetheir budgetary outlays on interestpayments. Financing the growing public deficitsby internalborrowing or expansionof money supply has exacerbatedthe inflationarypressures,generatedhigh real domestic interest rates, and added to the interestbill. ExternalEnvironment The ability of highly indebtedcountries to grow out of debt also depends on the markets for their main export products. The major debtors increasedthe volume of their exports substantiallysince 1982, but terms-oftrade Losses eroded the dollar gains. The value of their exports in nominal terms remains unchanged--at$150 billion a year. The GDP growth rates of the C-7 countries have be 'nreasonablyhigh, - 20 - but the spilloverto the highly indebtedcountrieshas been modest. The situationmay get worse. Instead of averaging 3 percent or more, the major industrialeconomiesare expectedto grow at 2.5 percent annually for some time. And if the United States is serious about reducing its huge trade deficit, the chances of its taking a large increase in exports from the highly indebtedcountriesare slim. Nor is it obvious that Japan and Garmany, always reluctant to stimulateand open their economies,will absorb substantially more exports from the highly indebtedcountries. Their combinedGDP exceeds the U.S. GDP by more than 60 percent,but they absorb far fewer exports of manufacturedgoods from Latin America. Another potent threat to the highly indebtedcountries is the growing market share of the newly industrializing economies in Asia. The price prospectsof debtor countries'major commoditiesare also unpromising. Real prices for most primary commoditieswill remain depressed because of structuraland cyclicaldemand factors. And the heavy subsidies the industrialcountriesgive to their relativelyhigh-cost producerswill continueto depress basic food prices. Another uncertaintyis the movement of real interestrates, which remain high. Almost two-thirdsof the highly indebtedcountries'debt is tied to variable interestrates, opening them to interestrate shocks. Every percentagepoint rise will add $3.5 billion to their debt servicing,creating larger demand for new capital inflows or debt reduction. The policies of the United States and its major trading partnerswill determine the interestrate, exchange rates, and the debt servicingcapacityof the highly indebted - 21 - countries. For example,a fall of the U.S. dollar could , ive interestrates way up and push the U.S. economy into a recession. A combinationof slower OECD growth, collapsingcommodity prices, and rising real interestrates could wreck the highly indebtedcountries'economic prospects. CommercialFinance The total likely to be availablefrom all known and identified sources of finance--privateinvestment,official transfers,bilateral, multilateral,and export credits--willnot exceed $6-7 billion a year. The financinggap during 1989-95 remains about $9-10 billion. Private commercial banks traditionallysuppliedthe bulk of financing to the highly indebted countries,but the constraintson their lending have intensifiedin recent years for five reasons. First, skepticismis growing about the near-term prospectsfor improvedcreditworthinessin the debtor countries. Bankets know that restoringcreditworthiness will be a long and uneven process. They know, too, that the domestic politicaldifficultiesof implementingreform programs reduce the likelihoodof success. To make matters worse, interestarrears are increasing,some because of unilateralmoratoria on debt service payments. Second, banks face intensifiedpressures,both regulatoryand competitive,to strengthentheir balance sheets. The regulatoryauthorities - 22 - in many industrialcountriesare adoptingmore conservativeguidelines, presentingbanks with stricterand more comprehensivecapital requirements. As a result, they are realigningtheir strategiesto strengthentheir capital base, restrainasset growth, focus on fee-basedactivities,cut exposure to developingcountries,and improve profitability. Third, the effect of depressed secondarymarket prices for developing country loans has damaged the share prices of banks that have large exposures. Prices in the secondarymarket suggest a reserve level of about 50 percent. If the reservesare increasedthat high, the regulatorycapital positionof the U.S. money-centerbanks will get worse, probably costing them substantiallosses. Moreover, banks face the prospectof additional provisioningon new lending,making participationin concerted new money packagesexpensiveand at best marginallyprofitable. Inter-country differencesmake matters worse. Even though the loan-lossprovisionsby the U.S. banks appear to be in line with other major banking systems, the burden and risk for the U.S. banks is greater. They have a higher concentrationof lending to highly indebtedcountriesand much weaker capitalization. Banks in Japan and the United States, unlike their Europeancounterparts,receive no major tax benefits for creatingreserves. The depreciationof the U.S. dollar has also reduced the relativeexposure of non-U.S. banks in developing countries. The reluctanceof U.S. money-centerbanks and Japanese banks to increasetheir exposure to highly indebtedcountries is thus understandable. Fourth, the long-termbusiness interestsof commercialbanks are once again diverging. The universal (or critical-mass)participationin concerted - 23 - lending--themodus operandiin the early years of the debt crisis--wasmade possibleby the banks' shared interestin protectingthe international financialsystemand in buying time to reduce exposures. Except for the large internationalbanks, this phase has ended, and the underlyingdifferencesin exposuresand businessstrategiesare determiningthe decisionsof individual banks to participatein lendingnew money. Many regionaland small-exposure banks are redirectingtheir lendingto traditionaldomesticand trade financing. Even among the larger banks, there are differencesin financial interestsand objectives. Those with multinationalcorporateclients--and,in some cases, with significantdomesticbankingoperationsin developing countries--canbe expectedto maintaindirect interestin improvingthe liquidityof specificdebtor countries. But other banks are trying to leave the debt-restructuring process,even at the cost of significantwrite-downs. Fifth, sustainingthe concertednew-moneyprocesswill require adaptationsand new approacheswhose successis as yet uncertain. Restructuringagreementshave consolidated most of (formerlyindependent)debt obligationsand establisheda uniformlegal standingfor all commercial creditors'claims. The sharingclauses in these agreementsmake it possible, however,for some creditorsto collectfull interestdue on their outstanding claims without contributingto the fresh-moneyloans that help provide the resourcesto pay that interest. This free rider (or recalcitrantbank) problemhas become much greateras the new-moneyparticipation rates fall./7 /7 In Colombia,for instance,only 112 of about 175 creditorbanks participatedin the January 1988 new-morney facility,which had to be reducedto completethe deal. - 24 - As a result, the ability and willingnessof the larger banks to continue to lend is subject to additionalstress. Exit instrumentsthus far have not sustained the burden sharing concept that was part of the original restructuring,for reasons that have perhaps less to do with the design of the instrumentthan with the extremel/ complex legal issuea surroundingattempts to close off the free-rideropportunity. It will thus be extremelydifficult--perhapsimpossible--togenerate aggregatenet flows for the highly indebtedcountries in the needed amounts exclusivelythrough the concertednew-money approach. Banks still hold about two-thirdsof total medium- and long-termclaims on the highly indebted countries,but the pressures not to lend will grow, forcing selectivityand strong reluctanceto accept exposure increasesthat in other circumstances might look reasonable(2 to 3 percent on average, roughly a quarter to a third of interestdue). The number of banks participatingin new-money packages is likely to narrow further,as is the number of countries for which such financingcan be arranged. Banks will likely concentrateany new exposure on countrieswhere their financialand long-termbusiness interestsare substantialand where the prospects for a successfulworkout are reasonably good. Smaller countries,particularlythose with weak adjustmentprograms, will continueto find it difficult to arrange concerted surport (though not necessarilycontinued restructuringsof principal). - 25 - SCOPE FOR ACTION The constraintsto mobilizingadequate flows of new money, especially from the commercialbanks, are likely to remain stron,. Part of the reason for this is that when a systemic paymentsproblem arises, markets tend to experience"revulsion,"and credit volumes become paralyzedby a neighborhood problem--goodand bad debtors are lumped together(Eichengreen1989). Colombiahas not had a debt problem as such in the 1980s, but its efforts to secure truly voluntary syndicatedloans has neverthelessbeen severely frustrated. If the net flows from commercialbanks are unlikely to rise significantly,what other means would meet the financingrequirementsof this group of countries? There is no clear-cut or simple answer. Needless to say, the debtor countriesthemselveshave to continuetaking primary responsibility for their fate through further adjustment,however painful and politically difficult it may be in the short term. The more flexibleand responsivetheir economies are, the more resilientthey become in facing unexpectedeconomic shocks and in improvingtheir creditwort%iness. Favorableeconomic policies and good economicmanagementwill attract both new project and trade financing,multilaterallending,export credits,and direct equity investment. So, sound policiesare essential for capital inflows anl for other alternativefinancing. The other possiblemeans for filling the financinggaps include reduction in the stock of debt or debt servicing followedby reflows of flight capital. And in cases where the debt reduction achieved is not sufficientto fill in the gap and the country is pursuing - 26 - sound adjustmentpolicies, consensualor sanctionedaccumulationof interest arrears could supplement. Debt Reduction Why is debt reduction then a preferredoption for the highly indebted countriesto fill in their external financingrequirementsin pursuit of a reasonablelevel of growth? The net negative transfershave lowered the investmentratios, which have reduced output growth and exports and in turn their capacityto fully service their debt. The reversal of the net negative transferscan be achieved by increasingnew flows of money but as we have examinedearlier Lhe probabilityof new money flows to highly indebted countries in the next five years or so, especiallyfrom the commercialbanks, is very limited. The other option is to reduce debt or debt servicing. The general argument in favor of debt reduction rather than new money is that many debtor countrieshave been unable to return to growth in the presence of very large debts. One reason for the persistenceof slow growth is that debt overhangacts as a tax on increases in current and future income. If for example,a country is able to increaseits exports as a result of policy reforms or more investment,a large share of the benefits is likely to accrue to creditorsrather than to the country itself. This will depress the returns to the country from fixed capital investmentand thereby weaken the incentive to invest even if finance is available. By reducing the creditors'share of the benefitsfrom the adoption of adjustmentpoliciesand by reducing the uncertaintysurroundingthe sustainabilityof adjustment,debt reduction encouragesinvestmentand the incentiveto implementbetter policieswhich, in - 27 - turn, boost exports and debt servicing capacity. In this case, debt reduction could make both debtors and creditorsbetter off. The IMF (1989) has found supportingempiricalevidence for the debtoverhanghypothesis. When external financialflows dried up after 1981, debtors were forced to run trade surplusesin order to service their debt. Adjustmentin the trade balance can be achieved by reducingthe consumptionoutput ratio or the investment-output ratio. In the highly indebted countries,the consumption-outputratio has not only failed to decline in proportionto the investment-output ratio but has actually risen on average between 1982-87. The second piece of evidenceto the existenceof these disincentives is the contrast in the behaviorof investment-output ratio between groups of countrieswith and without recent debt-servicingproblems. For countrieswith debt-servicingproblems, the average investmentratio fell from about 25 percent in 1980-81 to about 18.5 percent in 1987. By contrast, the group of countrieswithout debt-servicingproblemsexperiencedvery little change in its investment-outputratio, from about 28 percent in 1980-81 to 27.25 percent in 1987. A third indicationof this phenomenonis providedby changes in the compositionof investmentin the indebtedcountries. If debt service depends on overall macroeconomicperformance,disincentivesshould apply to both the private and public sectors. Empiricalestimates of disaggregatedbehaviorof investmentin a sample of debt-problemcountries show that both public and - 28 - private investmentratios drop from 1981 to 1984. The cumulativethrust of the evidence reviewed in the IMF study, that is the behaviorof the consumption-output ratio in the debt problem countries, the contrastbetween the investment-ontputratio in these countriesand in countrieswithout debt problems, and the sectoralevolution of investment between the public and private sectors are all broadly consistentwith the presenceof debt-overhangdisincentives. Debt-reductionmeasures can be divided in several categories: (a) exchangeof foreign debt against domesticasset (debt-equityconversion);(b) exchangeof foreign debt against another foreign asset at a discount; (c) debt buybacks,and (d) debt-servicingreduction through reduced interest rates. In the first approach, the original lender (or another party) has bought the debt at a discount on a secondarymarket, takes a loan to the country,and obtains in exchange local currency for the full face value of the officialexchange rate. This local currency is used for purchaseof local equity, relending,and so on. The advantage for lenders is that they find a use for their loans at a face value while the advantage to the borrower is reduced debt. But if the assets acquiredby the creditorare private and the debt is held by the government,which is the case in most highly indebted countries,this type of conversioncontributesto acceleratinginflationand higher real interestrates if the governmentis already facing fiscal deficits. In that case, to redeem the foreign debt, the government increases its internalborrowing or prints money. - 29 - The exchange of debt for other debt instrumentsat a discount such as bonds and exit bonds require that the new instrumentis a more securedasset, and the probabilityof the borrowerfully servicing this asset is larger than the old debt. This usually requires that the new asset is backed by collateralfor the principal or a guarantee for interestor both. To purchase the collateralthe country must have excess reserves that it can use or borrow, or obtain the resources from other sources. In the Mexican deal, althoughMexico provided collateralfor the ultimate repaymentof principai (zero coupon U.S. bonds it had bought for cash) the outcome was disappointing since the additionalvalue placed by creditorson the new instrumentdid not exceed the present value of the collateral. In the third type of operation,a country buys back its debt at a discount for cash. Bolivia and Chile are two such examples. Countries in debt difficultiesrarely have much cash lying about: the Bolivianoperation had to be financedby aid agencies. Chile was able to accumulatesome reservesin the Copper StabilizationFund due to unanticipatedincreasesin price of copper. In both cases, there were exceptionalcircumstancesthat facilitatedthe debt buybacks. In the literature,pursuasivearguments support this. The fourth type of debt reduction,that is reduced interestrates on exisitingdebt instruments,has not so far been put into practice. The reduced interestrate option is attractivefrom the debtors'viewpointas they receive a substantialrelief in cash flow. As the reduced interestrate is -30 - tied to economic performance,the fear of moral hazard is also, to a large extent, neutralized. A case-by-caseapproach to interestrate reduction -.. 3otiated in the frameworkof an agreed structuraladjustmentprogram can make a significantdent if the accountingand tax rules are modified to strengthenthe incentivesfor the commercialbanks. The present tax and accountingrules do not provide any incentivefor the banks to agree to reduction in interestrates. A variant of reduced interestrates that is attractiveis analogous to equity warrant attached to bond issues. if the country'smajor export commoditiesregisteran upward swing relativeto the threshold,the creditors participatein the gains proportionatelyand the interestrate reductiongranted earlier is brought closer to the market rate in accordancewith the coL.Ltry's payment capacity. It has also been suggestedthat reflows of flight capital the highly indebted countriescan also fill their externalfinancinggap. Estimates of flight capital vary greatly but the pool may be as large as $300 billion (Morgan Guaranty 1988). The conditionsunder which flight capital find its way back are not necessarilyconducive to the financialstabilityof the country. Those willing to bring ba(k their flight capital require higher risk premiumswhich result in high economy-widereal interestrates. They choose to keep the assets in a highly liquid form and do not always invest in the expansionof productivecapacity. And at the first possible indicationof political or economic uncertainty,these financialassets leave the shores of the country,accentuatingfinancialdestabilization. While foreign debt is guaranteedby the debtor'sgovernmentand thus safe from default risk, - 31 - domestic investmentsby residents face expropriationrisk (Khan and Haque 1985). This gives developing-countryresidentsfurther incentivesto place their funds in riskless savings accountsabroad, and it gives foreignersa comparativeadvantage in lending for domestic investment. If this expropriationasymmetry persists and uncertaineconomic policy environment persists there is little hope for substantialreflows of flight capital in the near term. Unless the country is able to eliminatedebt overhang,reduce economicuncertainty,and stabilizethe economicenvironmentthere is very little hope for significantinflows of flight capital. This finally leads to the question that has been the stumblingblock for the success of voluntar, debt reduction,that is free-ridingbanks that do not wish to participatein voluntary debt reductionhoping others will. These banks hold out expecting that the value of their claims will be strengthened and pushed back to par value by the exit of other creditors. The regulatory authoritiescan play an importantrole in this regard by treating the nonparticipantsdifferentiallyand harshly compared to those participatingin the debt reduction. The key to the future directionof debt reduction therefore lies in the way the sticks and carrots are allocated. The Brady Initiative The Brady Initiative,announced in March 1989, makes a significant departure from the existing debt strategyby allowing the use of financial resources of the InternationalFinancial Institutions(IFIs), that is, the World Bank and the IMF in support of debt and debt-servicereduction in - 32 - countries pursuing stronL adjustmentpolicies. Although this is a step in the right direction,the voluntary choice given to the commercialbanks for participationunder the existing ground rules and the limited resources likely to be available from the IFIs have raised some skepticismas to whether the size of debt reduction would be sufficientto offset the debt overhang. The voluntarymarket-basedcase-by-caseapproach is an adequate response if the tax, regulatory,and accountingrules provide incentivesto commercialbanks for participationand penalties for free-ridersor nonparticipants. The existing rules and precedentswere set to facilitatenew-money flows. Debt reduction,beyond a certain limited scale, would impose additionalfinancialcosts to the commercialbanks. The credit enhancementor the resources for debt buybacksprovided by IFIs would not be sufficientto make a significantdent in larger debtor countries. Thus the Brady Initiative may be able to assist the smaller countries,but it appears that the expectationsthat have been aroused about its impact in the larger countries are at present exaggerated. Ishrat Husain is Chief of the Debt and InternationalFinance Division,and Saumya Mitra is an economist in the Risk Managementand FinancialPolicy Department. The findings, interpretations, and conclusionsin this paper are the personalviews of the authors. They should not be attributedto the World Bank, its Board of Directors,its management,or any of its member countries. The authors are gratefulto Don Hanna for his assistancein the preparationof this study. - 33 - REFERENCES Eichengreen,Barry. 1989. "The U.S. Capital Market and Foreign Lending, 1920-1955." In Jeffrey D. Sachs, ed. DevelopingCountry Debt and Economic Performance. Vol. I. Chicago: Universityof Chicago Press. InternationalMonetary Fund. 1989. World EconomicOutlook. Supplementary note no. 1. Washington,D.C. (April). Khan, Mohsin, and Nadeem Haque. 1985. "ForeignBorrowingand Capital Flight." IMF Staff Papers 32(4):606-28. Morgan Guaranty Trust Company of New York. 1988. "LDC Debt Reduction: A CriticalAppraisal." World FinancialMarkets (7):1-12. U.S. Departmentof State. (date]. The Inter-AmericanSystem: Into the Next Century. Currenty Policy Series No. 1126. Washington,D.C.: GPO. PPR WorkingPaperSeries Contact IWa WPS240 The PublicRole in Private Post-SecondaryEducation: A Reviewof Issuesand Options Aughor Ake Blomqvist EmmanuelJimenez Date forpaper August1989 A. Bhalla 61059 July 1989 C. Cristobal 33640 WPS244 Do CaribbeanExportersPay Higher AlexanderYeats Freight Costs? July 1989 J. Epps 33710 WPS245 Developinga Partnershipof Indigenous and Land Use Peoples,Conservationists, PeterPoole Plannersin LatinAmerica August1989 S. Davis 38622 July 1989 S. Ainsworth 31091 August 1989 M. Raggambi 61696 August1989 S. King-Watson 33730 August 1989 S. King-Watson 33730 Ana-MariaArriagada WPS241 The Effectof Job Trainingon and PeruvianWomen'sEmployment Wages WPS242 A Multi-LevelModel of School Effectiveness MarlaineLockheed in a DevelopingCountry NicholasLongford WPS243 AvertingFinancialCrisis- Kuwait FawziHamadAl-Sultan WPS246 Causesof AdultDeaths in Developing RichardHayes Countries: A Reviewof Data and Thierry Mertens Methods GeraldineLockett LauraRodrigues WPS247 MacroeconomicPoliciesfor Structural CarlosA. Rodriguez Adjustment WPS248 PrivateInvestment,GovernmentPolicyand MansoorDailami ForeignCapitalin Zimbabwe MichaelWalton WPS249 The Determinantsof HospitalCosts: RicardoBitran-Dicowsky An Analysisof Ethiopia David W. Dunlop WilliamR. Cline WPS250 The BakerPlan: Progress, Shortcomings,and FutureEvolution WPS251 Patents, AppropriateTechnology, and North-SouthTrrade IshacDiwan Dani Rodrik WPS252 Do the SecondaryMarketsBelieve in Life After Debt VassilisA. Hajivassiliou JPR Wc- E- P6 Srles Ainhor DAM Contact fAboorDatepar WPS253 Public Debt,North and South HelmutReisen August1989 S. King-Watson 33730 WPS254 FutureFinancingNeedsof the Highly IndebtedCountries lshrat Husain SaumyaMitra August1989 S. King-Watson 33730 WPS255 The ExtemalDebt Difficultiesof Low IncomeAfrica CharlesHumphreys John Underwood August1989 S. King-Watson 33730 WPS256 Cash Debt Buybacksand the InsuranceValueof Reserves Swedervan Wijnbergen August1989 S. King-Watson 33730 WPS257 Growth,ExternalDebt, and the Real Swedervan Winbergen ExchangeRate in Mexico WPS258 The Role of VoluntaryOrganizations L David Brown in Developmen David C. Korten WPS259 DealingwithDebt: The 1930sand the 1980s BarryEkchengreen RichardPortes WPS260 Growth, DebtAccumulationand Sovereign Risk in a Small,OpenEconomy JagdeepS. Bhandari NadeemUl Haque StephenJ. Turnovsky WPS261 Inflation,ExtemalDebt and FinancialSector Reform: A QuantitativeApproachto ConsistentFiscalPolicy Swedervan Wijnbergen RobertoRocha RituAnand WPS262 EconomicPolicyand ExternalShocksin a SmallOpen Economy: the Irish Experience F. DesmondMcCarthy DermotMcAleese WPS263 How Has Instabilityin World Markets Affected AgriculturalExport Producers in DevelopingCountries? PeterHazell MauricioJaramillo AmyWilliamson August1989 WPS264 Case Studiesof Two IrrigationSystemsin Colombia: Their Performanceand Transferof Managementto Users Associations Herve Plusquellec WPS265 The Influenceof ImperfectCompetition in InternationalMarkets: SomeEmpirical Evidence AlexanderYeats C. Spooner 30464