SWP7S3 International Financing, for Developing Countries The Unfulfilled Promise Donald R. Lessard WORLD BANK STAFF WORKING PAPERS Nuimber 783 SERIES ON INTERNATIONAL CAPITAL AND ECONOMIC DEVELOPMENT Number 2 WORLD BANK STAFF WORKING PAPERS Number 783 SERIES ON INTERNATIONAL CAPITAL AND ECONOMIC DEVELOPMENT Number 2 Intemnational Financing for Developing Countries The Unfulfilled Promise Donald R. Lessard The World Bank Washington, D.C., U.S.A. Copyright ©) 1986 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing January 1986 This is a working document published informally by the World Bank. To present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The World Bank does not accept responsibility for the views expressed herein, which are those of the authors and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, area, or of its authorities, or conceming the delimitation of its boundaries, or national affiliation. The most recent World Bank publications are described in the annual spring and fall lists; the continuing research program is described in the annual Abstracts of Current Studies. The latest edition of each is available free of charge from the Publications Sales Unit, Department T, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from the European Office of the Bank, 66 avenue d'1ena, 75116 Paris, France. Donald R. Lessard, a consultant to the World Bank, is professor of international management at the Sloan School of Management of the Massachusetts Institute of Technology and at present visiting professor of international management at Stanford University. Library of Congress Cataloging-in-Publication Data Lessard, Donald R. International financing for developing countries. (World Bank staff working papers ; no. 783. Series on international capital and economic development ; no. 2) Bibliography: p. 1. Loans, Foreign--Developing countries. 2. International finance. I. Title. II. Series: World Bank staff working papers ; no. 783. III. Series: World Bank staff working papers. Series on international capital and economic development no. 2. HG3891.5.L473 1985 336.3'435'091724 85-29459 ISBN 0-8213-0681-2 FOREWORD This paper is one in a special series of World Bank Staff Working Papers on international capital and economic development. Prepared as background papers for World Development Report 1985, the series provides more detailed treatment and documentation of the issues dealt with in the Report. The papers cover a range of topics including a historical perspective on international capital and economic development; the effects of policies in industrial and developing countries on international capital flows, external debt, and economic development; and the role of official assistance, commercial bank lendLing, securities markets, and private direct investment in developing countries. Several studies of individual developing countries are also included in the series. The background papers draw on a large number of published and unpublished studies oE individual researchers, on World Bank policy analysis and research, and on reports of other orgarr- izations working on these issues. The papers are the work of individuals and the views and interpretations expressed in them do not necessarily coincide with the views and interpretations of the Report itself. I hope these detailed studies will supplement World Development Report 1985 in furthering understanding of the relationship between international capital and economic develop- ment. A complete list of the papers appears on the overleaf. Francis X. Colaco Staff Director World Development Report 1985 - iii - Papers in the Series Aliber, Robert. Financial Intermediation and the External Debt Crisis Batista, Paulo Noguiera, Jr. International Financial Flows to Brazil Since the Late 1960s: An Analysis of Debt Expansion and Current Payments Problems Blanchard, Olivier, and Lawrence H. Summers. Perspectives on High Real Interest Rates Throughout the World Codippily, Hilarian M.A. Ethiopia: International Financial Flows, 1967-83 Dornbusch, Rudiger. The Effects of OECD Macroeconomic Policies on Non-oil Developing Countries: A Review Fishlow, Albert. Capital Markets During the 19th Century and the Interwar Period: Lessons from the Past Hooper, Peter. The International Repercussions of the U.S. Budget Deficit Ibrahim, Tigani E. Kenya: Use of External Resources, 1963-1983 Ishiyama, Yoshihide, and Keiko Atsumi. Capital Outflows from Japan to Developing Countries Iqbal, Farrukh. Korea's Debt Accumulation, Use, and Management Strategies Johnson, John H. The Role of International Finance in Argentine Development, 1965-84 Lessard, Donald. International Financing for Developing Countries: The Unfulfilled Promise Llewellyn, David. International Financial Intermediation and the Role of Banks in Balance of Payments Financing Martone, Celso L. Macroeconomic Policies, Debt Accumulation, and Adjustment in Brazil, 1965-84 Muller, Patrice, and Robert Price. Public Sector Indebtedness and Long-Term Interest Rates O'Brien, Richard, and John Calverley. Private Banks and Developing Countries Pant, Chandrashekar. Yugoslavia: Economic Development and External Capital, 1970-1984 Rybczynski, T. M. Internationalization of Financial Arrangements and the Developing Countries: The Evolving Relationship Sachs, Jeffrey, and Warwick McKibbin. Macroeconomic Policies in OECD Countries and External Adjustment in Developing Countries Saini, Krishan G. Capital Market Innovations and Financial Flows to Developing Countries Sherbiny, Naiem A. Arab Financial Institutions and Developing Countries Swoboda, Alexander, and Hans Genberg. The Stages in the Balance of Payments Hypothesis Revisited Weigel, Dale, and Robert R. Miller. Foreign Direct Investment in Economic Development Wyplosz, Charles. International Aspects of the Policy Mix in Six OECD Countries Zietz, Joachim, and Alberto Valdes. The Costs of Protectionism to Developing Countries: An Analysis for Selected Agricultural Products - iv - Abstract International financing for developing countries has failed to fulfill its potential not only because bankers were lending too much or countries borrowing too much, but also because of the structure of the financial arrangements--overwhelmingly general obligation credit largely in the form of floating rate bank loans. Bank debt requires debt servicing that bears a perverse relationship to developing countries' ability to pay; it fails to shift risk to world capital markets in line with comparative advantage and, consequently, fails to shift responsibility for the selection and management of investments as well; and it concentrates the risk of default in major commercial banks that represent a small fraction of world financial markets and, as a result, increase the possibility that debt crisis in the developing countries will become a world debt crisis. Several steps to correct these shortcomings, including modifications in bank financing and an increased role of nonbank institutions, are outlined. i Table of Contents Page Introduction ............................................................. I I. The Potential Benefits of International Financing for Developing Countries .......................... . 3 Functions of International Financing ................... ... .3 Interactions with Pasi: Financing Actions .............. ... 10 The Correspondence of Financing Instruments and Financing Functions ................. 1. II. Shortcomings of Existing Patterns of International Finance ....... . ...... .12 What Constitutes Good International Finance .............. 12 Structural hortcomings of Loan Financing ................. 17 III. Why has this Inappropriate Pattern Developed and Why Does it Persist ......................... 23 Evolution of Financial. Institutions and Instruments in Developed Countries .................... .24 Borrower and Lender M.opia .......... .............. .... .26 Soverign Risk and the Reltive Enforceability of Elank Debt ..................... .. ..... 31 The Comparative Advantage of Banks in Creating Self-Ernforcing Contracts ... 33 The Role of the IMF arnd Other Multilateral Institutions in the Enforcement of Debt Claims ........ 36 V. Measures to Increase the Efficiency of International Financing for Developing Countries .37 Increased Role of Nonbank Institutions .40 Changes in the Role of Multilateral Financial Institutions. . 42 VI. Summary and Conclusions ..43 References ..... ... 45 - vii - Introduction For a developing country, international financing is a double-edged sword: It carries tremendous potential for increased economic welfare, but it also harbors many dangers. The experience of developing countries between 1973 and 1983, a period that may become known as the "decade of international debt," bears this out. On balance, the increased reliance of developing countries on international Einancing appears to have been beneficial, but there is little question that this financing has fallen far short of its potential and, in some cases, has made borrowing countries worse off. This paper examines the potential benefits to be derived from international financing, the reasons why it has failed to live up to this potential, and why its structure evolved as it did. Attention is also given to some proposals for change in (a) the structure of international finance for developing countries and (b) the behavior of borrowers and lenders within a given structure, in order to judge which proposals offer the greatest promise. A central theme here is that international financing for developing countries failed to fulfill its potential not only because bankers were lending too much or countries borrowing too much, but also because of the structure of the financial arrangements. An overwhelming proportion of international financing for developing countries is in the form of general obligation credit, and a high proportion of this credit has been in the form of floating-rate bank loans. From 1974 to 1983, more than 80 percent of the net external financing for nonoil developing countries was in the form of credit, and two-thirds of this amount consisted of bank loans. Although there is nothing wrong with bank credit per se, several of its characteristics make - 2 - it inappropriate for developing countries when it becomes such a large proportion of their external obligations. In particular, general obligation bank loans l.impose debt servicing requirements that bear a perverse relationship to a borrower's ability to pay, 2.fail to shift specific risks to a broad "world capital market pool" from particular countries that are unduly exposed, 3.do not require that Northern lenders, advisors, or trading partners take some responsibility for both the selection and execution of investment programs and projects, and 4.concentrate the consequences of potential defaults in a narrow sector of world capital markets. Overall, what we have seen is too much debt and not enough financing in other forms. On the positive side, we believe that a number of changes involving little or no incremental transfers could significantly increase the benefits of international financing to developing countries. Among these measures are changes, in the role of commercial banks and in the role of multilateral institutions (including the World Bank and the International Monetary Fund), reentry of nonbank financial intermediaries, and an increased role for direct investment and other forms of contractual involvement by nonfinancial corporations. Many of these shifts can and should be incorporated in the ongoing restructuring of the existing obligations of developing countries. I. The Potential Benefits of International Financing for Developing Countries The Functions of International Financing External financing enables developing countries to: 1. enhance the potential national income over time by investing in profitable projects that cannot be financed with domestic resources, 2. accelerate or delay domestic consumption 1- relative to anticipated national income, 3. smoothe domestic consumption in response to sharp fluctuations in income or required outlays, 4. shift risks associated with particular development strategies or economic ventures to foreign investors or governments, 2/ 5. shift responsibility for i:he selection or management of investments, and 6. obtain concessional transiEers of resources (which, in a strict sense, are a form of aid rather than finance). 3/ 1/ A more precise term would absorption--the combination of consumption and investment. 2/ Formally, these separate goals can be modeled as maximizing the utility of national consumption over time, where the utility of consumption at any point in time depends not only on its absolute level, but also on its level relative to previous and planned levels of consumption. This captures the effects of growing absorptive capacity over time as well as the adjustment of expectations. Since the country is treated as a unit, distributional considerations are ignored. See Sachs and Cooper (1984) for further discussion on this point. 3/ For a recent discussion of finance versus aid, see Leipziger (1984). -4- 1. Enhancing potential income overtime. Countries benefit from external financing if it enables them to undertake investments that could not have been financed out of domestic resources alone and whose social return is in excess of the cost of funds obtained, as illustrated in Figure 1A. The solid line represents the anticipated income path with domestic financing only, the dotted line indicates income attainable with the increased investment made possible by external financing, and the shaded area shows the cost of financing or "debt service." The area between the solid line and the shaded area is the gain in net national income resulting from external financing. The decision rule associated with this goal of enhancing income over time is to borrow abroad up to the point where the marginal cost of finance--the interest rate if one abstracts from uncertainty--is just equal to the marginal yield or return expected from the project or program being financed. This type of financing was used by the United States in the late 1860s, when, on the heels of the Civil War, it borrowed substantial amounts from Europe to build the transcontinental railroad. Another case is that of Brazil, which relied heavily on external sources in the mid-1960s to finance an export-oriented industrial economy. In both instances, foreign financing followed in the wake of basic institutional and technological changes that created major new investment opportunities. 2. Altering the time path of national consumption. Countries whose current income outstrips their "absorptive capacity" and threatens to disrupt their society, or that face an imminent decline in external revenues as resources are depleted, often delay expenditures by accumulating external financial claims, thus trading reduced present expenditures for increased potential future expenditures. Alternatively, countries that expect high future incomes from proven resource positions often borrow abroad to -5- Figure 1 Patterns of Rternal Finance A~~~~~~~~~~~ Debt service _ > / B Accumulation of reserves Spending of C reserves A-Gross income with A-Consumption (Absorption) / / external finance B-Net income with B-Income external finance C-Income with no external finance Time Time A.External finance to enhance B. External finance to postpone consumption potential income over time relative to pattern of income A A B~ ~ ~ ~~~~~~~~~~4 Borrowing _W 40| Borrowing *De service/accumulation , C Debt service of reserves Borrowing A-Consumption (Absorption) A-Consumption (Absorption) B-Income B-Income Time C. External finance to accelerate D. External finance to smoothe consumption consumption relative to pattern of income -6- accelerate expenditures in anticipation of these revenues. In either case, external financing allows a country to uncouple current expenditures from current income and shift expenditures to those years where they are most valuable. This shifting of expenditures is illustrated in Figure 1B for a country seeking to delay consumption, and in Figure IC for a country spending in anticipation of future revenues. The decision rule here is to borrow (lend) to the point where the marginal cost (expected return) of external finance is equal to the country's marginal rate of time preference for expenditures. Countries that have used external financing in this way are Saudi Arabia and other core OPEC countries that "stored" purchasing power abroad during the early years of their revenue boom. On the other hand, Mexico used external financing both to enhance income and to shift consumption. With the discovery of oil, Mexico had not only immense new investment opportunities, but also higher immediate consumption demands because of the sense of greater national wealth. Thus it shifted consumption forward in time while it also borrowed to finance investment. 3. Smoothing the path of national expenditures over time. With so- called balance of payments financing, developing countries use external financing to maintain national absorption in the face of shortfalls in export revenues that result from downturns in the world economy, shifts in terms of trade, domestic economic problems, or sudden increases in desired or required expenditures because of external or domestic pressures. 41 In contrast to the financing taken on the enhance income--which might be termed "true development 4/ Short-term borrowing is a substitute for drawing down national reserves. Therefore, reserves and "borrowing capacity" serve similar functions. - 7 - financing"--balance of payments financing is often viewed as something to be avoided. Although the need for such financing is often the result of economic mismanagement, certain benefits can be obtained from short-term "decoupling" of income and expenditures. Figure 1D illustrates how balance of payments financing can smoothe expenditures over time. The solid line represents national income in the absence of any short-term financing, the dotted line shows smoothed income, and the shaided area indicates debt service associated with the short-term financing. The decision rule associated with financing of this type--which is often difficult to implement--is to borrow in the short term to offset temporary declines in net resource flows and to use short-term surges in income to repay debt or accumulate reserves. The difficulty lies in distinguishing short-term shocks from basic shifts in a country's economic circumstances. 5/ Most oil-importing countries turned to balance of payments financing in the wake of the first and second oil shocks. The International Monetary Fund's Compensatory Financing Facility and the European Community's Stabex facility are examples of institutionalized mechanisms that seek to distinguish between export shocks and mismanagement and hence to reduce some of the dangers inherent in this type of financing. 4. Shifting risks to external investors. Another way to smoothe expenditure flows in the face of uncertain, fluctuating revenue streams is to 5/ In practice, it is very difficult to distinguish between short-term fluctuations in income and long-term declines in wealth. Thus, borrowing to deal with what is thought to be a temporary shortfall might result in an acceleration of consumption in the face of declining future income prospects. - 8 - exchange claims against the risky revenue stream from a country's endowment for claims promising a more stable and, in most cases, smaller revenue stream. 61 Financial mechanisms that perform this function include futures, contracts, equity interests in specific ventures, and bonds or other contracts indexed to variables whose future value is uncertain. When uncertainties exist regarding either future returns or financing costs, financing decisions must take into account not only the (expected) marginal cost of financing and the (expected) marginal return of the investment in question, but also the impact of the resulting allocation of risk on the country's well-being. Two principles are at work in this case. The first is related to the relative willingness of the borrower or lender to accommodate risk, the second to the relative ability of one party or the other to average out or modify the risks. In general, both the borrowing country and the lender prefer to avoid risks. As a result, the borrower would be willing to pay a premium (that is, a higher expected cost) to shift the risk to the lender, whereas the lender would demand a premium for taking on the risk. To the extent that the lender, either because of a greater ability to diversify a specific risk or a greater tolerance for bearing risk, demands a smaller premium than the maximum the borrower is willing to pay, it will have a comparative advantage in bearing 6/ The reason why a developing country should expect to accept a stable income stream that is smaller than the expected level of its initial risky stream is that investors in world capital markets are risk averse and demand a premium to take on the risk. The reason why shifting risk to foreign investors can make the country better off is that the premium demanded may be less than the implicit premium the country should apply to reflect its own risk aversion. The "outside" risk premium is likely to be less than the "inside" risk premium whenever outsiders have greater scope for reducing the risk in question through diversification than the country in question. -9- the risk in question, and both parties will be better off by transferring risk. The resulting decision rule is for a country to shift risk via financial mechanisms up to the point where the marginal risk premium demanded by external investors of financial institutions is just equal to the cost (in real resources) of reducing these risks by adopting alternative development strategies, or to the premium that the country is willing to pay to eliminate these uncertainties. 7/ 5. Shifting responsibility for the selection or execution of programs or projects. When a country's obligations are linked to the outcomes of specific projects or undertakings, as opposed to being backed by its general credit, foreign lenders or investors obtain a stake in the success of the project, enterprise, or program in question. This linkage may lead to improved performance and reduced risk when lenders or investors have some control over variables crucial'to a project's success. For example, if part or all of a project-specific obligation is tied to the performance of the project being financed, the vendor will have a greater interest in seeing that the project design is appropriate and its management is satisfactory. Similarly, if global obligations of a spec:ific borrowing country are linked to that country's volume of manufactured exports, lenders will have an interest in seeing that exports are not hampered by protectionism in industrialized countries. The other side of the coin, of course, is that as the investor's stake in a particular project or program grows, the lender will demand greater 7/ A rigorous statement of this decision rule requires an explicit specification of the national utility function with respect to the expected level and uncertainty of anticipated future income. - 10 - control. Equity investment, whether of a direct or portfolio nature, represents the extreme case of shifting risk, and consequently responsibility in the degree of control required by the investor. Debt involving repayment obligations that are fixed regardless of specific investment outcomes, even if nominally linked to a specific project or enterprise, is unlikely to fulfill this function since it will probably be viewed as a general obligation of the borrowing country. Only where the project is financed on a stand-alone basis, with no possibility of governmental bailout, will so-called project financing provide this link. 6. Obtaining concessional resource transfers. Most financial transactions involve the exchange of current real resources for claims (often uncertain, and sometimes explicitly contingent on future events) against future income. However, financing involves concessional transfers of resources whenever current resources are transferred without a (full) corresponding ttansfer of financial claims. The decision rule with concessional financing is to obtain as much as possible since it directly increases a developing country's potential level of expenditures without creating offsetting claims against future income. Although concessional financing has declined as a proportion of total financing of developing countries over the last decade, many countries appear to have continued to view external financing as something that should be utilized to the fullest extent possible, rather than something in which the current benefits must be carefully traded off against the resulting future costs. Interactions with Past Financing Actions At any point in time, a country's financing requirements and options are determined to a large extent by repayment obligations resulting from past - 11 - financing choices. Whenever external financing is sought to enhance income or to shift or smoothe the pattern of expenditures, these claims against current income must be taken into account. A country may have sufficient gross income to finance all attractive developmient projects, but repayment obligations may bring net income below the required level. Thus in order to maintain the desired investment pattern, it will have to roll over the existing financing. Similarly, a country's increasing gross current income may be sufficient to cover the desired current level of consumption, but repayment obligations may bring net income below the desired level, making rollover desirable. Finally, the existence of fixed external obligations will exacerbate the impact of sudden revenue shortfalls since the absolute decrease will represent a larger proportionate decrease of net revenues than of gross revenues. The flexibility of the timing of repayments will thus be a major factor in determixiing a country's gross external financing requirements. This will be true whether the flexibility takes the form of allowing deferral of repayments as a function of some set of external events, or actually changing the amount due, as in the case with contingent, risk-shifting financial contracts. The Correspondence of Financing Instruments and Financing Functions Each of the six functions of external financing described above are associated with specific forms of financing: debt financing, risk capital financing, and grant or subsidy financing. Debt financing involves claims with (substantially) fixed repayment requirements. It can transfer resources over time to enhance income and to shift or smoothe consumption. Risk capital financing involves claims whose repayment is contingent on specified future - 12 - outcomes. It shifts risk and, under some circumstances, responsibility. Subsidy financing involves unrequited transfers of resources in the form of grants, subsidies, guarantees, or concessional interest rates. Table 1 relates each of these functions to specific types of financing. The Changing Structure of Financing for Developing Countries Although the external financing used by developing countries involves a variety of instruments and institutions, debt financing plays a significant role. Further, an increasing fraction of debt financing is in the form of loans from private creditors, primarily commercial banks. Virtually all such debt carries floating interest rates and, until the most recent reschedulings, has been denominated mainly in dollars. From 1974 to 1983, debt from private sources accounted for nearly 50 percent and total debt for more than 80 percent of net external financing in the developing countries excluding the buildup in official reserves and the private holdings reflected in errors and omissions (International Monetary Fund 1985). II. Shortcomings of Existing Patterns of International Finance What Constitutes Good International Finance? Whether debt financing is appropriate for developing countries is a complicated issue. Its value depends on the extent to which debt financing enables borrowers to exploit investment opportunities, smoothe income, shift risks, and shift responsibility to terms that are mutually advantageous to borrowers and lenders. -- 13 - Table 1 Financial Mechanisms: A Three-Dimensional Typology Extent to Which Extent to Which Mechanism Transfers Resources Extent to Which Involves Concessional Over Time Shifts Risks Resource Transfer Not To a To a at Short Long Moder- Large Moder- Large All Term Term Little ately Extent Little ately Extent Commercial Bank Loan X X x Long-Term Bond X X X World Bank Loan X X X IDA Term Loan X X X IMF Compensa- tory Finance X X x Commodity Futures Contract X X X Commodity Price Linked Bond X X X Equity Investment X X X - 14 - Lenders, of course, would like to see no repayments crises. However, the absence of such crises does not imply that the level or structure of financing has been appropriate. It may merely reflect good luck, an overly conservative use of external financing by borrowers, or such high penalties in the case of default that borrowers are forced to make costly internal adjustments. Borrowers, in contrast, would like to see financing that supports uninterrupted growth. When the fundamental circumstances of a country change for the worse, however, there is no economic way to maintain consumption at previous levels, and financing should not be expected to bridge the entire gap. Even so, the fact that repayments crises result in deadweight costs (costs borne by the borrower that are not offset by equivalent gains to the lender) is not necessarily a bad thing. In the absence of an international legal system with enforcement powers, no international financing would be possible without the existence of penalties in the case of default, which, by definition, are deadweight loses. A general test of the efficiency of international financing for developing countries is impossible to devise since it would require knowing what would have taken place in its absence. It is possible, however, to examine various dimensions of existing financing to see if it performs the services that it should. For example, do net financing flows adjust positively or negatively in line with the borrowing country's income and, hence, shift risks in a mutually beneficial way? Would any alternative arrangements carry smaller deadweight costs if borrowers had to adjust to adverse developments in world trade or capital markets or in their own economies? - 15 - Under existing institutional arrangements, debt financing creates incentives for inappropriate choices on the part of both borrowers and lenders and may entail serious distributional consequences within borrowing or lending countries. Even if inappropriate behavior by borrowers and lenders is ignored, there are four principal reasons why the existing debt-dominated structure is less than ideal. In this section, we discuss the reasons why debt financing cannot be expected to perform many of the functions desired of international financing for developing countries. Why debt financing may create incentives for inappropriate behavior by lenders and borrowers is discussed in Section III. Structural Shortcomings of Loan Financing Consider, now, the main structural shortcomings of debt financing. 1. Debt service mismatch. Most developing countries experience fluctuations in revenues owing to world economic cycles, shifts in the terms of trade, and domestic political and economic events. International financing provides a basis for smoothing national consumption over time through borrowing in periods of low income and replenishing reserves or repaying debt in periods of high income. If a country already has substantial external obligations, however, debt service requirements will magnify the volatility of national income available for consumption and force an even greater reliance on international financing in order to obtain the same smoothing over time. This effect will be exacerbated to the extent that debt service requirements themselves vary perversely with national incomes, as appears to be the case at the present time. Most private internationaL lending is at floating rates, and total debt service in any period consists of interest at the current short-term - 16 - market rate (LIBOR) and the scheduled reduction in principal. As is well known, inflation tilts real debt service on any loan with a fixed nominal repayment schedule toward the present (see, for example, Lessard and Wellons 1979; Kincaid 1981; and Goodman 1982). In other words, whenever nominal interest rates rise to reflect anticipated inflation, the effective maturity of an outstanding loan is decreased and the required repayments are accelerated. Changes in real interest rates, of course, also change debt service requirements. In recent years, increases in interest rates have tended to coincide with decreases in the incomes of developing countries. Thus, these countries have faced the highest debt service requirements precisely when they were least able to pay. In the current world recession, for example, real interest rates are at an all-time high and nominal interest rates, which affect the time pattern of debt service requirements, remain at high levels. At the same time, export prices and volumes are dismal for most developing countries. This is clear from figure 2. 81 8/ From 1973 to 1983, for example, the correlation between the aggregate terms of trade for developing countries and debt service as a proportion of outstanding debt was -.61. (Debt service per dollar was calculated as LIBOR plus 0.14 on the assumption of seven-year average maturity.) This correlation, based on eleven annual observations, is nearly significant at the 10 percent level. Major components of these two series, the relative price of primary commodities and LIBOR, are available with greater frequency. (The relative price of primary commodities is obtained by deflating the IMF index of dollar prices of primary commodities by the U.S. GNP deflator lagged by one period to reflect survey delays in the latter measure). The contemporaneous correlation over the same period is -.42, significant at the .01 level, while the correlation of terms of trade with LIBOR lagged six months, which adjusts for the fact that interest payments are made in arrears, is -.53, also significant at the .01 level. (See IMF 1985 for further analyses along these lines.) - 17 - jjSure 2 Terms of Trade versus Debt Service 105 % 17 % ~~~% 1978 100 016 95 ~ 14 90 R RATIO O - 13 DEST SERVICE 85 - PAYMENT T TOTAL2 OUTSTANDING DEST , A - I I l- I I I 1973 1974 1975 1976 1977 1,978 1979 1980 1981 1982 1983 - 18 - In a world with perfect information ard complete enforceability, the part of the perverse variation in debt service due to fluctuations in inflation would not be a problem. Claims would be rolled over unless the present value of a borrowing country's future net exports fell short of the present value of outstanding claims. In other words, illiquidity would never be an issue and the only risk would be that of insolvency. Given limited information and enforceability, however, rolling over is not a sure thing, and the arbitrary shortening of maturities via increases in LIBOR, as well as the shortening of available maturities, can create problems for borrowers as well as for the system as a whole. 2. Limited risk sharing. Most commercial bank loans to developing countries (and all World Bank loans) involve explicit or implicit government guarantees. Thus, although the funds may be earmarked for a specific project or program, their repayment is not contingent on that project's outcome, and the risk of success or failure of the specific project or program is borne by the guarantor. This nonspecific nature of bank credit has two effects. First, it trivializes the role of private banks and public institutions in evaluating projects. Since all claims are general obligations of the sovereign, a loan to a good project is no better than a loan used to acquire arms or maintain consumption in the face of a reversal in the terms of trade. This also has important behavioral implications, as we shall see in the following section. Second, and probably much more important, it means that within the present system risks inherent in projects or strategies are shifted only through nonperformance at the country level. This is a costly and inefficient mechanism, resulting in limited risk sharing. In fact, repeated assertions of - 19 - bankers that few, if any, defaults are imminent are clear evidence that the system provides little risk shifting, especially given the radical fluctuations in trade and other project and strategy-specific risks incurred by developing countries. Ironically, much of the academic literature on external financing for developing countries stresses the strategic nature of risks of nonperformance, but downplays the issue of the extent to which the system succeeds in "passing through" exogenous risks faced by borrowers to investors with a comparative advantage in bearing such risks (see, for example, Eaton and Gersovitz 1981a, 1981b). Thus, the reduction of risk in the system is seen largely as a search for mechanisms to enhance the enforceability of claims. Given the nonspecific nature of bank loans, this would effectively preclude the shifting of any exogenous risks. Most credit to developing countries, whether provided by commercial banks, development banks (including the World Bank), or export finance agencies of industrialized countries, require the explicit or implicit guarantee of the national government or central bank of the borrowing country. 9/ As a result, the soundness of the loan depends primarily on the borrowing country's overall creditworthiness and not on the economics of any specific undertaking, or on the solidity of the project's private backers. There is little true project financing for developing countries whereby lenders have recourse only to the cash flows of the project in 9/ Even in the case of loans to specific corporations or banks, government guarantees are often demanded. When these are not obtained, the national government is often forced to assume responsibility for the obligation in case of nonperformance by the borrower, as was the case with Chile. - 20 - question. World Bank project loans, for example, are all guaranteed' by the national government. Although disbursement of funds may be tied to a particular undertaking, repayment of funds is a general obligation. The same is true of most commercial bank loans which, although possibly disbursed for specific projects, are, in fact, general obligations of the borrowing country. Not only do most loans to developing countries involve general obligations of the borrowing countries, but they also represent obligations that do not vary with the performance of the borrower's macroeconomy or its ability to service external debt. In essence, the bank contract says: "You will pay interest (at the current market rate) plus 1/n of principal per year, regardless of your situation." It is true that through refinancing and reschedulings such debt repayment obligations are modified to reflect a country's situation, but such flexibility after the fact is not a very good substitute for flexibility before the fact, since reschedulings are power plays that almost invariably result in deadweight costs. 3. Limited shifting of responsibility. In addition to not shifting specific risks from borrowing countries to lenders, this noncontingent character of loans to developing countries implies that specific lenders have little or no stake in the success or failure of specific undertakings since their recovery of the amount lent depends only on borrowers' overall payment performance, which, in turn, depends both on their overall ability to pay and on the penalties they face it they do not. Thus, the bank that lends money for the purchase of armaments is in the same position as the bank that funds an industrial project with an economic return of 30 percent. Similarly, the bank and firm that engineer, construct, and finance a pipeline through credit stand to gain the same amount - 21 - whether or not the ultimate petroleum discoveries are sufficient to justify constructing the pipeline. This is in sharp contrast to contingent financing--direct equity investment or quasi-equity investment such as production shares or incentive contracts--where success of the project is critical to the recovery of funds by the financier. Clearly, in such cases, the financier will have a strong interest in assuring that the project or program being financed is well conceived and executed. Furthermore, to the extent that the interests of industrial countries control. outcomes critical to a program's or project's success, forcing these countries to accept a stake will cause them to influence outcomes in ways that are favorable to the project. An example of this effect can be found in the recent investments of automakers in the manufacture of components in numerous developing countries. One of the greatest risks facing such operations is that northern markets will increase protection in the form of tighter local value-added requirements. Given that firms wiLl recover their investments in developing countries only if markets remain open for these products, they will lobby with their home governments to assure such continued access. If, in contrast, such plants had been constructed by developing country governments with bank financing, no one with political power in the industrialized countries would have the same interest, and the depree of protection probably would be higher. The same goes for natural resource projects. If U.S. producers still maintained major equity or quasi-equity stakes in, say, Chilean copper mines, the array of lobbying pressures in the recent attempt by U.S. producers to limit access of foreign producers would have been shifted further in favor of imports. - 22 - 4. Concentration of default impact. The fact that claims against developing countries are concentrated in commercial banks may magnify the impact of nonperformance on the system, and on the continued availability of financing for developing countries. Although the aggregate volume of claims against developing countries is a small fraction of total world financial claims, it is large relative to the capitalization of the key banks. As of 1983, for example, loans to six' developing countries (Argentina, Brazil, Chile, Mexico, Venezuela, and the Philippines, as reported by Bergsten, Cline, and Williamson 1984) accounted for more than 180 percent of the shareholders' equity of the nine major money-center banks in the United States. Over the same period, they accounted for a much smaller fraction of the capitalizations of major regional banks (47 percent), and by implication, for the rest of the U.S. banking system. In any case, developing country loans loom large relative to the risk capital of the banks making those loans. When the debt of developing countries is viewed in relation to the total value of financial claims outstanding in the world economy, though, they do not loom nearly as large. As of 1980, for example, the total market value of traded stocks and bonds in major markets was $5,290 billion (Ibbotsen and Siegel, 1983). If bank loans, currency, and marketable real estate holdings are included, the total exceeds $11,500 billion, placing the total obligations of developing countries at well under 10 percent. Although this percentage is by no means trivial, it does suggest that the world financial system could absorb the loss of a significant proportion of claims on developing countries if the effects of such a loss were spread throughout the system. The impact of a major default by a developing country on the world financial system can be better understood by comparing its potential magnitude - 23 - with financial losses experienced in recent years. A default of, say $250 billion is very large amount, but, in aggregate, it is small compared with the total value of financial claims outstanding in and among market economies. Observed variations in stock market values provide another useful comparisons. In the United States during 1973 and 1974, share values fell by almost 40 percent, or roughly $600 billion, yet the system did not collapse. IO/ A major default could force banks to limit additional credit or even to cease rolling over existing credits. The exposure of banks to such events bring lender country authorities with their varying political agendas into debt renegotiations, perhaps further distorting the incentives facing private lenders. III. Why has this Inappropriate Pattern Developed and Whr Does it Persist? How did the existing debt-dominated structure of international finance came to be, and if it is so inappropriate, why haven't financial markets and institutions attempted to overcome the limitations that we have 10/ Even if the banking authorities of industrialized countries were to intervene quickly in order to avert any run on banks that might result from such a default, the distributional impact of a $250 billion default on developing country debt would be quite different from that of a $250 billion drop in equity values. The former would be covered primarily with an initial monetary expansion (a tax on money-fixed holdings) and a later requirement for higher general tax revenues, whereas the latter would involve a loss in savings for retirement and the like, which probably would have a smaller impact on output in industrialized countries. There is no guarantee that banking authorities would, in fact, intervene to avoid a collapse of any element of the system. - 24 - outlined? At least five reasons. can be cited to explain the emergence of general obligation, floating-rate bank delt: 1. the pattern of evolution among the financial institutions and instruments in developed countries, 2. borrower and lender myopia, 3. the assertion of sovereignty by borrowers in reaction to previous intrusive financing modes, 4. the relative enforceability of bank debt in the presence of sovereign risk, and 5. the role of the IMF and other multinational institutions in enforcing debt claims. Each of these is discussed below. Evolution of Financial Institutions and Instruments in Developed Countries Over the period of greater growth in developing country indebtedness, financial patterns industrialized countries changed drastically. Increased interest rate volatility led to a decoupling of the term of lending and the term for which interest rates were fixed. Floating rate credit under long- term credit lines or rollover agreements became the rule rather than the exception and in some countries, most notably the United Kingdom, fixed rate instruments virtually disappeared. This decoupling of interest rates and maturity allowed commercial banks to play a much more aggressive role in long- term finance, a trend that given further impetus by the growth of offshore money-markets dominated by banks and the surge in funds placed with these institutions both on and offshore in the wake of OPEC's revenue windfall. - 25 - Developing countries in general simply rode along with these changes. The substantial list of Einancial innovations over the last decade contains none that were motivated by the special needs of developing countries (for a discussion of recent financial innovations, see Dufey and Giddy 1981). There are two reasons for 'this. First, developing countries remain small factors in the world financial markets. Their total obligations represent less than 10 percent of total outstanding financial obligations-- closer to 5 percent if commercial real estate is included in the financial portfolio. Second, most financial institutions are inherently conservative and therefore tend to introduce innovations mainly in their dealings with the most creditworthy borrowers--which developing countries have not been for some time. However, changes in the structure of developing country obligations cannot be attributed entirely to a passive following of market developments. These countries did shift to floating rate bank loans to a greater extent that private firms in industrialized countries, although not more than public borrowers in these countries. The comparison with private borrowers appears to be the more relevant one., Although most borrowing in developing countries has been undertaken by governments of has been guaranteed by them, much of this has gone to finance parastatal enterprises (see Gillis, Jenkins, and Lessard 1982 for a discussion of the role of public enterprise finance in the total foreign borrowing of developing countries). For much of the remainder, the government simply acts as a conduit to private firms or banks. This is in sharp contrast to European municipal governments, for example, whose offshore - 26 - borrowing is geared mainly toward financing investments in traditional infrastructure. A comparison of developing country financing with that of private firms in industrialized countries shows that the former involves a much higher proportion of debt financing. In general, firms in the United States have increased their reliance on debt financing over the last two decades--it reached a peak of 45 percent in the mid-1970s (see Taggart 1985). European firms made more use of debt, but they, too, appeared to reach peak levels in the mid-1970s. Borrower and Lender Myopia Inappropriate Borrower Behavior. A major problem with debt financing is that where there is uncertainty regarding future returns from the investments being financed, its explicit cost relative to expected project returns is likely to overstate the desirability of borrowing in order to invest and, as a result, countries are likely to overborrow. Diaz-Alejandro (1982) notes for example, that given the real interest rates prevailing in the 1970s "the price of either extravagance or sensible capital formation was Ilow. " When countries trade off increased investment against future debt service obligations, they should take into account the fact that investment returns will vary, but that debt will have to be serviced regardless of future income levels. Technically, this can be done by discounting the certainty equivalent of future benefits at the real interest rate on foreign borrowing. This certainty equivalent will be strictly less than the expected value, however, if swings in returns from the new investment contribute to the - 27 - variance of national income. Although real interest rates might have been low in the 1970s, the certainty equivalents of future project returns--especially for investments representing a "deepening" of exposure to a set of key risky variables affecting national income (for example, OECD income-dependent manufactures for Brazil, oil for Mexico)--should have been significantly lower than their expected values to reflect their national "systematic" risk. At the same time, the certainty equivalent of debt with LIBOR-linked payments would have exceeded that of riskless real debt since debt service on such obligations covaries negatively with the export revenues of many developing country borrowers. This potential borrower myopia will be reinforced if political leaders responsible for borrowing choices have relatively short time horizons. The "time bomb" nature of bank credit (that is to say, the potential for a mismatch of incremental obligations with incremental earnings is substantial but largely unpredictable) makes this type of financing particularly inappropriate in political settings where decisionmakers cannot or do not take a long view. A further problem created by the general obligation nature of bank claims is that, within a decentralized system, borrowing units typically view only the explicit costs of funds whereas society at large bears the contingent future costs of providing an explicit or implicit guarantee. This problem is particularly serious in developing countries with large public enterprises that have been granted substantial autonomy, ostensibly in order to increase the quality of economic decisionmaking by establishing clear responsibilities. - 28 - It has been a major factor in the borrowing behavior of almost all countries facing debt servicing difficulties. 1l/ Inappropriate Lender Behavior. Bankers have been accused of doing little analysis, but rather of following the pack and thus lending too much to countries currently in favor and too little to those out of favor. Whether or not this is true, at least two structural factors predispose individual banks to behave this way, even if they are rational in microeconomic terms. Certain aspects of a bank's internal organization and reward structure will also induce rational individuals to engage in behavior that is not rational for the bank. A perfectly rational bank may hold that analysis is a fruitless endeavor and that it pays to "lend with the pack" since it risk of lending to a particular country bears little or no relation to the quality of the projects it finances, and its risk depends in large part on the behavior of lenders of last resort, including its own central bank and the International Monetary Fund. In fact, its behavior will be very similar to that of the manager of PEMEX, Pertamina, or any other major "autonomous" public borrowers who is entitled to write his own guarantee--in this case a put option to the lender of last resort on the loan. Within banks, loan officers may have fairly short horizons and may be biased to overlend to risky cases if they generate substantial fee income. The fact that there is no secondary market for most loans, and that loans are not marked to market even when there is such a market, postpones the day of 11/ See Gillis, Jenkins, and Lessard (1982) and Baldwin, Lessard, and Mason (1983) for a discussion of inappropriate behavior when central governments grant implicit guarantees to individual decisionmaking units. Sachs and Cooper (1984) provide further reasons why the explicit cost of foreign borrowing to private borrowers will understate its cost to society. - 29 - reckoning and increases the likelihood of such behavior (see Herring and Guttentag 1984). Assertion of Economic Sovereignty by Developing Countries The increased availability of general obligation debt financing allowed developing countries to have much greater control over externally financed activities that had hitherto been the case. Direct foreign investment, the major alternative form of external financing available to developing countries permits foreign economic actors to intrude into the local economy and political systen. In many cases, these intrusions either have been or were viewed as a continuation of colonial or imperial patterns of foreign control with its perceived unfair distribution of economic gains. The other major source of financing was bilateral government aid or loans from the World Bank or regional development banks. The former typically carried its own economic and political strings, while the latter typically was restricted to projects favored by the institutions in question. Bank financing allowed the governments of developing countries much greater leeway, and in many cases made possible the financing of large parastatal organizations. Sovereign Risk and the Relative Enforceability of Bank Debt Sovereign risk is a key obstacle to international financing of any kind. It is the result of the limited enforceability of contracts with sovereigns, coupled with the discretionary powers of sovereigns. Any contract that bridges two jurisdictions requires that both sovereigns agree, explicitly or tacitly, to enforce the contract within their legal systems. Thus, whether the contract directly involves a sovereign or not, the sovereign will enter as a key element in its enforceability. - 30 - Contracts with sovereigns are difficult to enforce since a sovereign may reject a claim against it within its own territory. Moreover, courts typically will not recognize a claim against a foreign sovereign because of the doctrine of sovereign immunity. 12/ Even when courts do accept jurisdiction over foreign claims, however, the remedies that can be applied are limited. In the absence of an invasion to seize a country's assets, claims can only be exercised against assets present in the court's jurisdiction. Thus, liquidation is not an option, as it is with private borrowers. Even when the contract is with a private agent within the foreign country, many of the same obstacles to enforcement are present. In order to enforce the claims within the borrower's home country, the claimant must be granted access to the legal system. Judgments outside of the agent's home country are again limited to assets that can be seized within those jurisdictions. Problems arising from this limited enforceability are further complicated by the fact that governments have a great deal of discretion regarding policy choices that influence their own ability to meet any particular set of obligations, as well as that of domestic firms or banks. Many of these policies--for example, shifts in monetary policy, limits on exchange remittances, changes in competition policy, changes in taxes, and so on--could not be deemed a breach of contract. 12/ Sovereign immunity has been rolled back significantly in recent years. Courts in the United Kingdom have allowed claims against sovereigns where the contracts were for commercial purposes, and the U.S. Foreign Sovereign Immunity Act of 1976 removed immunity for sovereign's commercial transactions. - 31 - Sovereign risk is likely to be a larger factor in the international financing of developing countries than in transactions among industrialized countries for three reasons. First:, because of their limited domestic financial systems, developing count:ries have less scope for domestic financial intermediation. Thus, at least some of the savings-investment process is externalized. Furthermore, substantial fraction of the assets of the wealthiest individuals is usually outside of the country, and new funds are brought in through foreign loans. -3- Second, because of their greater likelihood of being "out of step" with the world economy, either in terms of growth prospects or risk exposures, developing countries tend to depend more on external financing and, hence, are more affected by obstacles to international transactions. Third, because of the often greater participation of the government in the national economy, coupled with the lesser degree of institutionalization of domestic financial and industrial interests, the governments of manly developing countries exercise grater discretion over the course of the economy and over factors determining the value of various claims on the economy. Thus the ability of the government to influence outcomes (termed moral hazard in the literature on economic risk sharing), coupled with limited ability to impose legal sanctions on nonperformance, means that contracts with developing countries have little economic value unless they are self- 13/ In fact, it is nor believed that the acquisition of foreign assets by wealthy individuals in several key Latin American countries equaled or exceeded the increase in the external, general obligation borrowings of these countries. - 32 - enforcing--that is, unless it is in the self-interest of the contracting parties to honor the contract under virtually all circumstances. Recognition of these two considerations has led academics who study international financing arrangements to focus on determinants of self- enforceability and, hence, on the supply of loans (see, for example, Eaton and Gersovitz 1981a, 1981b, and Sachs and Cohen 1982). Admittedly, these models explain much of what has happened, but the trouble is they have often been interpreted to imply that developing countries are "dishonest" borrowers that "oplan" to default on their obligations. A more neutral interpretation is that developing countries are amoral, economically rational actors who recognize that there will be circumstances under which it will not be in their best interest to meet their obligations. Factors Giving Rise to Self-Enforceability. Since self- enforceability requires that the (present discounted) economic value to a borrower of meeting its obligations must be equal or greater than the present value of not meeting them, it is most likely to hold for countries that would suffer a great deal if they did not pay, and be least likely to hold for those facing negligible penalties. The principal penalties that can be imposed on a sovereign, or that a sovereign cannot avoid having imposed on private firms operating within its jurisdiction, are the withholding of future finance and the blocking of commercial transactions that would put national assets at risk of seizure. The cost of these sanctions to a borrower depends, of course, on the importance of future trade and financing to the country in question. Eaton an Gersovitz argue that countries with volatile incomes may be more creditworthy than those with more stable incomes noting that although their ability to pay - 33 - would be lower under certain circumstances than countries with more stable incomes, their willingness to pay would be greater, precisely since their income variability makes foreign financing highly desirable. Sachs and Cohen show that countries with export-oriented development strategies are more creditworthy, not because they are somehow sounder and more stable than other countries, but because they are dependent on continued world trade in order to maintain their standard of living. The extensive defaults of Latin American countries in the 1930s can easily be explained from this perspective. Because of the world depression, the trade of these countries and, hence their ability to service their debts, decreased drastically. However, the loss of markets for their products also removed the penalties that normally would have been associated with such default, thus reducing their willingness to pay. Although it is hard to distinguish between ability and willingness in the event of nonperformance, it is useful to think of ability as being reflected in current cash flows (liquidity) and willingness in present value terms (insolvency). The Comparative Advantage of Banks in Creating Self-Enforcing Contracts. In cases where few externa.l sanctions can be imposed, contract enforceability is greatest when there are recurring transactions, as is well known. This means that nonperformance on any one contract will lead to a loss of many subsequent ones. In such cases, a party to a contract, say developing country borrower, builds a reputation over time by honoring its commitments and must weigh the loss of this reputation and its impact on future borrowing against the benefit of defaulting. The concentration of claims against developing countries in relatively few banks implies that ongoing trade and finance will involve a - 34 - fairly steady stream of transactions and, hence, high implicit costs to nonperformance. In contrast, a firm with a one-shot investment or a group of bondholders might have less of a continuing relationship and, as a result, a weaker position. Sovereign guarantees for credits to private borrowers within developing countries further reinforce this point, since they increase the continuity of transactions involving the sovereign. 14/ Since the enforceability of international financial claims against development countries depends upon sanctions that would be imposed on their future trade or finance in the case of default, it depends upon the potential for a collective response by lenders. Since all bank debt of a given country involves the same type of general obligations, there is little conflict of interest among banks. The banks themselves have evolved collaborative mechanisms to minimize any playing off of one bank against another by the borrower in a repayment crisis and to maximize the sanctions that would be imposed in the case of nonperformance. It would be much harder for holders of other types of claims on developing countries to create and maintain such mechanisms. With different types of claims, actions that result in nonperformance in one case, for example, would not necessarily result in performance in others (see table 2). 14/ Commercial banks, because of the sanctions they can bring to bear on specific borrowers, have at times forced countries to make good on nonperforming loans to private firms or banks even when there was no explicit government guarantee. In some cases, the governments in question assumed the loans; in others, they put into place special policies, such as dual exchange rates, that had the effect of transferring resources to the firms to allow them to meet their external obligations. - 35 - Table 2: Types of Nonperformance by Type of Claim Type of External Claim Foreign Home Foreign Home Portfolio Direct Currency Currency Currency Currency Equity in Equity in Government Government Debt Debt Local Foreign- Debt Debt of of Private Controlled Type of Risk Private Private Firm Local Firm Firms Firms Government default on external claims X X X X X X Exchange controls X X X X X Inflation/ devaluation X X a/ X b/ X a/ X a/ Adverse changes in fiscal policy X X a / X a/ X b Credit controls X a/ X a/ X a/ Changes in limits on foreign ownership ? X Limits on behavior of foreign-controlled firms X a. Indirect effect via impact on borrowinq firm's cash flows. b. Direct and indirect effect. Source: Lessard, et al. (1983). - 36 - The Role of the IMF and Other Multilateral Institutions in the Enforcement of Debt Claims The comparative advantage of banks in enforcing contracts with developing country borrowers is further strengthened by the role of the International Monetary Fund and, to a lesser degree, the World Bank. Both institutions have leverage because they represent a continuing source of financial support and because they play a special role in signaling to other financial institutions whether or not to cut off lending to a particular country. 15/ Because of the more complex nature of direct foreign investment, or even portfolio investment, private providers of these types of financing do not obtain similar help in the enforcement of their claims. 16/ In fact, the measures imposed by the IMF in conjunction with commercial banks often have the effect of imposing losses on claims denominated in the local currency claims against local private firms, or equity in many forms in order to increase the ability of the debtor country to meet its external, foreign- currency, general obligations. Thus it is not surprising that the structure of international financing for developing countries has shifted toward general obligation to sovereign lending. 15/ In fact, the rapid movement of the World Bank toward structural adjustment loans in the face of declining project lending possibilities can be interpreted as a move to maintain self-enforceability by holding out future net lending as a reward to countries that meet their current obligations. 16/ The United States and various other investor countries attempt to impose sanctions in the case of expropriation of direct holdings. However, these mechanisms are usually not very effective because it is difficult to determine damages and assign fault, and hence, obtain the collective response required to make the penalties effective. - 37 - V. Measures to Increase the Efficiency of International Financing for Developing Countries Some of the limitations of the current system can be addressed through innovation and change within the banking system as it now organized, whereas others will require that alternative institutions or instruments be developed or that substantial changes be made in banking itself. What I consider to be the most important changes are outlined below. 17/ Modifications in Bank Financing Smoothing Real Debt Service Patterns. The problem of the perverse variability of debt service could be ameliorated in several ways. The first and simplest would be for major international lenders to adopt debt service formulas on floating rate loans that call for roughly constant real debt service, (see, for example, Goodman 1982). An alternative that would provide for even smoother real payments, since it locks in real interest rates, is a price-level index-linked loan (for a recent discussion, see Williamson 1981). With such a loan, a real rate of interest would be contractually fixed, but the outstanding principal would be adjusted periodically for changes in some general price index. An important question thal: may arise with index-linked debt is the choice of the index, since different borrowers would want different base currencies or combinations of currencies. It is possible, however, that a large number of developing countries would find a standard combination--such as a price-level claims denominated in SDRs--attractive. 17/ This section draws substantialLy on Lessard and Wellons (1979). - 38 - Increasing Repayment Flexibility. Although the above measures would go a long way toward reducing the negative impacts of credit market fluctuations on debt service requirements, they do not provide developing countries with a safety valve in the case of difficulties resulting from world economic downturns, shifts in the terms of trade, or changes in local economic conditions. The IMF has already addressed this issue with its Compensatory Financing Facility (CFF), but commercial financing typically provides no flexibility. Totally flexible repayment terms on long-term debt are out of the question since loans would no longer be enforceable. If the flexibility were limited in nature, however, it might be acceptable to lenders. An example of such a mechanism would be a bond of calling for equal payments of principal in each year, but with a provision that in any one year the borrowing country could opt to repay some lesser amount, subject to provisions for catching up in future years. In essence, such a bond would provide a degree of automatic refinancing at the borrower's discretion. A bond with the timing of repayments linked to trade flows is another variation on this theme. In order to be enforceable, repayments under such a contract would have to be linked to some aggregate trade measures, exogenous to the borrower in question. Bailey (1983) has suggested linking repayments to a country's own net exports, and Diaz-Alejandro (1984) has proposed that CFF be expanded to deal with this problem Increased flexibility along various dimensions would help developing countries cope with specific risks, but only by postponing repayment obligations. Many risks, are not cyclical in nature, however, but represent permanent changes in the value of existing resources and facilities. In such cases, postponing payments will simply compound the problem. In contrast, - 39 - financing arrangements that E'XpliCitly shift risk are viable, whether or not these risks are cyclical. Increased Risk-shifting with General Obligation Financing. Although general obligation financing cannot, by definition, shift the risks of particular projects or enterprises to foreign providers of finance, it can be employed to lay off certain risks that affect the economy as a whole. Two specific innovations deserve particular attention: (a) commodity-price linked securities, and (b) trade-linked securities. Both deal with narrowly defined sets of risks relevant at the national as well as the enterprise level. Many developing countries depend, and will continue to depend, upon a small number of primary product exports for their foreign exchange earnings. Such countries could issue commodity-linked bonds. In addition to shifting some of the basic exposure of developing countries, such instruments should reduce contracting risks since they are narrowly drawn and primarily shift risks that are outside the control of the borrowers. Of course, investors would still face the risk of default, but this risk is not likely to be any greater than that accompanyinig straight bonds. Increased Nonrecourse Financing. The general obligation character of most bank financing is a major structural flaw in the existing system and gives rise to many of its behiavioral anomalies. Increased nonrecourse lending would provide lenders with more incentive to conduct proper analysis and would reduce the likelihood that borrowing errors would result from decentralized decisionmaking. There are, however, several obstacles to such a shift. First, individual banks wouldi probably be reluctant to give up their general claims without compensation is some form. The World Bank might play a role here by facilitating project loans and providing, for example, completion - 40 - guarantees in return for an increased flow of true project financing. Similarly, the Controller of the Currency could create a separate classification for project loans with escrowed export proceeds. A more difficult issue is whether a country could default on a money- fixed project loan without jeopardizing its overall credit standing, that is, whether it could obtain true nonrecourse financing. In most cases, this would be extremely unlikely; I believe that nonrecourse financing is most likely to be viable where instruments that share directly in project outcomes are employed. Increased Role of Nonbank Institutions Increased Project-Specific Risk Capital. International financing at a project or enterprise level is likely to be superior to nonspecific financing--especially debt--if some of the risks entailed could be borne more easily by foreign rather than by local investors, or if it is important to provide foreigners with a stake in the project because they have provided the technology or access to the market. The two primary existing mechanisms for international risk transfers between developing countries and world capital markets are direct foreign investment and portfolio investment in equity. Both mechanisms penetrate the national economy and involve substantial enforcement difficulties and compliance costs. Simpler, more narrowly defined, risk-shifting devices are likely to be superior. Consider alternative arrangements that may be used for financing the development of the oil reserves of a country that will be a significant oil exporter (see Blitzer, Lessard, and Paddock 1983 for an in-depth discussion of the various alternatives.) If a significant fraction of the production is to - 41 - be used in the domestic market, a major risk associated with direct or portfolio equity investment in the clevelopment of local oil production will be the pricing of the output in the domestic market. This pricing is a political outcome, however, and is likely to be influenced by the foreign ownership of the oil company. The profits of the domestic oil company are likely to be affected by a wide variety of local political choices, including labor policy, tax policy, and exchange rate policy. As a result, foreign investors are not likely to become involved unless they have considerable control over the domestic situation--this will be costly meddling from the perspective of the developing country in which the investment takes place. Portfolio investors are unlikely to be involved at all nlness a highly institutionalized domestic capital market exists that provides a set of national "bedfellows" to protect the interest of foreign shareholders. A production share is a less complex instrument that avoids many of the risks in the hands of the domest:ic government and yet provides a mechanism to lay off market price risks on a world economy. Nevertheless, it also involves an element of control, which, from the perspective of the domestic government, may not be desirable. A commodity-linked bond is even more narrowly defined and, hence, need not be tied to a specific project. Of course, it requires the existence oi a widely traded commodity for which an external price is readily available. Furthermore, it does not provide foreign investors with much of a stake in the national elements of a successful project--such as those associated with the discovery of oil, the development itself, and the management of the facilities once "on stream." Quasi-equity financing arrangements such as production shares often provide a desirable compromise between debt instruments that provide - 42 - foreigners with no stake in local operations, direct equity investment where foreigners assume total control, and portfolio foreign investment in the equity of local firms requiring that significant institutional preconditions be met. Changes in the Role of Multilateral Financial Institutions One of the principal reasons why general obligation debt has emerged as the dominant form of international financing for developing countries is that it is supported by the current institutional framework, including the IMF and the World Bank. The role of one or both of these institutions should be broadened to provide increased support for various forms of contingent financing along the lines of the support they already provide for general obligation debt financing. This could be done, for example, by having the World Bank exact borrower government guarantees of compliance within the terms of a particular contingent contract instead of guarantees of fixed payments irrespective of project outcomes. The World Bank, in turn, might guarantee the same performance to third parties. This would be an important step in increasing the flow of project financing and hence, in increasing the degree to which both commercial risk and responsibility are shifted from developing countries to firms and financial institutions with a comparative advantage in bearing these risks. An even stronger measure for these institutions would be to formally cofinance or otherwise lend their support for the enforcement of only those claims they deemed appropriate both in amount and form. Both measures should be given serious consideration. - 43 - VI. Summary and Conclusions International financing plays an important role in the economy of developing countries. It allows them to go ahead with attractive investments that they otherwise would have to forego or postpone, to shift income over time, to smooth out sharp fluctuations in income, and to lay off risk in accordance with comparative advantage in risk bearing. On the negative side, however, it can require onerous debt servicing at times of great economic stress. The central theme of this paper is that the current debt-dominated structure of international financing for developing countries fails to provide many of the important potential advantages of international financing and significantly increases the likelihood of costly debt crises. The debt crisis that developing countries face today is not just the result of bad luck with the world economy or of their overborrowing, but can be traced in part to the badly flawed structure of international financing. The flaws, in turn, can be attributed in part to desire of developing countries to minimize foreign involvement in their own economies, and to the role of the IMF and the World Bank, among others, in enforcing the general obligations of developing countries, but not project- or program-contingent financing. Bank debt requires debt servicing that bears a perverse relationship to the ability of developing countries to fulfill this obligation. It fails to shift risk to world financial markets in line with comparative advantage and, consequently, fails to shift responsibility for the selection and management of investments as well. Finally, it concentrates the risk of default in major commercial banks that represent a small fraction of world - 44 - financial markets and, as a result, increases the possibility that the debt crisis in the developing countries will become a world debt crisis. It is not enough merely to recognize that the existing structure of International financing is flawed. Concerted action on the part of borrowers, lenders, and the multilateral intermediaries is needed. Fortunately, significant changes can be made within the existing institutional structure without increasing the flow of concessional financing. First, the volatility of debt service on nonspecific credit can be reduced through innovation in the repayment pattern on floating rate debt. Now that the World Bank is shifting to floating rates, it should take the lead in promoting such innovation. Further, the IMF could insist that commercial banks adopt similar measures as part of any rescheduling agreement. Second, the ex ante flexibility of debt services should be increased to avoid the inevitable costly after-the-fact changes in debt terms while still maintaining discipline and appropriate incentives. Third, to the extent that developing country's activities are substantially concentrated in a few sectors, nonspecific financing arrangements should be exploited to shift risks such as commodity price or trade fluctuations to world financial markets. Fourth, developing countries should shift to project or enterprise-specific financing wherever it is important to shift key risks or provide foreign suppliers of funds with a stake in project outcomes in order to ensure that they will assist in project selection and management. The feasibility of project or enterprise financing can be increased by designing quasi-equity investments that expose foreign investors to a limited range or risks and, hence, reduce the required degree of foreign capital. - 45 - References Bailey, Norman A. 1983. "A Safety Net for Foreign Lending." Business Week, January 10. Baldwin, Carliss Y., Donald Et. Lessard, and Scott Mason. 1983. "Budgetary Timebombs? Controlling Government Loan Guarantees." Canadian Public Policy IX 3: 338-346. Bergsten, C. Fred, William R. Cline, and John Williamson. 1985. "Bank Lending to Developing Countries: The Policy Alternative." Institute for International Economics, April. Blitzer, Charles R., Donald R. Lessard, and James L. Paddock. 1983. "Risk Bearing and the Choice of Contract Forms for Oil Exploration and Development." Energy Journal 5 1: 1-27. Diaz-Alejandro, Carlos F. 1984. "International Markets in the 1980s." Journal of International Affairs 38 1: 11-20. _ 1982. "North-South Issues and the International Financial System." Paper presented at the Tel Aviv University Conference on the Future of the International Monetary System. October. Dufey, G., and I. Giddy. 1981. "Innovation in the International Financial Markets." Journal of International Business Studies (Fall): 35-51. Eaton, Jonathan, and Mark Gersovitz. 1981a. "Debt with Potential Repudiation: Theoretical and Empirical Analysis." Review of Economic Studies 48: 289-309. . 1981b. "Poor Country Borrowing in Private Financial Markets and the Repudiation Issue." Princeton Studies in International Finance. Gillis, Malcolm, Glenn P. Jenkins, and Donald R. Lessard. 1982. "Public Enterprise Finance: Towards a Synthesis." Public Enterprise in Less- Developed Countries ed. Leroy P. Jones et. al. New York: Cambridge University Press. Goodman, Laurie S. 1982. "An Alternative to Rescheduling LDC Debt in an Inflationary Environment." Columbia Journal of World Business 17 (Spring): 20-27. Ibbotsen, Roger and Lawson Siegel. 1983. "The World Wealth Portfolio," Journal of Portfolio Management (Winter) 5-7. International Monetary Fund. 1983. World Economic Outlook. Washington, D.C. - 46 - _ 1985. "Foreign Private Investment in Developing Countries." Occasional Paper no. 33. Washington, D.C. Kincaid, C. Russell. 1981. "Inflation and the External Debt of Developing Countries." Finance and Development 18 (December): 45-48. Leipziger, Danny M. 1984. "The Concessionality of Foreign Assistance," Finance and Development 21 (March): 44-46. Lessard, Donald R. and Philip Wellons. 1979. "Financing Development: Innovation and Private Financial Markets." Technical report, UNIDO. Lessard, Donald R., T. Bollier, R. Eckaus, and R. Kahn. 1983. "Country Risk, Capital Market Integration, and Project Evaluation: A Canadian Perspective." Working Paper, MIT, Sloan School of Management. Sachs, Jeffrey, and D. Cohen. 1982. "LDC Borrowing with Default Risk." NBER Working Paper no. 925. _ 1984. "Growth and External Debt under Risk of Debt Repudiation." Working Paper. Sachs, Jeffrey, and Richard Cooper. 1984. "Borrowing Abroad: The Debtor's Perspective." NBER Working Paper no. 1427. Taggart, Robert. 1985. "Secular Patterns in the Financing of U.S. Corporations" in B. Friedman, ed., Corporate Capital Structure in the United States, Chicago: University of Chicago Press. Williamson, John. 1981. "The Why and How of Funding LDC Debt." paper presented at the Second Interamerican Conference on Capital Markets, Caracas, Venezuela. The World Bank Headquarters European Office Tokyo Office 1818 H Street, N.W. 66, avenue d'lena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France l-I Marunouchi 3-chome Telephone: (202) 477-1234 Telephone: (1) 47.23.54.21 Chiyoda-ku, Tokyo 100, Japan Telex: WUI 64145 WORLDBANK Telex: 842-620628 Telephone: (03) 214-5001 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WASHINGTONDC ISSN 0253-2115/ISBN 0-8213-0681-2