sovereign debt structure for crisis prevention - world bank internet

O C C A S I O N A L PA P E R

Sovereign Debt Structure forCrisis Prevention

Eduardo Borensztein, Marcos Chamon, Olivier Jeanne,Paolo Mauro, and Jeromin Zettelmeyer

237

INTERNATIONAL MONETARY FUND

Washington DC

2004

237


237

Sovereign D

ebt Structure fo

r Crisis P

revention

2004

O C C A S I O N A L PA P E R 237

Sovereign Debt Structure for Crisis Prevention



Washington DC

2004

© 2004 International Monetary Fund

Production: IMF Multimedia Services DivisionFigures: Jorge Salazar

Typesetting: Alicia Etchebarne-Bourdin

Cataloging-in-Publication Data

Price: US$25.00(US$22.00 to full-time faculty members and

students at universities and colleges)

Please send orders to:International Monetary Fund, Publication Services

700 19th Street, N.W., Washington, D.C. 20431, U.S.A.Tel.: (202) 623-7430 Telefax: (202) 623-7201

E-mail: [email protected]: http://www.imf.org

recycled paper

Sovereign debt structure for crisis prevention/Eduardo Borensztein . . . [et al.]—Washington, D.C.: International Monetary Fund, 2004.

p. cm.—(Occasional paper); 237

Includes bibliographical references.ISBN 1-58906-377-5

1. Debts, public. 2. Financial instruments. I. Borensztein, Eduardo. II. Occa-sional paper (International Monetary Fund); no. 237.

HJ8011.S68 2004

Preface vii

I Overview 1

Two Views on the Status Quo 1Debt Structures with Existing Instruments: Emerging Market

Countries Versus Advanced Economies 3Ideas for Sovereigns from the Corporate Context: Explicit Seniority 3Expanding the Set of Instruments: Real Indexation 4Toward Better Sovereign Debt Structures: A Road Map 5

II Facts on Existing Public Debt Structures 7

Public Debt in Emerging Market Countries Versus Advanced Economies 7

Sovereign Versus Corporate Liability Structures 11

III Rendering Debt Structures Less Crisis Prone with ExistingInstruments 14

Problems with the Status Quo 14Determinants of Government Debt Structure 15Policy Implications 19

IV Explicit Seniority in Privately Held Sovereign Debt 23

Economic Role of Seniority 23Approaches and Obstacles in Implementing Explicit Seniority 25Conclusions 28

V Expanding the Set of Instruments: Indexation to Real Variables 29

Benefits of Indexation to Real Variables 29Real Variables Beyond the Control of the Country’s Authorities 31Real Variables Partially Within the Control of the Country’s

Authorities 38Obstacles for Variables Partly Within the Control of the Government 42Steps to Foster Acceptance 43Real Indexation: Which Variables for Which Countries? 44

VI Past and Future of Innovation in Sovereign Borrowing 46

Financial Innovation in Sovereign Borrowing: A Haphazard Process 46Road Maps for Future Innovation 48

VII Conclusions 49

Contents

iii

CONTENTS

Appendix Investors’ Attitudes Toward Growth-Linked and Other Innovative Financial Instruments for Sovereign Borrowers: Results of a Survey 51

References 56

Boxes

1. Debt Structure and Hedging 172. Creating Domestic Markets for Long-Term Domestic-Currency

Bonds: Country Experiences 203. Developing International Markets for Bonds in Emerging

Market Currencies 214. Enforcing Contractual Seniority 265. Effect on Borrowing Costs of a Switch to First-in-Time

Seniority 276. Proposals for Indexation to Real Variables 307. Benefits of GDP Indexation for Emerging Markets and

Advanced Economies 418. Previous Examples of Indexation to Real Variables 43

Text Tables

1. External Sovereign Debt: Currency Composition, 1980–2003 102. Structure of Domestically Issued Government Bonds at

End-2001 103. Structure of Total (Domestic and External) Central Government

Debt, 2001 124. Percentage Share of the Top Three Exports in Total Exports,

1990–99 325. Top Five Natural Disasters by Percent of GDP Lost 346. Small Countries: Types of Disasters, 1975–2002 367. Output Growth and Trading Partners’ Growth, 1970–2002 388. Introduction of Inflation-Indexed Securities by Sovereigns 47

Text Figures

1. Advanced Economies and Emerging Market Countries: Public Debt Stocks and Debt Composition 7

2. Structure of External Public Debt in Emerging Market Countries 83. Emerging Market Countries: Fixed- Versus Floating-Rate

Sovereign Bond Issues 84. Structure of Internationally Issued Debt: Maturity Composition 95. Emerging Market Countries: Structure of Public Debt 116. All Developing Countries: Public Sector Bonds and Loans

Issued in International Markets 137. Recent Crises: Impact of Exchange Rate Depreciation on Public

Debt-to-GDP Ratio 158. Share of Long-Term Local-Currency Bonds in Total Government

Domestic Bonds and Inflation History 169. Share of Long-Term Local-Currency Bonds and Financial

Liberalization 1810. Institutional Quality and Domestically Issued Long-Term

Local-Currency Debt 1911. Interest Savings over the Economic Cycle 40

iv

Contents

Appendix Tables

A1. Question 3: Obstacles to Growth-Linked Bonds 53A2. Question 4: Obstacles to Growth-Linked Bonds 54A3. Question 5: Obstacles to Commodity-Indexed Bonds 54A4. Question 6: Obstacles to Domestic-Currency Bonds 54

Appendix Figures

A1. Question 1: Premium over Plain Vanilla Bonds 52A2. Question 2: Premium over Plain Vanilla Bonds 53

v

The following symbols have been used throughout this paper:

. . . to indicate that data are not available;

— to indicate that the figure is zero or less than half the final digit shown, or that the itemdoes not exist;

– between years or months (e.g., 2001–02 or January–June) to indicate the years ormonths covered, including the beginning and ending years or months;

/ between years (e.g., 2001/02) to indicate a fiscal (financial) year.

“n.a.” means not applicable.

“Billion” means a thousand million.

Minor discrepancies between constituent figures and totals are due to rounding.

The term “country,” as used in this paper, does not in all cases refer to a territorial entity thatis a state as understood by international law and practice; the term also covers some territorialentities that are not states, but for which statistical data are maintained and provided interna-tionally on a separate and independent basis.

This Occasional Paper is intended to stimulate debate on the issue of sovereigndebt structures for crisis prevention. It was prepared under the general guidance ofRaghuram Rajan. The authors include Eduardo Borensztein, Marcos Chamon, OlivierJeanne, Paolo Mauro, and Jeromin Zettelmeyer. Work on the paper was led by PaoloMauro. The authors are grateful to Jonathan Ostry, Anna Gelpern, Sean Hagan,Simon Johnson, Thomas Laryea, and several other colleagues for helpful comments;to Priyanka Malhotra and Martin Minnoni for excellent research assistance; and toUsha David for editorial assistance. Special thanks to Leslie Payton-Jacobs of EMTAfor helpful suggestions and cooperation in circulating the survey, and to Kellett Han-nah for web services. Archana Kumar of the External Relations Department editedthe paper and coordinated its production.

The opinions expressed are solely those of the authors and do not necessarily re-flect the views of the International Monetary Fund or its Executive Directors.

Preface

vii

The way countries structure their public borrow-ing has long been considered an important de-

terminant of economic performance. This topic hasrecently received renewed attention as a result of notonly steep increases in public debt levels in emerg-ing market countries—and a number of highly visi-ble and damaging crises—but also pronouncedchanges in the composition of those debts.1 There isincreasing recognition that debt structure has impor-tant implications for both the frequency of crises andthe disruption they cause when they strike.2 Indeed,the official sector is beginning to give renewedprominence to the possible need for innovations inthe design of countries’ financial liabilities.3

The debate on government debt in the context ofpossible reforms of the international financial archi-tecture has thus far focused on crisis resolution.4This Occasional Paper seeks to broaden the debateby asking how government debt could be structuredto pursue other objectives, including crisis preven-tion, international risk-sharing, and facilitating theadjustment of fiscal variables to changes in domesticeconomic conditions. To that end, this paper consid-ers recently developed analytical approaches to im-proving the structure of sovereign debt using exist-ing debt instruments. It then reviews a number ofproposals—including the introduction of explicit se-niority and GDP-linked instruments—in the sover-eign context and discusses their pros and cons, andthe related practical challenges.

While recognizing that there is no easy substitutefor sound macroeconomic policies—fiscal policiesin particular—and that no amount of financial engi-neering could eliminate crises, this paper askswhether greater use of relatively underutilized fi-nancial instruments could help reduce the frequencyof damaging crises. After identifying commonsources of vulnerability, the paper takes a first passat identifying instruments and structures that couldhelp achieve a more resilient debt structure, and setsforth some preliminary considerations about theirfeasibility.

Two Views on the Status Quo

Developing a strategy for addressing possible in-efficiencies in existing debt structures requires anunderstanding of what may cause them. On this sub-ject, there are two views in the policy and academicdebate. The first, which underlies most proposals forreforming the “international financial architecture,”assumes that today’s array of instruments is inher-ited from historical accident and has persisted owingto inertia: the existing structures can be changed,though not without substantial effort, through re-forms involving coordination among market partici-pants. The second view argues that the status quo isan adaptation to deeper problems, such as difficul-ties in enforcing contracts in the international set-ting, lack of policy credibility, and weaknesses indomestic institutions. The outcome may well be in-efficient, but it cannot be improved without address-ing the underlying problems.

History and Inertia

The “architecture” analogy is one of a housewhose current form results from the way it wasbuilt in the past, in response to incentives or needsthat may have had little to do with those of its pre-sent inhabitants. Under this view, making a case forreform merely requires showing that the architec-ture gives rise to costly and inefficient outcomes.Of course, structures that are considered part of the

I Overview

1

Note: The authors of this section are Paolo Mauro and JerominZettelmeyer.

1International Monetary Fund and World Bank (2001 and2003); IMF, World Economic Outlook (September 2003, Chapter3); Reinhart, Rogoff, and Savastano (2003); Guidotti and Kumar(1991).

2International Monetary Fund (2003a); and Allen and others(2002).

3The Declaration of Nuevo León (Special Summit of the Ameri-cas, Monterrey, Mexico, January 2004) supports “the efforts of bor-rowing countries to work with the private sector to explore new ap-proaches to reduce the burden of debt service during periods ofeconomic downturns” (available via the Internet: www.summit-americas.org/SpecialSummit/declaration_monterrey-eng.htm).

4International Monetary Fund (2003b).

I OVERVIEW

architecture do not generally change by them-selves: this requires a reform effort. But the goodnews is that through such an effort, most structurescan be torn down and rebuilt, or at least renovatedand cleaned.

Changes to the status quo could however be diffi-cult to achieve for many reasons, especially a needfor coordination among market participants. For in-dividual market participants, it is hard to go againstmarket practice in drafting contracts. Moreover, re-forms often require mustering support from nationalparliaments, international bodies, or market partici-pants. A number of potential obstacles thus stand inthe way of contractual or financial innovation (Allenand Gale, 1994):

• Coordination problems and the need to ensure“critical mass” for new instruments. The appealof an innovation often depends on its simultane-ous adoption by many contracting parties. Forexample, learning to price new financial instru-ments may require excessive resources from theviewpoint of an individual investor, but may beworth the effort collectively for the potential in-vestor class. More generally, individual borrow-ers considering whether to issue a new financialinstrument will not take into account the benefitsfor other borrowers and investors that would re-sult from establishing a new asset class. And inthe absence of a concerted effort to guarantee aminimum critical mass, investors may be con-cerned about the possibility of limited liquidityfor the new instruments and thus demand a “nov-elty premium.”

• The highly competitive structure of financialmarkets. A private financial institution wouldhave to incur costs to develop a new type of fi-nancial instrument. However, it may be unable tomaintain a monopoly over the provision of thisinstrument for a long time: patents are still rarely(though increasingly) used for financial instru-ments, and imitation is relatively easy. Thus, theprivate incentive to develop the instrument in thefirst place may be low, even if its social benefitmay be high.

• The need for standards. To create a liquid sec-ondary market where investors can easily diver-sify their portfolio, it is important to have in-struments with the same features for allcountries or all firms issuing them. Moreover,for financial instruments where payments aredue when certain conditions are met, it is cru-cial to have verifiable standards for whetherthose conditions are met. For example, the mar-ket for credit default swaps remained small foryears but took off as soon as the standards for a

“credit event” were properly defined and be-came broadly accepted.5

• Signaling. Individual countries may be reluctantto issue new financial instruments or existing in-struments with new contractual features if theyfear that such innovations may be misperceivedas signs of weakness or lack of commitment togood policies.

Deeper Problems

An alternative view is that prevailing contracts andmarket practices result from the responses of credi-tors and sovereign debtors to deeper problems, in-cluding difficulties in enforcing contracts involvingsovereign borrowers, and the possibility of moralhazard (behavior that does not maximize the likeli-hood of repayment) on the part of debtors. Costlydebt crises may look inefficient ex post but are, inthis view, the only way to discourage defaults (Doo-ley, 2000; Dooley and Verma, 2001). Existing debtinstruments are seen as optimal because they implythat crises will occasionally occur to constrain or dis-cipline borrowing governments. Similarly, “risky”and seemingly inefficient debt structures heavilyweighted toward foreign-currency-denominated debtand short-term debt are rationalized as necessaryevils to reduce moral hazard on the part of policy-makers, or minimize debt dilution (Chamon, 2002;Jeanne, 2000, 2004; Tirole, 2002; and Sections II andIII).6 Thus, crisis-prone debt structures can be asymptom rather than the root cause of countries’ in-ability to commit to good policies; such inability mayin turn result from weak domestic institutions.

Under this view, attempts to reform the interna-tional financial architecture by changing outcomesbut without addressing underlying distortions couldwell be counterproductive. For example, restrictionsor taxes on short-term debt might seek to induce amove from short-term to long-term flows. However,their impact might be undone by international in-vestors’ shift toward other forms of debt that aresimilarly difficult to dilute, such as foreign-currencydebt. Alternatively, if the impact of the restrictionscannot be undone, they might end up reducing oreliminating capital flows altogether. As in OscarWilde’s Canterville Ghost, for the stain to ceasefrom reappearing on the carpet the next morning, itis not enough to apply the latest carpet cleaner. Theghost itself must be laid to rest.

2

5Credit default swaps are instruments giving the holder theright to sell a bond at its face value in the event of default by theissuer.

6The disciplining role of short-term and other risky forms ofdebt has also been emphasized in the corporate context(Calomiris and Kahn, 1991; Diamond, 1991).

Ideas for Sovereigns from the Corporate Context

Both interpretations of the status quo have somemerit, and this paper draws upon them in the subse-quent sections. The focus on underlying causes ofinefficiencies in existing debt structures leads to adiscussion of associated policy and institutional fail-ures, and remedies for them. Beyond this, though,and recognizing that crises are exceedingly costly,7this paper provides a preliminary analysis of the casefor innovations that could directly improve sover-eign debt structures, but may have been impeded inthe past primarily by inertia.

Debt Structures with ExistingInstruments: Emerging MarketCountries Versus Advanced Economies

In analyzing existing debt structures, two sets ofcomparisons provide insights into how debt struc-tures might be improved (Section II). First, a com-parison between debt structures in emerging marketcountries and advanced economies highlights char-acteristics that make advanced economies less crisisprone. Second, a comparison between sovereignsand corporates highlights the roles of equity and se-niority in corporate liability structures, with poten-tial applications in the sovereign context.

Compared with advanced economies, emergingmarket and developing countries find it relativelydifficult to issue long-term debt in their own curren-cies. Greater reliance on short-term and foreign-currency debt is associated with a higher frequencyof debt crises (Section III). Short-term debt (or debt indexed to short-term domestic interest rates) is associated with vulnerability to sudden changes inmarket sentiment: worsening perceptions of thecountry’s creditworthiness can quickly feed intohigher interest costs, often leading to vicious circles.Similarly, with relatively large shares of foreign-currency debt, depreciations can abruptly render acountry insolvent.

Only a handful of the largest economies issue debtdenominated in their own currency on internationalmarkets, perhaps reflecting in part their economicsize and the use of their currencies as a vehicle for in-ternational trade. Bonds issued internationally areotherwise relatively homogeneous, usually taking the

form of fixed-rate bonds with relatively long maturi-ties. By contrast, the composition of debt issued do-mestically varies considerably across countries. Fewemerging markets issue large amounts of long-termlocal-currency debt, even in their domestic markets.But a number of them have increasingly made use ofdomestically issued alternatives to foreign-currencydebt, including short-term debt, inflation-indexeddebt, and floating-interest-rate debt.

Emerging market countries’ difficulties in issuinglong-term local-currency bonds on the domesticmarket seem to result from deeper problems, such aslack of monetary and fiscal policy credibility, and re-lated worries about the possibility of inflation or out-right default. While the requisite credibility may takea long time to build, several emerging market coun-tries have recently begun issuing local-currencybonds with maturities of a few years, and have reliedon inflation-indexed bonds for longer maturities.Compared with floating-rate and foreign-currencydebt, CPI indexation is less likely to lead to debtcrises, because it tends to not amplify the effects ofadverse shocks. Moreover, the development of do-mestic private pension funds often creates a naturalbase of investors seeking the protection againstchanges in purchasing power that CPI indexationprovides.

Regarding debt issued internationally, some inter-national financial institutions (IFIs) have often beenamong the first parties to issue bonds denominated inthe currencies of emerging markets (usually in com-bination with exchange rate swaps with emergingmarket residents that issue in one of the world’s maincurrencies). Opportunities to raise funds at more fa-vorable rates have been, and should continue to be,the primary motivation for the IFIs’ involvement inthese operations: the IFIs have been able to tap newinvestor bases interested in holding assets denomi-nated in emerging market currencies but bearing nodefault risk. This said, contributions to the develop-ment of new financial markets that can later betapped by developing countries are a welcome by-product of such funding decisions by the IFIs.

Ideas for Sovereigns from theCorporate Context: Explicit Seniority

Partly as a result of contract enforcement issues,sovereign liability structures both in emerging mar-ket countries and in advanced economies are not asrich as those of corporations. A notable difference isa lack of an explicit seniority structure, which at thecorporate level exists either by statute or throughbond covenants. As a result, sovereign creditors tendto be more exposed to “debt dilution” than do their

3

7It is difficult to estimate the extent to which the costs to thedomestic economy result from default itself rather than other as-pects—such as bank runs or sudden drops in the exchange rate—with which defaults are typically associated. Nevertheless, de-faults are associated with widespread bankruptcies, sizable joblosses, and declines in domestic demand. In addition, the negativedomestic implications of a forced debt restructuring are perceivedto be so traumatic that policymakers will delay this option untilall other possibilities have been exhausted (IMF, 2002a).

I OVERVIEW

corporate counterparts (Section IV). Debt dilutionoccurs when new debt reduces the claim that exist-ing creditors can hope to recover in the event of a de-fault. Long recognized as a problem in corporatedebt, dilution seems to have recently become a sig-nificant problem in emerging sovereign debt mar-kets. For example, by issuing large numbers of newbonds to a wide base of creditors in the 1990s, Ar-gentina drastically reduced the value of the initialbondholders’ claims.

Debt dilution has undesirable consequences forboth debt structures and the amounts and terms atwhich sovereigns borrow. Its adverse effects on debtstructure stem from investors’ efforts to hold debtforms that are harder to dilute—such as short-termdebt or debt that is costly to restructure. Such instru-ments in turn make the debtor more vulnerable tocrises and render the impact of crises more severe.Dilution also increases the likelihood that highly in-debted countries will overborrow. Countries near de-fault may be able to place new debt with investorswithout facing prohibitive interest rates, as the newcreditors effectively obtain a share of the existingcreditors’ debt recovery value. At low debt levels, theopposite problem may occur, as the possibility of di-lution tends to raise interest rates unnecessarily.

In principle, debt dilution could be ruled out by anexplicit, “first-in-time” seniority structure giving pri-ority to earlier debt issues, because in the event ofbankruptcy the original creditors would be repaidfirst. First-in-time seniority would tend to reduce bor-rowing costs at low debt levels, but make borrowingmore expensive at high debt levels. In fact, if the prob-ability of a debt crisis were substantial, markets wouldexpect a new debt issue to be junior to most outstand-ing debt in the event of a crisis, and thus demand ahigher interest rate compared to the present system.The effect on borrowing costs would reward prudentborrowing behavior and discourage overborrowing.Explicit seniority could also improve debt structuresby reducing incentives to issue “crisis-prone” debtforms that are hard to dilute.

Explicit seniority would also entail risks, however.In particular, an unavoidable consequence of limit-ing dilution and making new borrowing harder athigh levels of debt is that this may prevent somecountries from accessing debt markets in situationsof illiquidity, in turn increasing the likelihood of liq-uidity crises. Another potential drawback is that se-niority could complicate debt pricing and, as a re-sult, make debt more expensive (at least untilmarkets became familiar with the new system). Un-certainty would be increased by the possibility thatsovereigns find ways to circumvent seniority whentheir borrowing levels are elevated, for example, byobtaining direct bank loans under different jurisdic-tions or providing collateral for subsequent loans.

Finally, explicit seniority could have consequencesfor sovereign debt restructurings, an issue that is notanalyzed in this paper.

Explicit seniority in sovereign debt could be im-plemented in a number of ways, including statutes atthe international level; national statutes in debtorcountries and issuing jurisdictions; debt contracts; orsome combination of the three. This paper exploresideas for a contractual implementation of explicit se-niority in general terms and describes some of theobstacles. The two main difficulties that arise in acontractual framework are how to ensure that thesovereign continues to apply the first-in-time senior-ity structure to all subsequent borrowing and how toenforce the priority structure in the event of restruc-turing. This paper suggests an approach to deal withthose issues, although this area clearly requires fur-ther work.

While this paper concludes that explicit seniority isa novel approach to improving debt structures that isworthy of further research, it is only a first pass at theissue, and further research is needed before arrivingat a definite conclusion. In fact, while seniority couldbe beneficial for countries with moderate debt levels,it may make market access more difficult for coun-tries with elevated levels of debt: although desirablein many circumstances to prevent overborrowing,this could present new policy challenges. Moreover,an overall judgment would depend on the effects ofseniority on crisis resolution, which is not taken uphere. Further analysis would also be needed on howto overcome potential legal and practical obstacles tointroducing contract-based seniority. Nevertheless,given the potential benefits of explicit seniority forcrisis prevention—and other enhancements to bondcontracts that would also mitigate debt dilution—thispaper calls for further analysis and discussion of theissue.

Expanding the Set of Instruments:Real Indexation

Another key difference between sovereigns andcorporates is that sovereigns lack equity, or equity-like instruments, whereby investors would share insovereigns’ fortunes and misfortunes. Although equitycould never be fully reproduced in the sovereign con-text, the risk-sharing benefits of equity might be mim-icked through currently underutilized financial instru-ments with payment terms indexed to real variablessuch as gross domestic product (GDP) (Section V).

Real indexation involves higher payments wheneconomic performance is relatively strong, and lowerpayments when economic performance is relativelyweak. For example, countries could issue bonds providing for lower payments when GDP growth is

4

Toward Better Sovereign Debt Structures: A Road Map

weak or in the event of a natural disaster. Real index-ation would thus tend to stabilize the debt-to-GDPratio, providing two main benefits: first, it would re-duce the likelihood of debt crises and, second, itwould reduce the need for procyclical fiscal policies.

Indexation to variables largely outside the controlof the authorities, such as commodity prices, naturaldisasters, or output of trading partner countries,might provide considerable insurance benefits,though only to limited groups of countries. Indexa-tion to variables partly within the control of the au-thorities, such as GDP or exports, could provide sub-stantial insurance benefits to a broad spectrum ofcountries, though its introduction would presentgreater challenges.

The cost of such insurance for borrowing coun-tries is likely to depend on the extent to which anumber of obstacles can be overcome. In addition tothe need for large-scale issuance to ensure marketliquidity, the main obstacles seem to relate to theneed for investors to be able to hedge the risk in-volved in holding such instruments; the potential foropportunistic mismeasurement by country authori-ties of variables partly within their control; and pos-sible difficulties in pricing complex instruments.

The requisite large scale for launching new typesof bonds could be attained in the context of a debtrestructuring or through international coordination.Should a number of emerging markets issue GDP-indexed bonds, international investors holding aportfolio of such bonds would find GDP risk to bewell diversified, because the correlation of growthrates across emerging markets is typically very low.Reforms that would help overcome obstacles relatedto potential mismeasurement include strengtheningthe independence of national statistical agencies.

Toward Better Sovereign DebtStructures: A Road Map

Improved debt structures should not be viewed asa substitute for sound policies. Sound policies notonly reduce the likelihood of debt crises directly butare also a prerequisite for better debt structures andpossible financial innovations that would in turnmake countries less prone to crises. Nevertheless,this paper argues that improved debt structuresmight play a role in ameliorating economic perfor-mance and making crises both less likely and lessdamaging.

Historically, financial innovation seems to havetaken place in a somewhat haphazard manner, andhas often been prompted by intervention on the partof policymakers (Section VI). Innovations in theareas described above are unlikely to be an excep-tion to this historical norm, especially because the

incentives for individual market participants to inno-vate are likely to be lower than for the group as awhole.

A potential road map for implementing the policysteps analyzed in this paper is likely to require ef-forts by a number of different actors, includingcountry authorities, international investors, the inter-national community, and researchers.

Sound macroeconomic policies are by far themost important prerequisite for more desirable debtstructures. Indeed, excessive reliance on “risky”types of debt is primarily a symptom, rather than acause, of a perception of risk on the part of investors.Sound policies and credibility are also a precondi-tion for issuing new forms of debt, such as instru-ments involving elements of real indexation, and forminimizing potentially adverse effects on localbanking systems that may be large holders of gov-ernment debt.

Beyond better policies, country authorities couldseek to create or deepen the market for local-currency-denominated debt by issuing, for example,local-currency-denominated bonds with shorter ma-turities, and inflation-indexed bonds for longer ma-turities. In doing so, they should be alert to opportu-nities provided by private pension systems thatcreate a natural demand for local-currency and infla-tion-indexed debt, and in some cases GDP-indexeddebt. In these endeavors, the authorities need to bemindful of sequencing: in countries where long-termlocal-currency-denominated debt is widely held as aresult of restrictions on capital flows or on the rangeof assets that banks and institutional investors canhold, it would be crucial to establish greater credibil-ity before lifting such restrictions.

There are advantages of using instruments withreturns indexed to real variables closely related to is-suing countries’ economic performance. For thosesmall countries that are especially vulnerable to nat-ural disasters, disaster insurance would seem to bedesirable if available at a reasonable cost. Greateruse of hedging against commodity price fluctuationswould also seem desirable for countries relying on asmall set of commodities in their export and revenuestructure. Larger, more diversified countries (bothadvanced and emerging) will be better hedgedagainst macroeconomic fluctuations if they issuebonds indexed to a key macroeconomic aggregate,such as GDP.

Financial market participants’ willingness to en-gage in a dialogue with the official sector, and sharetheir views, expertise, and concerns regarding poten-tial innovations is an indispensable ingredient forprogress in improving debt structures. Market partici-pants can only be expected to explore innovations thatmake good business sense for them. However, twosets of considerations suggest that market participants

5

I OVERVIEW

may collectively have an incentive to participate insuch a dialogue. First, the initial costs associated withinnovation (including learning costs) are lower whenshared by market participants as a group than if in-curred individually. Second, innovations—includingsome in which the official sector played a major role,such as the creation of Brady bonds—have occasion-ally helped expand the scope of financial markets,thereby generating business opportunities.

The IFIs should continue to track short-term debtand foreign-currency debt as indicators of vulnera-bility. They should also encourage countries to bor-row in local currency and with longer maturities,while recognizing that crisis-prone debt structurestypically result from underlying problems that them-selves need to be addressed. To the extent that highshares of short-term or foreign-currency debt reflectpolitical economy pressures (perhaps motivated bythe electoral calendar) on debt managers to attainshort-run interest cost “savings” at the expense ofundue increases in the risk of crises, conditionalitywith respect to debt structure could be considered,on a case-by-case basis. However, its desirabilitywould have to be weighed against the costs thatmight result, for example, from reducing capitalmarket access for countries where short-term andforeign-currency instruments are the only ways ofpreserving it—possibly in the context of an incipientliquidity crisis.

While the IFIs’ primary goal in deciding upon thecurrency composition of their own debt issuancemust remain the minimization of borrowing costs,market development may continue to be a welcomeby-product. The first bond issues in a currency unfa-miliar to international markets require substantial

additional preparatory work: the IFIs are wellplaced to work with the authorities toward that end,though the costs in terms of staff resources shouldnot be neglected.

If relatively underutilized instruments such as inflation- or GDP-indexed bonds are deemed desir-able, their emergence could be aided in a number ofways: international dialogue among potentially in-terested parties; strengthened independence of coun-tries’ statistical agencies; and technical assistance toimprove the quality and transparency of national in-come statistics.

A number of potential steps analyzed in thispaper—such as the creation of an international debtregistry to help monitor seniority features of sover-eign debt held by private agents—would take some-what longer to implement. The desirability and prac-tical feasibility of such innovations in theinstitutional framework could be further explored.

Additional research would seem especially desir-able in the following areas:

• the determinants and consequences of domesticdebt structures (including the collection of dataon domestic debt for a large number of coun-tries);

• empirical evidence on debt dilution and the theo-retical case for and against seniority in the sover-eign context;

• surveys of investors’ and borrowers’ attitudes to-ward financial innovation and obstacles relatedto it; and

• the development of pricing models for currentlyunderutilized financial instruments.

6

Public debt in emerging market countries differsin several respects from that in advanced

economies. First, average debt levels were tradition-ally equivalent to a lower share of GDP in emergingmarket countries than they were in advancedeconomies; the gap has closed in recent years, partlyas a result of reductions in the debt of advancedeconomies (Figure 1). Second, reliance on externallyissued debt has been far greater in emerging marketcountries than in advanced economies. Third, whilethe structure of external debt of emerging marketcountries is similar to that of advanced economies,the structure of their domestic debt—in terms of ma-turity, currency composition, and the prevalence ofindexed debt—is very different.1

The remainder of the section reviews such differ-ences in greater detail, first, by considering externaland domestic debt separately and, second, by offer-ing a consolidated view.

Public Debt in Emerging MarketCountries Versus Advanced Economies

Structure of International Debt

The international sovereign debt of emerging mar-ket countries consists mainly of medium-term andlong-term fixed-interest-rate bonds denominated inforeign currency. Bank loans used to be the mainform of financing during the 1970s and 1980s. Astock conversion of loans into bonds took placethrough the Brady deals; this helps explain the dropin loans, and rise in bonds, in the early 1990s (Figure2, first panel). When developing countries re-enteredinternational credit markets in the 1990s, they did somainly through bond issues (Figure 2, secondpanel). The prevalence of bond financing is notunique to emerging market countries: its relative im-portance has grown even more sharply in advanced

II Facts on Existing Public Debt Structures

7

Note: The authors of this section are Marcos Chamon, OlivierJeanne, and Jeromin Zettelmeyer.

1In this paper, “external” (or “international”) and “domestic”refer to the jurisdiction where the debt is issued.

45

55

65

75

85

Advanced economies3

All emerging markets2

Large emerging markets1

0220009896941992

Domestic and Foreign Debt Issued by Emerging Market Countries

Composition of Total Debt:Unweighted Average, 1992–2002

0

20

40

60

80

100DomesticExternal

Tran

sition

Econ

omies

Middle

East

and

Africa

Latin

AmericaAsiaAll

Emer

ging

Marke

ts2

Larg

e

Emer

ging

Marke

ts1G-7

Figure 1. Advanced Economies and Emerg-ing Market Countries: Public Debt Stocksand Debt Composition(In percent of GDP)

Sources: IMF, World Economic Outlook (September 2003); and IMFstaff estimates.

1Argentina, Brazil, Chile, China, Hungary, India, Indonesia, Israel,Korea, Malaysia, Mexico, Philippines, Poland, Russia, South Africa,Thailand,Turkey, and Venezuela.

2Countries listed in footnote 1 above plus Bulgaria, Colombia,Costa Rica, Côte d’Ivoire, Croatia, Ecuador, Egypt, Jordan, Lebanon,Morocco, Nigeria, Pakistan, Panama, Peru, Ukraine, and Uruguay.

3Australia, Austria, Belgium, Canada, Denmark, Finland, France,Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand,Norway,Portugal, Spain, Sweden,United Kingdom,and United States.

II FACTS ON EXISTING PUBLIC DEBT STRUCTURES

economies, where virtually all international borrow-ing is through bonds.

Most internationally issued public debt carries afixed interest rate—for both advanced economies andemerging market countries. Much of the emergingmarket debt issued in the 1970s and 1980s had afloating rate. Certain types of Brady bonds also had afloating rate, and the proportion of floating-ratebonds was relatively high until the mid-1990s: in1994, about 40 percent of JPMorgan’s EmergingMarkets Bond Index Global (EMBIG) carried a float-ing rate. As new bond issues began to take off in themid-1990s and emerging markets moved from bankloans to bond financing; they also moved from floating-rate to fixed-rate instruments (Figure 3).Currently, floating-rate bonds make up less than 5percent of the EMBIG.

Most international sovereign debt of emergingmarket countries is issued at medium-term (5–10years) or longer-term maturities. However, the aver-age maturity of emerging market countries’ debt hasdeclined in recent years, and is now lower than for

advanced economies. The share of the EMBIG con-sisting of debt with remaining maturities over 20years declined from 40 percent in the mid-1990s toabout 20 percent by the end of 1999, and has stabi-lized since then (Figure 4, top panel). The declinestems from two factors. First, the 30-year bonds is-sued through the Brady deals are gradually becom-ing less important in the stock of emerging marketcountries’ debt as new debt is issued. Second, ex-cluding the Brady bonds, the average maturity ofnew emerging market countries’ bond and loan is-sues is significantly lower than for advancedeconomies, and has declined substantially over thepast two decades (Figure 4, bottom panel).2

Most sovereign debt issued internationally by bothemerging market countries and advanced economiesis in foreign currency (Table 1). Only a handful ofadvanced economies issue a substantial share oftheir international sovereign debt in their own cur-rency. This does not imply that governments inemerging market countries and advanced economiesface the same constraints in the international debtmarket. Indeed, some advanced economies issuelocal-currency debt in their home markets and attractinternational investors to purchase it there.

8

2Certain types of short-term debt that have been widely cited asa source of fragility in some recent crises, such as the MexicanTesobonos in 1994 and the Russian GKOs in 1998, were issueddomestically and, hence, were not international debt in the senseused here.

Stocks of Privately Held Debt

New Issues

0

50

100

150

200

250

300

350

Loans

Bonds

2000969288841980

0

20

40

60

80

100

120

Loans

Bonds

032000969288841980

Figure 2. Structure of External Public Debtin Emerging Market Countries(In billions of U.S. dollars)

Sources:World Bank, Global Development Finance (2003); CapitalData, Bondware and Loanware.

0

20

40

60

80

100

Floating interest rate

Fixed interest rate

0220009896949290888684821980

Figure 3. Emerging Market Countries:Fixed- Versus Floating-Rate Sovereign Bond Issues(In billions of U.S. dollars)

Source: Capital Data, Bondware and Loanware.

Public Debt in Emerging Market Countries Versus Advanced Economies

Structure of Domestic Debt

The composition of government debt issued onthe domestic market is much more heterogeneousacross countries than is the composition of debt is-sued internationally. While there are differences be-tween emerging market countries and advancedeconomies, these are overwhelmed by differencesamong emerging markets. On average, emergingmarket countries rely on long-term local-currencydebt to a lesser extent than do advanced economies,but there is substantial variation among emergingmarkets: such debt is virtually absent in Latin Amer-ican countries, but represents more than one-half oftotal debt for several countries in emerging Asia andemerging Europe (Table 2).

More generally, emerging market countries displayremarkable differences in terms of the breakdown ofdomestic government bonds into five categories:local-currency long-term fixed-rate, local-currency

short-term fixed-rate, floating-rate (indexed to a do-mestic rate), inflation-indexed, and foreign-currency.Each of the first four categories represents more thanone-half of total debt in at least one country. There isalso significant heterogeneity within regions. In Asia,for example, some issuers (Indonesia and Malaysia)do not make use of long-term local-currency debtwhereas other issuers (India, Taiwan Province ofChina) rely almost exclusively on this form of debt.In Latin America, some countries (Venezuela) borrowmainly through interest-rate-indexed debt, whereas inothers (Chile) the majority of government debt is in-dexed to inflation.

A Consolidated View

Domestic and international debt markets are sub-stitute sources of finance and should be consideredjointly. International and domestic debt markets havebecome integrated, with residents and nonresidentsbeing active in both. For example, in both Argentinaand Uruguay, more than half of the internationaldebt was held by residents. Conversely, foreign resi-dents held more than 80 percent of Mexican domes-tically issued Tesobonos and Cetes before the 1994crisis (IMF, 1995) and a large share of RussianGKOs and OFZs issued before the 1998 crisis (about30 percent, according to IMF staff estimates). Sover-eigns seem able to choose whether to tap interna-tional debt markets or domestic debt markets. In ad-dition to market conditions, this choice depends ontwo sets of considerations. First, domestic publicdebt is likely to crowd out financing to the domesticprivate sector, because some firms may be con-strained to the domestic credit market owing to highcosts of borrowing abroad. Second, international in-vestors are less protected in domestic jurisdictions.

Obtaining an integrated picture of debt structurethat includes both domestic and international debt isdifficult owing to data limitations, especially for do-mestic debt. Since the 1994 Mexican crisis, the in-ternational community has made an effort to im-prove the availability of cross-country statistics.3This effort has focused primarily on external debt, inspite of the important role played by domestic debtin several recent emerging market crises (Mexico in1994; Russia in 1998; and Brazil in 1998). Neverthe-less, it is possible to establish some basic stylizedfacts on total (domestic plus international) govern-ment debt.

Although advanced economies have traditionallyrelied on domestic debt markets to a greater extentthan have emerging market countries, there seems tohave been some convergence in this regard in recent

9

3See IMF (2003c).

0

40

80

120

160>20 years 10–20 years 5–10 years <5 years

02200098961994

Emerging Market Countries:Remaining Maturity Structure of the EMBIG1

(In percent)

Average Maturity at Issue of Sovereign Debt Issues(In years)

4

6

8

10

12

14

Emerging market countries

Advanced economies

022000969288841980

Figure 4. Structure of Internationally IssuedDebt: Maturity Composition

Sources: JPMorgan, Emerging Market Bond Index Global(EMBIG); Capital Data, Bondware and Loanware.

1Percentages in first three years do not add up to 100 becauseof instruments with unknown remaining maturity.


10

Table 1. External Sovereign Debt: Currency Composition, 1980–20031

(Unweighted average, in percent)

Emerging Market AdvancedCountries2 Economies3

Domestic currency4 0.3 7.5

Foreign currency 99.7 92.5U.S. dollar 54.8 42.4Euro 14.6 6.5Japanese yen 14.0 14.5Deutsche mark5 11.7 11.1European Currency Unit (ECU)5 . . . 8.0Others6 4.6 10.0

Source: IMF, Bonds, Equities and Loans database.1All bond and loan issues, 1980–2003.2Argentina, Brazil, Chile, China, Hungary, India, Indonesia, Israel, Korea, Malaysia, Mexico, Philippines, Poland, Rus-

sia, South Africa,Thailand,Turkey, and Venezuela.3Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Nether-

lands, New Zealand, Norway, Portugal, Spain, Sweden, United Kingdom, and United States.4Includes euro issues (but not European currency unit issues) for European Monetary Union member countries.5Prior to the introduction of the euro.6Includes Italian lira, British pound, French franc, and Swiss franc.

Table 2. Structure of Domestically Issued Government Bonds at End-2001(In percent of total)

DomesticDomestic-Currency-Denominated Bonds

Foreign-____________________________________________

Government Indexed to Currency-_____________________Bonds/GDP Not indexed Domestic Denominated______________________(In percent) Long term1 Short term1 interest rate Inflation Bonds

Emerging market countries 28.2 41.5 18.6 26.4 7.2 6.3

Latin America 24.0 5.6 13.7 50.8 16.6 13.4Brazil 52.1 9.5 0.0 53.0 7.0 30.5Chile2 . . . 0.0 21.0 0.0 55.8 23.2Mexico 11.0 12.8 23.6 60.1 3.5 0.0Venezuela 9.0 0.0 10.0 90.0 0.0 0.0

Asia 26.6 52.4 16.5 22.2 7.8 1.1India 27.0 81.6 18.4 0.0 0.0 0.0Indonesia 34.0 24.6 0.0 30.9 38.9 5.6Malaysia 36.3 0.0 19.8 80.2 0.0 0.0Thailand 13.8 91.5 8.5 0.0 0.0 0.0Philippines 22.0 64.2 35.8 0.0 0.0 0.0

Europe and others 32.4 56.5 23.6 13.6 0.4 5.9Czech Republic . . . 41.1 58.9 0.0 0.0 0.0Hungary 27.0 56.0 23.0 21.0 0.0 0.0Poland 16.0 62.6 26.5 10.9 0.0 0.0Slovak Republic 29.0 86.8 13.2 0.0 0.0 0.0South Africa 35.0 92.4 5.2 0.0 2.4 0.0Turkey 55.0 0.0 14.5 49.9 0.0 35.6

Sources: IMF staff estimates; and JPMorgan, Guide to Local Markets (2002).1Short term is defined as an initial maturity of less than one year, and long term is defined as an initial maturity of more than one year.2For Chile, the shares refer to bonds issued by the central bank.The amount of bonds issued domestically by the central government is negligible.

Sovereign Versus Corporate Liability Structures

years (Figure 5). The domestically issued compo-nent of total public debt has increased at a quickerpace than the foreign component in emerging marketcountries, especially in the 1990s. This may reflectdevelopment of the domestic markets, though the in-crease in total debt also mirrors fiscal expansions.

Nevertheless, much of the increase in domesticallyissued debt has continued to take the form of foreign-currency debt and short-term debt. The differencesbetween advanced economies and emerging marketcountries in this respect have thus persisted (Table 3).Consolidating international and domestic debt, about47 percent of central government debt in emergingmarket countries was denominated in (or indexed to)a foreign currency in 2001, against about 5 percentfor advanced economies. Long-term local-currencydebt represented 76 percent of total debt in advancedcountries, 36 percent of total debt in emerging marketcountries, and 15 percent of total debt in Latin Amer-ica. While inflation-indexed public debt issues havebeen at the forefront of the development of domesticdebt markets in a few countries, foreign-currency is-sues have been key in most others.


Moving to a comparison between sovereigns andcorporates, the richer liability structure of corporatesis apparent along three dimensions. First, and by farthe most striking, a large share of corporate liabili-ties consists of outside equity.4 Moreover, corporatesmake use of financial instruments that combine debtand equity, such as convertible bonds—the holdercan convert these bonds into stocks at a predeter-mined exchange ratio at prespecified dates. In con-trast, sovereigns lack not only equity but also equity-like instruments that would make returns a directfunction of variables such as tax revenues, the fiscalbalance, or GDP.5 Second, corporations make exten-sive use of collateralized (secured) debt as well asseniority distinctions within unsecured debt. In con-trast, sovereigns issue comparatively little secureddebt, and unsecured debt is not formally prioritized.Third, corporate bond contracts sometimes include

covenants that place restrictions on future financingdecisions of the firm, by placing limits on total in-debtedness, or restricting the issuance of debt at thesame or higher levels of seniority. No such restric-tions exist at the sovereign level, at least in debt con-tracts with private creditors.6

Use of Secured and Subordinated Debt

Corporations in advanced economies typicallyissue liabilities belonging to several classes with dif-ferent priority in the event of liquidation or bank-ruptcy reorganization. These include secured debt,7ordinary unsecured debt, subordinated debt, preferredstock, and common stock. Secured debt gives the titleto a pledged asset to the debtor. The remaining liabil-ity classes define an absolute priority ranking: in theevent of liquidation, each of them is to be repaid onlyif the higher ranking class was repaid in full.8 Both

11

6IMF programs typically restrict total nonconcessional debt(see below in this section).

7Including capital leases, in which the contract promises afixed payments stream, and the leased capital asset can be seized(repossessed) by the creditor in the event of default.

8In U.S. bankruptcy reorganizations under Chapter 11, absolutepriority is frequently violated: equity holders often receive securi-ties of positive market value even though some debt holders arenot fully repaid (Franks and Torous, 1989; Weiss, 1990). How-ever, deviations from absolute priority among debt holders seemto be rare.

4In a sample of 5,000 U.S. industrial corporations surveyed byBarclay and Smith (1995), the average ratio between debt andcommon stock was about 1 in 3, that is, common stock made upabout 75 percent of total firm value on average.

5Informally speaking, a country’s currency might be viewed ashaving a few of the features of equity. Investors holding currency,or other local-currency-denominated assets, share in the fortunesof the issuing country: as the real exchange rate is correlated witheconomic performance, real returns tend to be higher when thecountry’s economic growth is relatively strong.

15

20

25

30

35

40

Domestically issued

Internationally issued

022000989694921990

Figure 5. Emerging Market Countries:Structure of Public Debt(In percent of GDP)

Source: IMF staff estimates.


12

Table 3. Structure of Total (Domestic and External) Central GovernmentDebt, 2001(In percent of total)

Long-Term Total Debt as a Foreign-Currency Domestic- Share of GDP

Debt1 Currency Debt1 (In percent)

Emerging market countries 47.1 35.7 48.8

Latin America 67.9 15.2 37.0Argentina 96.8 . . . 53.7Brazil 43.82 3.33, 4 66.22, 5

Chile 92.7 0.0 15.6Mexico 35.6 57.56 22.6Venezuela 70.6 0.0 27.0

Asia 29.3 53.7 56.5China 17.7 82.3 24.0India 14.5 69.97 65.17

Indonesia 46.0 51.06 90.9Malaysia 16.7 0.0 69.2Philippines 47.4 17.68 64.9Thailand 33.6 78.8 24.8

Others 47.6 17.4 51.3Poland 34.8 34.8 39.3Russia 90.3 . . . 50.0South Africa 14.4 61.23, 4, 8 46.8Hungary 30.1 44.09 52.1Turkey 68.2 0.0 68.5

Advanced economies6 5.6 75.9 51.8Australia 0.53 85.53 10.1Belgium 1.3 80.4 101.2Canada 6.03 41.53 40.3Denmark 12.1 69.0 52.5Finland 16.3 66.7 45.4France 0.0 92.0 49.4Germany 0.0 97.0 35.3Italy 3.2 86.8 102.6Japan 0.0 74.8 121.6Netherlands 0.0 81.0 43.1New Zealand 18.53 48.23 31.2Norway 3.73 60.53 18.4Portugal 8.6 87.4 58.9Spain 4.13 83.83 47.2Sweden 20.1 55.9 51.6United Kingdom 1.4 60.5 38.6United States 0.0 58.8 33.1

Sources: IMF staff; OECD, Central Government Debt Yearbook 1992–2001; and websites of the country authorities.1In percent of total central government debt for emerging market countries and total central government mar-

ketable debt for advanced economies.2Includes debt held by the Central Bank.3Based on residual maturity.4Only marketable domestic-currency bonds.5Consolidated government debt.6Includes debt indexed to inflation and domestic interest rates.7Includes debt owed to National Small Savings Funds.8Includes debt with maturities of three years or more.9Data for 2002.


secured and subordinated debt make up significantportions of the corporate debt stock.9

In contrast, sovereign liabilities generally fall intojust two classes—secured debt and unsecured debt.Within the unsecured debt class, there is no distinc-tion between ordinary debt and subordinated debt.Secured sovereign claims are generally collateralizedby future receipts, such as oil revenue or other exportreceivables (Chalk, 2002; and IMF, 2003d). To serveas collateral, future receipts need to be removed fromthe direct control of the sovereign, that is, the transac-tions associated with future receipts need to occurunder foreign jurisdiction. For this reason, collateral-ized future receipts typically involve export revenuesaccruing to the government (usually through a publicenterprise), rather than domestic tax revenues.

Secured debt is a far smaller proportion of totaldebt for sovereigns than it is for corporations. Of the79 developing and emerging market countries thathad at least one public sector (including public en-terprises) international loan or bond outstanding onJanuary 1, 2003, about half (39 countries) owed col-lateralized loans or bonds. However, the face valueof collateralized debt was only 6.2 percent of theface value of total debt outstanding (7.1 percent inthe group of countries that had some collateralizeddebt). The share of collateralized loans or bonds intotal loans and bonds was higher than 25 percent inonly 9 countries. In the postwar era, secured sover-eign debt is a relatively recent phenomenon: the firstmodern collateralized bond was issued by Mexico in1988 (Figure 6).10

Financing Restrictions in Debt Contracts

Corporate bond contracts in the United States haveoften contained “negative covenants” that restrict fu-

ture financing decisions (Smith and Warner, 1979;Asquith and Wizman, 1990; and Goyal, 2003). Theseinclude clauses that place restrictions on net worth ortotal debt, possibly excepting subordinated debt.These clauses protect creditors from dilution throughadditional debt issued in the same seniority class (seeSection IV). Creditor protections of this type—inparticular, restrictions on future debt issues—aregenerally absent from sovereign debt.11

Some elements of IFI conditionality could be in-terpreted as analogous to negative covenants, in thesense that—among other purposes—they serve toprotect the financial interests of IFI creditors by re-stricting the borrower’s financing decisions (e.g., bylimiting the fiscal deficit or placing limits on exter-nal debt). Analogous conditions or covenants do notexist in privately held debt.

13

9Secured debt constituted 53 percent of the debt of the averageU.S. industrial firm (38 percent if leases are excluded), ordinarydebt 35 percent, and subordinated debt 12 percent (Barclay andSmith, 1995). Large firms tend to issue more ordinary and subor-dinated debt, whereas smaller firms tend to issue more secureddebt. About 75 percent of firms issued liabilities in at least threepriority classes, that is, common stock plus at least two debtclasses (or preferred stock and at least one debt class). The resultsare similar for the United Kingdom (Lasfer, 1999).

10Sovereigns made greater use of collateralized bonds in thepre–World War I era. Collateral usually took the form of infra-structure (especially railways), raw materials, or, especially whenbonds were issued following a debt restructuring, tax revenues(Mauro and Yafeh, 2003).

11The main exception is the “negative pledge clause” in sover-eign bond contracts and official debt, which prohibits new collat-eralized debt unless the incumbent debt holders are given anequal claim on the collateral.

0

20

40

60

80

100Uncollateralized bonds and loansCollateralized bonds and loans

0220009896949290888684821980

Figure 6. All Developing Countries: PublicSector Bonds and Loans Issued inInternational Markets1

(Total issues, in billions of US. dollars)

Source: Capital Data, Bondware and Loanware.1Developing countries defined to include emerging markets.

Bonds and Loans issued by sovereigns and public sector enter-prises, excluding project financing.

Existing debt structures in emerging market coun-tries seem to rely excessively on risky forms of

debt—such as short-term and foreign-currencydebt—which may amplify the economic cycle, in-crease the likelihood of crises, and make crises moredifficult to manage. Increases in risky forms of debtmay be the result of worsening debt sustainability,but they also reinforce the rise in vulnerability.

Problems with the Status Quo

Short-Term Debt

Empirical studies have found short-term debt to bea leading indicator of vulnerability to internationalfinancial crises (Bussière and Mulder, 1999; and Ro-drik and Velasco, 1999). Theoretical models haveput forward two types of mechanisms to explain thisempirical association.

First, short-term debt may make governmentsmore vulnerable to debt rollover crises. Indeed, thisseems to have been an important factor in triggeringthe recent crises in Mexico (1994) and Russia(1998). In the extreme case of a pure liquidity crisis,investors stop lending to the government simply be-cause they expect others to do the same. If the aver-age maturity of the debt is low, the government isthen at the mercy of self-fulfilling creditor panicsthat can be triggered by shifts in market sentiment(Sachs, 1984; Alesina, Prati, and Tabellini, 1990;Cole and Kehoe, 2000; and Chamon, 2003). In theless extreme but probably more realistic case wherethe crisis mixes elements of illiquidity and insol-vency, the government would be vulnerable to apiece of bad news, whose real impact would be am-plified by creditors’ unwillingness to roll over theirclaims (Jeanne, 2004; in the corporate context, seeDiamond, 1991).

Second, short-term debt can give rise to viciouscircles stemming from the two-way interaction be-

tween debt levels and interest rates. If the debt has ashort maturity or bears a floating interest rate,changes not only in the international interest ratesbut also in the country’s own creditworthiness willaffect the interest bill relatively quickly. A sovereignwith a high level of short-term debt may thus find it-self trapped in a bad equilibrium in which high inter-est payments lead to a high probability of default,which in turn increases the default risk premium andthe interest rate (Calvo, 1988).

More generally, the relatively short average matu-rity of debt in emerging markets may also amplifythe economic cycle. In emerging market countries,economic downturns are typically associated withincreases in interest rates because of increases in thedefault and devaluation risk premiums, thus reduc-ing the scope for countercyclical fiscal policies.1

Foreign-Currency Debt

The vulnerabilities created by significant levels ofdebt denominated in (or indexed to) a foreign cur-rency have also been evident in several recent crises,an aspect emphasized, for both private and publicdebt, in the balance sheet approach to crises (Allenand others, 2002). In fact, the depreciation of the localcurrency has often led to a sharp increase in govern-ment debt as a share of domestic GDP or fiscal re-ceipts (Figure 7). The contribution of this revaluationwas especially large in Argentina and Uruguay—twocountries where the fraction of debt denominated inforeign currency and the depreciation of the local cur-rency were substantial. By contrast, it was small inKorea, where the government had little foreign-currency debt, and in Turkey, where the real deprecia-tion of the currency was moderate.

Not only does foreign-currency debt make crisesmore severe but it also reduces the scope for domes-tic policies to alleviate the impact of crises. The

III Rendering Debt Structures Less Crisis Prone with Existing Instruments

14

1GDP growth is negatively correlated with the spread on foreign-currency bonds issued internationally by emerging mar-ket countries (Eichengreen and Mody, 1998). This effect is likelyaugmented, for domestic currency borrowing, by the expected de-preciation premium.

Note: The authors of this section are Marcos Chamon andOlivier Jeanne.

Determinants of Government Debt Structure

revaluation of government debt amplifies the initialfiscal problem and reduces the ability of the govern-ment to implement policies that might mitigate thedisruption in the private sector (Jeanne andZettelmeyer, 2002). In addition, monetary policycannot be used to inflate foreign-currency debt away.The government is faced with the well-known andmuch-debated dilemma of choosing between raisingthe interest rate and letting the currency depreciate,with both options having adverse effects on domes-tic balance sheets.

Such amplification mechanisms resulting fromhigh shares of foreign-currency debt create the poten-tial for vicious circles and can thus make countriesmore vulnerable to crises in the first place. Imbal-ances in domestic balance sheets would lead investorsto attack the local currency; with the resulting depre-ciation, balance sheets would in turn deteriorate evenfurther (Krugman, 1999; Aghion, Bacchetta, andBanerjee, 2001; and Jeanne and Zettelmeyer, 2002).


Why do governments in emerging market coun-tries and their lenders settle on debt that seems to beunduly crisis prone, even though they are the firstones to suffer the costs in a crisis? Lack of credibilityof monetary and fiscal policies seems to be an impor-tant factor. Other factors may also be at work, includ-ing the nature of the domestic investor base and char-acteristics of domestic financial regulation (in thecase of domestic debt), as well as the country’s eco-nomic size (in the case of international debt).

Credibility of Monetary and Fiscal Policies

Governments in many emerging market countriescannot borrow on the same terms as advancedeconomies because of lack of credibility of mone-tary and fiscal policies. Unsustainable policies leadcreditors to anticipate that the government will ex-propriate them in one way or another—directlythrough a default or indirectly through inflation ifdebt is denominated in local currency. Thus, not onlydo governments face higher and more volatile inter-est rates, and—from time to time—loss of marketaccess, but they are also under pressure to issue debtinstruments that are more prone to crises.

Lack of credibility plays an especially importantrole in the period leading up to crises, as govern-ments tend to shift the composition of their debt to-ward shorter maturities and foreign-currency de-nomination. Notable examples include Mexico’sshift to Tesobonos in 1994 and Brazil’s switch to-

ward short-term, foreign-currency, and floating-ratedebt in 1998.2 These examples are consistent withthe view that short-term debt is a symptom, ratherthan a cause, of an impending crisis.3 With a loom-ing crisis, investors would usually prefer to holdshort-term debt for two reasons:

• Dilution. First, investors holding short-term debtmay expect the government to repay them beforethe default actually takes place (Rogoff, 1999).Thus issuing short-term debt dilutes the out-standing long-term debt.4

15

2In the period leading up to Mexico’s 1994 crisis, both thestock of Tesobonos and the interest rate differential between thepeso-denominated Cetes and the dollar-linked Tesobonos rosesharply (IMF, 1995, pp. 53–69, Figures I–6 and I–7).

3Indeed, systematic panel regressions show that, while highershares of short-term debt are associated with greater likelihood ofdebt crises, that association is no longer significant when takinginto account that greater use of short-term debt is itself influencedby other factors, including low credibility (Detragiache andSpilimbergo, 2001).

4Anticipating dilution, investors could be reluctant to lend on along-term basis or be willing to do so only at high interest rates(Bolton and Jeanne, 2004). In fact, the government could useshort-term debt to postpone the necessary adjustment at the ex-pense of long-term creditors. However, such dilution might turnout to be desirable, even for long-term creditors, if it bought timeto permit the necessary adjustment and avoided a default.

0

20

40

60

80

100

120Contribution of exchange rate depreciationIncrease in the public debt-to-GDP ratio

Argen

tina

2000

–02Uru

guay

2000

–02Tu

rkey

2000

–02

Braz

il

2000

–02

Russia

1997

–99

Indon

esia

1996

–98

Korea

1996

–98Tha

iland

1997

–98

Figure 7. Recent Crises: Impact ofExchange Rate Depreciation on PublicDebt-to-GDP Ratio1

(In percentage points of GDP)

Sources: IMF, Country Reports and International Financial Statistics.1The contribution of exchange rate depreciation is measured as

the increase in the precrisis debt-to-GDP ratio that results fromsetting the precrisis real exchange rate to its postcrisis level.

III RENDERING DEBT STRUCTURES LESS CRISIS PRONE

• Discipline. Second, investors may expect that ifthe government deviated from desirable policies,it would soon face higher interest rates or a roll-over crisis (Diamond, 1993; Diamond and Rajan,2001; and Jeanne, 2004). Short-term debt may bethe best option when the need for discipline out-weighs the expected costs—to both the borrowerand its lenders—resulting from the possibility ofa rollover crisis.

High-inflation episodes durably change the struc-ture of government debt. Indeed, no country that hasexperienced hyperinflation has a significant fractionof its government debt in long-term local-currencyinstruments (Table 3). Furthermore, in an admit-tedly limited cross section of countries for whichdata are available, the average inflation rate in1970–90 is negatively associated with the share oflong-term local-currency instruments in governmentdomestic debt in 2001 (Figure 8; Jeanne, 2003).

The persistence of foreign-currency and indexeddebt—in some cases long after disinflation or fiscaladjustment have been achieved—may reflect twofactors. First, it often takes decades for countries togain anti-inflationary credibility: some countriesmay even be trapped in a situation where they can-

not prove that they can be trusted with long-termlocal-currency debt because of creditors’ fears thatsuch debt would be inflated away (Jeanne, 2003).Second, private agents may get used to a certaintype of instrument, and impediments to financialinnovation may hamper the transition to a differentfinancial structure.

The prevalence of foreign-currency or inflation-indexed debt may be the result of past disinflationand fiscal stabilization efforts. When inflation ishigh because of an underlying fiscal problem, theauthorities may be caught in a vicious circle inwhich inflationary expectations imply high interestrates on local-currency debt; in turn these make itmore difficult to stabilize the fiscal and monetarysituation (Calvo, 1988). The government can seekto break such vicious circles by borrowing in for-eign currency, or with indexed debt, at lower inter-est rates. If a lower interest rate is not sufficient to make the debt dynamics sustainable, and a fiscaladjustment is required, foreign-currency debt andinflation-indexed debt can make the government’scommitment to the adjustment more credible, be-cause they cannot be inflated away (Calvo andGuidotti, 1990).

A relatively large share of foreign-currency or in-dexed debt may result from investors’ and borrowers’attempts to protect themselves from uncertainty in anenvironment of high and variable inflation (Ize andParrado, 2002). With long-term domestic-currencydebt, the future real burden of debt is very uncertainat the time of issue, and could be unsustainably highif inflation turns out to be much lower than expected.This risk is especially relevant for countries with animperfectly credible fixed exchange rate peg andhigh domestic interest rates. Borrowing at a high in-terest rate in domestic currency effectively involves abet that the currency will be devalued—a bet that canturn out to be very costly in the event the devaluationdoes not occur. In such an environment, foreign-cur-rency debt, CPI-indexed debt, or floating-rate debtmay be less risky than long-term local-currency debt(Jeanne, 2003).

From the point of view of borrowing countries,the relative desirability of local-currency debt,foreign-currency debt, and CPI-indexed debt canbe assessed by looking at the unit in which domes-tic output is the most stable (Box 1). For mostemerging markets, relatively volatile inflation im-plies that GDP is far more stable when expressed in foreign-currency terms than in local-currencyterms. (The opposite holds for the G-7 countries;see also Fontenay, Milesi-Ferretti, and Pill, 1997;and Missale, 1999.) Thus a simple hedging motivemay help explain the greater reliance on foreign-currency debt in emerging market countries. Interestingly, for both the G-7 and emerging mar-

16

Share of Long-Term Local-Currency Bonds in Total Government Domestic Bonds, 2001 (In percentage points)

Log of Average Inflation, 1970–900 1 2 3 4 5 6 7–20

0

20

40

60

80

100 China

Thailand

South Africa

Slovak RepublicIndia

Philippines

Hungary

Poland

Brazil

Indonesia

Mexico

Chile

Turkey

VenezuelaMalaysia

Figure 8. Share of Long-Term Local-Currency Bonds in Total Government Domestic Bonds and Inflation History

Sources: IMF, International Financial Statistics; and JPMorgan.

ket countries, the volatility of GDP is smallestwhen GDP is expressed in terms of CPI-indexedunits, suggesting that CPI-indexed bonds may provide substantial advantages to both groups ofcountries.

Finally, governments may borrow in foreign cur-rency to reduce nominal interest payments (whichare high in local-currency terms owing to the infla-tion premium) and thus the headline fiscal deficit(Blejer and Cheasty, 1991). Typical measures of

the public deficit take into account the flow of in-terest payments but not the changes in the realvalue of the principal due to currency depreciationor inflation.5

17


Box 1. Debt Structure and Hedging

In assessing the relative merits of local-currencydebt, foreign-currency debt, and inflation-indexed debt,a key criterion is which type of debt results in the low-est probability of the debt-to-GDP ratio exceeding agiven threshold. This is equivalent to asking whetherthe volatility of output is lowest when output is ex-pressed in terms of local currency, dollars, or the con-sumer price index. In fact, debt commits the borrowerto repay a fixed quantity in terms of some unit—thelocal currency, a foreign currency, or, for inflation-indexed debt, a price index. For example, a 10-yeardollar-denominated zero coupon bond issued in 2004commits the government to repay a certain quantity D$

of dollars in 2014. Assuming that this is the only debtissued by the government, the debt-to-GDP ratio in2014 is given by D$/Y$, where Y$ is the country’s GDP in 2014 expressed in terms of dollars. Viewed from2004, the principal debt repayment due is known (with no uncertainty): the likelihood that the debt-to-

GDP ratio will exceed a given threshold is therefore en-tirely determined by the volatility of Y$. More gener-ally, debt denominated in unit A is preferable to debtdenominated in unit B if output is less variable whenexpressed in unit A than in unit B. Thus one way of as-sessing the relative desirability of local-currency debt,dollar debt, and CPI-indexed debt is simply to comparethe volatility of output expressed in terms of local cur-rency, dollars, and the consumer price index. In the ex-ercise conducted below, volatility is defined more pre-cisely as the standard deviation of those changes in the10-year growth rate of output (in each of the threeunits, considered in turn) that cannot be predicted on the basis of past output growth. For advancedeconomies, or G-7 countries, the volatility of output islower when denominated in local currency than it iswhen denominated in foreign currency, but for emerg-ing market countries, this is reversed, owing to theirhigher and more volatile inflation.

5Several countries where inflation is an important considera-tion report an “operational” fiscal balance that includes the effectof inflation on the debt principal.

0

50

100

150

200

250

Dollar Indexed

Inflation Indexed

Local Currency

Emerging Market Countries

0

5

10

15

20

25

Dollar Indexed

Inflation Indexed

Local Currency

Advanced Economies

SD of Cumulative GDP Growth

SD of Cumulative GDP Growth

Standard Deviation of Cumulative 10-Year Unexpected Growth(In percentage points)

Source: IMF, International Financial Statistics.Note:The sample of emerging markets consists of 11 countries. (None experienced annual inflation rates higher than 100 percent in any year

during 1955–2000).The result that GDP is more stable when expressed in foreign-currency terms rather than in local-currency terms becomeseven more striking when countries that experienced inflation rates above 100 percent are included in the sample. The sample of advancedeconomies consists of 21 countries. Unexpected output growth is the difference between actual growth and the growth predicted by an autore-gressive (AR (1)) process.The sample period is 1965–2000. For a given year, growth is predicted for the subsequent 10 years based on data forthe previous 20 years.

Domestic Investor Base, Financial Regulation, and Pension Systems

A large base of domestic investors may be expectedto make it easier for a country to absorb shocks tocapital flows and, more specifically, for the govern-ment to issue debt domestically and in local currency,if this is not precluded by a history of high inflation(IMF, 2003e). Indeed, countries with a larger domes-tic investor base—as proxied by the ratio of bank de-posits to GDP—are found to have a smaller share offoreign-currency bonds in total (private and public)bonds (Claessens, Klingebiel, and Schmukler, 2003).The type of pension system has important conse-quences for the size of the domestic investor base,and thus the development of the domestic governmentdebt market. Prefunded pension systems are likely toinduce significant domestic savings available for in-vestment in government domestic debt, though inpractice the relationship between pension systemsand the share of domestic debt is not straightforward,probably because of the presence of other determi-nants of debt structure.

Governments may also create a captive market fortheir debt—including long-term debt—through fi-nancial regulation, moral suasion, or direct control of

financial institutions. Many countries have regula-tions that prevent domestic pension funds from in-vesting more than a fraction of their portfolios in foreign assets: such regulations can be quite con-straining, even in the case of OECD countries (Fischer and Reisen, 1994). Moreover, in financiallyrepressed countries, the government may induce do-mestic banks to buy its debt, especially where thegovernment controls a large share of the banking sys-tem. Finally, the presence of capital controls has beenfound to be associated with the share of local-currency borrowing in the domestic credit market(Hausmann and Panizza, 2002).

Indeed, the share of long-term local-currency in-struments in total government domestic debt is highnot only in financially liberalized countries with rel-atively strong policy credibility, such as many ad-vanced economies, but also in countries with lowerdegrees of financial liberalization (Figure 9). Thismay be tentatively interpreted in terms of threestages in the development of domestic governmentdebt markets. In a first stage, in some developingcountries, financial repression forces residents tohold long-term nominal local-currency debt. Ascountries develop and move to a second stage, how-ever, they often ease restrictions even before estab-lishing strong credibility: investors thus may shift toother forms of debt. In a third stage, countries attaincredibility while refraining from restrictions, and in-vestors hold long-term local-currency debt voluntar-ily, as in advanced economies.

Political Economy Determinants

The determinants of government domestic debtstructures are rooted in the domestic political econ-omy. Explanations under the heading of “lack of pol-icy credibility” are ultimately related to differentways in which the government can reduce the bur-den of its obligations to creditors. The risk of non-payment is in turn determined by the stability of thegovernment and the weight that creditors carry in thepolitical arena. Great Britain in the seventeenth cen-tury is an especially interesting example, because itwas the first country to develop a government debtmarket comparable in size to those observed now inadvanced economies. It has been argued that thiswas made possible by the new system of checks andbalances set up after the Glorious Revolution, whichgave creditors more control over the government(North and Weingast, 1989). A domestic creditorconstituency may also help ensure that the govern-ment will respect creditor rights.

Political economy considerations have some im-plications for the link between financial repressionand the development of the domestic debt market.Even if long-term nominal debt were the result of fi-

18


Financial Liberalization Index, 1996

Share of Long-Term Local-Currency Bonds in Total GovernmentDomestic Bonds, 2001(In percentage points)

China

India

Philippines

Thailand

South Africa

Singapore

Hong KongSAR

IsraelIndonesia

Mexico

ChileTurkey

MalaysiaBrazil

Venezuela

0 0.2 0.4 0.6 0.8 1.0–20

0

20

40

60

80

100

Figure 9. Share of Long-Term Local-CurrencyBonds and Financial Liberalization

Sources: Abiad and Mody (2003); IMF, International FinancialStatistics; and JPMorgan.

Note:The financial liberalization index refers to 1996: the latestavailable from Abiad and Mody (2003). Although the quadraticterm in the regression corresponding to the figure is statisticallysignificant, the small number of observations suggests the U-Shapeshown above should be interpreted as merely suggestive.

nancial repression, this would not necessarily implythat the government would seek to expropriate itscreditors. Defaulting on a debt that is held by domes-tic banks would likely generate a financial melt-down. Thus, the government’s incentives to avoid adefault are likely greater if domestic banks hold thepublic debt.

Measures of political stability and rule of law arepositively correlated with the domestic public debtas a share of GDP (Figure 10). This is consistentwith previous findings of a significant correlationbetween the size of the domestic local-currencybond market (including both private and public debt)and political economy variables, such as rule of lawand democracy (Burger and Warnock, 2003; andClaessens, Klingebiel, and Schmukler, 2003).

Finally, historical accidents may contribute toexplaining the development of domestic debt mar-kets in local currency: changes in financial struc-ture that occurred because of a transitory eventoften persist long after the event. For example,Canada and Australia developed their domesticlong-term local-currency debt market during WorldWar I because the government had large financingneeds that could no longer be fulfilled by borrow-ing on London’s financial markets (Bordo, Meiss-ner, and Reddish, 2003).

International Debt Market

It has been argued that many countries, bothemerging and advanced, are unable to issue in theirown currency on international markets at reasonablecost, owing to an unwillingness of international in-vestors to bear exchange rate risk. In existing empir-ical studies, issuance of bonds in local currency isnot found to be closely related to countries’ policiesor institutions (Hausmann and Panizza, 2002). Nordo these variables seem to help explain the (admit-tedly limited) variation across countries in the struc-ture of sovereign debt issued on international mar-kets. However, countries’ economic size seems to beassociated with whether they issue bonds in theirown currency on international markets. This is sug-gested by both systematic cross-country regressionson modern data (Eichengreen, Hausmann, andPanizza, 2002) and historical evidence. In the nine-teenth century, for example, local-currency debt is-sued by a few “emerging markets” of the time, suchas Russia, was actively traded on the secondarymarket in London (Flandreau and Sussman, 2002).By contrast, the local-currency debt of countriesjudged to have better creditworthiness, such as theNordic countries, was not traded in London. Therole of the ruble as a vehicle currency for a largenumber of international trade and finance transac-tions may help explain this difference.

Policy Implications

Domestic Debt Markets

To make public debt structures less crisis prone,a key long-term policy objective is to develop adeep domestic market for government debt, espe-cially for long-term local-currency instruments. Is-suing long-term local-currency debt requires mone-tary and fiscal credibility. In part, this can begained through a combination of policy success—such as stabilizing from high inflation and reducingpublic debt levels—and institutional reforms thatcreate an expectation that stabilization gains will besustained. The latter include medium-term fiscaland monetary policy frameworks that constrain fu-ture policy choices, such as longer-term fiscal andinflation targets, and central bank independence. In the long run, such institutions may be a betterfoundation for policy credibility than the disciplinethat comes from risky forms of debt (Falcetti andMissale, 2002).

19

Policy Implications

Percentage of Long-Term Local-Currency Debt

Governance

–50 –30 –10 10 30 50–1.0–0.8

–0.6

–0.4

–0.2

0

0.2

0.4

0.6

0.8

Figure 10. Institutional Quality andDomestically Issued Long-Term Local-Currency Debt(Component orthogonal to per capita GDP)

Sources: JPMorgan, Guide to Emerging Markets (2001); andKaufmann, Kraay, and Mastruzzi (2003).

Note:The scatter plot shows the association between an indi-cator of governance and the size of the local-currency debt mar-ket, taking into account that both these variables are associatedwith GDP per capita.The horizontal axis reports the residuals ofa regression of long-term local-currency debt on GDP per capita.The vertical axis reports the residuals of a regression of an insti-tutional index on GDP per capita.The correlation is significant atthe 5 percent level.The correlation between governance and theshare of long-term local-currency is highly significant when notcontrolling for GDP per capita.


It has been argued that, even with stabilization andinstitutional reforms, acquiring sufficient credibilityto issue long-term local-currency debt may take sev-eral years or even decades (Caballero, Cowan, andKearns, 2003, based on a comparison between Aus-tralia and Chile). However, a number of countrieshave recently been able to issue nonindexed long-term debt almost immediately after stabilizing at low

levels of inflation and reforming their monetary pol-icy frameworks (Box 2). In addition, governmentsthat are unable to issue nonindexed long-term debtcan become less reliant on risky debt forms by issu-ing inflation-indexed debt.

The main advantage of inflation-indexed debt isthat it breaks the automatic link between govern-ment solvency and large exchange rate swings that

20

Box 2. Creating Domestic Markets for Long-Term Domestic-Currency Bonds: Country Experiences

A number of emerging markets—including Colom-bia, India, Singapore, South Africa, Taiwan Province ofChina, and Thailand—have traditionally issued long-term domestic debt in local currency. This reflects longhistories of low or moderate inflation, in some casescombined with financial repression. Recently, however,some countries have begun to issue long-term, nonin-dexed local-currency debt, in spite of having experi-enced high inflation during the 1970s or 1980s, and lib-eralizing their capital markets.

Chile last experienced high inflation in the 1970sand inflation has remained moderate since the early1980s. Until recently, more than half of the public debtwas inflation indexed, reflecting a desire to promote in-dexation as a way of avoiding dollarization, and to con-duct monetary policy in terms of “real” policy rates anda “real” term structure of interest rates. As part of thatstrategy, indexed bonds of 8–20-year maturity were issued in the early 1990s. With inflation in the low single-digit levels (since 1999) and the adoption of aformal inflation targeting regime, the policy interestrate switched to nominal targets (August 2001). Subse-quently, the central bank issued two-year and five-yearnonindexed peso bonds.

Israel stabilized from high inflation in 1985. Infla-tion initially fell to moderate levels (10–20 percent)and later—after 1995—to single digits. In 1985, about40 percent of public sector debt was in foreign cur-rency, while the rest was inflation-indexed local-currency debt. By 2002, foreign-currency debt had dis-appeared, the share of inflation-indexed debt was lessthan 40 percent, and nonindexed local-currency debtmade up the remaining debt stock. This transformationtook place in two stages. First, foreign-currency debtwas reduced and substituted by inflation-indexed debt, which peaked at 80 percent in the 1990s. Second,inflation-indexed debt was gradually substituted bynonindexed local-currency debt. A two-year nonin-dexed bond was first introduced in 1995. Subsequently,

the maturities of local-currency bonds were graduallylengthened: 5-year, 7-year, and 10-year bonds were in-troduced in 1998, 2000, and 2001, respectively. Aver-age maturity in 2002 stood at about six years.

Mexico had its last high inflation episode in 1987–88. Inflation fell to single-digit levels by 1993–94, roseto moderate levels after the 1994–95 crisis, and fellagain to single-digit levels by 2000. In the decade be-tween 1989 and 1999, domestic debt consisted mainlyof short-term local-currency debt and floating-ratedebt—except for the well-known sharp increase inshort-term foreign-currency debt leading up to the 1994crisis—with inflation-indexed debt generally taking athird place. This began to change in 2000, when theMexican government issued 3-year and 5-year nonin-dexed bonds, followed by 7-year and 10-year bonds in2002, and, finally, a 20-year bond in 2003.

Poland stabilized from near-hyperinflationary levelsin 1991. Inflation hovered at about 30 percent until1996, then dropped to 10–20 percent and finally (in1999) to single-digit levels. Until 1992, all of Poland’slocal-currency debt, which was small relative to its foreign-currency debt, consisted of short-term treasurybills. In 1992, Poland introduced an inflation-indexedone-year bond and a three-year floating-rate bond. In1994—with inflation still at around 30 percent—2-yearand 5-year nonindexed local-currency bonds were intro-duced, followed by a 10-year floating-rate bond in 1995,and a 10-year nonindexed bond in 1999. The share offoreign-currency debt in domestic debt fell from about20 percent in 1994 to essentially zero in 2002.

These experiences share a common pattern. Nonin-dexed bonds of more than five-year maturity were is-sued almost immediately after fiscal stabilization, thedecline of inflation to low single-digit levels, and theadoption of formal inflation-targeting regimes (Israel in1991; Chile, Poland, and Mexico in 1999) and formalcentral bank independence (Chile in 1989; Mexico in1994; Poland in 1998). Moreover, all countries under-took major pension reforms in the direction of fullyfunded systems. This suggests that the development ofa long-term local-currency bond market requires sub-stantial and credible reforms but is feasible without along period of credibility building.

Sources: Galindo and Leiderman (2003); Herrera andValdés (2003); Werner (2003); IMF Country Reports; and na-tional statistical sources.

Policy Implications

may result from external shocks or losses in investorconfidence. Indeed, inflation-indexed debt exposesthe borrower to less uncertainty than does foreign-currency debt (Box 1). Moreover, CPI-indexed debtprovides as much policy discipline and investor pro-tection as do short-term debt and foreign-currencydebt.

A potential objection to issuing inflation-indexeddebt is that it may lead to an economy-wide cultureof sweeping CPI indexation, up to and includingwage contracts. This would reduce real wage flexi-bility and weaken the effectiveness of stabilizationpolicies. The example of Chile has often been men-tioned in this context, but in that case a broader useof CPI indexation seems to have been specificallyencouraged (Herrera and Valdés, 2003). In general,there is no automatic link between the currency ofdenomination of public and private liabilities and thedenomination of other contracts in the economy. An-other potential problem in some cases is the need fortimely and reliable measurement of the CPI.

An alternative form of indexation that requiresless statistical capacity is indexation to a market-determined domestic interest rate, that is, floating-rate debt. However, floating-rate debt implieshigher debt repayments during bad times, whereasinflation-indexed debt is usually either acyclical or provides a slight hedge. From the perspective of decoupling government solvency from shifts inconfidence and external shocks, floating-rate debtshares many of the disadvantages of foreign-currency debt.

Pension reforms and financial regulation oftenhave important implications for the development ofthe domestic debt markets. A shift toward a fullyfunded pension system is often especially signifi-cant, because pension funds have a natural interestin debt securities carrying a relatively low defaultrisk, and denominated in CPI units or local cur-rency. Regulation that induces pension funds to in-vest a significant portion of assets locally may fur-ther enhance that interest. Of course, regulation

21

Box 3. Developing International Markets for Bonds in Emerging Market Currencies

A considerable share of the relatively rare interna-tional bonds denominated in emerging market curren-cies have not been issued by sovereigns, but rather byinternational organizations or multinational corpora-tions that were able to reduce their borrowing costs byissuing in those emerging market currencies and, whenthey did not have a natural hedge, engaging in currencyswaps. This process began in the mid-1980s with thecurrencies of countries such as Italy, Portugal, andSpain and in recent years has involved the currencies ofemerging market countries such as Brazil, Chile, theCzech Republic, Hungary, Mexico, Poland, the SlovakRepublic, and South Africa. Such opportunities for sav-ings may be the result of the ability to tap an investorbase that seeks to hold bonds denominated in an emerg-ing market currency but are essentially default free.Regulation or differences in tax treatment occasionallyhave also played a role.

Beginning in 2001, the World Bank has allowed itsborrowers to request a local-currency swap. That re-quest is considered on a case-by-case basis, dependingon whether current market conditions allow for a mutu-ally advantageous outcome to be achieved. Followingsuch a swap, the World Bank would face net liabilitiesin hard currency (as it normally does) while the emerg-ing market that engaged in the swap would face net lia-bilities in its own currency. While there has been someinterest in these local-currency swaps, no transactionshave taken place to date, perhaps owing to initial learn-ing barriers.

Recent proposals have sought to generalize this typeof approach so as to fully exploit its benefits for emerg-ing market countries. Levy-Yeyati (2003) argues that

the role of the IFIs in issuing debt denominated inemerging market currencies could be expanded. Hepoints out that a number of emerging market residentsmove their money abroad for a number of reasons, in-cluding a desire to decrease their vulnerability to possi-ble government default. While those residents are un-willing to hold the credit risk of their own country, theymay be willing to hold its currency risk because muchof their consumption basket is in their local currency.The IFIs could issue risk-free local-currency debt, per-haps making it more attractive by indexing it to localinflation. The IFIs could then lend to the country with-out creating any currency mismatches in their balancesheets. The crowding-out effect on domestic creditmarkets may not be large, as some of those fundswould have fled the country anyway.

Eichengreen, Hausmann, and Panizza (2002) suggestthat the IFIs could issue in a basket of local-currencyunits indexed to the CPI of a number of emerging mar-ket countries. The pooling of currencies would helpthat market achieve a critical mass. The IFIs could si-multaneously arrange a series of swaps with emergingmarket countries that had issued hard currency debt.However, if the swaps were arranged directly withemerging market country governments, the IFIs wouldface a default risk. If the swaps were arranged throughan intermediary (which would then absorb, but alsocharge for, the default risk), the transactions costs couldbe large, partly because the swap markets betweenemerging market currencies and the world’s major cur-rencies are not very liquid. For a comprehensive cri-tique of these proposals, see Goldstein and Turner(2004).


that forces investors to hold long-term local-currency debt may also generate a temptation to reduce the debt burden through inflation or depre-ciation, and makes it easier for governments tooverborrow. This temptation is somewhat mitigatedby a large constituency of local-currency debt hold-ers, who may pressure the government to followsound fiscal and monetary policies. Moreover, ifdomestic banks hold large amounts of public debt,the government may be reluctant to provoke a crisisin which sovereign default would turn into a bank-ing crisis. While the costs of financial repressionwill typically exceed its benefits, a greater share oflong-term local-currency debt may be a welcomeby-product.

While governments are seeking to establish credi-bility and develop domestic debt markets, they canundertake steps that help insulate the economy fromexternal and confidence shocks, even in the presenceof substantial shares of short-term debt or foreign-currency debt. Three such steps are mentionedbriefly here. First, countries can avoid a bunching ofrepayment obligations, maintain adequate reservelevels, and manage reserves appropriately.6 Second,

there is a trade-off between low debt levels and a lesscrisis-prone debt structure: with low debt levels, thepublic sector would remain solvent even after a de-preciation; this reduces the urgency to move awayfrom foreign-currency-denominated debt. Third, theimportance of public sector vulnerabilities for theeconomy as a whole depends on conditions in othersectors, especially mismatches in the private finan-cial sector and the flexibility of exports in respond-ing to a depreciation.

Debt Issued on International Markets

Over the past two decades, some international fi-nancial institutions (IFIs) may have helped foster thecreation of a market for internationally issued bondsin emerging market currencies. The IFIs have oftenbeen among the first parties to issue bonds denomi-nated in emerging market currencies that had notpreviously been used in international bonds (Box 3,p. 21). The IFIs’ main motivation for issuing bondsdenominated in emerging market currencies, typi-cally in combination with exchange rate swaps, hasbeen to reduce their borrowing costs. At the sametime, however, many residents of emerging marketsmay have borrowed internationally in their own cur-rency as the counterparts in such swaps.

22

6International Monetary Fund and World Bank (2001); andIMF (2000a).

A s described in Section II, the absence of ex-plicit seniority is a striking difference between

sovereign and corporate debt. This section arguesthat explicit seniority in sovereign debt—and, moregenerally, contractual innovations that help protectcreditors from the consequences of future additionalborrowing by the debtor country—could potentiallyplay a useful role by promoting safer debt structures,discouraging overborrowing, and lowering interestcosts for countries with moderate debt levels. How-ever, this section is only a first pass at the issue, anda more thorough analysis of its consequences in dif-ferent situations as well as its practical feasibilityand impact on crisis resolution is warranted.1

Economic Role of Seniority

The traditional argument for the existence of se-nior debt is that it prevents debt dilution (Fama andMiller, 1972). Debt dilution is analogous to the dilu-tion of equity through new equity issues. When newdebt is issued, the recovery value of the debt has tobe shared among more creditors in the event of in-solvency; thus, dilution leads to a reduction in the re-payment expected by each claim holder. New credi-tors are compensated for this effect through higherinterest rates. But the initial creditors are not: forthem, dilution means a capital loss. Anticipating thispossibility, they may either require higher interestrates at the outset or refuse to lend altogether. Thus,the possibility of debt dilution may ultimately back-fire on the debtor. But if the original creditors weresenior—in other words, if creditors that lent earlier

had priority over those that lent later—they would beless concerned about subsequent debt issues, be-cause in the event of bankruptcy they would be re-paid first. Thus, first-in-time seniority may serve asan antidote to debt dilution.

At the corporate level, the case for senior debt isparticularly strong when managers pursue their ownagendas (Hart, 1995; Hart and Moore, 1995). Self-interested managers may have incentives to over-borrow—for example, to finance projects that givethem private benefits, to make the firm as large aspossible, or to keep their jobs in a situation when itwould in fact be efficient to liquidate the firm. If ex-isting debt can be diluted, it is easy to overborrow,because new capital receives a share of the debt re-covery value corresponding to its share in total debtoutstanding. This is not the case when the new debtis junior: the presence of senior debt can thus serveas a disciplining device.

The dilution problem has become increasingly rel-evant to emerging market debt. For dilution to play arole, two conditions must be met. First, there must bea substantial chance of default or restructuring: aslong as repayment remains safe, new debt issues arenot a cause of concern for existing creditors. Second,debt must be issued to different groups of creditorsover time. This would not be the case, for example, ifemerging market governments kept borrowing fromessentially the same group of banks, as happenedduring the 1970s and 1980s. However, in a case suchas Argentina in the 1990s, which issued 156 bonds toa wide base of customers, dilution of earlier debt bylater issues may have been intense. The same is truefor other recent default or restructuring cases. For ex-ample, in the first half of 1998, Russia and Ukrainecontinued accessing capital markets through newdebt issues at rapidly rising interest rates, therebysharply diluting the existing debt stock.

One sign that the dilution problem is relevant inemerging market debt is that creditors seem to bemaking efforts to protect themselves against dilution.For example, one of the two Eurobonds that creditorswere offered in Ecuador’s 2000 debt exchange con-tained a “principal reinstatement” clause, which pro-vided for an automatic upward adjustment in principal

IV Explicit Seniority in Privately HeldSovereign Debt

23

Note: The author of this section is Jeromin Zettelmeyer.1Gelpern (2004) argues that explicit seniority may result in a

more predictable and less complicated debt restructuring process.Establishing a fixed priority ranking among creditors eliminatesone aspect of the creditor collective action problem: the “race tothe courthouse,” whereby creditors seek an advantage by beingthe first to litigate. However, it can also be argued that an explicitseniority ranking would complicate debtor-creditor negotiations,because senior creditors have an incentive to agree quickly to alarge haircut that leaves junior creditors with nothing (seeZettelmeyer, forthcoming, for a discussion).

in the event of a default. The face value of the bond-holder’s claim was to increase by a given amount inthe event that Ecuador defaulted on the new bondsafter the restructuring (30 percent if a default occurredin the first three years after issuance, and graduallydeclining to zero after 10 years). Thus, incumbentbondholders received (temporary) protection from di-lution that might result from new debt issuance.

The dilution problem could be eliminated if, in theevent of a debt restructuring, creditor claims wereserved in the order in which the debt was issued.With this “first-in-time” seniority, initial creditorswould be repaid (possibly in full), whereas the mostrecent creditors would receive much less, and possi-bly nothing.2 If credible and enforceable, this couldhave desirable effects on the structure, cost, andamount of public borrowing. Explicit seniority couldbenefit the debt structure by reducing the incentiveto issue debt forms that are hard to dilute, such asshort-term foreign-currency debt. It could also lowerthe costs of borrowing at moderate levels of debt, be-cause creditors would not have to worry about debtdilution in the future. Finally, by eliminating thepossibility of issuing debt at the expense of previouscreditors, seniority could reduce the incentives tooverborrow. This said, countries with high levels ofdebt would find tighter market access and highercosts if they were to adopt a first-in-time seniorityrule. This is an inherent drawback of any mechanismthat reduces debt dilution and needs to be taken intoconsideration when evaluating the possible adoptionof a seniority rule.

The following subsections elaborate on these pointsbefore turning to the question of how seniority-likefeatures could in practice be introduced into sovereigndebt. A number of obstacles and difficulties are dis-cussed. While it is too early to say whether such diffi-culties could be resolved, the section suggests thatconsideration could be given to creating an interna-tional debt registry that would make it easier to moni-tor both total indebtedness and the contractual termsunder which public debt is issued. This would be anecessary (albeit not sufficient) step for the develop-ment of an explicit seniority structure based on debtcontracts, and may also have benefits of its own.

Effects of Seniority on the Quantity and Price of Debt

The benefits of making incumbent sovereign debtsenior are most obvious when governments are bi-

ased toward excessive borrowing.3 Diluting existingdebt makes it easier for politicians to finance activi-ties that may not be in the taxpayer’s best interest.Examples include consumption or investment thatbenefit special interests, fiscal expansions ahead ofelections, or—instead of politically costly reforms orrestructurings—“gambles for redemption,” wherebythe dilution of existing debt may allow new govern-ment borrowing even when it is generally knownthat the country is insolvent. This may postpone acrisis, but the default, when it finally happens, willbe much larger.

Seniority could also help curb overborrowing thatarises even without a political bias toward excessiveborrowing, purely as a consequence of the inabilityof the debtor to commit to not dilute the existingdebt (Sachs and Cohen, 1982; Kletzer, 1984; Detra-giache, 1994; and Eaton and Fernandez, 1997). Sup-pose that a country would like to borrow up to agiven debt level. Once it has done so, it will gener-ally have an incentive to borrow some more, becausethe dilution effect implies that new debt can beplaced relatively cheaply. This will lead to excessiveborrowing from the country’s original perspectiveand—as it is anticipated by creditors—to higher in-terest rates. As a result, the country is worse off thanif it had been able to commit to the original debtlevel. Seniority can rectify this problem by acting asa substitute for commitment: the debtor can crediblypromise to the initial creditors that it will refrainfrom additional borrowing, because seniority elimi-nates the possibility of dilution.

Senior debt could also reduce borrowing costs forcountries with low levels of debt; these countriesmight end up borrowing more. At present, the possi-bility of future debt dilution makes emerging mar-ket countries’ debt relatively expensive even atmoderate levels. As a result, governments that wantto “play it safe” and keep their borrowing costsdown have an incentive to borrow less than desir-able, that is, below the level at which they wouldwant to borrow if they could commit to not dilute(Bolton and Jeanne, 2004; and Zettelmeyer, forth-coming). With debt seniority acting as a bindingpromise to not dilute, countries with levels of debtthat are low relative to sustainable levels would beable to borrow more, at equal or lower interest ratesthan would prevail in the absence of seniority.

As suggested above, any mechanism that limitsdebt dilution could impede debt dilution when it is infact desirable. Consider a situation in which a (sol-

IV EXPLICIT SENIORITY IN PRIVATELY HELD SOVEREIGN DEBT

24

2In a variant of first-in-time seniority, time units could be cal-endar years, for example. All debt issued within a given yearwould have equal priority. Debt issued in year t would have ab-solute priority over debt issued in year t – 1.

3The IMF’s World Economic Outlook (September 2003, Chap-ter 3) suggests that overborrowing—defined as public debt that ishigher than what can be repaid based on a country’s fiscal trackrecord—is widespread among developing countries and emergingmarket borrowers.

vent) country suffers an adverse shock that would re-quire large additional financing. Issuing junior debtcould be prohibitively expensive, because the proba-bility of a debt crisis is now higher. In such a case, itmight be in the interest of the incumbent creditors toallow dilution—in other words, allow new debt is-sues at the same, or even a higher, level of seniorityas the existing debt—in order to provide the liquiditythat might stave off a debt crisis. This is similar tothe logic of “debtor in possession” financing mecha-nisms in corporate bankruptcy, whereby creditorsmay waive their seniority in order to allow the firmto access new money. A mechanism for waiving se-niority in such situations could be useful. However,coordinating creditors on waiving seniority may bedifficult in practice.

Collateralized debt shares some features withlegally senior debt, but also suffers from some specific—and potentially serious—disadvantages.Like legally senior debt, fully collateralized debtcannot be diluted and may discourage overborrow-ing and lower the costs of responsible borrowing.4However, issuing collateralized debt is possibleonly for countries that have assets abroad or largecash flows originating in other jurisdictions. Moreimportant, promoting collateralized debt may becounterproductive if the basic problem is an inher-ent government bias to overborrow. With standarddebt, the possibility of default would deter investorsfrom lending to an irresponsible government. Incontrast, collateralized debt could make it possiblefor lending to continue, effectively at the expense offuture governments and generations.

Effects of Seniority on Debt Structure

Introducing seniority could have an impact notonly on the quantity and price of debt but also on itsstructure. In the absence of explicit seniority, credi-tors may have an incentive to seek “de facto senior-ity” by opting for debt instruments that are hard todilute, such as short-term debt (Sachs and Cohen,1982; Kletzer, 1984; Chamon, 2002; Bolton andJeanne, 2004). If maturities are short, lenders canrefuse to roll over (or roll over at higher interestrates) as soon as they perceive an attempt to dilute.In contrast, if maturities are long, lenders are cap-tive, and will suffer a capital loss.

Dilution fears could also create an incentive toborrow in debt forms that are relatively difficult torestructure, for example, through international bondsrather than domestic bonds or bank loans. If a coun-

try is able to repay some debt, but not all, it may de-fault selectively on the instruments that can be rene-gotiated more easily. Once everyone has realizedthis, however, it will only be possible to issue hard-to-restructure instruments (Bolton and Jeanne, 2004;Lipworth and Nystedt, 2001).5

The implication of the bias toward short maturitiesis to make debt more crisis prone. The bias towarddebt instruments that are relatively hard to restruc-ture deepens crises when they occur and impedes re-structurings that could have avoided major defaults.Seniority could help prevent costly crises by remov-ing these biases.

Approaches and Obstacles inImplementing Explicit Seniority

An explicit, first-in-time seniority structurecould arise through one of three mechanisms. First,new statutes at the international level, created by aninternational treaty or an amendment of the IMF’sarticles (IMF, 2002b; and Bolton and Skeel, 2003).Second, national legislation whereby debtor gov-ernments would commit to repay debt—in theevent of default—in the order in which it was is-sued. For such commitment to be credible in theeyes of investors, changes to the law would have tobe hard to make, and the domestic judiciary wouldneed to be strong and independent. Third, contrac-tual provisions protecting bondholders from dilu-tion by future debt issues; such provisions wouldbe enforced by the courts of the country where thedebt was issued. The remainder of this section fo-cuses mainly on the third approach, because itseems promising and raises complex issues thathave not been analyzed before.

The contractual approach would loosely follow theexample of creditor protections in corporate bondcovenants. Specifically, the debt contract between aninitial creditor and the debtor could contain acovenant prohibiting the debtor from issuing any sub-sequent debt unless future creditors agreed to be con-tractually subordinated to the initial creditor’s claim.To enforce this covenant, senior creditors would needto be given the power to declare a default and accel-erate if debtors fail to ensure that future creditors aresubordinated. The next creditor’s contract would theninclude a similar clause with respect to future debt,and so on. This sequence of contracts might generatea priority structure that gives senior creditors a legal

Approaches and Obstacles in Implementing Explicit Seniority

25

4On public sector collateralized borrowing, see IMF (2003d);and Chalk (2002).

5Debt restructurings in the 1990s seemed to reflect an implicit,though evolving, de facto seniority structure (Zettelmeyer,forthcoming).

basis to sue junior creditors in the event that seniorityis violated in a default (Box 4).6

An explicit first-in-time seniority structure in sov-ereign debt could conceivably be undermined by debtissues that are formally subordinated but have shortermaturities and thus require earlier repayment. Indeed,to the extent that this is possible, first-in-time senior-ity might have perverse effects on the debt structureat high levels of debt, as it may lead sovereigns indistress to issue debt of even shorter maturities thanunder the present system. In the contractual ap-proach, a possible solution could be an intercreditorprovision to relinquish payments received 90 daysprior to a formal default to the senior creditor. Alter-natively, bond covenants could stipulate the suspen-sion of payments to junior creditors—or their redi-rection to a trustee—prior to a formal default, uponobserving prespecified signs of debt distress (e.g.,when debt service capacity indicators cross certainthresholds). Finally, covenants could be added thateffectively give senior creditors veto power over theissuance of short-term debt, by granting them theright to ask for early repayment.7

As is often the case with financial innovation,first-in-time seniority would add complexity to thedebt instruments. Financial market participantswould need to keep track of the seniority status of the various bonds when pricing them. Methodsto price bonds with different degrees of senioritywould have to be developed. It is possible that thisadditional complexity would make sovereign debt less attractive and raise borrowing costs, atleast until markets became accustomed to the newsystem.

A further caveat is that highly indebted countriesmight not benefit from a switch to a first-in-timeseniority regime. A new creditor would rank juniorto all preexisting claims, and would hence need tobe compensated by a higher interest rate, or couldeven refuse to lend at all if the risk of default weresufficiently high (Box 5). A similar problem mightapply to countries that are liable to suffer fromlarge adverse shocks that would increase their fi-nancing needs significantly. For many countries,however, this would be just a transitional problem.Under a pari passu regime, countries tend to over-borrow because dilution lowers the costs of obtain-ing additional debt. Under a seniority regime,countries would have an incentive to maintainlower debt levels but, in some cases, they may needto reduce their debt levels before they are able to


26

Box 4. Enforcing Contractual Seniority

Establishing a feasible contractual priority structurebased on time of issue requires solving two enforce-ment problems. First, assuming that outstanding claimsdefine a consistent legal priority structure, this structuremust be enforceable in the event of a restructuring. Sec-ond, a mechanism must be found that ensures that thepriority structure is defined consistently. In particular,debtors have to be prevented from issuing new claimsin contravention of earlier contracts, that is, claims thatare not explicitly subordinated to those of previouscreditors. The resolution of both these problems willmost likely require the establishment of both debtor-creditor and intercreditor obligations.

Resolving the first problem—enforcing contractualpriority in a default—requires giving senior creditors thelegal basis to sue junior creditors who receive paymentsin contravention of their order of priority. To provide cer-tainty, this legal basis is likely to require privity of con-tract between the senior creditor and the junior creditors.Specifically, the junior creditor would need to enter intoa contract with the senior creditor that provides that the

junior creditor will not receive any payments from thedebtor until the senior creditor is paid in full.

Making sure that the above framework is establishedalso presents some challenges. Contrary to the case inwhich existing creditors agree to subordinate themselvesto future creditors, as in the case of debtor-in-possession-type financing (Buchheit and Gulati, 2002), existingcontracts must contain covenants that ensure that futurecreditors subordinate themselves in a way that gives se-nior creditors comfort that their priority can be enforced.The basic instrument for achieving this would be to givesenior creditors the power to declare a default and accel-erate if debtors fail to ensure that future creditors aresubordinated to them. However, this is a fairly blunt in-strument. Notably, if new creditors are not subordinatedand previous creditors accelerate, they will not be able toenforce the seniority of their claims with respect to themost recent group of new creditors. This said, the threatof acceleration and technical default may conceivablyprovide sufficient discipline to the debtor for a consistentcontractual priority structure to get off the ground.

6The IFIs (and possibly other official creditors) would be ex-cluded from an explicit priority structure for privately held debt.This would not imply giving them legal seniority—rather, theirseniority status would remain legally indeterminate with regard tothe debt instruments whose seniority was governed by the newsystem. First-in-time seniority in privately held sovereign debtwould thus have no bearing on the issue of IFI seniority.

7An analogous problem is that first-in-time seniority could beundermined through collateralized debt issues. Again, this prob-

lem would have to be dealt with by explicitly limiting or rulingout such issues through provisions in bond contracts.

secure a reduction in their borrowing costs. Thesecountries would have to gradually reduce their debtburdens to levels that are viewed as reasonably safebefore issuing senior debt. Alternatively, they couldincorporate seniority clauses that do not become ef-fective immediately, but instead stipulate that thedebt will become senior to new debt issued after aspecified time interval (say, five years). This wouldallow time to adjust from high debt levels, and pro-vide a commitment to doing so.

Before concluding, it is worth emphasizing thatthere may be alternative contractual ways of reduc-ing the dilution problem while not establishing afull-fledged legal priority structure in the event ofdefault. One possibility would be to give creditorscontractual options that provide them with somecontrol over the debtors’ subsequent borrowing be-

havior or allow them to renegotiate in the event ofnew borrowing. At the most basic level, contractscould include put options allowing debtors to askfor early repayment in the event that certain limitson total debt are exceeded. Alternatively, futuredebt issues could trigger changes in the paymentterms of existing bonds in a way that offsets theloss of value inflicted upon their holders by the newborrowing. These alternative approaches may go aconsiderable way toward addressing the dilutionproblem without raising the legal difficulties offull-fledged first-in-time seniority. Because they do not have the effect of making the outstandingdebt stock legally senior relative to future debt is-sues, they might also avoid possible transitionalproblems associated with full-fledged first-in-timeseniority.

Approaches and Obstacles in Implementing Explicit Seniority

27

Box 5. Effect on Borrowing Costs of a Switch to First-in-Time Seniority

As argued earlier, the main effect of first-in-time se-niority on the quantity of countries’ borrowing wouldbe to reduce incentives to overborrow. After countriesand markets have adjusted to the new regime, onewould expect to see fewer countries at or close to un-sustainable debt levels. In addition, borrowing costswould be lower, because spreads would no longer re-flect the risk of future debt dilution.

However, the effect of a switch to first-in-time se-niority on impact—for given outstanding debt stocks—might be to either raise or lower borrowing costs, de-pending on the size of countries’ existing debt stocks.Countries with low debt levels would see their borrow-ing costs fall, whereas countries with high debt levelswould see borrowing costs rise, and might even be cutoff from additional net borrowing altogether. This isbecause at low debt levels, a creditor buying an extraunit of debt under first-in-time seniority would expectto be senior in the event of a default (because defaultwould only occur after substantial accumulation of sub-sequent debt). In contrast, under the present regime thiscreditor would expect to rank equally, in the event ofdefault, as the holders of debt issued subsequently. Forvery large outstanding debt levels, and a correspond-ingly high probability of a debt crisis in the near future,the opposite is true: a new creditor would expect torank junior to most outstanding debt, and would conse-quently want to be compensated by higher interest ratesthan under the present system.

The argument is illustrated in the figure below. Thisassumes that the country in question has a minimum re-covery value of debt, denoted D, and a maximum sus-tainable debt level of D

_. Under first-in-time seniority,

an extra debt unit issued when actual debt is lower thanD is risk free (rS(D) = r for D < D), as the new creditorwill have first claim to the certain recovery value. Incontrast, under the present system, even the very firstdebt unit issued is risky (rP(0) > r), because in the event

of default its holder would have to share the recoveryvalue with the holders of all subsequently issued debt.

As D approaches D_

, borrowing costs under senioritywould go to infinity, because in the event of default,any new unit of debt would be junior to all other debtunits and would receive nothing. In contrast, rP(D

_) is

finite because new debt continues to enjoy a positiveclaim to the debt recovery value; as such, investors ex-pect a positive return even when default is certain. Con-sequently, the marginal borrowing cost curves rS(D)and rP(D) must intersect at some debt level between Dand D

_. (The corresponding average borrowing cost

curves—not depicted here—intersect at a higher D thanshown in the figure below, because borrowing costs onthe inframarginal units would be lower under senioritythan under the present system.)

Cost of borrowingan extra

debt unit, r

Debt outstanding, D

First-in-timeseniority,

rs (D)Present system,rP(D)

D D

Internationalrisk-free rate, r

Marginal Borrowing Costs


Conclusions

Explicit seniority in privately held sovereign debt,or other possible enhancements in debt contracts thatsignificantly reduce the scope for debt dilution, couldhave benefits in terms of discouraging overborrowing,lowering the debt costs of responsible borrowers, andreducing the incentives to adopt risky debt structures.These benefits would need to be balanced against apotential cost, namely, the possibility that countrieswith high levels of debt would find it more difficult toaccess international capital markets than they dounder the present system. Countries with more mod-erate debt levels—but which nevertheless pay a sub-stantial risk premium—could thus stand to gain themost from adopting a seniority regime. Other coun-tries might need to find ways to reduce their debt tosafer levels before the benefits could be secured.

This said, the analysis of both the feasibility andoverall effects of seniority-like features in sovereigndebt would need to be extended before reaching amore definitive judgment on the issue. In particular,further analysis would be needed in three areas:(1) the impact of seniority on crisis resolution (an as-pect not addressed in this paper); (2) the legal and

operational feasibility of a contractual senioritystructure; and (3) the desirability and feasibility ofenhancements in bond contracts that might protectbondholders from dilution without the need to createa legal seniority structure.

One possible step that would not seem to dependon how the verdict on the usefulness and practicalityof explicit seniority ultimately turns out is the cre-ation of an international official debt registry pub-lishing the terms of all public debt contracts.8 Whilesuch a registry might be a necessary step toward acontractual seniority structure, it could also be ofvalue more generally. Greater transparency in thelevel and structure of sovereign debt might con-tribute to the efficiency of the debt market, andwould play a helpful role in the context of debt re-structuring operations. Although the costs of estab-lishing a debt registry should be evaluated carefully,the public good aspect of information is an argumentfor officially sponsoring such a registry.

28

8Such a registry has been proposed on several occasions in thepast—for example, by Allen (1988); Kaeser (1990); and Eaton(2002).

Most proposals for reform of the international fi-nancial architecture have taken the set of avail-

able financial instruments as given. This section askswhether greater use of underutilized instrumentscould be beneficial in providing countries with in-surance and reducing the likelihood of crises. Debtsustainability hinges on developments in real vari-ables, such as exports and GDP. Adverse shocks tothese variables often prompt debt crises.1 For somecountries, such shocks routinely take the form of ad-verse developments in commodity prices, naturaldisasters, or declines in imports by trading partners.For others, changes in economic growth are moredifficult to relate to specific events or sectors of theeconomy.

Debt indexed to real variables has been proposedas a way of mitigating changes in debt sustainabilitythat might result from real shocks (Box 6). This sec-tion explores the role of debt (or insurance) contractsindexed to real variables, which include

• variables that are largely outside the control ofthe country’s authorities and in many cases canbe measured in a relatively straightforwardmanner—such as commodity prices, naturaldisasters, or imports by a country’s main tradingpartners; and

• broader measures of economic activity that arepartly within the control of the country’s author-ities and are typically measured by the country’sown statistical agencies—such as the country’sown exports or GDP.

The relative merits of indexing to variables inthese two groups will depend on individual country

characteristics such as the sources of shocks, the re-liability of the statistical information, and the per-ceived credibility of the authorities. In fact, while in-dexation to broader measures such as GDP wouldlikely provide greater insurance benefits, potentialinvestors might be concerned about the authorities’incentives to tamper with GDP data or even under-take less-growth-oriented policies, as suggested bysome studies (e.g., Krugman, 1988). As indexationto variables outside the control of the authorities hasbeen extensively treated in many previous studies,this section provides more detailed informationabout indexation to variables partly within the con-trol of the authorities, whose properties and chal-lenges are relatively underexplored.

Benefits of Indexation to Real Variables

Indexing bond repayments to real variables suchas GDP or exports, or some of their key determi-nants (such as natural disasters, commodity prices,or trading partners’ total output or imports), wouldtend to promote debt sustainability (or raise the levelof sustainable debt) by stabilizing variables such asthe debt-to-GDP ratio—thereby providing a numberof benefits to borrowing countries:

• If the economy falls onto a path of persistentweak growth, the smaller increases in the debt-to-GDP ratio resulting from indexation wouldreduce the likelihood of default and debt crises.

• Governments would find it easier to maintain asmooth path for tax rates and essential public ser-vices despite fluctuations in economic growth.The need to pay higher interest rates in years ofrapid growth might even make it more difficultfor governments to boost noninterest spendingunsustainably during times of economic boom.

• By acting as an “automatic-stabilizer,” real in-dexation would reduce pressures on govern-ments to engage in procyclical fiscal policy.Emerging markets are often forced to tighten

V Expanding the Set of Instruments:Indexation to Real Variables

29

Note: The authors of this section are Eduardo Borensztein andPaolo Mauro.

1Slow growth underlies many debt crises, including the LatinAmerican debt crisis of the 1980s and the debt crisis of the highlyindebted poor countries (HIPCs) in the 1980s and 1990s. Thegrowth slowdown that began in Argentina in 1998 contributed toits recent debt crisis—though vulnerabilities had built up previ-ously (IMF, 2003f). Several studies find that slow economicgrowth or high debt help predict debt crises (Detragiache andSpilimbergo, 2001; Easterly, 2001; Kraay and Nehru, 2003; andManasse, Roubini, and Schimmelpfennig, 2003).

fiscal policies during economic downturns in anattempt to maintain credibility and access to in-ternational financial markets, as suggested bythe sudden stops literature (Calvo, 2003).2 Butreal indexation may be helpful more generally togovernments that are seeking to stabilize thedebt-to-GDP ratio, whether because of legal orconstitutional constraints, agreements such asthe Stability and Growth Pact, or inability to bor-row beyond a certain level.

This section focuses on the insurance benefits ofreal indexation for the borrowing countries and theircitizens. Yet, real indexation can be viewed more gen-erally as a desirable vehicle for international risk-sharing and as a way of avoiding the disruptions aris-ing from formal default. It thus has a number ofpotential benefits for international investors:

• International investors would also benefit from alower frequency of formal default, which oftenresults in costly litigation/renegotiation and isthus disruptive even to large private financial in-stitutions that might be considered risk neutral.

• Citizens of lending countries would appreciatethe ability to invest in assets whose return islinked to other countries’ fortunes. As countries’incomes are far from being perfectly correlated,this would provide a welcome diversification opportunity.

30

V EXPANDING THE SET OF INSTRUMENTS

Box 6. Proposals for Indexation to Real Variables

A first wave of interest in indexing debt to GDP, ex-ports, or key commodity prices emerged in the after-math of the debt crisis of the 1980s. Bailey (1983) sug-gested the conversion of debt into proportional claimson exports. Lessard and Williamson (1985) made thecase for real indexation of debt claims. Krugman (1988)and Froot, Scharfstein, and Stein (1989) considered therelative merits of indexing debt to variables out of thedebtor country’s control (such as commodity prices)versus variables partially under the country’s control(exports or GDP). At the time, a majority view withinthe academic community seemed, on balance, to em-phasize the moral hazard costs rather than the insurancebenefits of indexing to exports or GDP. The decline incommodity prices was fresh in people’s minds as one ofthe causes of the debt crisis and—with commoditiesstill representing a significant share of production andexports for some of the countries most affected—index-ing to commodity prices seemed like a better idea.

A second wave of interest originated from Shiller’s(1993, 2003) proposal to create “macro markets” forGDP-linked securities. In Shiller’s proposal, these wereto be perpetual claims on a fraction of a country’s GDP.Barro (1995) shows that bonds ought to be indexed toconsumption and government expenditure in a modelof optimal debt management where the governmentseeks to smooth tax rates over time. But he suggestsGDP-indexed bonds as a more realistic alternative withfewer problems related to moral hazard and measure-ment. This section analyzes a possibility related toShiller’s proposal: a bond whose coupon payments areindexed to GDP growth is equivalent to a plain vanillabond combined with a security whose payoff dependson deviations of the issuing country’s growth rate froma baseline. While Shiller security markets would have

to be set up from scratch, the possibility analyzed inthis section might be easier to implement: it would re-quire introducing an indexation clause in otherwisestandard sovereign bonds.

For emerging market economies, the case for contin-gent debt contracts has received new impetus after thefinancial and debt crises of the 1990s. Caballero (2003)recommends that countries issue bonds with contingen-cies to commodity prices and other external variablesof relevance to them (e.g., Chile should issue bonds in-dexed to the price of copper). Haldane (1999) arguesthat emerging markets would benefit from indexingdebt to commodity prices. Daniel (2001) argues thatmany governments would benefit from hedging oilprice risk through existing financial instruments andmarkets, and that international institutions should en-courage them to explore this possibility. Drèze (2000a)suggests the use of GDP-indexed bonds as part of astrategy to restructure the debt of the poorest countries.Varsavsky and Braun (2002) make the case for restruc-turing Argentina’s debt into GDP-indexed bonds.

Related proposals have also been made for advancedcountries. Some investment banks in Sweden proposedto introduce GDP-indexed bonds in the mid-1990s.The idea received some support in official circles, butfell by the wayside partly because the National DebtOffice at the time was focused on promoting greateruse of inflation-indexed bonds (Englund, Becker, andPaalzow, 1997). Obstfeld and Peri (1998) suggest thatindividual governments in the European Union shouldissue perpetual euro-denominated liabilities indexed todomestic nominal per capita GDP growth. They arguethat nominal rather than real indexing would protectsecurities holders against inflation. Drèze (2000b)makes a similar proposal for the EMU countries.

2Gavin and Perotti (1997) find that during deep recessions thefiscal surplus typically increases in Latin American countries,whereas it falls in OECD countries. They also find that publicspending is far more procyclical in Latin American countries than itis in OECD countries—a result confirmed by Talvi and Végh(2000) and the World Economic Outlook (September 2003, Chapter3) for emerging markets and developing countries more generally.

Real Variables Beyond the Control of the Country’s Authorities

• Finally, for the case of GDP indexation, finan-cial market participants might be interested inthe opportunity to take a position on countries’future growth prospects. This is already possibleto some extent through countries’ stock markets,but these are often not representative of theeconomy as a whole, especially in emergingmarket countries.

Real Variables Beyond the Control ofthe Country’s Authorities

Some shocks that have a major economic impactare largely beyond the control of the country’s au-thorities. For the vast majority of countries, fluctu-ations in world commodity prices are essentiallygiven. Of course, countries may wish to diversifytheir production, export, and revenue structures,but this takes time and may not always be desir-able. Similarly, there is little countries can do to avoid natural disasters, though investments inpreparedness and relocation of activities to less disaster-prone areas help mitigate the conse-quences of disasters. Finally, except for a few verylarge economies, countries typically must take asgiven the economic performance of their tradingpartners.

Bonds whose repayments are linked to such vari-ables are worth considering. Bonds whose repay-ments are indexed to commodity prices have beenused, although rarely, since the 1800s. Insuranceagainst natural disasters (including through cata-strophe bonds—whose repayment is waived in theevent of a catastrophe) is a much more recent inno-vation that is already widespread in the private sec-tor, but has only been used by a handful of sover-eigns. Bonds whose repayment is linked to tradingpartners’ performance have not been previouslyused (or advocated), but their use would seem to bedesirable.

Commodity Price Shocks

Shocks to commodity prices have important ef-fects on debt sustainability for several countrieswhere commodities are a sizable share of exportsand GDP, and export taxes are a substantial share oftotal revenues. Shocks to prices of key imports, no-tably oil, are perhaps even more important formany developing countries, though insuring againstthese shocks has received less attention than shocksto exports. A number of studies suggest thatchanges in commodity prices (or, more generally,terms-of-trade shocks) can have a large impact onrevenues and economic growth, though the exact

magnitude of the estimated impact differs consider-ably across studies.3

Most advanced countries have well-diversifiedproduction and export structures. The percentageshare of the top export in total exports of goodsamounts to more than one-third only for Iceland(fish) and Norway (oil) (Table 4).4 Even for ad-vanced countries for which commodities are oftenconsidered significant, such as Australia, the share ofthe top three exports in total exports is no more thana quarter.

For several developing countries and a few emerg-ing markets, a single product or a few products con-stitute an overwhelming portion of total exports. Ofthe 27 emerging market countries for which data areavailable, 6 have more than 25 percent of exports inone commodity: Chile (copper, 27 percent), Colom-bia (oil, 29 percent), Egypt (oil, 27 percent), Kenya(tea, 28 percent), Venezuela (oil, 80 percent), andZimbabwe (tobacco, 34 percent). However, of theemerging market countries with the largest bondmarket capitalization (Argentina, Brazil, Korea,Mexico, and Russia—on the basis of the EMBIGweights for 2000–03), only Russia’s share of the topthree exports is above one quarter.

As might be expected, the importance of com-modity price shocks for economic performance dif-fers considerably across countries. Collier and Dehn(2001) report that large adverse shocks to commod-ity prices are significantly associated with a declinein output growth for commodity-producing coun-tries, though not for other countries. For the typicalcommodity producer, the worst shocks—in the low-est 2.5 percent of the shock distribution—are associ-ated with a 7 percent output decline in the year of theshock and a cumulative 14 percent decline in theyear of the shock and the following three years.5Shocks to commodity prices have widely differenteffects on tax revenues depending on the countryunder consideration. For example, changes in oilprices of reasonable magnitudes affect the revenuesof oil-producing countries by several percentagepoints of GDP (IMF, 2000b).

31

3Using a broad panel of countries, Easterly and others (1993)find that a favorable terms-of-trade shock of 1 percentage pointper annum was associated with an increase in the annual growthrate of real GDP per capita by 0.42 percentage points in the 1970sand 0.85 percentage points in the 1980s. Studies based upon vec-tor autoregressions find that, for the typical emerging marketcountry, only a small share of output fluctuations can be attrib-uted to terms-of-trade shocks (Hoffmaister and Roldós, 1997).

4The data refer to 1999 and are drawn from UNCTAD (2001).Products are defined at the Standard Investment Trade Classifica-tion (SITC-3) level.

5Panel regressions estimated by IMF staff, though not reported inthis paper for the sake of brevity, yield smaller and less significantcoefficients for both commodity-producing and other countries.


32

Tabl

e 4.

Perc

enta

ge S

hare

of t

he T

op

Thr

ee E

xpo

rts

in T

ota

l Exp

ort

s,19

90–9

9

Tota

l Sha

reof

Top

3

Cou

ntry

Prod

uct

1Ex

port

Sha

rePr

oduc

t 2

Expo

rt S

hare

Prod

uct

3Ex

port

Sha

rePr

oduc

ts

Adv

ance

d ec

onom

ies

19.6

10.9

7.4

37.8

Icel

and

Fres

h fis

h36

.3A

lum

inum

15.6

Pres

erve

d an

d sm

oked

fish

15.0

66.8

Nor

way

Petr

oleu

m a

nd c

rude

oils

37.6

Gas

8.1

Fres

h fis

h5.

951

.6Si

ngap

ore

The

rmio

nic

cells

20.1

Dat

a pr

oces

sing

equ

ipm

ent

17.3

Part

s an

d ac

cess

orie

s 8.

846

.2Fi

nlan

dPa

per

and

pape

rboa

rd19

.7Te

leco

m e

quip

men

t17

.1W

ood

and

woo

d pr

oduc

ts3.

740

.5H

ong

Kon

g SA

RO

uter

,kni

tted

gar

men

ts14

.9O

uter

,wom

en’s

garm

ents

11.7

The

rmio

nic

cells

9.

235

.7Ir

elan

dD

ata

proc

essi

ng e

quip

men

t13

.5Pa

rts

and

acce

ssor

ies

9.0

Nitr

ogen

com

poun

ds8.

831

.3Sw

eden

Tele

com

equ

ipm

ent

14.6

Pape

r an

d pa

perb

oard

8.0

Spec

ial t

rans

actio

ns,o

ther

s6.

228

.8C

ypru

sPh

arm

aceu

tical

pro

duct

s11

.8Ve

geta

bles

8.5

Frui

t an

d nu

ts

7.3

27.6

New

Zea

land

Mea

t an

d ed

ible

mea

t of

fals

12.7

Milk

and

cre

am7.

8Fr

uit

and

nuts

4.

625

.0C

anad

aPa

ssen

ger

vehi

cles

14.5

Spec

ial t

rans

actio

ns,o

ther

s6.

0Pa

rts

and

acce

ssor

ies

4.3

24.8

Emer

ging

mar

ket

coun

trie

s35

.39.

97.

052

.2Ve

nezu

ela

Petr

oleu

m a

nd c

rude

oils

80.8

Alu

min

um3.

3Va

riou

s fo

rms

of ir

on1.

285

.2Ph

ilipp

ines

Spec

ial t

rans

actio

ns,o

ther

s51

.2T

herm

ioni

c ce

lls

11.4

Dat

a pr

oces

sing

equ

ipm

ent

9.0

71.6

Col

ombi

aPe

trol

eum

and

cru

de o

ils28

.8C

offe

e12

.3C

oal,

ligni

te a

nd p

eat

7.2

48.3

Ken

yaTe

a28

.4C

offe

e10

.4R

efin

ed p

etro

leum

pro

duct

s8.

247

.0Jo

rdan

Fert

ilize

rs,c

rude

25.6

Fert

ilize

rs,m

anuf

actu

red

10.6

Phar

mac

eutic

al p

rodu

cts

10.3

46.5

Chi

leC

oppe

r27

.2O

res,

met

al c

once

ntra

tes

12.1

Frui

t an

d nu

ts

7.1

46.5

Zim

babw

eU

nman

ufac

ture

d to

bacc

o33

.4Va

riou

s fo

rms

of ir

on6.

8C

otto

n5.

846

.0Is

rael

Pear

ls,p

reci

ous

ston

es30

.1Te

leco

m e

quip

men

t11

.1D

ata

proc

essi

ng e

quip

men

t4.

245

.5Pe

ruG

old,

nonm

onet

ary

20.1

Cop

per

12.2

Ore

s,m

etal

con

cent

rate

s10

.342

.7Eg

ypt

Ref

ined

pet

role

um p

rodu

cts

27.4

Petr

oleu

m a

nd c

rude

oils

8.4

Cot

ton

6.8

42.6

Dev

elop

ing

econ

omie

s50

.916

.910

.077

.9A

lger

iaG

as41

.7Pe

trol

eum

and

cru

de o

ils39

.7R

efin

ed p

etro

leum

pro

duct

s15

.096

.5Fa

roe

Isla

nds

Fres

h fis

h65

.9Pr

eser

ved

and

smok

ed fi

sh14

.9C

rust

acea

ns a

nd m

ollu

sks

5.5

86.4

St.L

ucia

Frui

t an

d nu

ts63

.6A

lcoh

olic

bev

erag

es12

.7O

uter

,kni

tted

gar

men

ts5.

882

.2O

man

Petr

oleu

m a

nd c

rude

oils

74.4

Pass

enge

r ve

hicl

es

3.9

Ref

ined

pet

role

um p

rodu

cts

2.6

80.8

Syri

an A

rab

Rep

ublic

Petr

oleu

m a

nd c

rude

oils

64.3

Vege

tabl

es5.

9Fr

uit

and

nuts

4.

574

.6G

rena

daSp

ices

53.1

Mea

l,w

heat

or

mes

lin fl

our

10.8

Fres

h fis

h10

.274

.0St

.Vin

cent

and

the

Gre

nadi

nes

Frui

t an

d nu

ts47

.8M

eal,

whe

at o

r m

eslin

flou

r15

.0R

ice

11.2

74.0

Togo

Cot

ton

33.5

Fert

ilize

rs,c

rude

26.5

Con

stru

ctio

n m

ater

ial

11.6

71.7

Beliz

eSu

gar

25.9

Pres

erve

d fr

uits

23.8

Cru

stac

eans

and

mol

lusk

s20

.069

.6M

acau

Out

er,k

nitt

ed g

arm

ents

38.8

Out

er,w

omen

’s ga

rmen

ts16

.2U

nder

garm

ents

13.7

68.7

Sour

ce:U

NC

TAD

(20

01).

Not

e:W

ithin

eac

h gr

oup,

coun

trie

s ar

e ra

nked

by

the

tota

l sha

re o

f the

top

thr

ee e

xpor

ts.T

en c

ount

ries

with

the

hig

hest

exp

ort

shar

es a

re r

epor

ted.

The

cou

ntry

gro

up a

vera

ges

refe

r to

the

who

le s

ampl

e.


Both the IMF and the World Bank have long pro-vided facilities specifically designed to help coun-tries adjust to large terms-of-trade shocks. The IMFhas used a number of facilities including the Com-pensatory Financing Facility (CFF), established inthe 1960s to assist countries experiencing either asudden shortfall in export earnings or an increase incereal import costs caused by fluctuating worldcommodity prices. Yet, limits to administrative ca-pacity imply that only a few large terms-of-tradeshocks will lead to loans in the context of programssupported by the IFIs. More important, with shocksto commodity prices being highly persistent (Cashin,Liang, and McDermott, 2000), these loans facilitateadjustment to adverse shocks and the resulting lowerincome levels, but they are not designed to provideinsurance or maintain income levels through state-contingent transfers.

Countries seem to be reluctant to insure themselvesagainst commodity risks through financial instru-ments. Until the 1980s, government and multilateralinterventions aimed at price stabilization in commod-ity markets were commonplace. A consensus amongacademics and policymakers seems to have emergedsince then for a shift away from market interventionand toward management of commodity risks throughfinancial instruments (see, e.g., Claessens and Dun-can, 1993; Dehn, Gilbert, and Varángis, forthcom-ing).6 In the context of that shift, researchers in acade-mia and international institutions, notably the WorldBank, spurred a debate on the merits of commodity-price-linked bonds. Such bonds could be viewed as away for countries to hedge against changes in com-modity prices even in cases where futures or forwardsof a sufficiently long maturity are either absent or toocostly. Nevertheless, commodity-price-linked bondshave been used only in a limited number of instances.

Recent efforts to provide sovereigns with market-based insurance against commodity fluctuations donot appear to be attracting much interest. Since Sep-tember 1999, the World Bank has offered risk man-agement products linked to IBRD loan exposures,including swaps that seek to hedge against fluctua-tions in interest rates, currencies, and commodityprices. However, swaps that hedge against commod-ity price fluctuations have not been used by WorldBank clients to date. Nor do sovereigns make muchuse of hedging opportunities even where active andliquid markets exist for futures and forwards oncommodities with maturities extending beyond afew years, as is the case for oil (Daniel, 2001).

Sovereigns’ reluctance to use financial instru-ments to hedge against commodity price fluctuationsis not fully understood, though it may be related tothe following factors:

• Insurance—even if incomplete—is already pro-vided, to some extent, by the IFIs.

• Prices obtained by some countries for the partic-ular quality of the commodities they producemay not be closely correlated with those ob-served on international exchanges.

• Regarding commodity-price-linked bonds morespecifically, international investors often ex-press a preference for separating exposure tocountry risks (through standard bonds) from ex-posure to commodity price risks (through exist-ing markets for commodity price forwards andfutures).

• Without hedging, the borrowing country doesnot need to share its gains on the upside; and itmay force its lenders (through default) to sharethe downside.

• It may be politically difficult to pay for hedging,especially in the event that—with hindsight—this turns out to have been the wrong decision.This factor is not unique to sovereigns and maybe one of the reasons why hedging remainsmore limited than might be expected even in thecorporate world.7

Countries wishing to reduce volatility resultingfrom commodity price shocks could of course con-sider taking steps to diversify their economies. How-ever, this process may take many years and somecountries may even judge that their comparative ad-vantage really lies in a few products. Hedgingagainst commodity price fluctuations through finan-cial instruments, including commodity-price-linkedbonds, would therefore seem desirable for severalcountries.

Natural Disasters

Natural disasters take a massive human and eco-nomic toll on virtually all countries. Their impactrelative to the size of the economy tends to behigher, the lower the degree of economic develop-ment and the smaller the size of the country underconsideration. This is confirmed by considering thetop five most devastating disasters, ranked by the di-rect loss of capital stock in percent of GDP, for vari-

33

6Shocks to commodity prices tend to be highly persistent, im-plying that schemes to stabilize commodity export earningswould be exceedingly costly (Cashin and others, 2000). Davisand others (2001) suggest that the record of stabilization funds forrenewable resources is mixed.

7What would the president of an oil company say—after a risein oil prices—to the treasurer who had hedged against the pricechange in the futures market? See Hull (2002, p. 74).


ous groups of countries: advanced, emerging market,developing, and small (Table 5).8 For advancedcountries, the direct loss of capital stock attributableto specific natural disasters usually does not exceeda few percentage points of GDP. For emerging mar-kets, this direct impact can occasionally be equiva-lent to more than 10 percentage points of GDP. Forsmall developing countries, especially small islands,the impact can occasionally be equivalent to more

than a year’s worth of output. The impact on the fis-cal deficit and the trade deficit can also be equivalentto several percentage points of GDP (Freeman,Keen, and Mani, 2003). In many instances, the im-pact on the GDP growth rate is not commensuratewith the direct loss in the capital stock, suggestingthat indexation to GDP in these cases would not pro-vide sufficient insurance.

A few countries are routinely affected by the sametype of natural disaster: a few small islands are re-peatedly hit by cyclones; other countries are particu-larly prone to earthquakes, or floods (Table 6). How-ever, perhaps a majority of countries that are proneto natural disasters are affected by a variety of disas-ters, rather than a single type of event. This makes iteven more difficult in these countries to estimate thelikelihood of disasters and to obtain insurance con-tracts (or bond indexation clauses) than in countrieswhere disasters always tend to be of the same type.

34

Table 5. Top Five Natural Disasters by Percent of GDP Lost

Direct Change inEconomic Loss Type of GDP Growth Rate

Country Year (In percent of GDP) Disaster Measure (In percent)

Advanced economiesItaly 1980 4.5 Earthquake 7.2 Richter –2.6Australia 1982 3.3 Drought . . . –5.7Greece 1978 2.6 Earthquake . . . 4.5Japan 1995 2.5 Earthquake 7.2 Richter 0.5Spain 1983 2.4 Flood . . . 0.6

Emerging market countriesZimbabwe 1982 29.3 Drought . . . –9.7El Salvador 1986 27.3 Earthquake 7.5 Richter –1.2Dominican Republic 1998 13.8 Hurricane 210 kph –0.8El Salvador 2001 10.9 Earthquake 6.1 Richter –0.3Chile 1985 9.1 Earthquake . . . –0.9

Developing countries (excluding small countries)

Honduras 1998 38.0 Hurricane 270 kph –2.0Belize 2000 34.4 Hurricane 215 kph 6.7Jamaica 1988 28.2 Hurricane . . . 4.3Nepal 1987 24.6 Flood . . . –2.8Guatemala 1976 22.9 Earthquake 7.5 Richter 5.4

Small countriesSt. Lucia 1988 365.0 Hurricane . . . 10.0Mongolia 1996 171.6 Forest fire 80,000 km2 –3.7Vanuatu 1985 146.3 Cyclone . . . –4.6Samoa 1991 138.9 Cyclone 167 kph 2.1Dominica 1979 100.8 Hurricane . . . –30.1

Source: Emergency database (EM-DAT), compiled by Office of Foreign Disaster Assistance (OFDA) and Centre for Research on Epidemiology of Dis-asters (CRED).

Note: The definition of advanced countries follows that of the IMF’s World Economic Outlook. The set of emerging market countries is the union ofthose defined as such by the International Finance Corporation, Global Stock Market Factbook 2002, and the JPMorgan Emerging Market Bond IndexGlobal. Small countries are defined as those with GDP below $5 billion in 2002.The remaining countries are included in the set of developing countries.Direct economic loss is calculated using CRED’s measure of direct damage to physical infrastructure as a result of the natural disaster.

8Data are only available for the direct loss of capital in a sam-pled area affected by the disaster, and such estimates could be bi-ased, particularly if they are made very soon after a disasterstrikes. The overall effect on the economy is harder to gauge. Theimpact on income is likely greater in the immediate aftermath ofthe disaster, as demand by economic agents affected by the disas-ter declines and economic activity is disrupted more generally.Later on, however, growth may accelerate somewhat, as the econ-omy rebuilds its capital stock and catches up toward its previouslevels of activity.


Most natural disasters have a temporary impact onthe fiscal deficit and the economy’s growth rate. In theaftermath of a disaster, spending jumps to alleviatehumanitarian emergencies and to begin reconstructinginfrastructure. As far as the real economy is con-cerned, for most large disasters considered in Table 4,growth in the year of the disaster declined by a fewpercentage points but soon returned to its long-runtrend, typically within one or two years following thedisaster. Caselli and Malhotra (2004) find that, inmost instances, natural disasters have no permanentimpact on a country’s long-run growth path, thoughthey have a permanent effect on the level of income.From the narrow perspective of debt sustainability,most natural disasters may be viewed as creating re-payment difficulties in the aftermath of the disaster,but not as affecting the country’s ability or willingnessto meet its external obligations on a permanent basis.

This set of facts has implications for whethercountries are likely to be interested in insuring them-selves against natural disasters, and for the types ofinsurance contracts that can be more efficient.Larger, more advanced economies are typically ableto cope with natural disasters on their own, by lettingthe debt-to-GDP ratio rise in the aftermath of thedisaster and gradually reducing it over a number ofyears, through slightly higher taxes or lower spend-ing on other items. For emerging market countries,and especially developing countries, it is often diffi-cult to muster sufficient funds from the private capi-tal markets to cope with disasters. In many instances,the international community has therefore appropri-ately stepped in with aid and new lending. For itspart, the IMF provides emergency assistance tocountries that have experienced a natural disaster orare emerging from conflict.9 At the same time, somenatural disasters, even if seriously damaging, fail tocapture international attention. And even when aid isprovided by the international community, it is insuf-ficient to prevent declines in income levels, andlending merely helps smooth the adjustment to alower income level. Thus, disaster-prone countriesmight be expected to have a strong interest in obtain-ing insurance from the private sector. Both the WorldBank and the Inter-American Development Bankhave advocated greater use of insurance against nat-ural disasters for some countries, and have sought tohelp country clients obtain such insurance from themarkets (Gilbert and Kreimer, 1999; Inter-AmericanDevelopment Bank, 2002).

Despite the existence of markets for insuranceagainst natural disasters, at present only a handful ofcountries use them. Insurance is available both di-

rectly from insurance companies and through innov-ative financial instruments such as catastrophebonds, which waive some or all of the principal andinterest repayments in the event of a prespecifiedcatastrophe, or weather derivatives, which providepayouts in the event of temperatures or rainfallabove or below prespecified trigger levels over a cer-tain period. Sovereigns rarely opt for disaster insur-ance, even for their own property (Freeman, Keen,and Mani, 2003). Sovereigns rarely issue catastrophebonds, though it is somewhat more common for pri-vate companies to do so.

The possible reasons for why countries do notoften insure themselves against natural disasters in-clude the following:

• In the case of small and poor countries, insur-ance companies may consider the size of the po-tential market too limited for them to incur thecost of estimating the likelihood of disasters.

• Insurance premiums may be high because of po-tential moral hazard: contracts normally providefor payoffs based upon actual losses rather thanthe intensity of the natural phenomenon, whichcould be measured, say, by the number of pointson the Richter scale in the case of earthquakes.This tends to reduce countries’ incentives to in-vest in preventive measures aimed at reducingthe physical destruction that would result fromdisasters.

• Countries may expect the international commu-nity to step in with generous aid and financialsupport in response to disasters.

The international community could help countriesmeet a number of prerequisites to obtain insurancecontracts against natural disasters from the privatesector at a reasonable premium. These include his-torical data on previous events; appropriate infra-structure, such as weather stations, necessary forcontracts based on specific measurements, such asrainfall or temperatures; and historical correlationsof actual losses with the scale of events. Technicalassistance in these areas, especially to small coun-tries, might also be helpful.

Countries that are prone to natural disasterscould be encouraged to consider insurance throughthe private markets. To mitigate possible disincen-tives, the international community could emphasizeits readiness to step in with help even for countriesthat have obtained insurance from the private sec-tor. One possibility might be for the internationalcommunity to commit to providing emergency as-sistance, perhaps on a concessional basis, to coun-tries deemed to have undertaken appropriate mea-sures to mitigate the impact of possible disasters(Freeman, Keen, and Mani, 2003).

35

9Emergency loans are subject to the basic rate of charge andmust be repaid within 3!/4–5 years (without expectation of earlyrepayment).


36

Tabl

e 6.

Sm

all C

oun

trie

s:Ty

pes

of D

isas

ters

,197

5–20

02

Cum

ulat

ive

Cum

ulat

ive

Dam

age

GD

P 20

02

Dam

age

(Mill

ions

of 2

002

(Mill

ions

of

(Per

cent

N

umbe

r of

C

ount

ryU

.S.d

olla

rs)

U.S

.dol

lars

)of

GD

P)D

isas

ters

Mos

t Fr

eque

nt T

ypes

of D

isas

ters

Afg

hani

stan

601.

6..

...

.56

21 fl

oods

,19

eart

hqua

kes,

6 sl

ides

,3 d

roug

hts,

4 co

ld w

aves

Alb

ania

35.8

4,82

00.

712

5 flo

ods,

3 ea

rthq

uake

s,1

cold

wav

e,1

aval

anch

eA

ntig

ua a

nd B

arbu

da10

6.9

720

14.8

76

hurr

ican

es,1

dro

ught

Arm

enia

148.

52,

360

6.3

42

flood

s,1

eart

hqua

ke,1

dro

ught

Barb

ados

145.

52,

440

6.0

54

hurr

ican

es,1

floo

d

Beliz

e34

5.5

840

41.1

94

hurr

ican

es,3

floo

ds,1

col

d w

ave,

1 tr

opic

al s

torm

Beni

n11

.92,

700

0.4

1710

floo

ds,4

dro

ught

s,2

wild

fires

,1 w

inds

torm

Bhut

an3.

757

00.

64

2 flo

ods,

1 cy

clon

e,1

wild

fire

Bosn

ia a

nd H

erze

govi

na16

3.6

...

...

41

win

dsto

rm,1

slid

e,1

flood

,1 d

roug

htBu

rkin

a Fa

so..

.3,

140

...

179

drou

ghts

,6 fl

oods

,2 in

sect

infe

stat

ions

Buru

ndi

...

630

...

84

flood

s,3

drou

ghts

,1 w

inds

torm

Cam

bodi

a16

0.6

3,66

04.

413

8 flo

ods,

3 dr

ough

ts,2

fam

ines

Cap

e Ve

rde

Isla

nds

5.0

620

0.8

106

drou

ghts

,1 t

ropi

cal s

torm

,1 c

yclo

ne,1

vol

cano

,1 fa

min

eC

entr

al A

fric

an R

epub

lic0.

21,

060

0.0

115

flood

s,2

win

dsto

rms,

2 w

ildfir

es,2

dro

ught

sC

had

112.

81,

970

5.7

2613

dro

ught

s,7

flood

s,4

inse

ct in

fest

atio

ns,2

win

dsto

rms

Com

oros

65.1

250

26.0

74

cycl

ones

,2 v

olca

noes

,1 d

roug

htC

ongo

0.1

3,01

00.

05

4 flo

ods,

1 dr

ough

tD

jibou

ti10

.059

01.

714

7 dr

ough

ts,6

floo

ds,1

win

dsto

rmD

omin

ica

128.

125

051

.27

7 hu

rric

anes

Eritr

ea6.

1..

...

.5

3 dr

ough

ts,1

win

dsto

rm,1

inse

ct in

fest

atio

n

Fiji

813.

61,

830

44.5

2517

cyc

lone

s,3

flood

s,2

drou

ghts

,2 e

arth

quak

es,1

win

dsto

rmG

abon

...

4,96

0..

.1

1 flo

odG

ambi

a,T

he..

.35

0..

.15

9 dr

ough

ts,3

inse

ct in

fest

atio

ns,3

floo

dsG

eorg

ia2,

425.

03,

450

70.3

94

eart

hqua

kes,

2 flo

ods,

2 dr

ough

ts,1

win

dsto

rmG

rena

da29

.141

07.

14

2 hu

rric

anes

,1 t

ropi

cal s

torm

,1 fl

ood

Gui

nea

12.9

3,21

00.

48

3 flo

ods,

3 dr

ough

ts,1

tor

nado

,1 e

arth

quak

eG

uine

a Bi

ssau

...

200

0.0

126

flood

s,3

inse

ct in

fest

atio

ns,1

win

dsto

rm,1

wild

fire,

1 cy

clon

eG

uyan

a31

.171

04.

45

2 dr

ough

ts,2

floo

ds,1

slid

eH

aiti

289.

33,

550

8.1

3219

floo

ds,5

hur

rica

nes,

4 dr

ough

ts,2

win

dsto

rms

Kir

ibat

i..

.50

...

11

drou

ght

Kyr

gyzs

tan

239.

11,

600

14.9

83

slid

es,3

ear

thqu

akes

,1 c

old

wav

e,1

flood

Lao

P.D.R

.38

7.6

1,86

020

.823

12 fl

oods

,7 d

roug

hts,

2 ty

phoo

ns,2

win

dsto

rms

Leso

tho

...

800

...

145

win

dsto

rms,

5 dr

ough

ts,1

fam

ine,

3 flo

ods

Libe

ria

60.1

560

10.7

51

win

dsto

rm,1

slid

e,1

cold

wav

e,1

flood

,1 d

roug

htM

aced

onia

,For

mer

Yug

osla

v R

epub

lic o

f40

8.8

3,73

011

.05

2 flo

ods,

1 w

ildfir

e,1

cold

wav

e,1

fam

ine


37

Mad

agas

car

2,49

5.5

4,56

054

.731

22 c

yclo

nes,

6 dr

ough

ts,2

floo

ds,1

inse

ct in

fest

atio

nM

alaw

i37

.31,

930

1.9

2314

floo

ds,6

dro

ught

s,2

fam

ines

,1 e

arth

quak

eM

auri

tani

a54

.099

05.

525

12 d

roug

hts,

8 flo

ods,

3 in

sect

infe

stat

ions

,1 w

inds

torm

,1 t

orna

doM

auri

tius

1,44

5.8

4,56

031

.715

14 c

yclo

nes,

1 dr

ough

tM

icro

nesi

a,Fe

dera

ted

Stat

es o

f11

.8..

...

.4

2 tr

opic

al s

torm

s,1

typh

oon,

1 dr

ough

t

Mol

dova

195.

81,

630

12.0

74

flood

s,2

win

dsto

rms,

1 dr

ough

tM

ongo

lia2,

953.

51,

090

271.

016

7 w

inte

r st

orm

s,3

flood

s,3

wild

fires

,2 d

roug

hts,

1 w

inds

torm

Moz

ambi

que

442.

03,

620

12.2

4115

dro

ught

s,14

floo

ds,5

win

dsto

rms,

5 cy

clon

es,1

slid

e,1

fam

ine

Nam

ibia

1.1

2,87

00.

08

7 dr

ough

ts,1

floo

dN

ethe

rlan

ds A

ntill

es17

.12,

740

0.6

22

hurr

ican

es

New

Cal

edon

ia0.

4..

...

.7

7 cy

clon

esN

icar

agua

2,55

6.4

2,61

097

.932

8 flo

ods,

4 dr

ough

ts,4

vol

cano

es,3

ear

thqu

akes

,3 w

ildfir

es,4

sto

rms

Nig

er14

.12,

180

0.6

2110

dro

ught

s,7

flood

s,3

inse

ct in

fest

atio

ns,1

win

dsto

rmPa

pua

New

Gui

nea

582.

62,

930

19.9

3310

ear

thqu

akes

,6 v

olca

noes

,5 s

lides

,5 fl

oods

,3 d

roug

hts

Rw

anda

...

1,74

0..

.9

4 flo

ods,

4 dr

ough

ts,1

ear

thqu

ake

Sam

oa61

9.5

260

238.

36

3 cy

clon

es,1

win

dsto

rm,1

wild

fire,

1 flo

odSã

o To

mé

and

Prín

cipe

...

50..

.2

2 dr

ough

tsSe

yche

lles

1.8

700

0.3

21

flood

,1 t

ropi

cal s

torm

Sier

ra L

eone

9.9

780

1.3

43

win

dsto

rms,

1 flo

odSo

lom

on Is

land

s29

.432

09.

212

6 cy

clon

es,2

ear

thqu

akes

,2 d

roug

hts,

1 ts

unam

i,1

win

dsto

rm

St.K

itts

and

Nev

is32

7.8

360

91.1

76

hurr

ican

es,1

floo

dSt

.Luc

ia1,

552.

266

023

5.2

84

hurr

ican

es,2

tro

pica

l sto

rms,

1 w

inds

torm

,1 s

lide

St.V

ince

nt a

nd t

he G

rena

dine

s46

.336

012

.99

4 hu

rric

anes

,4 fl

oods

,1 v

olca

noSw

azila

nd85

.81,

200

7.2

108

drou

ghts

,1 c

yclo

ne,1

floo

dTa

jikis

tan

706.

31,

210

58.4

2512

floo

ds,5

slid

es,3

ear

thqu

akes

,2 d

roug

hts,

2 w

inds

torm

s

Togo

...

1,36

0..

.6

4 flo

ods,

2 dr

ough

tsTo

nga

112.

614

080

.49

7 cy

clon

es,1

win

dsto

rm,1

ear

thqu

ake

Vanu

atu

304.

323

013

2.3

2714

cyc

lone

s,8

eart

hqua

kes,

2 w

inds

torm

s,1

slid

e,1

volc

ano

Zam

bia

22.7

3,76

00.

614

7 dr

ough

ts,5

floo

ds,2

inse

ct in

fest

atio

ns

Sour

ce:C

ente

r fo

r R

esea

rch

on E

pide

mio

logy

of D

isas

ters

(C

RED

) an

d O

ffice

of F

orei

gn D

isas

ter A

ssis

tanc

e (O

FDA

),EM

-DAT

.N

ote:

Smal

l cou

ntri

es a

re d

efin

ed a

s th

ose

with

GD

P be

low

$5

billi

on in

200

2.D

isas

ters

in E

M-D

AT a

re d

efin

ed a

s th

ose

natu

ral d

isas

ters

tha

t ha

ve c

ause

d 10

or

mor

e fa

talit

ies,

affe

cted

100

or

mor

e pe

o-pl

e,le

d to

an

appe

al fo

r in

tern

atio

nal a

ssis

tanc

e,or

res

ulte

d in

a d

ecla

ratio

n of

a s

tate

of e

mer

genc

y.T

he d

ata

are

base

d on

rep

orts

sen

t to

CR

ED a

nd c

ompi

led

by e

xter

nal o

rgan

izat

ions

suc

h as

the

var

ious

UN

org

aniz

atio

ns,c

ount

ry g

over

nmen

ts,a

id-d

isbu

rsin

g ag

enci

es,a

nd r

eins

uran

ce c

ompa

nies

.


Changes in Total Imports by Main Trading Partners

As noted above, many countries would not derivemajor insurance benefits from instruments hedgingspecific risks, such as natural disasters or changes incommodity prices. For these countries, especiallyhighly open economies, changes in total imports oroutput by main trading partners might constitute animportant determinant of economic performance. Itis therefore worth considering whether it would bedesirable to let a country’s bond repayments dependon an index of total imports by the country’s maintrading partners.

Regressions of individual countries’ annual outputgrowth on trade-weighted partner growth may helpgauge the extent to which indexation to partnergrowth might stabilize the debt-to-GDP ratio for thevarious countries (Table 7).10 For many advancedeconomies, output growth is clearly related to devel-opments in partner countries. The relationship ismore tenuous—with lower average R2 coefficients—for emerging market countries and developing coun-tries.11 A possible interpretation of this finding is thatadvanced countries tend to export services and manu-

factures with a relatively high technological content,for which demand shocks are more relevant; by con-trast, emerging market countries and developingcountries tend to export commodities, for which sup-ply shocks are more relevant, and manufactures withlower technological content, for which demand mightincrease when the advanced economies experiencelow growth and redirect their demand toward cheapergoods. Developments in trading partners’ outputgrowth are found to be a key determinant of eco-nomic growth also over longer time horizons: usingfive-year panel regressions, Arora and Vamvakidis(2004) find that a 1 percentage point increase in trad-ing partners’ growth raises a country’s growth rate byas much as 0.8 percentage points, with the effectbeing somewhat stronger in more open economies.

In sum, several countries would derive consider-able benefits from insurance or indexation to realvariables largely beyond the control of the nationalauthorities, such as natural disasters, commodityprices, or trading-partner output growth. Neverthe-less, many countries, including several among themain emerging markets, seem unlikely to derive sig-nificant benefits from such insurance or indexation.

Real Variables Partially Within theControl of the Country’s Authorities

Indexing to broader indicators of a country’s eco-nomic performance would provide a closer match toits ability to repay. The closest indicators would beGDP or exports; related indicators such as industrialproduction or electricity consumption could also be

38

Table 7. Output Growth and Trading Partners’ Growth, 1970–2002

Slope Coefficient R2

Advanced economies (49)Mean 0.886 0.281Median 0.914 0.266

Emerging market countries (28)Mean 0.805 0.139Median 0.841 0.068

Developing economies (23)Mean 0.542 0.084Median 0.517 0.043

Sources: IMF, International Financial Statistics and Direction of Trade Statistics.Note:The regressions use annual data for 1970–2002, when available. Countries with fewer than 20 observations

are not considered.The definition of advanced economies is derived from the IMF’s World Economic Outlook.The setof emerging market countries is the union of those defined as such by the International Finance Corporation, GlobalStock Market Factbook 2000, and the JPMorgan Emerging Market Bond Index Global.The remaining countries are in-cluded in the set of developing countries.The number of countries in each group is reported in parentheses.

10Trading-partner output growth is constructed as a weightedaverage of output growth for all trading partners for which dataare available. For year t, the weights are computed as the share ofexports to each country and are 11-year moving averages cen-tered on year t.

11Similarly, the relationship is statistically significant in 18 outof 23 advanced economies, but there are only 9 out of 28 emerg-ing market countries and 9 out of 49 developing countries forwhich data are available for at least 20 years.

Real Variables Partially Within the Control of the Country’s Authorities

considered. The macroeconomic benefits that bor-rowing countries might derive from real indexationcan be illustrated by a number of simple numericalexamples, which, for the sake of brevity, are pre-sented here for the case of GDP.

A Simple Example

Consider the case of a country whose real GDPhas been growing for many years at 3 percent and itis expected to continue doing so. Assume that thiscountry can issue regular, plain vanilla bonds at, say,7 percent interest. That country could consider issu-ing a dollar-denominated, floating-rate bond with acoupon rate that varies according to the performanceof the domestic economy. Specifically, the couponrate could equal

coupont = max [r + (gt – g–), 0], (1)

where

gt = actual growth rate of GDP,

g– = baseline growth rate of GDP,

andr = 7 percent.

The baseline growth rate of GDP is agreed uponby the contracting parties prior to the bonds’ issue:in this case, 3 percent could be a reasonable base-line. The coupon rate might be expected to include asmall insurance premium in addition to the 7 percentcharged for plain vanilla bonds, but this insurancepremium is set to zero in this example. Yearlycoupon payments will be reduced by 1 percentagepoint for every percentage point by which real GDPgrowth falls short of its 3 percent trend—but thecoupon has a minimum of zero. In years whengrowth turns out to be 1 percent, the coupon will be5 percent. In years when growth turns out to be 5percent, the coupon will be 9 percent.

This exercise and those that follow are based onthe relatively simple formula above. Continuity—with small changes in realized growth resulting insmall changes in coupon payments—seems desir-able to minimize incentives to misreport. The needfor symmetry—with the coupon varying in propor-tion to the gap between actual and baseline GDPgrowth on both the upside and the downside—maybe a more open question. Many institutional bondinvestors are required to hold assets that pay a posi-tive interest rate, suggesting the need for a zero orpositive minimum for the coupon rate. Borrowersmight prefer to include a cap on coupon payments,which is omitted in this example. The link from thegrowth rate to coupon payments could follow morecomplicated formulas, but simplicity is likely to becrucial in helping sell this type of instrument. In-

dexation could apply to the principal, but it seemspreferable to apply it to the coupon, because thisyields interest savings during times of weak eco-nomic growth, thereby reducing the need for fiscalpolicies to be procyclical. At any rate, mutual agree-ment between borrowers and lenders would seemthe most natural way of determining the exact formof the contract.

When GDP growth turns out lower than usual,debt payments due will also be lower than in the ab-sence of indexation, helping maintain the debt-to-GDP ratio at sustainable levels, and avoiding whatcould be a costly and politically difficult adjustmentin the primary balance at a time of weak economicperformance. Conversely, when GDP growth turnsout higher than usual, the country will pay more thanit would have without indexation, thus reducing itsdebt-to-GDP ratio less than it would have otherwise.In sum, this insurance scheme keeps the debt-to-GDP ratio within a narrower range. For this insur-ance, the borrowing country will pay a small pre-mium above the interest rate that it would ordinarilybe charged.

Using the specific form of the contract in (1), thefollowing example may be considered. Suppose thatbeginning in 1990, half of the total government debtof Mexico and Argentina consisted of these GDP-indexed bonds. And suppose that the average growthrate over the 20 years prior to the beginning of thecontract is chosen as the baseline growth rate. Forthe purpose of this example, assume that the compo-sition of the debt has no impact on the behavior ofany of the other variables in the economy: variablessuch as the GDP growth rate and the overall deficitbehave exactly as they did in 1991–2002. Whatcoupon rates would have been paid on the bonds,and what would have been the interest savings (orextra costs) for these two countries?

Mexico (Figure 11, top panel) grew relativelyrapidly in the 20 years prior to 1990, by 4.4 percenton average, compared to about 3 percent in1991–2002. This would have resulted in an averagecoupon rate of 5.9 percent compared to the expected7 percent. In good years, Mexico would have paidhigher-than-average coupon rates. However, duringthe Tequila crisis of 1995, when output contractedby more than 6 percent, the coupon rate would havefallen to its minimum of zero. Mexico would haveobtained a large reduction in the interest bill, leavingmore room to avoid procyclical fiscal measures.Large interest savings would also have applied dur-ing the sudden slowdown of 2001–02.

In Argentina (Figure 11, lower panel), growthover the 20 years prior to 1990 averaged only 0.9percent per year. In 1991–2002, despite two majorcrises, the average growth rate rose to 2.3 percent.This would have resulted in coupon rates of 8.8 per-

39


cent on average. The coupon rate would have fallensharply in 1995 (Tequila crisis) and again beginningin 1999 (Brazil crisis), though it would have hit the minimum of zero only in 2002. Interestingly,Argentina defaulted on foreign debt in 2002 and ac-tually did not make any interest payments in thatyear. Argentina would have made roughly no netsavings on its interest bill: it would have paidhigher-than-average coupon rates in the years ofrapid growth (the early 1990s), but would havemade sizable savings in its interest bill since 1999.

Avoiding Procyclical Fiscal Policy

The example above illustrates how GDP indexa-tion of bond repayments could reduce the need forcountries to engage in procyclical fiscal policy.When GDP growth is below trend, the governmentwill be able to have a lower primary surplus (higherprimary spending and lower taxes) with indexationthan without it; conversely, when GDP growth isabove trend, the government will need to have a

higher primary surplus (lower primary spending andhigher taxes) than without indexation. Thus GDP in-dexation of bond repayments tends to make forsmoother paths of the primary surplus, taxes, andprimary spending, over the cycle. The advantagescould be large for both emerging market countriesand advanced economies (Box 7).

While this paper focuses on emerging marketcountries, GDP indexation could also help advancedeconomies to avoid procyclical fiscal policies, inparticular where the government faces constraintson its deficit level. The constraints could arise fromlegal or constitutional provisions (as for some of theU.S. states) or a concerted policy commitment suchas the Stability and Growth Pact of the EuropeanUnion. To gauge the benefits of GDP-indexed bondsfor advanced economies, Borensztein and Mauro(2004) consider what would have happened to thedebts, total deficits, and primary deficits of some ofthe European Monetary Union (EMU) countries hadthey been subject to a 3 percent of GDP limit on thefiscal deficit beginning in 1980, and then simulatewhat would have happened to their primary deficitshad the debt been indexed to GDP. Adjusting theprimary balance to meet the 3 percent total deficitceiling would have significantly curbed the abilityof some of these countries to conduct countercycli-cal fiscal policy. Growth-indexed debt, however,would have largely offset the impact of the deficitceiling, helping preserve the countercyclicality offiscal policy.

Having argued that GDP-indexed bonds mightprovide substantial insurance benefits, it remains tobe shown whether the insurance premium countriesmight expect to pay would be sufficiently small forthe insurance to be attractive to them.

Diversifiability of Growth Across Countriesand the Insurance Premium

Pricing financial instruments that do not yet existis a difficult task, but considerations loosely basedon the Capital Asset Pricing Model (CAPM) suggestthat the insurance premium on GDP-indexed bondsissued by emerging market countries would likely besmall. In fact, income growth rates are not highlycorrelated with possible measures of a “world mar-ket portfolio,” and the CAPM implies that only thesystematic portion of risk is reflected in expected re-turns, because unsystematic risk can be diversifiedaway by investors.12

40

12This is in line with studies that show large, unrealized gainsfrom international risk sharing resulting from the relatively lowcorrelation of income growth rates across countries, at a varietyof horizons (Athanasoulis, Shiller, and van Wincoop, 1999).

–1.0

–0.5

0

0.5

1.0

1.5

2.0

0299 20019795931991 –9

–6

–3

0

3

6

9

12

GDP growth rate

Coupon rateMexico

Argentina

2001 02999795931991–3

–2–1012345

–12

–8–4

048

121620

GDP growth rate

Coupon rate

Interest bill savings

Interest bill savings

Figure 11. Interest Savings over theEconomic Cycle(Interest bill savings as percent of GDP, left scale; Coupon and GDPgrowth rates in percent, right scale)

Source: IMF, World Economic Outlook.

Real Variables Partially Within the Control of the Country’s Authorities

Simple regressions of individual countries’ GDPgrowth rates on various proxies for the return on the“world market portfolio” (including world real stockreturns, U.S. real stock returns, world GDP growth,and U.S. GDP growth) yield relatively low R2 coeffi-cients.13 For emerging market countries, the highestR2 coefficient is 0.20 (Russia) using world stock re-turns; beta coefficients range from –0.6 (Bulgaria) to0.43 (Russia), with an unweighted average of 0.032(Borensztein and Mauro, 2004). Comovement acrosscountries is somewhat higher for advanced countries,and marginally lower for developing countries.

To illustrate the implications of the equationabove for the pricing of GDP-indexed bonds, the fol-lowing example assumes that the relevant portfolio

for investors is the world stock market, the risk-freerate of return is 3 percent, the expected return on themarket portfolio is 8 percent and, taking the case ofMexico from the (unreported) individual country regressions, that the β of the country’s growth rate with respect to the return on the world stockmarket is 0.072. Then, the indexation premium willbe (8 – 3) x 0.072, that is, 0.36 percentage point ayear—fairly small compared with the spreads oftenobserved in emerging markets. This premium is inexcess of the rate that the country pays on plainvanilla bonds, that is, it is in addition to the premiumthat compensates for default risk. It is likely, how-ever, that default risk would decline significantly if acountry were to convert a large portion of its debtinto indexed bonds such as these. For simplicity, theabove assumes that the default risk is uncorrelatedwith the GDP growth risk. The appeal of indexedbonds is even greater when this assumption is re-laxed, letting default risk rise when growth falls.

41

Box 7. Benefits of GDP Indexation for Emerging Markets and Advanced Economies

How much additional room would countries havehad for countercyclical fiscal policy if their debt hadbeen indexed to GDP at the beginning of the 1990s?To address that question, a simple exercise is con-ducted for 20 advanced economies and 25 emergingmarket countries. For each country, it is assumed that,in 1991, the entire debt stock was indexed to GDP;each year, the interest rate on the entire debt under theindexed contract would equal the implied interest rate(from the interest bill and the previous year’s debtstock) observed in the actual data plus an indexationterm equal to the difference between actual growth andbaseline growth (with a minimum of zero); the base-line growth in the contract was the average growth ratein 1980–2001, which could be viewed as resultingfrom a mix of adaptive expectations and perfect fore-sight; and total deficits (and thus the total debt path)

and economic growth were exactly as observed duringthe simulation period. Then it is estimated what theprimary balance would have been with indexation in1992–2001, and the resulting correlation between theprimary balance and the GDP growth rate—a sum-mary measure of the government’s ability to conductcountercyclical fiscal policy. (For example, if primaryspending is more expansionary during recessions, thatcorrelation will be higher.) This correlation is thencompared to the same measure based upon actualdata.

The correlation between the primary balance and theGDP growth rate would have been substantially higherwith indexation than it was in the actual data, and theincrease in correlation would have been slightly morepronounced for emerging market countries than itwould for advanced economies (see the table).

Correlation Between Primary Balance and Real GDP Growth

Emerging Markets Advanced Economies_______________________ ___________________________Without With Without With

indexation indexation indexation indexation

Mean 0.30 0.77 0.40 0.64Median 0.37 0.80 0.45 0.74

Sources: OECD, Analytical database; and IMF, International Financial Statistics, Government Finance Statistics, and Country Reports.Note: Emerging market countries include Argentina, Brazil, Bulgaria, Chile, China, Colombia, Côte d’Ivoire, Ecuador, Hungary, India, Indonesia,

Korea, Lebanon, Malaysia, Mexico, Morocco, Nigeria, Pakistan, Peru, Philippines, Poland, South Africa, Turkey, Uruguay, and Venezuela. Advancedeconomies include Australia,Austria, Belgium, Canada, Denmark, Finland, France, Germany, Iceland, Ireland, Italy, Japan, Luxembourg, Netherlands,New Zealand, Norway, Spain, Sweden, United Kingdom, and United States.

13The sample period is 1970–2001. The broad findings are un-changed, focusing on subperiods that might be viewed as charac-terized by greater financial and trade globalization or more fre-quent emerging market crises.


Obstacles for Variables Partly Withinthe Control of the Government

Despite its potential advantages, real indexationhas been used in practice only in a limited number ofcases (Box 8). If such advantages were really signif-icant, as the analysis above suggests, why has realindexation not been more widespread? Part of theanswer, of course, is that creating markets for new fi-nancial instruments is never easy, as argued in Sec-tion II. However, indexation to real variables that arepartly within the control of the government, such asGDP, presents an additional set of practical and con-ceptual obstacles that are discussed below.

• Possibility of lower incentives for growth-promoting policies. Bonds indexed to variablesthat are partly affected by government policiescould reduce the authorities’ incentives to pursuegrowth-promoting policies. The extent to which agovernment (whether benevolent or kleptocratic)would—for a number of years—pursue lessgrowth-oriented policies with indexed bonds thanit would with standard bonds (on which it coulddefault) is an open question, but investors mightreasonably be concerned about that possibility.

• Potential misreporting. When repayments arelinked to economic indicators produced by thedebtor country, the authorities might be temptedto tamper with the measurement of those indica-tors. How strong the temptation would be, andwhether the authorities would place their reputa-tion (and possibly market access) at stake arealso open questions.14 Nevertheless, the poten-tial for misreporting could make investors reluc-tant to hold instruments of this type. Of course, itis high growth rather than low growth that is typ-ically considered a success and gets politiciansreelected, but the incentives might be reversed ifthe new instruments were to constitute a largefraction of a country’s external debt.

• Data revisions. Even for advanced countries, re-visions compared with initial data releases forvariables such as GDP can be substantial. In-vestors might perceive potential data revisions asan unwelcome source of uncertainty, and mightbe concerned about the possibility that debtorcountries might use the revisions opportunisti-cally to reduce repayments.

• Lags. The benefits of real indexation in reducingthe procyclicality of fiscal policy depend onwhether repayments are linked to variables thattruly reflect the current state of the economy. Al-though the state of the economic cycle tends topersist in time and GDP data become availablewith only a few months’ delay in many coun-tries, debtor countries might be concerned aboutlinking repayments to macroeconomic indicatorsthat, in practice, become available with signifi-cant lags.

• High volatility of returns. While many interna-tional investors already invest in emerging mar-ket financial instruments with very volatile re-turns, such as stocks, some bond investors mightbe reluctant to accept the additional volatility ofreturns resulting from the variable coupon pay-ments associated with real indexation. Moreover,many institutional investors may be preventedfrom doing so by the constraints placed on therange of assets that they are allowed to invest in.

• Complexity and difficulty in pricing. While somefinancial market players thrive on dealing withand pricing complicated financial instruments,many investors, especially bond investors, areoften reluctant to buy instruments that are diffi-cult to understand and price. Indeed, for bond in-vestors in particular, there appears to be a trendaway from complex instruments toward simplebonds, as shown by several swaps to retire Bradybonds, and a decline in the prominence of floating-rate bonds. While a generally acceptedpricing formula does not seem to be necessaryfor a market to operate, lack of a well-estab-lished pricing model for GDP-indexed bondsmight hamper their acceptance by investors. In-deed, if investors were to apply a very high dis-count rate to future uncertain payments, borrow-ing countries would find GDP-indexed bondstoo costly to be issued.

• Politicians’ short horizons. As real indexationwould have a significant impact only for rela-tively long-term bonds, politicians currently inpower—given their short horizons—might be re-luctant to pay an insurance premium today in ex-change for benefits that might only be reaped bytheir successors.

• Inability to recall the bonds. Real indexation isunlikely to be consistent with the callability ofbonds. Suppose that a GDP-indexed bond werecallable: should GDP growth turn out better thanexpected, the interest rate faced by the country onstandard bonds would presumably fall (becausethe country would now look more solvent); theborrower would then have an incentive to recall

42

14It is not clear whether episodes of cheating on the data wouldnecessarily kill a market: stock markets, for example, have sur-vived many scandals of this type. Misreporting is always a possi-bility that investors are aware of, and it is presumably reflected inasset prices.

Steps to Foster Acceptance

the indexed bonds and issue plain vanilla bonds ata lower interest rate. At present, less than 5 per-cent of all emerging market bonds are effectivelycallable. However, the ability to recall a bond isanother important form of insurance—in this case,against fluctuations in interest rates. Therefore,the appeal of GDP-indexed bonds may depend onthe relative importance of uncertainty over interestrates and uncertainty over GDP growth.

The importance of such obstacles was assessedthrough a systematic survey of market participantsconducted by IMF researchers in collaboration withthe Emerging Markets Traders Association (EMTA)and the Emerging Markets Creditors Association(EMCA). Respondents identified liquidity and thepotential for mismeasurement of GDP as the key ob-

stacles in using growth-linked instruments. The sur-vey and the results are described in further detail inthe appendix.

Steps to Foster Acceptance

Overcoming many of the obstacles that mightmake it difficult for real indexation to emerge re-quires credibility, which is ultimately to be providedby the potential issuers. Nevertheless, a number ofadditional steps could be considered if these bondsare deemed useful:

• Improving the quality and timeliness of the data.Ensuring that the macroeconomic indicators ac-curately reflect the state of the economy is cru-

43

Box 8. Previous Examples of Indexation to Real Variables

A handful of emerging market countries have al-ready issued bonds with elements of real indexation:Mexico has issued bonds indexed to oil prices; andvarious Brady bonds issued by Mexico, Nigeria,Uruguay, and Venezuela to commercial banks in ex-change for defaulted loans in the early 1990s were is-sued with value recovery rights (VRRs) that were de-signed to provide additional payments in the event ofan increase in prices of commodities such as oil. How-ever, the indexation formulas were exceedingly com-plex, and the characteristics of each country’s formu-las differed widely. Moreover, there were restrictionson the tradability of the bonds and the detachability ofthe VRRs (http://www.emta.org/ndevelop/primer.pdf).Loans combined with protection (through swaps) fromcommodity price fluctuations have also been madeavailable by the World Bank to member countries, be-ginning in September 1999, though interest has thusfar been limited.

Costa Rica, Bulgaria, and Bosnia and Herzegovinahave issued bonds containing an element of indexationto GDP. These bonds, which were issued as part ofBrady restructuring agreements, contain clauses orwarrants (value recovery rights) that increase the pay-off to bondholders if GDP (or GDP per capita) of thedebtor country rises above a certain level. In the case ofBulgaria, the bonds provided for a GDP “kicker” suchthat, once real GDP exceeds 125 percent of its 1993level, creditors will be entitled to an additional 0.5 per-cent in interest for every 1 percent of real GDP growthin the year prior to interest payment. At the same time,these bonds were callable by the issuer and even at thetime of issue it was widely expected that Bulgariawould repay the principal and refinance it, should thekicker appear likely to be triggered by rapid economicgrowth. Indeed, Bulgaria has already swapped a por-tion of its indexed bonds for newly issued, nonindexedbonds. In any case, indexed bonds are very much ex-ceptions, and, in the few instances when they have been

issued, the indexation clause was set so far “out of themoney” that it was unlikely ever to be triggered.

Going beyond sovereigns, one set of bonds (for a fewhundred million dollars and a maturity of four to sixyears) recently restructured by the city of Buenos Airesincludes indexation of principal repayments to thecity’s revenues. This seems quite close to the notion ofequity for a public entity (the proceeds from the right tocollect taxes). Indexation to revenues, however, wouldseem unlikely to gain widespread acceptance, owing toreduced incentives to collect revenues.

Finally, moving away from bonds and toward pureclaims to the indexation component, a market for deriv-atives on indicators of real economic activity has re-cently been developed. In September 2002, GoldmanSachs and Deutsche Bank successfully completed thefirst-ever parimutuel auctions of economic deriva-tives—specifically, options on the U.S. Bureau ofLabor Statistics release of change in U.S. NonfarmPayroll data for September 2002 (www.longitude.com/html/news_oct04_2002.html). In April 2003, the samefirms hosted pari-mutuel auctions for three- and six-month options on the European Harmonized Index ofConsumer Prices (Risk Magazine, March 2003). (Seealso Baron and Lange, 2003.)

While there are no precedents of IFIs linking repay-ments to measures of economic activity, the IFIshave—in exceptional cases—made loan disbursementsexplicitly conditional on economic activity. The IMF’sStand-By Arrangement with Mexico in 1986 included agrowth contingency such that, should GDP growth fallbelow a benchmark level, the authorities would be al-lowed to implement an additional public investmentprogram, financed by additional loans from the WorldBank and commercial banks. The growth contingencymechanism was activated in 1988 (Boughton, 2001, pp.441–50). The use of adjustors—especially for changesin commodity prices—in IMF programs also plays asimilar role.


cial for countries to reap the full benefits of realindexation. This is especially important to en-sure that indexation acts as an accurate “auto-matic stabilizer.”

• Ensuring the integrity of the data. Investors’ mainconcern may be that errors or revisions could beused opportunistically in order to reduce debtpayments. While the existence of substantialmarkets for CPI-indexed bonds in many countriessuggests that data integrity issues are not insur-mountable, it may be more difficult to accuratelyestimate real GDP than consumer prices. To over-come these problems, the importance of macro-economic indicators could be reemphasized inthe current drive toward increased transparencyand improved data quality, which involves effortssuch as the Reports on the Observance of Stan-dards and Codes (ROSCs).15 Governments couldlikewise strive to guarantee the independence ofstatistical agencies. Alternatively, GDP-linkedcontracts could be based on indicators producedby independent sources not affiliated with thegovernment, or the quality of the data could beassessed by independent reviewers.16

• Drafting a sample indexation clause. As in thecase of collective action clauses, a sample index-ation clause could be designed for possible inclu-sion in bond contracts. To avoid uncertainty andscope for opportunistic tampering with the data,such a clause could provide a clear definition ofthe variables determining payments due. The def-inition could include the agency responsible forproducing the data, the time of data release andcoupon payment, and a statement that method-ological changes would not be taken into consid-eration for the purposes of determining paymentsdue. Alternatively, an outside party could vet thatany methodological change stemmed from purelytechnical motives. A clear method for dealingwith revisions could also be established. Onepossibility would be to state that data revisionswould be ignored, and to establish that couponpayments for each given date would be based onGDP as estimated on a predetermined date.

• Drawing on commitment to sound policies.Countries whose commitment to sound policiesis underpinned by rules or formal agreements(e.g., the Stability and Growth Pact) might be es-pecially good candidates for real indexation. In

fact, not only would such countries be morelikely to derive benefits from less procyclical fis-cal policies but they would also find it easier topersuade markets that real indexation would notresult in inappropriate policies.

• Building on existing systems of peer monitoringof data. Countries that have already shown theycan agree on common statistical standards to de-fine and monitor GDP data for an important pur-pose (such as European countries involved in theMaastricht convergence process and the Stabilityand Growth Pact) might find it relatively easy topersuade markets of the reliability of theirmacroeconomic data.

• Meeting demand by institutional investors. Op-portunities to place nonstandard bonds are some-times created by the investment objectives of institutional investors. For example, inflation-indexed bonds are often considered to be attrac-tive to private pension funds. Some advancedeconomies’ public pension systems de facto tendto index pension benefits to GDP. Private pen-sion plans, which may seek returns close to thatbenchmark, might therefore be interested in in-vesting in GDP-indexed bonds issued by thegovernment.

Real Indexation:Which Variables forWhich Countries?

Which variables are more likely to make real in-dexation attractive to both borrowing countries andinternational investors? Individual country charac-teristics are most relevant in determining borrowers’interest. Tailor-made contracts could be providedover the counter, especially to small countries, byglobal private financial institutions. However, somedegree of standardization of contracts across coun-tries would be key for the emergence of a dedicatedclass of traders and investors, and would be espe-cially important for the creation of an active and liq-uid secondary market on an organized exchange.

In considering which variables or set of variablesthey could index to, individual countries would seekto identify the variables that would best help pre-serve debt sustainability while attracting interest onthe part of international investors. In choosingamong variables that are largely outside the controlof the authorities, the main criterion is likely to bewhich contingency best adjusts the value of debt tothe country’s repayment capacity. Insurance againstnatural disasters or changes in commodity pricesmay be desirable for a number of countries, includ-ing several small developing countries and thoseemerging market countries whose production struc-

44

15On best practices with respect to data revisions, see Carson,Khawaja, and Morrison (2004).

16Payments arising from inflation-indexed securities issued byBrazil’s federal government are based upon an inflation measureproduced by the Getúlio Vargas Foundation.

Real Indexation:Which Variables for Which Countries?

tures rely on one or a few commodities. For largerand more diversified economies at a higher stage ofeconomic development, indexing to trading-partneroutput growth is more likely to be relevant.

When considering variables that are partly withinthe control of the authorities, considerations relatedto the credibility of policies and measurement arelikely to weigh more heavily. While the exercisesabove have considered the case of GDP indexation,the volume of exports is also a relevant measure ofeconomic performance and external repayment ca-pacity, and the debt-to-exports ratio is a closelywatched indicator. Indeed, for many developingcountries, data on exports might well be more reli-able than data on GDP. At the same time, govern-

ment policies may affect trade openness more di-rectly than they affect GDP, and indexation to acountry’s overall exports might reduce incentives toundertake policies promoting trade openness. Otheralternatives include industrial production or electric-ity consumption, which in some countries are highlycorrelated with GDP and yet possibly harder to tam-per with than GDP data. However, the extent towhich these variables represent good proxies foroverall economic activity varies widely across coun-tries. All in all, GDP—the most comprehensive andgenerally accepted measure of a country’s income—would seem the most natural candidate if the bondsof many different countries were to be indexed to thesame economic variable.

45

A s argued in Section II, several obstacles to fi-nancial innovation involve the need to coordi-

nate the actions of many economic agents. This sug-gests that financial innovation may be expected totake place in discrete steps, rather than gradually,which at times are hard to predict. Indeed, financialinnovation in practice turns out to look like a ratherhaphazard process. New financial instruments do notseem to be adopted as the end-product of a system-atic search, or a gradual evolution leading to supe-rior forms of finance. Instead, innovation seems toresult from historical accident, a constellation ofspecial circumstances, or strong intervention on thepart of policymakers.1

Financial Innovation in SovereignBorrowing: A Haphazard Process

While the bulk of sovereign borrowing has histori-cally taken the form of plain vanilla debt, sovereignborrowers have displayed considerable creativityover the years. In 1782, the State of Virginia issuedbonds linked to the price of land and slaves. In 1863,the Confederate States of America issued “cotton”bonds payable in pounds sterling or French francs butconvertible into cotton at a predetermined price. Thiswas an excellent hedge: if cotton prices went up,higher value would be transferred from the borrowerto the lender—as investors requested payment in theform of cotton—just as revenues rose for the Confed-erate States, a major cotton producer (Barone andMasera, 1997). In the pre–World War I era of bond fi-nance, it was common for the emerging market coun-tries of the day to issue bonds simultaneously in morethan one major exchange, with coupons payable inany one of a few currencies, at the discretion of theinvestor. For example, in 1913 China issued a bondwith coupons payable in sterling, rubles, marks,francs, or yen (Flandreau and Sussman, 2002). “Gold

clauses,” effectively indexing payments to the priceof gold, were widespread in the United States in thenineteenth century through 1933 (Kroszner, 1999).

Despite such creativity, innovations in sovereignborrowing have been limited, and regularities intheir timing and form are hard to discern on the basisof macroeconomic fundamentals. Leadership and in-tervention on the part of the official sector often un-derlie the timing and nature of financial innovation.This point may be illustrated by referring to threehistorical experiences: first, the history of the intro-duction of inflation-indexed bonds in a variety ofcountries; second, the episode of syndicated bankloans in the 1970s; and third, the recent introductionof collective action clauses.

• Inflation-indexed bonds. The case for inflation-indexed bonds has been made at various stagesduring the past couple of centuries by many of the leading economists of their day (Baroneand Masera, 1997). Yet, countries’ experiencesdiffer widely on the timing and circumstancesunder which they introduced inflation-indexedbonds (Table 8). A few sovereigns began issuing inflation-indexed bonds several decades ago;others have done so more recently; the vast ma-jority have never issued indexed bonds at all.There does not appear to be much associationbetween country characteristics and the intro-duction of inflation-indexed bonds. Countriesthat have issued indexed debt are at variousstages of development, including both advancedeconomies and emerging market countries. (De-veloping countries with relatively limited statis-tical capacity have not issued such bonds.) Whilesome countries, such as Brazil, Israel, and theUnited Kingdom, began issuing inflation-in-dexed bonds during periods of high inflation,others have experienced high inflation withoutresorting to these bonds, and several, such asCanada, Sweden, and the United States, beganissuing indexed bonds during periods of low in-flation. Moreover, for inflation-prone countries,indexed bonds are not necessarily issued in theproximity of inflation peaks. Similarly, while

VI Past and Future of Innovation inSovereign Borrowing

46

Note: The author of this section and the next is Paolo Mauro.1Borensztein and Mauro (2004) provide greater detail and fur-

ther evidence on the issues addressed in this section.

Financial Innovation in Sovereign Borrowing: A Haphazard Process

some countries introduced inflation-indexedbonds around the time they promoted the use ofprivate pension plans, there is no simple relation-ship with pension system or reform.

• Syndicated bank loans. One reading of history isthat the largest innovation in how emerging mar-kets countries have financed themselves duringthe past two centuries—namely, the temporaryswitch to syndicated bank loans in the 1970s—largely owed to official encouragement and im-plicit guarantees. Indeed, bond issues, now the

predominant form of financing for emergingmarkets, have historically been the norm: finan-cial flows to emerging markets took this form al-most exclusively during the first golden era ofglobal financial integration, until World War Ibrought such flows to an end (Bordo, Eichen-green, and Kim, 1998; Mauro, Sussman, andYafeh, 2002). When international capital startedflowing again toward emerging markets, in theaftermath of the first oil price shock, it did so inthe form of syndicated bank loans. Why did sav-ings by residents of oil-rich countries ultimately

47

Table 8. Introduction of Inflation-Indexed Securities by Sovereigns

Average CPI InflationRate in Three Years Indexed Public Debt Outstanding in 1999______________________________________

Prior to Introduction In millions of In percent of totalPeriod of Issue (In percent) U.S. dollars government debt

Argentina 1972–1989 18.6 0 0Australia 1985–1988 8.4 . . . . . .

1993– 3.8 27,860 29.5Brazil 1964– . . . 45,291 19.6Canada 1991– 4.6 6,636 1.5Chile 1956– 39.6 14,960 62.0Colombia 1967– 13.7 4,9491 13.21

Czech Republic 1997– 9.3 150 1.7Finland 1945– . . . 0.7 0France 1998– 1.7 3,994 0.6Greece 1997– 7.6 197 0.2Hungary 1995– 23.2 394 3Iceland 1955– 4.3 4942 11.52

India . . . . . . 166 0.2Ireland 1983– 18.6 260 1.1Israel 1955– 32.7 79,037 80.2Italy 19833 18.5 0 0

20034 2.2 11,938 0.9Japan 20045 –0.6 900 0.06

Mexico 1989– 110.7 2,528 8.4New Zealand 1977–1984 14.2 . . . . . .

1994– 1.4 361 2.3Norway 1982– 9.8 30 0.1Poland 1992–2000 292.2 07 0.07

Sweden 19523 8.2 . . . . . .1994– 5.4 15,475 12.5

Turkey 1994– 80.8 8,561 24.3United Kingdom 1975– 8 10.7 . . . . . .United Kingdom 1981– 13.2 55,288 12United States 1997– 2.8 57,014 0.8

Sources: Campbell and Shiller (1996); Kopcke and Kimball (1999); Price (1997); Deacon and Derry (1998); official websites of country authorities; andIMF staff estimates.

1January 2003.2February 2003.3Only one issue of inflation-indexed bonds.4One issue in September 2003, indexed to euro area inflation measured by Eurostat.5One issue in March 2004.6From March 2004.7From April 2000.8Index-linked national savings certificates.

VI PAST AND FUTURE OF INNOVATION IN SOVEREIGN BORROWING

make their way to oil-importing emerging mar-kets through banks in advanced economiesrather than directly through bonds? Official en-couragement and implicit guarantees of banklending by the advanced economies led oil-richcountry residents to place their savings primarilyin a handful of the largest and most prestigiousbanks in the advanced economies. In turn, thesebanks found it profitable to on-lend such fundsto emerging markets—at least until the debt cri-sis struck in the early 1980s. The return to bondsin the 1990s through the Brady deals of coursealso involved considerable intervention and sub-sidies on the part of the official sector.

• Collective action clauses. Finally, collective ac-tion clauses (CACs) had failed to emerge in NewYork law contracts until recently, even thoughthe case for their introduction had been made fora number of years as part of the debate on re-forming the international financial architecture.Once the international community mustered suf-ficient consensus about the desirability of CACs,they were adopted in sovereign bonds issued byseveral countries within a few months.2 This isone example of how innovation may take placein the context of international coordination.

Road Maps for Future Innovation

Financial innovation in sovereign markets seemsto be a somewhat haphazard process, and many ben-eficial financial innovations require intervention tobe successful. Some of the instruments described inthis paper seem to have desirable properties, and

with some improvement in statistical resources andcredibility of data and policies, they might attract in-vestors’ interest.

While the process of financial innovation alwaysfaces significant hurdles, externalities would seemespecially strong in the case of real indexation tovariables that are partly within the control of the au-thorities. It would thus be difficult to develop a mar-ket for indexed bonds of this type through a gradualapproach. A small initial issue would not be very at-tractive because it would do little to reduce the like-lihood of a debt crisis. The holders of contingentdebt would be implicitly subsidizing the holders ofnoncontingent debt. Two factors would substantiallyincrease the chances of success: large scale wouldhelp both to reduce the probability of default signifi-cantly and to ensure sufficient liquidity on secondarymarkets; international coordination would help bothto create a dedicated class of investors and to pro-vide opportunities for risk diversification.

Two scenarios seem plausible for new types ofbonds to be issued successfully:

• A large-scale launch by one country swapping asubstantial proportion of the existing debt, possi-bly in the context of a debt restructuring. Indeed,historically, restructurings have provided oppor-tunities for innovation: for example, value recov-ery rights (VRRs) were incorporated into Bradybond deals as investors looked for an upside op-portunity to recoup previous losses. Other coun-tries might follow, perhaps with large one-offswaps, not necessarily in the context of debt ser-vicing difficulties.

• A launch of new bonds of the same type by sev-eral countries with some degree of—possibly in-formal—coordination. The recent introductionof collective action clauses by several issuersseems to constitute a precedent in this respect.

48

2See IMF (2003g).

The debt structures of countries have an impor-tant influence on their economic performance

and vulnerability to crises. In particular, excessivereliance by emerging market countries on short-termdebt and foreign-currency debt exposes them to risksof rollover crises and sharp increases in the debt bur-den resulting from exchange rate changes. Ofcourse, risky debt structures are often themselvessymptoms of underlying institutional and policyweaknesses that need to be convincingly addressed.But beyond this, the paper has argued that there arevaluable lessons to be extracted for improving sover-eign debt structures from liability structures in thecorporate sector. In particular, debt with different de-grees of seniority, and instruments with equity-likefeatures, could help to reduce the vulnerabilities in-herent in current sovereign debt structures. Threekey messages emerge from the analysis:

• First, credibility of fiscal and monetary policiesis a central prerequisite to buttress investors’willingness to hold long-term local-currencybonds. Credibility, in turn, depends on both thequality of institutions and a reputation for soundpolicymaking. Without supporting reforms,building such a reputation can take many years,but the combination of macroeconomic stabi-lization with institutional and structural reformscan accelerate this process, as demonstrated bythe experience of several emerging marketcountries, including Chile, Israel, Mexico, andPoland, in the last decade. Soon after bringingtheir inflation rates into the single digits and un-dertaking reforms of their monetary and fiscalframeworks, these countries successfully issuedunindexed local-currency bonds with medium-term maturities. While initially relying on inflation-indexed bonds, which played a helpfuland important role in the transition, most ofthese countries graduated to routinely issuingnonindexed long-term local-currency debt. Thissuggests that emerging market countries can im-prove their debt structures relatively quickly, aslong as they show clear commitment to soundpolicies.

• Second, progress in overcoming the problem ofdebt dilution in the sovereign context could re-duce the cost of borrowing and increase marketaccess for low-debt countries, and help preventcrises that result from overborrowing and riskydebt structures in high-debt countries. In the cor-porate context, debt dilution is addressed throughmethods that include debt covenants and explicitseniority. This paper has argued for considerationof analogous innovations in the sovereign con-text, in order to curb incentives for overborrow-ing, reduce costs of borrowing at low levels ofdebt, and limit the bias toward risky types ofdebt, such as short-term debt. This said, measuresthat reduce the scope for debt dilution are alsolikely to have some drawbacks: for example,making borrowing harder at high levels of debtmay not always be desirable, especially if it exac-erbates the risk of liquidity crises. In addition,some open questions remain, including the con-sequences of explicit legal seniority for crisis res-olution and potential legal obstacles to the imple-mentation of first-in-time seniority. While thiscautions against making strong policy recom-mendations at the present time, the possible ben-efits of explicit seniority in the sovereign contextseem to warrant further attention to the issue.

• Third, this paper has argued that instruments withequity-like features, which provide for lower pay-ments in the event of adverse shocks and weakeconomic performance, could help sovereigns toimprove debt sustainability and international risksharing. Disaster insurance could benefit smallcountries prone to frequent natural disasters. In-dexation to commodity prices might confer bene-fits for commodity-producing countries. GDP-indexed bonds would likely provide substantialinsurance benefits to a broader range of countries,including the advanced economies and the mainemerging market countries, though they presentgreater implementation challenges. In principle,GDP-indexed bonds could be issued relativelyquickly, especially by countries with trusted andindependent statistical offices. Whether these

VII Conclusions

49

bonds would attract sufficient investor interest atreasonable cost to borrowing countries remains anopen, empirical question. In particular, potentialconcerns of investors about complexities and dif-ficulties in pricing would need to be addressed.Market acceptance and the requisite liquiditycould be sought through international coordina-tion or a large swap, possibly in the context of arestructuring. Countries could seek to ensure theindependence of their statistical agencies, andtechnical assistance efforts could be stepped up inthis area. More ambitiously, methods could besought whereby outside parties could provide an

independent view on whether countries’ data arebeing systematically distorted.

The analysis in this paper suggests that progress inameliorating debt structures could yield substantialbenefits in economic performance and internationalrisk sharing, while reducing the frequency of crisesand the damage they entail. While sound policies re-main a precondition for securing better sovereigndebt structures, renewed attention to innovativestructures that may have become possible as a resultof the increased sophistication of financial marketscould be well rewarded.

VII CONCLUSIONS

50

IMF researchers, in collaboration with the Emerg-ing Markets Traders Association (EMTA) and the

Emerging Markets Creditors Association (EMCA),conducted a survey study of market participants’ at-titudes toward innovation in emerging markets’debt instruments. The survey focused on growth-linked bonds, but also included questions aboutcommodity-indexed and local-currency securities.The survey was distributed among the members ofEMTA and EMCA, thus reaching a broad range ofbond market participants, including participantsfrom both the “buy side,” such as asset managersand proprietary trading desk managers, and the“sell side,” such as research strategists and ana-lysts, as well as both “crossover” investors, whohold emerging market bonds only occasionally, inresponse to perceived profit opportunities, andemerging-market-dedicated investors.

Sample and Distribution Method

A link to the web-based survey was distributed viae-mail by EMTA and EMCA intermediaries to about1,000 potential respondents at EMTA and 30 atEMCA, indicating that it had been designed by IMFresearchers, who would not have access to respon-dent names or other individual information. Individ-ual passwords allowed each respondent to fill out thesurvey only once.1 An option was also provided tofill out the survey and return it via e-mail or fax, anda few respondents chose to do so.

Research Strategy and Response Rate

In designing a survey of this type—to be sentwithout any incentives to busy financial market par-ticipants—there is a trade-off between length and thelikely response rate. The design was somewhat morecomprehensive and complex in both presentation ofbackground information and seeking of potential an-swers. Growth-linked bonds in particular are morecomplicated than other existing instruments andneeded to be presented clearly to financial marketparticipants who were not familiar with them. Fewerresponses based on a clearly formulated setup wereconsidered to be preferable to more responses basedon relatively limited information. Similarly, a keyobjective was considered to assess the relative im-portance of a fairly large number of potential obsta-cles to financial innovation.

Consequently, the number of responses turned outrelatively low, at 28. This response rate should beviewed bearing in mind that the main questions cov-ered uncharted territory and required considerableanalysis by the respondents. An alternative interpre-tation of the low response rate is that potential re-spondents may have been generally dismissive of theidea of growth-linked bonds and decided that com-pleting the survey was not worth their time. Accord-ing to that interpretation, the results reported belowwould reflect selection bias in favor of growth-linked bonds, that is, they would provide an exces-sively optimistic picture of investors’ views regard-ing growth-linked bonds.

Respondents were asked to indicate their profes-sion and areas of expertise, for example, buy side ver-sus sell side, or dedicated investor versus crossoverinvestor. Despite its small size, the set of actual re-spondents spans a broad range of professions andareas of expertise. In presenting the results, we high-light any systematic differences in the responses thatseem to depend on the type of respondent, though

Appendix Investors’ Attitudes TowardGrowth-Linked and OtherInnovative Financial Instrumentsfor Sovereign Borrowers: Resultsof a Survey

51

Note: The authors of this appendix are Eduardo Borenszteinand Paolo Mauro.

1This protected the integrity of the survey against the possibil-ity of a single individual providing multiple entries. The selectionof passwords for individual respondents was conducted by inter-mediaries. This information was not available to the researchers.

APPENDIX

such results need to be interpreted with special cau-tion owing to the small sample size.

Survey Results

Overall Attitudes Toward Growth-LinkedBonds

Questions 1 and 2 sought to assess respondents’overall attitudes toward growth-linked bonds. Theypresented two alternative designs for growth-linkedbonds, and asked respondents to price such bonds.

Question 1 asked respondents to consider the caseof an emerging market sovereign borrower (“Emerg-ingLand,” EL) that had successfully tapped interna-tional financial markets for a number of years, wascurrently not experiencing major problems, butwhose bonds were trading at substantial spreadsabove U.S. treasuries. In the example, EL had expe-rienced real GDP growth of 3 percent on averageover the past 15 years, with a maximum of 7 percent,and a minimum of negative 8 percent; and averagegrowth and volatility of GDP could be expected tobe similar in the next decade. Respondents wereasked to assume that 10-year eurobonds (U.S. dollar-denominated) issued by EL (“plain vanilla bonds,”PV bonds) with a coupon of 7 percent currentlytrade at a spread of 400 basis points above U.S. trea-suries. EL was said to be contemplating issuance of

a growth-indexed bond (“Growth Bond”) with a 10-year maturity and with a coupon of 7 percent plusthe difference between real GDP growth during thatyear and 3 percent. However, coupon payments wererestricted to be non-negative:

Coupon = 7 percent + (real GDP growth – 3 percent), with a minimum of zero.

Respondents were asked what premium theywould require to hold a growth-linked bond ratherthan the plain vanilla bonds offering the same ex-pected coupon payment. The growth-linked bondwas designed to pay higher coupons in years whengrowth was higher than average, and lower couponsin years when growth was lower than average, with aminimum of zero. Respondents were given the fol-lowing options: (1) spreads more than 50 basispoints lower than PV bonds; (2) spreads 10–50 basispoints lower than PV bonds; (3) same spreads as PVbonds; (4) spreads 10–50 basis points higher thanPV bonds; (5) spreads 50–100 basis points higherthan PV bonds; (6) spreads 100–200 basis pointshigher than PV bonds; (7) spreads 200–300 basispoints higher than PV bonds; (8) spreads more than300 basis points higher than PV bonds; and (9) un-willing to purchase regardless of the spreads.

As Figure A1 shows, there was a wide variety ofanswers: some respondents said that they would ac-cept spreads that were lower or the same as those onPV bonds, whereas others said that they would beunwilling to purchase such bonds regardless of thespreads. The median answer was a premium of be-tween 100 and 200 basis points. Buy-side respon-dents indicated somewhat higher premiums on aver-age than sell-side respondents, as did “dedicated”emerging market investors compared to those whoidentified themselves as “crossover” investors.

Question 2 asked for the required premium whenthe growth-linked bond had a different specification,which ensured a minimum positive coupon paymentregardless of the economic performance of the bor-rowing country at any given time. The yearly couponhad a minimum of 3.5 percent, and an extra payoff inyears of positive growth, according to the followingformula:

Coupon = 3.5 percent + real GDP growth, with a minimum of 3.5 percent.

Respondents were told that this bond carried thesame expected coupon as the bond in question 1. Re-spondents displayed greater propensity to hold thebond in question 2, with the mean premium over aplain vanilla bond being just over 100 basis points(Figure A2). This compares with a mean of just above150 basis points for the bond in question 1. Again,

52

Percent of Responses

0

5

10

15

20

25

30

35

Unwilli

ng

to Pu

rchase

More

than

300

200 –

300

100 –

200

50 –1

00

10 –5

0

Lower

or Sa

me

Figure A1. Question 1: Premium over PlainVanilla Bonds (In basis points)

Source: IMF staff.

Appendix

sell-side participants and crossover investors appearedmore willing to hold this bond than did buy-side mar-ket participants and dedicated investors.

Main Obstacles to Growth-Linked Bonds

Questions 3 and 4 sought to gauge the relative im-portance of a number of obstacles to the introductionof growth-linked bonds or, equivalently, the sourcesof premiums that investors would demand to holdgrowth-linked bonds rather than plain vanilla bonds.

Question 3 asked respondents whether any of a setof five hypothetical changes to the status quo wouldlead them to reduce the premiums they required tohold growth-linked bonds. The hypothetical changesas well as the corresponding mean and median an-swers are reported in Table A1: 1 is very importantand would lead respondents to reduce the spreads by50 basis points or more; 2 would lead respondents toreduce the spreads by 20–50 basis points; 3 wouldlead respondents to reduce the spreads by 10–20basis points; and 4 is irrelevant.

As shown in Table A1, among the factors thatwould make respondents more willing to holdgrowth-linked bonds, they highlighted the issuanceof a large volume of growth-linked bonds in the con-text of a debt restructuring operation and methodsaimed at buttressing the integrity of the GDP data.

Question 4 was of a qualitative nature and askedrespondents to consider which obstacles made themreluctant to hold growth-linked bonds. The potentialobstacles as well as the corresponding mean and me-

dian answers are reported in Table A2: 1 is very im-portant, 4 is not important.

Among the factors that made investors reluctant tohold growth-linked bonds, respondents pointed mostoften to uncertainty about future liquidity in marketsfor these bonds and concerns about the integrity of

53

Table A1. Question 3: Obstacles to Growth-Linked Bonds

Hypothetical Change Mean Median

The United States is planning to issue growth bonds at about the same time. 3.35 4

Five other major emerging market sovereigns are planning to issue growth bonds at about the same time. 2.95 3

A reliable economic consultancy firm announces it will provide free software with a formula for pricing growth bonds. 3.52 4

A well-respected international consortium reports a study showing that the GDP data provided by the country are reliable, and announces it will monitor GDP data quality annually. 2.78 3

Growth bonds covering at least 50 percent of the country’s debt are issued in the context of a negotiated restructuring of EL’s debt. 2.42 2

Percent of Responses

0

5

10

15

20

25

Unwilli

ng

to Pu

rchase

More

than

300

200 –

300

100 –

200

50 –1

00

10 –5

0

Lower

or Sa

me

Figure A2. Question 2: Premium over PlainVanilla Bonds (In basis points)

Source: IMF staff.

APPENDIX

54

Table A2. Question 4: Obstacles to Growth-Linked Bonds

Potential Obstacle Mean Median

Uncertainty about future liquidity of growth bonds. 1.72 2

Complexity/difficulty in pricing. 2.23 2

Uncertainty about integrity of GDP data reported by EL. 1.73 1

Concern that EL will have fewer incentives to promote economic growth. 3.20 3

Variable coupon instead of fixed coupon. 2.91 3

Table A3. Question 5: Obstacles to Commodity-Indexed Bonds


It is too difficult to forecast commodity prices beyond three to five years. 2.54 3

You invest in many countries and only a few countries are heavily dependent on a single commodity. It is not worth your time to learn about commodity prices. 3.17 3

You are not interested in direct exposure to commodity price fluctuations, even if many of the countries you invest in are heavily dependent on commodities. 2.71 3

You are interested in exposure to commodity price fluctuations, but prefer to obtain it directly through forwards or futures linked to commodity prices. 2.36 2

Table A4. Question 6: Obstacles to Domestic-Currency Bonds


You are concerned about the possibility of a rise in inflation in EL. 1.91 1

You are concerned that EL’s central bank could intervene in foreign exchange markets and pursue an unfavorable exchange rate close to the time when bond payments are due. 2.00 2

You are concerned about your ability to hedge exposure to EL’s currency because of an illiquid NDF market. 2.04 2

Appendix

GDP data that must be provided by the issuing sov-ereign. Those concerns were less important for dedi-cated emerging markets investors than they were forcrossover investors. These results are consistent withthe answers to question 2, and highlight the impor-tance of both data reliability and market liquidity forpotential investors in growth-linked bonds.

Commodity-Indexed Bonds

Question 5 sought to identify the reasons whycommodity-indexed bonds have not been used morefrequently. It asked respondents to consider the caseof an emerging market that was heavily dependenton exports of a single commodity and sought toissue commodity-indexed bonds, that is, bondswhose return was indexed to the price of that com-modity. Again, the question asked respondents toidentify which potential obstacles made them reluc-tant to invest in commodity-indexed bonds. The po-tential obstacles as well as mean and median an-swers are reported in Table A3: 1 is very important, 4is not important.

Respondents indicated that difficulties in forecast-ing commodity prices beyond a three-to-five-year horizon and a preference to obtain exposure tocommodity prices directly through commodity de-rivatives made them reluctant to hold commodity-

indexed bonds. In additional comments, some fundmanagers noted that they lacked a mandate to investin commodities.

Domestic-Currency Bonds

Finally, respondents were asked to indicate therelative weight they attached to possible obstacles toholding domestic-currency-denominated bonds.

Question 6 asked respondents to indicate whichobstacles made them reluctant to invest in bonds de-nominated in EL’s currency. The potential obstaclesas well as the corresponding mean and median an-swers are reported in Table A4: 1 is very important, 4is not important.

Thus the factors that seem to make respondentsmore reluctant to invest in domestic-currency bondsinclude concerns about exchange rate manipulationand an unexpected rise in inflation. In additionalcomments, some respondents also cited concerns re-garding the convertibility of the domestic currency(with the Russian GKOs case being recalled) and thedomestic legal jurisdiction of local currency bonds.Interestingly, crossover investors seemed more will-ing to invest in local-currency instruments: their re-sponses revealed uniformly less concern with all thepossible factors that were suggested as deterrents toundertaking such investment.

55

References

Abiad, Abdul, and Ashoka Mody, 2003, “Financial Re-form: What Shakes It? What Shapes It?” IMF Work-ing Paper 03/70 (Washington: International MonetaryFund).

Aghion, Philippe, Philippe Bacchetta, and Abhijit Baner-jee, 2001, “Currency Crises and Monetary Policy in aCredit-Constrained Economy,” European EconomicReview, Vol. 45, No. 7, pp. 1121–50.

Alesina, Alberto, Alessandro Prati, and Guido Tabellini,1990, “Public Confidence and Debt Management: AModel and a Case Study of Italy,” in Public DebtManagement: Theory and History, ed. by RudigerDornbusch and Mario Draghi (Cambridge: Cam-bridge University Press), pp. 94–118.

Allen, Franklin, and Douglas Gale, 1994, Financial Inno-vation and Risk Sharing (Cambridge, Massachusetts:MIT Press).

Allen, Mark, 1988, “A Complementary Approach to theDebt Problem” (unpublished; Washington: Interna-tional Monetary Fund).

———, Christoph B. Rosenberg, Christian Keller, BradSetser, and Nouriel Roubini, 2002, “A Balance SheetApproach to Financial Crisis,” IMF Working Paper02/210 (Washington: International Monetary Fund).

Arora, Vivek, and Athanasios Vamvakidis, 2004, “TradingPartners: How Much Do They Matter for Growth?”(unpublished; Washington: International MonetaryFund).

Asquith, Paul, and T.A. Wizman, 1990, “Event Risk,Covenants and Bondholder Returns in LeveragedBuyouts,” Journal of Financial Economics, Vol. 27,pp. 195–213.

Athanasoulis, Stefano G., Robert J. Shiller, and Eric vanWincoop, 1999, “Macro Markets and Financial Secu-rity,” Federal Reserve Bank of New York EconomicPolicy Review, Vol. 5, No. 1, pp. 21–39.

Bailey, Norman, 1983, “A Safety Net for Foreign Lend-ing,” Business Week, January 10.

Barclay, Michael, and Clifford Smith, 1995, “The PriorityStructure of Corporate Liabilities,” Journal of Fi-nance, Vol. 50, No. 3, pp. 899–917.

Baron, Kevin, and Jeffrey Lange, 2003, “From Horses toHedging,” Risk Magazine, February. Available via theInternet: http://www.risk.net.

Barone, Emilio, and Rainer Masera, 1997, “Index-LinkedBonds from an Academic, Market and Policy-MakingStandpoint,” in Managing Public Debt, ed. by Marcello De Cecco, Lorenzo Pecchi, and GustavoPiga (Cheltenham, United Kingdom: Edward Elgar),pp. 117–47.

Barro, Robert, 1995, “Optimal Debt Management,” NBERWorking Paper No. 5327 (Cambridge, Massachusetts:National Bureau of Economic Research).

Blejer, Mario I., and Adrienne Cheasty, 1991, “The Mea-surement of Fiscal Deficits: Analytical and Method-ological Issues,” Journal of Economic Literature, Vol.29, No. 4, pp. 1644–78.

Bolton, Patrick, and David Skeel, 2003, “Inside the BlackBox: How Should a Sovereign Bankruptcy Frame-work Be Structured?” (unpublished; Princeton, NewJersey: Princeton University).

Bolton, Patrick, and Olivier Jeanne, 2004, “Structuringand Restructuring Sovereign Debt: The Role of Se-niority” (unpublished; Washington: InternationalMonetary Fund).

Bordo, Michael, Barry Eichengreen, and Jongwoo Kim,1998, “Was There Really an Earlier Period of Interna-tional Financial Integration Comparable to Today?”NBER Working Paper No. 6738 (Cambridge, Massa-chusetts: National Bureau of Economic Research).

Bordo, Michael, Christopher Meissner, and Angela Re-dish, 2003, “How ‘Original Sin’ Was Overcome: TheEvolution of External Debt Denominated in DomesticCurrencies in the United States and the British Do-minions,” NBER Working Paper No. 9841 (Cam-bridge, Massachusetts: National Bureau of EconomicResearch).

Borensztein, Eduardo R., and Paolo Mauro, 2004, “TheCase for GDP-Indexed Bonds,” Economic Policy(April), pp. 165–216.

Boughton, James, 2001, Silent Revolution: The Interna-tional Monetary Fund, 1979–1989 (Washington: In-ternational Monetary Fund).

Buchheit, Lee C., and G. Mitu Gulati, 2002, “SovereignBonds and the Collective Will,” Emory Law Journal,Vol. 51, No. 4, pp. 1317–64.

Burger, John D., and Frank E. Warnock, 2003, “Diversifi-cation, Original Sin, and International Bond Portfo-lios,” International Finance Discussion Paper No. 755(Washington: Board of Governors of the Federal Re-serve System).

Bussière, Matthieu, and Christian Mulder, 1999, “ExternalVulnerability in Emerging Market Economies: HowHigh Liquidity Can Offset Weak Fundamentals andthe Effects of Contagion,” IMF Working Paper 99/88(Washington: International Monetary Fund).

Caballero, Ricardo, 2003, “Coping with Chile’s ExternalVulnerability: A Financial Problem,” in CentralBanking, Analysis, and Economic Policies, Vol. 6(Santiago: Banco Central de Chile).

56

References

———, Kevin Cowan, and Jonathan Kearns, 2003, “Dol-lar-Risk, Banks and Fear-of-Sudden-Stop: Lessonsfrom Australia and Chile,” paper presented at theInter-American Development Bank conference, “Fi-nancial Dedollarization: Policy Options,” December.

Calomiris, Charles W., and Charles M. Kahn, 1991, “TheRole of Demandable Debt in Structuring OptimalBanking Arrangements,” American EconomicReview, Vol. 81, No. 3, pp. 497–513.

Calvo, Guillermo, 1988, “Servicing the Public Debt: TheRole of Expectations,” American Economic Review,Vol. 78, No. 4, pp. 647–61.

———, 2003, “Explaining Sudden Stop, Growth Col-lapse, and BOP Crisis: The Case of DistortionaryOutput Taxes,” Staff Papers, International MonetaryFund, Vol. 50 (Special Issue), pp. 1–20.

———, and Pablo Guidotti, 1990, “Indexation and Matu-rity of the Government Bonds: An ExploratoryModel,” in Public Debt Management: Theory and His-tory, ed. by Rudiger Dornbusch and Mario Draghi(Cambridge: Cambridge University Press), pp. 52–93.

Campbell, J.Y., and R.J. Shiller, 1996, “A Scorecard forIndexed Government Debt,” in NBER Macroeconom-ics Annual 1996, ed. by Ben S. Bernanke and Julio R.Rotemberg, pp. 155–97.

Carson, Carol S., S. Khawaja, and T. Morrison, 2004, “Re-visions Policy for Official Statistics: A Matter ofGovernance,” IMF Working Paper 04/87 (Washing-ton: International Monetary Fund).

Caselli, Francesco, and Priyanka Malhotra, 2004, “Na-tional Disasters and Growth: From Thought Experi-ment to National Experiment” (unpublished: Wash-ington: International Monetary Fund).

Cashin, Paul A., Hong Liang, and C. John McDermott,2000, “How Persistent Are Shocks to World Com-modity Prices? Staff Papers, International MonetaryFund, Vol. 47, pp. 177–217.

Chalk, Nigel A., 2002, “The Potential Role for Securitiz-ing Public Sector Revenue Flows: An Application tothe Philippines,” IMF Working Paper 02/106 (Wash-ington: International Monetary Fund).

Chamon, Marcos, 2002, “Why Can’t Developing Coun-tries Borrow from Abroad in Their Currency?” SocialScience Research Network Electronic Library Work-ing Paper.

———, 2004, “Can Debt Crises Be Self-Fulfilling?” IMFWorking Paper 04/99 (Washington: InternationalMonetary Fund).

Claessens, Stijn, and Ronald C. Duncan, 1993, ManagingCommodity Price Risk in Developing Countries (Bal-timore, Maryland: Johns Hopkins University Press).

Claessens, Stijn, Daniela Klingebiel, and Sergio Schmuk-ler, 2003, “Government Bonds in Domestic and For-eign-Currency: The Role of Macroeconomic and Institutional Factors” (unpublished; Washington:World Bank).

Cole, Harold, and Patrick Kehoe, 2000, “Self-FulfillingDebt Crises,” The Review of Economic Studies,Vol. 67, pp. 91–116.

Collier, Paul, and Jan Dehn, 2001, “Aid, Shocks, andGrowth,” Policy Research Working Paper 2688(Washington: World Bank).

Daniel, James, 2001, “Hedging Government Oil PriceRisk,” IMF Working Paper 01/185 (Washington: In-ternational Monetary Fund).

Davis, Jeffrey, Rolando Ossowski, James Daniel, andSteven Barnett, 2001, Stabilization and SavingsFunds for Nonrenewable Resources, IMF OccasionalPaper No. 205 (Washington: International MonetaryFund).

Deacon, Mark, and Andre Derry, 1998, Inflation-IndexedSecurities (London: Prentice Hall).

Dehn, Jan, Christopher L. Gilbert, and Panos Varangis,“Agricultural Commodity Price Volatility,” in Manag-ing Economic Volatility and Crises: A Practitioner’sGuide (forthcoming; Cambridge: Cambridge Univer-sity Press).

Detragiache, Enrica, 1994, “Sensible Buybacks of Sover-eign Debt,” Journal of Development Economics, Vol.43, pp. 317–33.

———, and Antonio Spilimbergo, 2001, “Crises and Li-quidity: Evidence and Interpretation,” IMF WorkingPaper 01/2 (Washington: International MonetaryFund).

Diamond, Douglas W., 1991, “Debt Maturity Structureand Liquidity Risk,” Quarterly Journal of Economics,Vol. 106, No. 3, pp. 709–37.

———, 1993, “Seniority and Maturity of Debt Contracts,”Journal of Financial Economics, Vol. 33, No. 3, pp.341–68.

———, and Raghuram Rajan, 2001, “Banks, Short-TermDebt, and Financial Crises: Theory, Policy Implica-tions and Applications,” Journal of Monetary Eco-nomics, Proceedings of the Carnegie-Rochester Con-ference on Public Policy, Vol. 54, pp. 37–71.

Dooley, Michael P., 2000, “Can Output Losses FollowingInternational Financial Crises Be Avoided?” NBERWorking Paper No. 7531 (Cambridge, Massachusetts:National Bureau of Economic Research).

———, and Sujata Verma, 2001, “Rescue Packages andOutput Losses Following Crises,” NBER WorkingPaper No. 8315 (Cambridge, Massachusetts: NationalBureau of Economic Research).

Drèze, Jacques H., 2000a, “Globalisation and Securitisa-tion of Risk Bearing,” CORE (unpublished; Univer-sité Catholique de Louvain, Belgium). Available viathe Internet: www.core.ucl.ac.be/staff/dreze.html.

———, 2000b, “Economic and Social Security in theTwenty-First Century, with Attention to Europe,”Scandinavian Journal of Economics, Vol. 102, No. 3,pp. 327–48.

Easterly, William R., 2001, “Growth Implosions and DebtExplosions: Do Growth Slowdowns Cause PublicDebt Crises?” Contributions to Macroeconomics, Vol.1, No. 1, Article 1 (Berkeley, California: ElectronicPress). Available via the Internet: http://www.bepress. com/bejm/contributions/vol1/iss1/art1.

———, Michael Kremer, Lant Pritchett, and Lawrence H. Summers, 1993, “Good Policy or Good Luck?Country Growth Performance and TemporaryShocks,” Journal of Monetary Economics, Vol. 32,pp. 459–83.

Eaton, Jonathan, 2002, “Standstills and an InternationalBankruptcy Court,” paper presented at the Bank of

57

REFERENCES

England conference, “The Role of the Official andPrivate Sectors in Resolving International FinancialCrises,” July.

———, and Raquel Fernandez, 1997, “Sovereign Debt,”in Handbook of International Economics, ed. byGene Grossman and Kenneth Rogoff (Amsterdam;New York and Oxford: Elsevier, North-Holland),pp. 2031–77.

Eichengreen, Barry, Ricardo Hausmann, and Ugo Panizza,2002, “Original Sin: The Pain, the Mystery, and theRoad to Redemption,” in Debt Denomination and Fi-nancial Instability in Emerging Market Economies,ed. by Barry Eichengreen and Ricardo Hausmann(Chicago: University of Chicago Press).

Eichengreen, Barry, and Ashoka Mody, 1998, “What Ex-plains Changing Spreads on Emerging-Market Debt:Fundamentals or Market Sentiment?” NBER Work-ing Paper No. 6408 (Cambridge, Massachusetts: Na-tional Bureau of Economic Research).

Englund, Peter, Torbjorn Becker, and Anders Paalzow,1997, Public Debt Management (Statsskuldspolitiken,in Swedish with English summary), Statens Of-fentliga Utredningar, Vol. 66 (Stockholm: Ministry ofFinance).

Falcetti, Elisabetta, and Alessandro Missale, 2002, “PublicDebt Indexation and Denomination with an Indepen-dent Central Bank,” European Economic Review,Vol. 46, pp. 1825–50.

Fama, Eugene F., and Merton H. Miller, 1972, The Theory of Finance (New York: Holt, Rinehart andWinston).

Fischer, Bernhard, and Helmut Reisen, 1994, “PensionFund Investment from Aging to Emerging Markets,”Policy Brief No. 9 (Paris: OECD Development Centre).

Flandreau, Marc, and Nathan Sussman, 2002, “Old Sins:Exchange Clauses and European Foreign Lending inthe 19th Century,” in Debt Denomination and Finan-cial Instability in Emerging Market Economies, ed.by Barry Eichengreen and Ricardo Hausmann(Chicago: University of Chicago Press).

Fontenay, Patrick de, Gian Maria Milesi-Ferretti, andHuw Pill, 1997, “The Role of Foreign-Currency Debtin Public Debt Management,” in Macroeconomic Di-mensions of Public Finance, Essays in Honour ofVito Tanzi, ed. by Mario I. Blejer and Teresa Ter-Minassian (London and New York: Routledge),pp. 203–32.

Franks, Julian, and Walter Torous, 1989, “An EmpiricalInvestigation of U.S. Firms in Reorganization,” Jour-nal of Finance, Vol. 44, No. 3, pp. 747–69.

Freeman, Paul, Michael J. Keen, and Muthukumara Mani,2003, “Dealing with Increased Risk of Natural Disas-ters: Challenges and Options,” IMF Working Paper03/197 (Washington: International Monetary Fund).

Freeman, Paul, Leslie A. Martin, Joanne Linneroot-Bayer,Reinhard Mechler, Georg Pflug, Koko Warner,2003, Disaster Risk Management (Washington: Inter-American Development Bank).

Froot, Kenneth A., David S. Scharfstein, and Jeremy Stein,1989, “LDC Debt: Forgiveness, Indexation, and In-vestment Incentives,” Journal of Finance, Vol. 44,No. 5, pp. 1335–50.

Galindo, Arturo, and Leonardo Leiderman, 2003,“Living with Dollarization and the Route to Dedol-larization,” paper presented at the Inter-AmericanDevelopment Bank conference, “Financial Dedollar-ization: Policy Options,” December.

Gavin, Michael, and Roberto Perotti, 1997, “Fiscal Policyin Latin America,” in NBER MacroeconomicsAnnual, ed. by Ben Bernanke and Julio Rotenberg(Cambridge, Massachusetts: MIT Press), pp. 11–61.

Gelpern, Anna, 2004, “Building a Better Seating Chart ofSovereign Restructurings” (unpublished; Washington:Council on Foreign Relations), June.

Gilbert, Roy, and Alcira Kreimer, 1999, “Learning fromthe World Bank’s Experience of Natural Disaster Re-lated Assistance,” Urban and Local GovernmentWorking Paper No. 2 (Washington: World Bank).

Goldstein, Morris, and Philip Turner, 2004, ControllingCurrency Mismatches in Emerging Markets (Wash-ington: Institute for International Economics).

Goyal, Vidhan K., 2003, “Market Discipline of Bank Risk:Evidence from Subordinated Debt Contracts” (un-published; Hong Kong: University of Science andTechnology).

Guidotti, Pablo, and Manmohan Kumar, 1991, DomesticPublic Debt of Externally Indebted Countries, IMFOccasional Paper No. 80 (Washington: InternationalMonetary Fund).

Haldane, Andy, 1999, “Private Sector Involvement in Fi-nancial Crisis: Analytics and Public Policy Ap-proaches,” Financial Stability Review, Issue No. 7,pp. 184–202. Available via the Internet: http://www.bankofengland.co.uk/fsr/fsr07.htm.

Hart, Oliver, 1995, Firms, Contracts, and Financial Struc-ture (Oxford: Clarendon Press).

———, and John Moore, 1995, “Debt and Seniority: AnAnalysis of the Role of Hard Claims in ConstrainingManagement,” American Economic Review, Vol. 85(June), pp. 567–85.

Hausmann, Ricardo, and Ugo Panizza, 2002, “The Mys-tery of Original Sin: The Case of the Missing Apple,”in Debt Denomination and Financial Instability inEmerging Market Economies, ed. by Barry Eichen-green and Ricardo Hausmann (Chicago: University ofChicago Press).

Herrera, Luis, and Rodrigo Valdés, 2003, “Dedollarization,Indexation and Nominalization: The Chilean Experi-ence,” paper presented at the Inter-American Develop-ment Bank conference, “Financial Dedollarization:Policy Options,” December.

Hoffmaister, Alexander W., and Jorge E. Roldós, 1997,“Are Business Cycles Different in Asia and LatinAmerica?” IMF Working Paper 97/9 (Washington: In-ternational Monetary Fund).

Hull, John C., 2002, Fundamentals of Futures and OptionsMarkets (Upper Saddle River, New Jersey: PrenticeHall).

International Monetary Fund, 1995, International CapitalMarkets Report: Developments, Prospects, and PolicyIssues (Washington: International Monetary Fund),August.

———, 2000a, “Debt- and Reserve-Related Indicators ofExternal Vulnerability.” Available via the Internet:http://www.imf.org/external/np/pdr/debtres.

58

References

———, 2000b, “The Impact of Higher Oil Prices on the Global Economy.” Available via the Internet:http://www.imf.org/external/pubs/cat/longres.cfm?sk&sk =3865.0.

———, 2002a, “Sovereign Debt Restructurings and theDomestic Economy: Experience in Four RecentCases.” Available via the Internet: http://www.imf.org/external/NP/pdr/sdrm/2002/022102.pdf.

———, 2002b, “The Design of the Sovereign Debt Re-structuring Mechanism—Further Considerations.”Available via the Internet: http://www.imf.org/external/np/pdr/sdrm/2002/112702.htm.

———, 2003a, “Sustainability Assessments—Review ofApplication and Methodological Refinements.” Avail-able via the internet: www:imf.org/external/np/pdr/sustain/2003/061003.htm.

———, 2003b, “Crisis Resolution in the Context of Sov-ereign Debt Restructuring—A Summary of Consider-ations.” Available via the Internet: www.imf.org/external/np/pdr/sdrm/2003/012803.htm.

———, 2003c, External Debt Statistics, Guide for Com-pilers and Users (Washington: International Mone-tary Fund). Also available via the Internet:http://www.imf.org/external/pubs/ft/eds/Eng/Guide/index.htm.

———, 2003d, “Assessing Public Sector Borrowing Col-lateralized on Future Flow Receivables,” June. Avail-able via the Internet: http://www.imf.org/external/np/fad/2003/061103.htm.

———, 2003e, Global Financial Stability Report, WorldEconomic and Financial Surveys (Washington: Inter-national Monetary Fund).

———, 2003f, “Lessons from the Crisis in Argentina.”Available via the Internet: http//www.imf.org/exter-nal/np/pdr/lessons/100803.htm.

———, 2003g, “Collective Action Clauses: Recent Devel-opments and Issues.” Available via the Internet:http://www.imf.org/external/np/psi/2003/032503.pdf.

———, and World Bank, 2001, Guidelines on Public DebtManagement (Washington: International MonetaryFund and World Bank).

———, 2003, Guidelines for Public Debt Management:Accompanying Document and Selected Case Studies(Washington: International Monetary Fund and WorldBank).

Ize, Alain, and Eric Parrado, 2002, “Dollarization, Mone-tary Policy, and the Pass-Through,” IMF WorkingPaper 02/188 (Washington: International MonetaryFund).

Jeanne, Olivier, 2000, “Foreign-Currency Debt and theGlobal Financial Architecture,” European EconomicReview, Papers and Proceedings, Vol. 44, pp. 719–27.

———, 2003, “Why Do Emerging Market EconomiesBorrow in Foreign Currency?” IMF Working Paper03/177 (Washington: International Monetary Fund).

———, 2004, “Debt Maturity and the International Finan-cial Architecture,” IMF Working Paper 04/137 (Wash-ington: International Monetary Fund).

———, and Jeromin Zettelmeyer, 2002, “Original Sin:Balance Sheet Crises and the Roles of InternationalLending,” IMF Working Paper 02/234 (Washington:International Monetary Fund).

Kaeser, Daniel, 1990, “Pour un système équitable dedésendettement,” speech given at a Swissaid Forumon Solutions to the Debt Crisis, October, availablefrom the author; shorter versions of the speech werepublished in Schweizerische Handelszeitung, April 5,1990, and Domaine Public, November 1, 1990.

Kaufmann, Daniel, Aart Kraay, and M. Mastruzzi, 2003,“Governance Matters III: Governance Indicators for1996–2002,” World Bank Policy Research WorkingPaper 2196 (Washington: World Bank).

Kletzer, Kenneth M, 1984, “Asymmetries of Informationand LDC Borrowing with Sovereign Risk,” EconomicJournal, Vol. 94, No. 374, pp. 287–307.

Kopcke, Richard W., and Ralph C. Kimball, 1999, “Infla-tion-Indexed Bonds: The Dog That Didn’t Bark,”Federal Reserve Bank of Boston, New England Eco-nomic Review, pp. 3–24.

Kraay, Aart, and Vikram Nehru, 2003, “When Is Debt Sus-tainable?” (unpublished; Washington: World Bank).

Kroszner, Randall, 1999, “Is It Better to Forgive Than toReceive? Evidence from the Abrogation of GoldIndex Clauses in Long-Term Debt During the GreatDepression” (unpublished; Chicago: University ofChicago). Available via the Internet: http://gsbwww.uchicago.edu/fac/randall.kroszner/research.

Krugman, Paul, 1988, “Financing vs. Forgiving a DebtOverhang,” Journal of Development Economics, Vol.29, pp. 253–68.

———, 1999, “Balance Sheets, The Transfer Problem,and Financial Crises” in International Finance andFinancial Crises, Essays in Honor of Robert P. Flood,ed. by Peter Isard and Andrew Rose (Dordrecht,Netherlands: Kluwer).

Lasfer, M. Ameziane, 1999, “Debt Structure, AgencyCosts and Firm’s Size: An Empirical Investigation”(unpublished; London: City University BusinessSchool).

Lessard, Donald R., and John Williamson, 1985, “FinancialIntermediation Beyond the Debt Crisis” (unpublished;Washington: Institute for International Economics).

Levy-Yeyati, Eduardo, 2003, “Una propuesta para des-dolarizar,” La Nación, October 19.

Lipworth, Gabrielle, and Jens Nystedt, 2001, “Crisis Res-olution and Private Sector Adaptation,” Staff Papers,International Monetary Fund, Vol. 47 (Special Issue),pp. 188–214.

Manasse, Paolo, Nouriel Roubini, and Axel Schim-melpfennig, 2003, “Predicting Sovereign DebtCrises,” IMF Working Paper 03/221 (Washington: In-ternational Monetary Fund).

Mauro, Paolo, Nathan Sussman, and Yishay Yafeh, 2002,“Emerging Market Spreads: Then Versus Now,”Quarterly Journal of Economics, Vol. 117, No. 2,pp. 695–733.

Mauro, Paolo, and Yishay Yafeh, 2003, “The Corporationof Foreign Bondholders,” IMF Working Paper 03/107(Washington: International Monetary Fund).

Missale, Alessandro, 1999, Public Debt Management (Ox-ford: Oxford University Press).

North, Douglass, and Barry Weingast, 1989, “Constitu-tions and Commitment: The Evolution of InstitutionsGoverning Public Choice in Seventeenth-Century

59

REFERENCES

England,” Journal of Economic History, Vol. 49,No. 4, pp. 803–32.

Obstfeld, Maurice, and Giovanni Peri, 1998, “RegionalNonadjustment and Fiscal Policy,” in EMU: Prospectsand Challenges for the Euro (Special Issue of Eco-nomic Policy), ed. by David Begg, Jürgen von Hagen,Charles Wyplosz, and Klaus F. Zimmermann.

Price, Robert, 1997, “The Rationale and Design of Infla-tion-Indexed Bonds,” IMF Working Paper 97/12(Washington: International Monetary Fund).

Reinhart, Carmen, Kenneth Rogoff, and Miguel Savastano,2003, “Debt Intolerance,” Brookings Papers on Eco-nomic Activity: 1, Brookings Institution, pp. 1–74.

Rodrik, Daniel, and Andrés Velasco, 1999, “Short-TermCapital Flows,” NBER Working Paper No. 7364(Cambridge, Massachusetts: National Bureau of Eco-nomic Research).

Rogoff, Kenneth, 1999, “International Institutions for Re-ducing Global Financial Instability,” Journal of Eco-nomic Perspectives, Vol. 13, pp. 21–42.

Sachs, Jeffrey, 1984, “Theoretical Issues in InternationalBorrowing,” Princeton Studies in International Finance, Vol. 54 (Princeton, New Jersey: PrincetonUniversity).

———, and Daniel Cohen, 1982, “LDC Borrowing withDefault Risk,” NBER Working Paper No. 925 (Cam-bridge, Massachusetts: National Bureau of EconomicResearch).

Shiller, Robert J., 1993, Macro Markets: Creating Institu-tions for Managing Society’s Largest Economic Risks(Oxford: Clarendon Press).

———, 2003, The New Financial Order: Risk in the 21stCentury (Oxford and Princeton, New Jersey: Prince-ton University Press).

Smith, C.W., Jr., and J. Warner, 1979, “On FinancialContracting: An Analysis of Bond Covenants,”Journal of Financial Economics, Vol. 7, No. 2,pp. 117–61.

Talvi, Ernesto, and Carlos Végh, 2002, “Tax Base Variabil-ity and Procyclical Fiscal Policy,” NBER WorkingPaper No. 7499 (Cambridge, Massachusetts: NationalBureau of Economic Research).

Tirole, Jean, 2002, Financial Crises, Liquidity, and the In-ternational Monetary System (Princeton, New Jersey:Princeton University Press).

United Nations Conference on Trade and Development(UNCTAD), 2001, COMTRADE, Geneva.

Varsavsky, Martín, and Miguel Braun, 2002, “¿Cuánto in-terés tienen que pagar los bonos argentinos?” LaNación, February 4. Available via the Internet:http://www.lanacion.com.ar/02/02/04/de_371608.asp.

Weiss, Lawrence A., 1990, “Bankruptcy Resolution:Direct Costs and Violation of Priority of Claims,Journal of Financial Economics, Vol. 27, No. 2,pp. 285–314.

Werner, Alejandro, 2003, “Undoing and Avoiding Dollar-ization,” paper presented at the Inter-American De-velopment Bank conference, “Financial Dedollariza-tion: Policy Options,” December.

Wilde, Oscar, 1906, The Canterville Ghost: An AmusingChronicle of the Tribulations of the Ghost of Canter-ville Chase When His Ancestral Halls Became theHome of the American Minister to the Court of St.James (Boston: J.W. Luce).

Zettelmeyer, Jeromin, “The Case for an Explicit SeniorityStructure in Sovereign Debt,” IMF Working Paper(forthcoming; Washington: International MonetaryFund).

60

Occasional Papers

Recent Occasional Papers of the International Monetary Fund

237. Sovereign Debt Structure for Crisis Prevention, by Eduardo Borensztein, Marcos Chamon, Olivier Jeanne,Paolo Mauro, and Jeromin Zettelmeyer. 2004.

236. Lessons from the Crisis in Argentina, by Christina Daseking, Atish R. Ghosh, Alun Thomas, and TimothyLane. 2004.

235. A New Look at Exchange Rate Volatility and Trade Flows, by Peter B. Clark, Natalia Tamirisa, andShang-Jin Wei, with Azim Sadikov and Li Zeng. 2004.

234. Adopting the Euro in Central Europe: Challenges of the Next Step in European Integration, by Susan M.Schadler, Paulo F. Drummond, Louis Kuijs, Zuzana Murgasova, and Rachel N. van Elkan. 2004.

233. Germany’s Three-Pillar Banking System: Cross-Country Perspectives in Europe, by Allan Brunner, JörgDecressin, Daniel Hardy, and Beata Kudela. 2004.

232. China’s Growth and Integration into the World Economy: Prospects and Challenges, edited by EswarPrasad. 2004.

231. Chile: Policies and Institutions Underpinning Stability and Growth, by Eliot Kalter, Steven Phillips,Marco A. Espinosa-Vega, Rodolfo Luzio, Mauricio Villafuerte, and Manmohan Singh. 2004.

230. Financial Stability in Dollarized Countries, by Anne-Marie Gulde, David Hoelscher, Alain Ize, DavidMarston, and Gianni De Nicoló. 2004.

229. Evolution and Performance of Exchange Rate Regimes, by Kenneth S. Rogoff, Aasim M. Husain, AshokaMody, Robin Brooks, and Nienke Oomes. 2004.

228. Capital Markets and Financial Intermediation in The Baltics, by Alfred Schipke, Christian Beddies, SusanM. George, and Niamh Sheridan. 2004.

227. U.S. Fiscal Policies and Priorities for Long-Run Sustainability, edited by Martin Mühleisen and Christo-pher Towe. 2004.

226. Hong Kong SAR: Meeting the Challenges of Integration with the Mainland, edited by Eswar Prasad, withcontributions from Jorge Chan-Lau, Dora Iakova, William Lee, Hong Liang, Ida Liu, Papa N’Diaye,and Tao Wang. 2004.

225. Rules-Based Fiscal Policy in France, Germany, Italy, and Spain, by Teresa Dában, Enrica Detragiache,Gabriel di Bella, Gian Maria Milesi-Ferretti, and Steven Symansky. 2003.

224. Managing Systemic Banking Crises, by a staff team led by David S. Hoelscher and Marc Quintyn. 2003.

223. Monetary Union Among Member Countries of the Gulf Cooperation Council, by a staff team led by UgoFasano. 2003.

222. Informal Funds Transfer Systems: An Analysis of the Informal Hawala System, by Mohammed El Qorchi,Samuel Munzele Maimbo, and John F. Wilson. 2003.

221. Deflation: Determinants, Risks, and Policy Options, by Manmohan S. Kumar. 2003.

220. Effects of Financial Globalization on Developing Countries: Some Empirical Evidence, by Eswar S.Prasad, Kenneth Rogoff, Shang-Jin Wei, and Ayhan Kose. 2003.

219. Economic Policy in a Highly Dollarized Economy: The Case of Cambodia, by Mario de Zamaroczy andSopanha Sa. 2003.

218. Fiscal Vulnerability and Financial Crises in Emerging Market Economies, by Richard Hemming, MichaelKell, and Axel Schimmelpfennig. 2003.

217. Managing Financial Crises: Recent Experience and Lessons for Latin America, edited by Charles Collynsand G. Russell Kincaid. 2003.

216. Is the PRGF Living Up to Expectations?—An Assessment of Program Design, by Sanjeev Gupta, MarkPlant, Benedict Clements, Thomas Dorsey, Emanuele Baldacci, Gabriela Inchauste, Shamsuddin Tareq,and Nita Thacker. 2002.

215. Improving Large Taxpayers’ Compliance: A Review of Country Experience, by Katherine Baer. 2002.

214. Advanced Country Experiences with Capital Account Liberalization, by Age Bakker and Bryan Chapple.2002.

61

OCCASIONAL PAPERS

213. The Baltic Countries: Medium-Term Fiscal Issues Related to EU and NATO Accession, by JohannesMueller, Christian Beddies, Robert Burgess, Vitali Kramarenko, and Joannes Mongardini. 2002.

212. Financial Soundness Indicators: Analytical Aspects and Country Practices, by V. Sundararajan, CharlesEnoch, Armida San José, Paul Hilbers, Russell Krueger, Marina Moretti, and Graham Slack. 2002.

211. Capital Account Liberalization and Financial Sector Stability, by a staff team led by Shogo Ishii and KarlHabermeier. 2002.

210. IMF-Supported Programs in Capital Account Crises, by Atish Ghosh, Timothy Lane, Marianne Schulze-Ghattas, Alesv Bulírv, Javier Hamann, and Alex Mourmouras. 2002.

209. Methodology for Current Account and Exchange Rate Assessments, by Peter Isard, Hamid Faruqee,G. Russell Kincaid, and Martin Fetherston. 2001.

208. Yemen in the 1990s: From Unification to Economic Reform, by Klaus Enders, Sherwyn Williams, NadaChoueiri, Yuri Sobolev, and Jan Walliser. 2001.

207. Malaysia: From Crisis to Recovery, by Kanitta Meesook, Il Houng Lee, Olin Liu, Yougesh Khatri, NataliaTamirisa, Michael Moore, and Mark H. Krysl. 2001.

206. The Dominican Republic: Stabilization, Structural Reform, and Economic Growth, by a staff team led byPhilip Young comprising Alessandro Giustiniani, Werner C. Keller, and Randa E. Sab and others. 2001.

205. Stabilization and Savings Funds for Nonrenewable Resources, by Jeffrey Davis, Rolando Ossowski,James Daniel, and Steven Barnett. 2001.

204. Monetary Union in West Africa (ECOWAS): Is It Desirable and How Could It Be Achieved? by PaulMasson and Catherine Pattillo. 2001.

203. Modern Banking and OTC Derivatives Markets: The Transformation of Global Finance and Its Implicationsfor Systemic Risk, by Garry J. Schinasi, R. Sean Craig, Burkhard Drees, and Charles Kramer. 2000.

202. Adopting Inflation Targeting: Practical Issues for Emerging Market Countries, by Andrea Schaechter,Mark R. Stone, and Mark Zelmer. 2000.

201. Developments and Challenges in the Caribbean Region, by Samuel Itam, Simon Cueva, Erik Lundback,Janet Stotsky, and Stephen Tokarick. 2000.

200. Pension Reform in the Baltics: Issues and Prospects, by Jerald Schiff, Niko Hobdari, Axel Schimmel-pfennig, and Roman Zytek. 2000.

199. Ghana: Economic Development in a Democratic Environment, by Sérgio Pereira Leite, AnthonyPellechio, Luisa Zanforlin, Girma Begashaw, Stefania Fabrizio, and Joachim Harnack. 2000.

198. Setting Up Treasuries in the Baltics, Russia, and Other Countries of the Former Soviet Union: An Assess-ment of IMF Technical Assistance, by Barry H. Potter and Jack Diamond. 2000.

197. Deposit Insurance: Actual and Good Practices, by Gillian G.H. Garcia. 2000.

196. Trade and Trade Policies in Eastern and Southern Africa, by a staff team led by Arvind Subramanian, withEnrique Gelbard, Richard Harmsen, Katrin Elborgh-Woytek, and Piroska Nagy. 2000.

195. The Eastern Caribbean Currency Union—Institutions, Performance, and Policy Issues, by Frits van Beek,José Roberto Rosales, Mayra Zermeño, Ruby Randall, and Jorge Shepherd. 2000.

194. Fiscal and Macroeconomic Impact of Privatization, by Jeffrey Davis, Rolando Ossowski, ThomasRichardson, and Steven Barnett. 2000.

193. Exchange Rate Regimes in an Increasingly Integrated World Economy, by Michael Mussa, Paul Masson,Alexander Swoboda, Esteban Jadresic, Paolo Mauro, and Andy Berg. 2000.

192. Macroprudential Indicators of Financial System Soundness, by a staff team led by Owen Evans, AlfredoM. Leone, Mahinder Gill, and Paul Hilbers. 2000.

191. Social Issues in IMF-Supported Programs, by Sanjeev Gupta, Louis Dicks-Mireaux, Ritha Khemani,Calvin McDonald, and Marijn Verhoeven. 2000.

190. Capital Controls: Country Experiences with Their Use and Liberalization, by Akira Ariyoshi, Karl Haber-meier, Bernard Laurens, Inci Ötker-Robe, Jorge Iván Canales Kriljenko, and Andrei Kirilenko. 2000.

Note: For information on the titles and availability of Occasional Papers not listed, please consult the IMF’s Publications Catalog or contact IMFPublication Services.

62

O C C A S I O N A L PA P E R



237


Washington DC

2004

237


237

Sovereign D

ebt Structure fo

r Crisis P

revention

2004

sovereign debt structure for crisis prevention - world bank internet

Documents