UnitedNationsConferenceonTradeandDevelopment
COMPENDIUMON
DEBTSUSTAINABILITY
ANDDEVELOPMENT
UnitedNations
NewYorkandGeneva,2009
CompendiumonDebtSustainabilityandDevelopment
Note
SymbolsofUnitedNationsdocumentsarecomposedof
capitalletterscombinedwithfigures.Mentionofsucha
symbol indicates a reference to a United Nations
document.
Thedesignationsemployedandthepresentationofthe
materialinthispublicationdonotimplytheexpression
ofanyopinionwhatsoeveronthepartoftheSecretariat
oftheUnitedNationsconcerningthelegalstatusofany
country, territory, city or area, or of its authorities, or
concerning the delimitation of its frontiers or
boundaries.
Material in this publication may be freely quoted;
acknowledgement, however, is requested (including
reference to the document number). It would be
appreciated if a copy of the publication containing the
quotation were sent to the Publications Assistant,
Division on Globalization and Development Strategies,
UNCTAD,PalaisdesNations,CHͲ1211Geneva10.
UNITEDNATIONSPUBLICATION
UNCTAD/GDS/DDF/2008/1
Copyright©UnitedNations,2009
Allrightsreserved
ii
CompendiumonDebtSustainabilityandDevelopment
CONTENTS
COMPENDIUMONDEBTSUSTAINABILITYANDDEVELOPMENT ..................................................... i
CONTENTS ................................................................................................................................. iii
CHAPTERI.OVERVIEW:DEBTSUSTAINABILITYINTHEORYANDPRACTICE..................................... 1
A.
B.
C.
D.
E.
F.
G.
Introduction ................................................................................................................................... 1
DefinitionandDimensionsofDebtSustainability ......................................................................... 3
CountryStudies.............................................................................................................................. 5
InstitutionalFrameworkforDebtManagement............................................................................ 7
CreditRatingAgencies ................................................................................................................... 8
GlobalRulesforInternationalFinanceandTrade ......................................................................... 9
ConclusionsandFutureTasks ...................................................................................................... 11
CHAPTERII.DEBTSUSTAINABILITYASSESSMENT:THEIMFAPPROACHANDALTERNATIVES......... 17
A.
B.
C.
D.
Introduction ................................................................................................................................. 17
WhatisDebtSustainability? ........................................................................................................ 18
ApproachestoAssessingDebtSustainability:ACriticalReview ................................................. 23
ReviewandConclusions............................................................................................................... 36
CHAPTERIII.THEMECHANICSOFDEBTSUSTAINABILITYANALYSIS.............................................. 45
A.
B.
C.
D.
E.
Introduction ................................................................................................................................. 45
DebtIndicatorsandEarlyWarningofCrisis................................................................................. 46
ThePresentValueofFutureIncome ........................................................................................... 48
TheFinancingGap........................................................................................................................ 50
DevelopmentPolicyͲBasedApproachtoDebtSustainability ...................................................... 58
CHAPTERIV.ANANALYTICALFRAMEWORKFORDEBTSUSTAINABILITYANDDEVELOPMENT ...... 63
A.
B.
C.
D.
E.
F.
Introduction ................................................................................................................................. 63
Debtandthe“FinanceGap”Model............................................................................................. 66
SustainableDebtLevels ............................................................................................................... 71
FiscalConsequencesofExternalDebt ......................................................................................... 77
DebtVulnerabilityandExternalShocks ....................................................................................... 78
Conclusions:PrinciplesforDebtManagementinDevelopmentStrategies ................................ 81
CHAPTERV.THEDEBTEXPERIENCESOFUGANDA,KENYAANDBOLIVIA...................................... 89
A.
B.
C.
D.
E.
F.
Introduction ................................................................................................................................. 89
Uganda ......................................................................................................................................... 90
Kenya’sDebtExperience.............................................................................................................. 96
Bolivia’sDebtExperience........................................................................................................... 101
DebtExperiencesCompared...................................................................................................... 108
ConcludingRemarks................................................................................................................... 110
CHAPTERVI.CASESTUDIES:ARGENTINAANDTHEREPUBLICOFKOREA.................................... 115
A.
B.
C.
D.
Introduction ............................................................................................................................... 115
LessonsfromtheArgentineCrisisandDefault.......................................................................... 116
ExternalDebtManagementoftheRepublicofKoreaduringtheCrisesof1979Ͳ1980
and1997Ͳ1998 ........................................................................................................................... 123
ConcludingRemarks................................................................................................................... 130
iii
CompendiumonDebtSustainabilityandDevelopment
CHAPTERVII.APPROPRIATEINSTITUTIONALSETTINGSFORPUBLICDEBTMANAGEMENT.......... 143
A.
B.
C.
Introduction ............................................................................................................................... 143
TheContextofPublicDebtManagement.................................................................................. 144
TheRoleandOrganizationofaDMO ........................................................................................ 148
CHAPTERVIII.CREDITRATINGAGENCIESANDTHEIRPOTENTIALIMPACTON
DEVELOPINGCOUNTRIES........................................................................................................ 165
A.
B.
C.
D.
E.
F.
Introduction ............................................................................................................................... 165
CreditRatingAgenciesintheInternationalFinancialSystem ................................................... 166
CRAs’ProceduresandMethods................................................................................................. 168
ImpactofRatings ....................................................................................................................... 172
PublicPolicyConcerns ............................................................................................................... 176
Conclusions ................................................................................................................................ 180
CHAPTERIX.PURSUINGSUSTAINABLEDEVELOPMENTSTRATEGIES:
THECASEOFTHEBALANCEOFPAYMENTRULESINWTO.......................................................... 191
A.
B.
C.
D.
E.
F.
Introduction ............................................................................................................................... 191
ExchangeControlsandConvertibility ........................................................................................ 194
TradeRestrictionsforBalanceͲofͲPaymentsPurposes............................................................. 199
TradeFinancingandEquity........................................................................................................ 202
TheGeneralAgreementonTradeandServices(GATS),BalanceofPayments,andDebt
Sustainability .............................................................................................................................. 204
Conclusions ................................................................................................................................ 209
CHAPTERX.RISKASSOCIATEDWITHTRENDSINTHETREATMENTOFSOVEREIGNDEBT
INBILATERALTRADEANDINVESTMENTTREATIES .................................................................... 211
A.
B.
C.
D.
E.
Introduction ............................................................................................................................... 211
SovereignDebtinBilateralTradeandInvestmentTreaties ...................................................... 211
ImplicationsforSovereignDebtProblemsofIncludingNationalTreatmentandMFN
TreatmentinFTAs ...................................................................................................................... 213
InvestorͲStateLawsuitsandSovereignDebt ............................................................................. 216
ConcludingRemarks................................................................................................................... 217
iv
CompendiumonDebtSustainabilityandDevelopment
CHAPTERI
OVERVIEW:DEBTSUSTAINABILITY
INTHEORYANDPRACTICE
AndrewCornford
(FinancialMarketsCenter)
A. Introduction
Debt sustainability, which concerns the feasibility for a country of meeting its debtͲrelated financial
obligations during a period beginning with the present, has proved an elusive concept. This is not
surprisinginviewofitsdependenceonanintrinsicallyuncertainfuture.Interestintheconditionsfordebt
sustainabilitybuildsonanearliertraditionofworkondebtmanagementwherethefocuswasoncountry
riskandonthelikelihoodandconsequencesofdebtdefault.
Theshiftinfocusfromcountryrisktodebtsustainabilityreflectsthesearchofnationalandinternational
policymakers for rules for external debt management which have a good theoretical justification and a
reasonable track record of application. The problems of debt management have traditionally also been
closelyrelatedtoconsiderationofseveralotherissuesinvolvingexternaldebt.Therecentshiftinfocusis
much less evident in the way in which these issues are approached. This is true, for example, of
considerationofexternaldebtpolicyasanimportantelementofglobalregimesforinternationalfinance
and trade. In spite of the lack of a direct link to debt sustainability some features of these regimes are
takenupinthepapersinthiscollectionowingtotheirimportancetotheframeworkofinternationalrules
withinwhichexternaldebtmanagementiscarriedout.
Traditionalcountryriskanalysishadtwodimensions,politicalriskandtransferrisk.1Thefirstreferstothe
determinantsofthepoliticalwillandthesecondtotheeconomiccapacitytomeetobligationsondebts
incurredthroughsovereignborrowingaswellasthroughthecrossͲborderliabilitiesofprivateinstitutions
operatingwithinthecountry’sfrontiers.Thetwodimensionsarenotcompletelydistinctsinceeconomic
capacity depends partly on a country’s willingness to take the policy measures required to meet debt
1
GoodoverviewsofthetraditionalanalysisofcountryriskareFriedman(1983)andKrayenbuehl(1988)
1
Overview:DebtSustainabilityinTheoryandPractice
obligations, and this willingness in turn reflects a balancing of political costs and benefits. Nevertheless
analysisunderthetwoheadingscoverslargelydifferentsubjects.
Thesubjects taken upunderpoliticalriskcomprise acountry’sconstitutional andpolitical environment,
thequalityofGovernmentandthelevelofcorruption,inequalitiesofincomesandwealth,literacyrates,
demographic structures, and ethnic and religious differences. Transfer risk concerns subjects not
necessarily less complex but mostly more easily quantifiable. Some of these factors can be classified as
having a substantially domestic origin such as fiscal and monetary policy, the exchangeͲrate regime,
access to natural resources, the use of funds acquired through foreign borrowing, the tax system, and
exchange controls for current and capital transactions. Other factors are external. These include trade
barriers to a country’s exports, commodity prices, interest rates and other conditions in international
financialmarkets,shippingcosts,theavailabilityofconcessionalfinancing,andnaturaldisasters.
Transfer risk varies with the availability and terms of external financing and with changes in the other
determinantsofaccesstoforeignexchange.Intheassessmentoftransferriskamajorroleisplayedby
quantitativeindicatorssuchasthefollowing:
x
The debt service ratio: interest and principal with a maturity of at least one year divided by
receiptsofforeignrevenueduringaperiodT;
x
Thedebt/GNPratio:externalpublicandprivatedebt(normallyexcludingthatwithamaturityof
lessthanoneyear)dividedbyGNP;
x
Theinterestserviceratio:interestpayments(normallyexcludingthoseondebtwithamaturityof
lessthanoneyear)dividedbyexportsofgoodsandservicesduringaperiodT,which,ifsubtracted
from the debt service ratio, indicates the percentage of foreign exchange receipts required to
serviceprincipal;
x
Thereserves/importsratio:officiallypublishedreservesdividedbyimportsduringaperiodT;
x
Theliquiditygapratio:anumeratorconsistingofdebtwithamaturityofuptooneyearminusthe
balance on current account divided by the sum of export receipts and unilateral transfers. The
ratioindicatestheliquiditygapwhichneedstobecoveredbyshortͲtermborrowing;
x
CurrentͲaccountbalance/GNP;
x
The compressibility ratio: nonͲessential imports as a percentage of total imports, an indicator
whichinprincipaldependsonclassifyingpartof importsas basic needs (energy,food,essential
inputs and investment goods) on the basis of knowledge of the economy’s requirements but
whichinpracticeisoftenbasedonaruleofthumbsuchas25percentofimports.
Until the late 1970s analysts tended to focus primarily on mediumͲterm indicators of transfer risk.
However, owing to countries’ greater use of international financial markets to meet their external
financingneedsandexperienceofthedebtcrisisofthe1980s,theyincreasinglydevotedgreaterattention
toindicatorsbearingonliquidity(forexample,thereserves/importsratio,theliquiditygapratio,andthe
compressibilityratio).
However, actual experience of countries’ debt problems has indicated limits to the usefulness of the
commonly used indicators. These limits are partly due to lack of information concerning aspects of
countries’positionswithanimportantbearingontheircapacitytomeettheirexternalobligations.During
theAsiancrisisof1997Ͳ1998,forexample,statisticsforofficialreservesdidnotinclude theauthorities’
commitments in the forward exchange markets or to private sector financial institutions which in both
cases reduced the foreign exchange available to meet external debt service. Moreover, the traditional
indicatorsofcountryriskaredesignedfortheassessmentofriskandaremuchlesswellsuitedtobetools
fordebtmanagement.
2
CompendiumonDebtSustainabilityandDevelopment
To meet the needs of the latter attempts began to be made to give a more precise meaning to the
conceptofdebtsustainability.Theseweredesignedtoprovideaconceptcapableofcontributingtopolicy
undereachofthethreemajorheadingsofdebtmanagement,namelyavoidanceoffinancialcrises,debt
management once a debt crisis appears imminent or is under way, and postͲdefault policy and
rescheduling.Atthesametime,partlyinresponsetothegrowingcomplexityofmanycountries’external
commitmentsandtothegreateravailabilityofinstrumentsformanagingthem,attentionhasincreasingly
beendevotedtocountries’overallexternalbalancesheets,andtotheproblemsandopportunitieswhich
theypresentforpolicymakers.
Work on sustainability accompanied parallel attempts to investigate the theoretical underpinnings of
conditionsfortheenforcementofcrossͲborderdebtcontracts,andhaslikewisebeenmarkedbyinterest
indevelopingamorerigorousconceptualframeworktoreplaceitsmoreadhocpredecessor.2However,
thepapersinthiscollectionpointtotheprobablyinsolubledifficultiesconfrontingtheattempttodevelop
aconceptofdebtsustainabilitycapableofservingasaphilosopher’sstoneforpolicymakers.
B. DefinitionandDimensionsofDebtSustainability
InChapterII,Wyploszprovidesanextensivereviewofthekeyconceptsinvolvedindifferentdefinitionsof
debtsustainability.Theseareasfollows:
x
Thresholdlevelofdebt/GDPratio;
x
Solvency, i.e. the condition that future surpluses on current account are sufficient to cover
interestobligationsandrepaymentsofprincipal;
x
Debt serviceability, i.e. solvency plus the additional condition of no illiquidity, which denotes
inabilitytoservicedebtsatparticularmomentsintime;
x
Solvency plus avoidance of the need for a major correction in the form of large cuts in public
expenditureorlargeincreasesintaxationrequiredfordebtservice;
x
Networth,i.e.theconditionthatthepresentvalueofcurrentͲaccountsurpluseslesscurrentdebt
isnotdecreasingovertime;
x Debtstationarity,i.e.theconditionthatthedebt/GDPratiodoesnotincreasewithoutbounds.
Wyploszpointsoutthatowingtothedependenceofeachoftheseconceptsonaninherentlyuncertain
future they cannot be used to construct universally applicable rules for debt sustainability, an attempt
which he characterizes as “mission impossible”. Thus rules using the concepts as a base for policy
prescriptions will necessarily be arbitrary and imprecise. Wyplosz elaborates the implications of this
impossibilitythroughanexaminationofproceduresforDebtSustainabilityAssessment(DSA)designedby
the IMF and the World Bank’s International Development Association (IDA) to formalize the notion of
prudentdebtstrategiesreceptiveoacountry’sdevelopmentneeds.
The startingͲpoint for the IMF’s DSA is a baseline fiveͲyear forecast combined with stress testing for
adverse shocks. To allow for the dependence of the probability of debt distress on countryͲspecific
economic and political conditions this technical exercise is combined with a Country Policy and
InstitutionalAssessment(CPIA)developedbytheWorldBank.TheCPIAgeneratesanindexofgovernance
quality based on 20 component indicators, and countries are classified into three groups according to
theirCPIAindex,thosewithindexesofhigherqualitybeingpermittedhigherdebt/GDPthresholds.
2
For a concise account of the development of the new conceptual framework for the analysis of crossͲborder debt see
SturzeneggerandZettelmeyer(2006:chapter2).
3
Overview:DebtSustainabilityinTheoryandPractice
AsWyplosznotes,thisprocedureincludesanumberofarbitrarychoicesregardingscenariosanddoesnot
allowformutuallyreinforcingeffectsduetocorrelationsamongshocksorforalternativepolicyresponses
to shocks. Moreover, it is not designed to take account of the potential of borrowing for actually
accelerating economic growth, except indirectly to the extent that a good CPIA index is likely to be
associatedwithanincreaseinthispotential.Thus,perhapsunsurprisingly,theIMF’sDSAhasbeenatthe
centreofdiscussionsondebtpolicybetweentheFundandnationalGovernments.
In view of the intractable nature of DSA, Wyplosz proposes limiting the exercise to less ambitious
objectives.Forthispurposeheprefersafocusontheevolutionofdebtlevelsandontherequirementsof
avoiding debt distress (concerns which were at the centre of the traditional approach to country risk
describedinsectionI,thoughthisisnotmentionedbyWyplosz).Inthecaseofcountrieswithaccessto
internationalfinancialmarketsimportantindicatorsforDSAaretherisk premiums in thetermsoftheir
borrowing. Moreover Wyplosz stresses that any procedure for DSA be open, and that it should include
experts other than those of the multilateral financial institutions themselves. DSA should also
accommodatethefactthatdebtaccumulation canbealegitimatepartofdevelopment policy.Wyplosz
acknowledges that his proposal can only lead to avoidance of debt distress under plausible, normal
conditions.Thepossibilityofdebtdistressinresponsetoexceptionaleventsissimplytobeacceptedasa
factoflife,andonewhoseconsequencesaretobedealtwithasandwhenitoccurs.
WyploszisnotsuggestingthattheelementsoftheIMFapproachtoDSAhavenovalue.Buttheyshould
bepartof theframeworkforpolicydiscussionand notamechanicalguide topolicyconclusions,asthe
IMFitselfincreasinglyrecognizes.
TheanalyticsoftheIMFapproachandofotherschematicapproachestodebtsustainabilityaremorefully
developed in Chapter III and IV by TranͲNguyen and Tola (henceforth TranͲNguyen) and Fitzgerald
(Fitzgerald). TranͲNguyen’s results include “templates” for debt sustainability based on alternative
national accounting identities as points of departure as well as conclusions concerning the longͲrun
stability – and thus feasibility – of time paths for debt. The paper also develops simple frameworks for
analyzing the relation between debt and growth. In a similar spirit Fitzgerald also explores the use of
nationalaccountingidentitiestoderivesimple“golden”rulesfordebtsustainabilityaswellasconstraints
onfiscalpolicywhichtakeaccountofaccesstoexternalfinancing.
Fitzgerald provides a critical review of “financing gap” models which were long widely used as an
analytical framework for discussion of debt sustainability. These models place economic growth at the
centre of the exercise. The objective of the planning authority is to maximize GDP growth subject to
constraints imposed by domestic savings, import capacity, and the fiscal ceiling determined by tax
revenueandaccesstosovereignborrowing.
Theshortcomingsofthesemodelsaretheirdependenceonstableandexogenouslygivenrelationships.
An alternative approach explored by Fitzgerald, which draws on assumptions now common in
macroeconomics,involvestakinginvestmenttobedeterminedbyintertemporalmaximizationsubjectto
relationshipsbetweenGDP,thecapitalstockdividedbetweenthatwhichisdomesticallyandthatwhichis
externallyfinanced,depreciation,thecostofnewinvestment,nationalincomedefinedasthedifference
betweenoutputandinterestcosts,andexternalconstraintsaccordingtowhichimportsaredetermined
bynationalincomeandexportsbytheproductivityoftheexportsector.Thisapproachcangeneratean
expression for the optimal debt level as a function of the allocation of financing to different major
categoriesofinvestment.
TranͲNguyen reviews recent literature on earlyͲwarning indicators of currency and debt crises. This
literatureisanaturaldevelopmentoftheearlierapproachtocountryriskdiscussedabove.However,toa
greaterextentthanearlierwork,themorerecentliteraturemakesuseofeconometricanalysis.Moreover
4
CompendiumonDebtSustainabilityandDevelopment
italsoincludesnewindicatorssuggestedbyrecentexperienceoffinancialcrisessuchasindicatorsofthe
fragilityofthefinancialsector.
AmajorconclusionofTranͲNguyenisthattheapproachessurveyedsufferfromtheshortcomingthatthe
concept of debt sustainability is not integrated into a framework which also includes a country’s
developmentstrategyandtheimpliedgrowthtrajectory.AlthoughTranͲNguyen’ssuggestionsastosuch
integration are limited to simple debt and growth analytics, the subject would be natural candidate for
inclusion in the more open, less ruleͲbound procedures to analysis of debt sustainability proposed by
Wyplosz.
C. CountryStudies
Ofthefivecasestudiesincludedinthiscollectionofpapersthree,Uganda,KenyaandBolivia,wereoflowͲ
incomecountrieswhoseexternaldebtwaslargelytheresultofpublicbilateralandmultilateralfinancing,
whiletheremainingtwo,ArgentinaandtheRepublicofKorea,werecountrieswhosedebtcrisesreflected
abreakdownintheiraccesstointernationalfinancialmarkets.
1. LowͲIncomeCountries
The experiences of Uganda, Kenya and Bolivia share key common features in the form of failure to
generatesustainablegrowthandpovertyreductionandcontinuingvulnerabilitytotheimpactofhigher
interest rates and of lower prices on their commodity exports. All three countries undertook reform
programs consisting of tighter macroeconomic policy and price liberalization. But the programs did not
addressmajorweaknesses.Forexample,theydidnotincludeabroadeningofthetaxbase,andthetariff
reductions adopted actually harmed progress towards this objective. The Governments of Uganda and
Kenya are still heavily dependent on foreign aid for the financing of their expenditures. Moreover
substantialproportionsofeconomicactivityandexportsinallthreecountriesremainconcentratedina
limitednumberofunprocessedprimarycommodities.
Therewerealsoimportantdifferencesbetweenthethreecountries’experiences.
UgandaandBolivia,which(unlikeKenya)arebothHIPCcountries,illustratebothgeneralweaknessesof
thisinitiativeandflawsmorespecificallyapplicabletothesituationsofthetwocountries.Thefirstsetof
weaknesses included inadequate analytical bases, which reflected dependence on unrealistic country
scenarios and failure to take proper account of vulnerability to exogenous shocks. The second included
toonarrowadefinitionofdebtsustainability,failuretoallowforthewayinwhichpostͲHIPCborrowing
couldspeedilyreversegainsinacountry’sdebtposition,andtheinappropriatenessofloansasopposedto
grantsforthefinancingofprogramsofpovertyalleviation.
DuringtheperiodcoveredbythecasestudiesUgandaandBoliviaachievedanadequatetechnicalcapacity
fordebtmanagement.Kenyaontheotherhandstilllacksanadequatesystemforthispurpose.
2. LowͲIncomeCountries
The main focus of the studies of the Republic of Korea and Argentina are their recent currencyͲcumͲ
banking crises, for the former in 1997Ͳ1998 and for the latter in 2000Ͳ2001. For the Republic of Korea,
thereisalsoareviewofanearlierdebtcrisisin1979Ͳ1980whichtheGovernmentsucceededinridingout
withoutrecoursetothedeflationarymeasuresusuallycharacteristicofpolicyresponsesinsuchcases.
5
Overview:DebtSustainabilityinTheoryandPractice
Therecentcrisesforbothcountriesweredominatedbydevelopmentsaffectingthecapitalaccountafter
periods in which the liberalization of capital transactions led to greater integration into international
financialmarkets.Inbothcasestherewerelargefluctuationsofcapitalflows–largeinflowspriortothe
crises followed by large outflows. In both cases the IMF’s policy prescriptions were ill suited to dealing
withthecrises.
TheoriginoftheRepublicofKoreacrisiswasanunfavorableshiftinitsexportmarketsbeginningin1995.
Thisledtoinventoryaccumulationandlossesamongstthecountry’slargeindustrialgroupsthatinturn
provokedareassessmentofthesegroups’prospectsamongsttheforeigninvestorsandlendersonwhich
thegroupswereincreasinglycomingtodepend.Asaresultofbankruptciesaccompaniedbyrevelations
concerningpoorcorporategovernanceandcorruption,during1997foreigners’flightfromthecountry’s
stock market accelerated and its banks faced increasing difficulties in rolling over shortͲterm interbank
loans.TheinitialfinancingpackageagreedlateintheyearbetweentheGovernmentandtheIMF,which
includedawiderangeofconditionsincludingmacroeconomicstringencyandliberalizationoffinancialand
labormarkets,failedtostemthecrisisasinterestratesreached40percentandthecurrencycontinuedto
depreciate.
A second package was accompanied by an agreement with creditor banks to loan extension and to a
lengthening of maturities in return for government guarantees on private debt. This sufficed to turn
aroundmarketsentiment,andasharprecoveryineconomicgrowthfollowedin1999.Thisexperienceof
theRepublicofKoreahasledcommentatorstoquerytheappropriatenessofthedeflationaryfiscaland
monetary conditions of the first package as a response to what was principally a capitalͲaccount crisis
ratheronecharacterizedbymacroeconomicimbalances.
The paper on Argentina locates the source of its crisis in the exchangeͲrate regime and the impact on
externalͲdebt dynamics of interest rates required to manage the country’s capital account. This
interpretation is at variance with the view of the country’s Governments during the preͲdefault period
whereby problems were seen to be due to fiscal mismanagement, which called for a policy response
consistingofaseriesofpackagesoffiscaltightening.
AsArgentina’scrisisgotunderway,thereweremarkeddisagreementsbetweenthenewGovernmentand
theIMFastoappropriatepolicymeasures.TheIMF’srecommendationsincludedallowingtheexchange
ratetofloatfreelyandanapproachtothebankingcrisiswhichwouldhaveentailedbankliquidations.The
Government’spolicies,whichaccompaniedthebeginningofaneconomicrecoveryfromthefirsthalfof
2002,includedexchangecontrolsandrestrictionsoncapitaloutflowsaspartofapolicyofmanagingthe
exchange rate, export taxes designed to capture for the Government some of the profits due to
devaluation, and a flexible monetary policy aimed at assisting the recovery of the banking sector. The
Government also resisted pressure from foreign Governments and the IMF to improve the terms of its
offerondebtrestructuringtoitsexternalcreditors.
Owing to the lack of the required data it is not possible to conduct a controlled experiment to test the
validityofthenowincreasinglywidelyheldbeliefastoinappropriatenessofstandardfeaturesofpolicy
programsassociatedwithIMFpolicypackages.However,theRepublicofKorea’sdebtcrisisof1979Ͳ1980
doesprovideacasestudyofthesuccessfulapplicationofadifferentpolicyapproach.
Thecrisisbeganin1979afteryearsofrapidgrowthpoweredbyaninvestmentboom.Majorfeaturesof
thecrisiswereasharpincreaseinthecurrentͲaccountdeficit,asevererecession,andariseininflationof
consumerpricestoanannualrateofalmost30percent.InsuchcircumstancesthestandardIMFpolicy
prescription would probably have involved macroeconomic stabilization through fiscal and monetary
tightening and allowing the exchange rate to float. The view of the Republic of Korea Government,
however,wasthatacceleratinginflationwasthesourceofdeterioratingincomedistribution,laborunrest,
anddecliningexportcompetitiveness.ThepoliciesadoptedincludedaoneͲoffdevaluationfollowedbya
6
CompendiumonDebtSustainabilityandDevelopment
managed float under which the Won was tied to a basket of major international currencies,
macroeconomic policies giving priority to stopping the economic downturn, and continued recourse to
externalfinancingofadecliningcurrentͲaccountdeficitdespiteanalreadyhighlevelofexternaldebt.
The economic recovery which followed is likely to have reflected the effects of an improvement in the
external economic environment as well the policies pursued. Commentators also attribute a significant
role to capital controls which prevented capital flight. Conditions associated with different countries’
currencyͲcumͲdebtcrisesareofcourseneverthesame.Nonetheless,theRepublicofKoreaexperienceof
the early 1980s deserves a place in the template of the menu of policy measures for debt crisis
management.
D. InstitutionalFrameworkforDebtManagement
Whatever the approach adopted by a developing country to debt sustainability, properly developed
institutionsfordebtmanagementarerequired.ThetasksoftheseinstitutionsassetoutinChapterVIIby
JaimeDelgadilloCortez(Delgadillo)includethefollowing:
x
x
x
x
x
x
x
x
Theproductionofreliabledebtdata;
Developmentofthedomesticfinancialmarket;
Ensuringadequatefinancingfordevelopmentalandsocialneeds;
EnsuringcompliancewithdebtͲserviceobligations;
Controllingcontingentliabilities;
Meetingtherequirementsofnegotiationswithcreditors;
Performingcost/riskanalysis;
Designingstrategiesfordebtsustainability.
For this purpose design of the institutional framework for debt management has to focus on the
following:
x
x
x
x
x
Governance;
Clarityoftherolesofthedifferentinstitutionsdealingwithdebtmanagement;
Specificationoftheobjectives;
Coordinationofpublicdebtmanagementwithotherpublicpolicies;
The organizational structure of the principal body responsible for debt management, the Debt
ManagementOffice(DMO);
Transparencyandaccountability.
x
Delgadillo discusses and exemplifies different options under these two headings for this institutional
framework.
ThenetworkofrelationsdescribedbyDelgadilloofwhichtheDMOisthecentreincludetheministryof
finance, the central bank, the national/planning or development office, creditors, international
organizations (which may themselves be among the country’s creditors), the public and private entities
whichareasourceofguaranteesandinsurancefortradefinance,etc.,andmajorparticipantsindomestic
financialmarkets.
Theinstitutionalframeworkfordebtmanagementcanbeexpectedtoevolveinresponsetothechanging
profileofacountry’sexternalliabilitiesanditsleveloffinancialdevelopmentaswellastotheincreasingly
comprehensive approach to management of a country’s external assets and liabilities which is now
receiving greater attention (see below). Inter alia, this approach may entail closer working relations
between the DMO as described by Delgadillo and those responsible for the regulation of financial
7
Overview:DebtSustainabilityinTheoryandPractice
institutions and thus for oversight of the currency and maturity risks associated with these institutions’
balancesheets.
E. CreditRatingAgencies
The ratings industry has its origins in firms which were established in midͲnineteenthͲcentury United
States to provide merchants with information on the creditworthiness of their customers. Ratings
originallyreferredtothecapacityofanobligortomeetpaymentsdueonaparticularfinancialobligation
after taking into consideration the creditworthiness of guarantors, insurers, and other forms of credit
enhancement.Butratingsmaynowrefertoissuers,includingcountries,aswellasissues.
Assessment of developing countries’ creditworthiness long relied on financial institutions’ own systems
for this purpose, guidance from their regulators, services providing information on country risks, and
rankingsofcountrycreditriskprovidedbypublicationssuchasinstitutionalInvestorandEuromoney.The
growthintheimportanceofcreditratingagenciesinrecentyearsreflectstherequirementsofthegrowth
of international capital markets which has led to increasingly widespread need for creditworthiness
assessments: borrowers are seeking ready recognition from investors; investors require an accessible
vehicle for assessing the quality of securities; and banks find ratings a useful marketing tool for selling
papertocustomers(Fight,2004:46).
SincethemidͲ1990stheperformanceoftheagencieshasbeencriticizedonseveralgrounds,asdiscussed
inChapterVIIIbyElkhoury:theirslownesstoreacttochangesincreditworthinessandthentheirtendency
onoccasiontooverͲreact;theiruseofuntransparentratingmethods;theirprivilegedregulatoryposition;
theirlackofaccountability;andthevulnerabilityoftheiroperationstoconflictsofinterest.
x
Criticsviewedtheagencies’responsetotheAsianfinancialcrisisof1997Ͳ1998ascharacteristicof
theirtendencytoslownesstoreactfollowedbyoverͲreaction.
x
Althoughtheagenciesmakeknownthefactorstakenintoaccountasinputstotheirratings,their
assignmentofweightstothesefactorsisopaque.
x
Theagencies’privilegedregulatorypositionisduetoinstitutionalinvestors’needforaratingof
investmentgradebyanofficiallyrecognizedagencyforthesecuritiesinwhichtheyarepermitted
to invest as well as to other regulatory exemptions accorded to such securities. In the United
States such recognition is reserved for Nationally Recognized Statistical Rating Organizations
(NRSROs), a designation conferred on only a limited number of agencies including the major
three,Moody’s,Standard&Poor’sandFitch.
x
Theagenciesarenotaccountablefortheirmistakesortheirabuseofpower.
x
Conflicts of interest may arise owing to the agencies’ involvement in the structuring of
instrumentstheyrate,theirprovisionofconsultancyservicestoissuers,thepotentialforpressure
to purchase agencies’ consultancy services in return for an improved rating, and the use of
aggressivesalestacticstoinduceanissuerto“solicit”andthuspayforaratingwhichithadnot
initiallyrequested(an“unsolicited”rating).
Elkhoury reviews some recent official initiatives to deal with these criticisms. These include the Credit
RatingAgencyReformActpassedbytheUnitedStatesCongressinSeptember2006,whichtightensthe
procedural requirements for NRSRO registration and certification, and strengthens the authority of the
Securities Exchange Commission over NRSROs; and a Code of Conduct issued in December 2004 by the
International Organization of Securities Commissions (IOSCO), whose objectives include ensuring the
integrityoftheratingprocessandachievinggreatertransparencyregardingratingsmethodology.
8
CompendiumonDebtSustainabilityandDevelopment
However, these steps are unlikely to satisfy the credit agencies’ growing band of critics. The agencies’
operationshavereturnedtothespotlightinconnectionwithratingsaccordedtotranchesofsecuritized
assetsduringthecreditcrisiswhichbeganinthesummerof2007.Againspecialattentionisfocusedon
the opacity of the methods underlying their ratings, their lack of accountability, and the potential for
conflictsofinterestarisingfromtheirroleasprovidersof“exanteopinions”and“structuringadvice”as
wellasratingsinthecaseofstructuredfinancing.3
The consequences may be more stringent rules for agency certification, minimum standards for the
trainingandqualificationsofagencies’analysts,andincreasedtransparencyregardingtheiroperations.If
one looks further into the future, a large increase and the number of credit rating agencies worldwide
seems quite likely. Inter alia, such an increase would be a natural consequence of the role accorded to
credit rating agencies in determining weights for credit risk in the determination of banks’ minimum
regulatorycapitalunderBasel2,whichmorethan100countriesarenowplanningtointroduce.
F. GlobalRulesforInternationalFinanceandTrade
Discussion of global rules in connection with external debt typically focuses mainly on arrangements
capable of making debt cries less likely and of facilitating their management and resolution. Subjects
include bankruptcy mechanisms for sovereign, and sometimes also private, crossͲborder debt,
improvementsintermsandfundingforIMFcrisislending,andpreͲcrisisinterventioninthemarketsfor
internationaldebt.4Buttheframeworkwithinwhichcountriesmanagetheirexternaldebtisalsoaffected
by developments elsewhere affecting rules for trade and trade finance, balanceͲofͲpayments measures
andforeigninvestment.
The importance of the latter set of rules was recognized in the Declaration on the Contribution of the
WorldTradeOrganizationtoAchievingGreaterCoherenceinGlobalEconomicPolicyMakingadoptedat
the time of the establishment of the WTO. This Declaration acknowledged the links between economic
policies as follows: “Successful cooperation in each area of economic policy contributes to progress in
other areas. Greater exchange rate stability…should contribute towards the expansion of trade,
sustainablegrowthanddevelopment,andthecorrectionofexternalimbalances.Thereisalsoaneedfor
anadequateandtimelyflowofconcessionalandnonͲconcessionalfinancialandrealinvestmentresources
todevelopingcountriesandforfurthereffortstoaddressdebtproblems,tohelpensureeconomicgrowth
anddevelopment.”Suchcoherenceinglobalpolicymakingrequiresthat“theinternationalinstitutionsin
eachoftheseareasfollowconsistentandmutuallysupportivepolicies”.
Twopapersinthiscollectiontakeupsomespecificinternationalrulesaffectingnationalpoliciesinareas
characterized by interfaces between trade, investment and external financing. Chapter IX by Howse
discusses the applicability of WTO rules to exchange restrictions and to measures directed at exports,
importsandthebalanceofpaymentsfromthepointofviewoftheircompatibilitywithnotionsoffairness
and equity. In Chapter X, Caliari examines the risks to national autonomy regarding debt policy which
couldresultfromprovisionsconcerninginvestmentinrecenttradeandinvestmenttreaties.
1. TradeandBalanceͲofPaymentsMeasuresunderWTORules
Howse takes as his startingͲpoint the commitment of United Nations Member States in the Millennium
Declaration to “an open, equitable, ruleͲbased, predictable and nonͲdiscriminatory multilateral trading
and financial system”. Whilst acknowledging that the concept of equity in international trade and in
3
Differencesbetweencreditratingagencies’rolewithrespecttostructuredfinance,ontheonehand,andbondissues,onthe
other,aredescribedinCommitteeontheGlobalFinancialSystem(2005).
4
ForasurveyofsuchproposalsseeSturzeneggerandZettelmeyer(2006:chapter12).
9
Overview:DebtSustainabilityinTheoryandPractice
financialrulesandinstitutionslacksagenerallyaccepteddefinition,Howseshowsthatinternationalrules
(including those of the WTO and the IMF Articles of Agreement) incorporate fairness, a concept closely
related to equity. Moreover one important ingredient of equity in international instruments concerning
trade,financeanddevelopmentisthenotionthatrulesshouldbeadjustedtothedevelopmentneedsof
countries’differentsituations.Anotheringredientisthatofpeople’srightto“voiceandparticipation”,i.e.
their right not to have a vision of development forced on them or decided by others. The Millennium
Declarationalsocontainsadistributionalcomponentsincetheconceptofglobalsolidarityrequiresthat
“globalchallengesmustbemanagedinawaythatdistributesthecostsandburdensfairlyinaccordance
withbasicprinciplesofequityandsocialjustice”.
Concerning exchange restrictions Howse notes the generally accepted view of the intent of GATT/WTO
provisionsforgoodstradethat,regardlessoftheeffectoftherestrictionsontradetransactions,theydo
notimposedisciplinesgoingbeyondthoseoftheIMF.However,suchrulingsarepermissibleforexchange
restrictionsnotendorsedbytheIMF,ascopewhichHowsebelieveshasbeenusedbytheGATTandthe
WTOasthebasisforexcessivelynarrowinterpretationofcountries’rightofrecoursetotrademeasures.
More generally Howse questions the apparent presumption of GATT/WTO case law that exchange
restrictionsnotendorsedbytheIMFentailviolationofGATT/WTOrules.
The GATT/WTO provisions for goods trade leave no scope for rulings on exchange controls applying to
capital as opposed to current transactions. However, the corresponding provisions for balanceͲofͲ
payments restrictions in the case of services trade under the General Agreement on Trade in Service
(GATS) could lead to challenges to capital controls on the ground that they are inconsistent with a
country’s specific commitments interpreted in combination with general GATS obligations. Howse
believesthatguidelinesshouldbedrawnupforsuchcasesbyinstitutionswithamandatetotakeaccount
ofequityinthetradeandfinancialsystems.
HowsealsodiscussestwootherrecentWTOrulingssuggestingashifttomorerestrictiveinterpretationof
provisions with a bearing on the compatibility of WTO rules with the principles of equity, voice and
participation.
The first ruling involved a case in which the United States challenged India’s continuing use of trade
restrictionsforbalanceͲofͲpaymentsreasonsinpursuitofdevelopmentpoliciesunderGATTArticleXVIII.
HeretheWTOAppellateBodyruledthatremovalbyIndiaofitsbalanceͲofͲpaymentsrestrictionswould
not require a change in its development policies since the objectives of these policies could equally be
achievedbymacroeconomicmeasures.HowsetakestheviewthatthisrulingisnotinaccordwiththeselfͲ
declaratorycharacterofGATTArticleXVIII.
InthesecondrulingtheAppellateBodydecidedagainstBrazil’suseofofficialsupportforthefinancingof
aircraft exports on the basis of arguments which included use of the benchmarks of the OECD Export
Credit Arrangement. As Howse points out, this Arrangement is an agreement reached through
negotiations involving the organization’s restricted membership which takes no account of structural
differencesbetweenthefinancialmarketsofdevelopinganddevelopedcountries.
2. DebtandBilateralTradeandInvestmentTreaties
Caliaridrawsattentiontotherisksforpolicytowardsexternaldebtwhichareinvolvedintheextensionof
the definition of investment to include debt instruments observed in some recent bilateral trade and
investmentagreements(suchastheUnitedStatesͲChileFreeTradeAgreementandtheCentralAmerica
Free Trade Agreement). These risks result from the association of investment in such treaties with the
obligationsofNationalandMostͲFavouredͲNation(MFN)Treatment.NationalTreatmentguaranteesnonͲ
discriminatory treatment of domestic and foreign firms. Under MFN Treatment each party to the
10
CompendiumonDebtSustainabilityandDevelopment
agreement binds itself to extend to the all others the same concessions as those accorded to the most
favoredparty.
Extended to sovereign debt, National and MFN Treatment could restrict a Government’s flexibility
regarding postͲcrisis measures such as those involving the distribution of losses between domestic and
foreigncreditorsandsupporttodomesticasopposedtoforeignbanksaspartoftherestructuringofthe
financial sector. They could also reduce the leverage of the debtor country during negotiations on the
restructuring of its external debt. In extreme cases they might even make it difficult to prioritize the
servicingofdomesticdebtincurredtomeettheGovernment’swages,salariesandpensionsobligations.
MoreoverapplicationofMFNTreatmenttoexternaldebtmighthavetheanomalousandalmostcertainly
unacceptable effect of according seniority in meeting debt obligations to the parties covered by the
agreementascomparedthosetopartiesnotsocovered.
The risks described cover in the first instance only countries covered by treaties whose definition of
investmentincludesdebt.However,precedentsbasedonbilateraltradeandinvestmenttreatiesarealso
often included among demands submitted by participants and as part of proposed frameworks
agreementsduringmuchbroadernegotiationsontradeandfinance.
G. ConclusionsandFutureTasks
Theprincipalfocusofthepapersinthiscompendiumistheneedforaconceptofdebtsustainabilitymore
systematic than the piecemeal indicators of country risk previously used for external debt assessment.
Problemsrelatedtodebtsustainabilityareexaminedthroughtheprismofaseriesofcountrystudies.The
papers also discuss the institutional framework at national level for debt assessment and management
andtheroleofcreditratingagenciesaswellasimportantfeaturesofglobalrulesbearingondeveloping
countries’autonomyregardingpoliciesfortheexternalsector.Theconcludingremarkswhichfolloware
limitedtoselectedfeaturesoftheconceptualframeworkforassessingdebtsustainabilityandofpolicies
designed to contribute to the achievement of such sustainability. They include suggestions as to some
possibledirectionsforfuturework.
1. MacroeconomicPolicy
Likeothercasestudiesofexternaldebtmanagement,thoseinthiscollectionhighlighttheimportanceof
appropriate macroeconomic policy to successful debt management. The contents of such policy
necessarily vary among countries owing to differences in both economic conditions and Governments’
objectives. Thus the experiences of Argentina and the Republic of Korea reviewed in the case studies
illustratethatsuccessfulmacroeconomicpoliciesinacontextofdebtcrisisdonotfollowageneralmodel
but rather consist of measures geared to countryͲspecific circumstances and based on countryͲspecific
balancingofthebenefitsandcostsofalternativeoptions.
The case studies also draw attention to the special vulnerability of lowͲincome countries to external
shocks.Thisisduetotheirlessdiversifiedstructuresofproduction,inparticulartheconcentrationoftheir
exportsinalimitednumberofprimarycommodities.Theproblemscaused bysuch concentrationare a
staple feature of the literature of development economics. The studies in this collection emphasize the
threatposedbythisvulnerabilitytotheachievementofdebtsustainability.
Onelessondrawnistheneedforthepoliciestowardsexternaldebtwhichaccommodatetheinvestment
requiredfordiversifyingacountry’sproductivebase.Thislessonconcernsnotonlythedebtmanagement
ofborrowingcountriesbutalsothetermsandconditionsoffinancingagreedwithofficialcreditors.Future
11
Overview:DebtSustainabilityinTheoryandPractice
work on policies in this area might also include different techniques available for hedging commodity
exportreceiptsandareexaminationofpricestabilizationatthenationallevelthroughmarketingboards.
Inthiscontextitmayalsobeworthrevisitingissuesclassifiedbyanearlierliteratureundertheheadingof
centralbankingprinciplesforanexporteconomy.Thisliterature,towhichamajorcontributorwasRaul
Prebisch,firstSecretaryͲGeneralofUNCTAD,onthebasisofhisexperienceasmanageroftheArgentine
Central Bank from 1935 to 1943, concerned the implications for appropriate monetary, and more
generally macroeconomic, policy of frequently observed differences between the external and internal
balanceofcommodityͲdependentcountries,ontheonehand,andofindustrialcountries,ontheother.5
Intheformergroupofcountriesamacroeconomicupswingtendedtobeassociatedwithamorepositive
balanceoftradeandexternalpayments,andthusariseinreservesofforeignexchange;andconversely
recession or depression tended to accompany a negative external balance and a contraction of foreign
reserves.Bycontrasttheexternalbalanceofindustrialcountriestendedtodeteriorateduringeconomic
upswingsandtoimproveduringdownswings.The monetarypolicyproposedforcommodityͲdependent
countries in response to their circumstances involved restriction during the upswing with the aim of
accumulatingreserves,whichwouldpermitamoreexpansionarypolicyandthefinancingofcontracyclical
measuressuchas public worksduringthedownswing. Theunderlyingideas ofthisliteraturecouldwell
have continuing relevance for the macroeconomic framework for policy towards external debt in
commodityͲdependentcountries.
2. TowardsaMoreInclusiveApproachtoDebtSustainability
Thiscollectionofpapershasachievedgreaterconceptualclarityconcerningdebtsustainabilitybutcannot
provide definitive, comprehensive guidelines for assessment and policymaking. The papers point to the
need for more flexible approaches to the subject which also take account of essential connections
betweenthemanagementofexternaldebtanddevelopmentstrategy.Conclusionswhichcanbedrawn
fromthecollectionincludethefollowing.
x
x
x
Assessment of debt sustainability will continue to require quantitative indicators as well as
analysisofqualitativefactorstraditionallyincludedintheassessmentofcountryrisk.
Amoreinclusiveviewofdebtsustainabilitywillsuggestnewindicatorsofcountryriskanddebt
sustainabilityinadditiontothetraditionalonessurveyed.
Assessment and policymaking should include discussion between the different parties – debtor
countries, creditors and international organizations Ͳ to resolve legitimate differences between
viewsastowhatconstitutesustainabledebtlevelsforacountry.
Adevelopmentalperspectiveshouldbeanintegralpartoftheapproachtodebtsustainability.Thisimplies
thatconsiderationofdebtsustainabilityshouldnotbeabstractedfromtherequirementsofdevelopment
strategy. Such an approach to debt sustainability requires the involvement not only of those with
responsibility for external financing and debt management but also of other policymakers who are
responsiblefordecisionsregardingdevelopmentstrategy.
These conclusions as to an appropriate framework for assessment of debt sustainability would be
consistent with treatment of the subject as part of a comprehensive approach to monitoring and
managementofacountry’sexternalassetsandliabilitiesoutlinedbelow.
5
TheargumentisexplainedinmoredetailinWallich(1950:chapterXV).
12
CompendiumonDebtSustainabilityandDevelopment
3. NationalBalanceSheetsandExternalDebt
Atthecentreoftheconceptualapproachestodebtsustainabilityreviewedinthiscollectionofpapersare
current and future receipts and outflows which determine the funds available for debt service. Such a
focuson“earningpower”(touseterminologycommonamongaccountantsandfinancialanalysts)could
beusefullysupplementedbyinformationcontainedinthebalancesheetsofacountry’sGovernmentand
firms,especiallytheirexternalliabilities.Aprofileofnationalexternalliabilitiesservesastheanalogueat
thisleveltoacorporation’scapitalstructurewhichprovidesnotonlyaguidetotheinstitution’sfunding
butalsoenablesittoindexandcontroldifferentfinancialrisks.6
A focus on external liabilities and assets would be a natural extension to debt sustainability of
recommendationsintheReportoftheWorkingGrouponCapitalFlowsoftheFinancialStabilityForumof
April 2000 (Financial Stability Forum, 2000). These recommendations responded to terms of reference
which included evaluation of prudential policies, regulations and risk management that might help to
reducesystemicrisksassociatedwiththebuildͲupofexternalindebtedness.
TheReportwasoneofmanyinternationalinitiativesundertakenintheaftermathofthefinancialcrisesof
the 1990s which involved mainly emergingͲmarket (i.e. middleͲincome) developing and transition
economies.Nevertheless,manyoftherecommendationsconcerningdatacollectionandanalysisandthe
managementofrisks couldapplyequallytolowͲincomedeveloping countries.Therecommendations of
theReportaredirectedatthepublicsector,thebankingsector,andthenonͲbankfinancialandcorporate
sectors. For lowͲincome countries the recommendations of greatest immediate interest are those
directed at the public sector (though the other recommendations can be expected to assume greater
importance with the development of their institutional infrastructure). This is partly due to the greater
relative importance of sovereign borrowing in such countries’ external liabilities. But it also reflects the
likelihoodoflessdevelopedaccessinlowͲincomecountriestoinformationconcerningassetsandliabilities
ofentitiesintheprivatesector.
The Report argues that detailed profiles of external balance sheets can make a major contribution to
monitoringandmanagingacountry’sexposuretodifferentfinancialrisks.Sectoraldataarenotonlypart
ofthisprofile(thoughapart,asjustexplained,whoseimportancevariesfordifferentsectorsaccordingto
a country’s level of development) but help to identify linkages capable of facilitating transfers of risk
exposurebetweendifferentsectors.
Forthepublicsectortherecommendationsaredesignedtotranscendthenarrowerfocusofpublicdebt
management still found in many countries. The aim of the profile of assets and liabilities should be to
enabletheformulationofastrategy balancing expectedcosts andriskscontainedin thepublicsector’s
external assets and liabilities. This process can benefit from the development of new vulnerability
indicators(forwhich,thoughtheReportdoesnotdiscussthis,accountingindicatorsusedaspartofthe
analysisofthefinancialstatementsoffirmscanoftenprovideusefulmodels).
Theimportanceofextendingtheprofilesofexternalassetsandliabilitiestothebankingsectorreflectsits
strategiceconomicroleandthedangerthatintheeventofafinancialcrisisitsproblemsarecapableof
inflictingeconomyͲwidedamage.Thedevelopmentofprofilesforthissectorwilloftenbenefitfromthe
factthatevenindevelopingcountriesfinancialreportingbybankstoregulatorsandshareholdersisofa
relatively high quality, though progress may still be required regarding the information necessary for
assessment of liquidity and foreignͲcurrency risk – two risks which assume special significance in crises.
Extension of the profiles of external assets and liabilities to the nonͲbank financial sector results in
coverage of institutions often more loosely supervised than banks or not supervised at all which were
nonethelessamajorsourceofvulnerabilityinsomecountriesintheAsianfinancialcrisis.
6
ForanilluminatingdiscussionoftherolecapitalstructureforbothcountriesandcorporationsseePettis(2001:chapter6).
13
Overview:DebtSustainabilityinTheoryandPractice
No more than the other techniques discussed in this collection of papers can national balance sheets
provide all the information required for the analysis of debt sustainability and the prevention and
containment of debt crises. They can, however, provide a framework for the further development of
conceptsclarifyingdebtsustainabilityaswellasforthemanagementoftherisksassociatedwithexternal
debt.
14
CompendiumonDebtSustainabilityandDevelopment
References
CommitteeontheGlobal FinancialSystem (2005).The Roleof RatingsinStructuredFinance:Issuesand
Implications.Basel,BankforInternationalSettlements,January.
FightA(2004).UnderstandingInternationalBankRisk.Chichester,JohnWiley.
FinancialStabilityForum(2000).ReportoftheWorkingGrouponCapitalFlows,April.
Friedman IS (1983). The World Debt Dilemma: Measuring Country Risk. Washington, DC, Council for
InternationalBankingStudies,andPhiladelphia,RobertMorrisAssociates.
Krayenbuehl TE (1988). Country Risk Assessment and Monitoring, 2nd edition. Cambridge, WoodheadͲ
Faulkner.
PettisM(2001).TheVolatilityMachineEmergingEconomiesandtheThreatofFinancialCollapse.Oxford,
OxfordUniversityPress.
SturzeneggerFandZettelmeyerJ(2006).DebtDefaultsandLessonsfromaDecadeofCrises.Cambridge,
MassachusettsandLondon,TheMITPress.
Wallich HC (1950). Monetary Problems of an Export Economy the Cuban Experience 1914Ͳ1947.
Cambridge,Massachusetts,HarvardUniversityPress.
15
CompendiumonDebtSustainabilityandDevelopment
CHAPTERII
DEBTSUSTAINABILITYASSESSMENT:
THEIMFAPPROACHANDALTERNATIVES
CharlesWyplosz
(GraduateInstituteofInternationalandDevelopmentStudiesandCEPR)7
A. Introduction
Debtsustainabilityisavexingissue.Itsimportanceisimmediatelyobviousbuttheconceptescapesany
easy definition. This situation is not unheard of in economics; price stability and full employment are
examples of other crucially important policy objectives that cannot be simply defined. Yet, while price
stability or full employment can both be measured with a reasonable degree of precision, debt
sustainabilitycannotevenbemeasureddirectly.
Every country, therefore, must grapple as best it can with the issue of debt sustainability. Private
borrowersareinthesamesituationasGovernments–forpublicdebts–andstates–forexternaldebts–
withonebigdifference:aprivatedefaultispromptlysanctionedaccordingtopreciselegislationunderthe
controlofcourts,whilepublicandexternaldebtdefaultsarefollowedbylitigationandnegotiationswithin
fuzzy legal rules and uncertain enforcement mechanisms. Uncertainty about the consequence of public
and external debt defaults is a source of perverse incentives to default (formally called moral hazard)
reflectingunwillingnessasopposedtoinabilitytopay.8
Officiallenderscannotavoiddealingwiththedebtsustainabilityissue.Themultilateralorganizationsand
theParisClubhavelongdealtwiththeissueonacasebycasebasis.Theirstatedruleofprocedurewasto
encourageborrowingcountriestoadoptprudentstrategies,whilebeingreceptivetotheirdevelopment
needs. “Prudent” and “receptive” are subjective attributes, however, which inevitably lead to
7
I am indebted to AnhͲNga TranͲNguyen for suggesting the topic and providing me with much knowledge about debt
sustainability analysis. Many useful comments were provided at the UNCTAD Expert Meeting Debt Sustainability and
DevelopmentStrategiesonOctober26Ͳ28,2005.Alltheviewsexpressedherearemine,asaretheerrors.
8
ThisdistinctionisintroducedinBulowandRogoff(1989).
17
DebtSustainabilityAssessment:theIMFApproachandAlternatives
controversies. It is natural to try and escape such controversies by designing systematic and therefore
universallyapplicableprocedures.Indeed,theWorldBank’sInternationalDevelopmentAssociation(IDA)
andtheIMFhaverecentlystartedtoformalizetheirdebtsustainabilityassessment(DSA)procedures.IDA
lending is now informed by a battery of criteria developed within the Country Policy and Institutional
Assessment (CPIA) approach, while the IMF and the World Bank have put in place a standardized DSA
proceduredesignedtoberoutinelyusedaspartofitssurveillanceandlendingoperations.
This paper examines the DSA procedure. The next section explains why it is mission impossible. Noting
thatsustainabilityisaforwardͲlookingconcept,itarguesthatanypracticaldefinitionisarbitrary,andthat
any sustainability indicator will be both arbitrary and too imprecise to serve as a tool for policy
prescription. Section C then examines the IMF’s procedure, intended to deal with this impossibility
principle by being both simple and transparent. Because of the “mission impossible” nature of the
exercise,however,theprocedureseemstobeevolvingtowardsmorecomplexity.Indeedsimplicitymay
comeattheexpenseofprecision,whichcallsforincreasingcomplexity.Inaddition,giventheIMF’sown
definition of sustainability, the procedure requires adopting, formally or informally, the CPIA approach
developedbytheIDA,asourceofopacity.ThesectionalsoreviewsotherDSAapproaches,somewhich
emphasizesimplicityatthecostofprecision,whileothersgofurtherinthedirectionofcomplexityatthe
costoftransparency.Arguingthatsimplicityandtransparencyindeedareessentialtomaketheprocedure
acceptable, Section D develops a series of principles that lead to a simpler, less ambitious and less
systematicprocedurethatseekstoreplacearbitraryjudgmentswithaframeworkfordialoguebetween
theofficiallendersandtherecipientcountries.
B. WhatisDebtSustainability?
1. Definitions
Debt sustainability is accepted that aims at answering a deceptively simple question: when does a
country’sdebtbecomesobigthatitwillnotbefullyserviced?Thequestioncanbeappliedtotheexternal
debtortothepublicdebt.Theanalyticsareidenticalonceitisnotedthattheexternaldebtislinkedto
theevolutionoftheprimarycurrentaccountbalanceinthesamewayasthepublicdebtislinkedtothe
primarybudgetbalance.Thisdistinctionwillbeblurredinthepresentsectionbyreferringto“debt”and
“primarybalance”,withoutspecifyingwhetheritappliestopublicorexternaldebtsandbalances.
The IMF’s own definition of sustainability is: a debt “is sustainable if it satisfies the solvency condition
withoutamajorcorrection[…]giventhecostsoffinancing”(IMF,2002,p.5).Solvency,inturn,needsto
be defined. Debt solvency is achieved when future primary surpluses are large enough to pay back the
debt, principal and interest. More technically, solvency requires that the current debt plus the present
discounted value of all expenditures does not exceed the present discounted value of all revenues (or,
equivalently, that the current debt not exceed the present discounted value of future revenues net of
nonͲinterestexpenditures).
The solvency definition is clear cut and has long been formalized, but raises many implementation
difficulties.Thesustainabilitydefinition,asstated,isvague.
SolvencyIssues
Solvency,andsustainabilityasaconceptthatbuildsuponsolvency,isentirelyforwardͲlooking.Itisfuture
balancesthatmatter,notthepastandnotjustthecurrentdebtlevel.Hugedebtscanbepaidback,and
smalldebtsmaynotbesustainable.Theoutcomedependsonwhattheprimarybalancewilllooklikein
thefuture,including theverydistantfuture.Infact,mostGovernmentsareindebtedforeverandmany
externaldebtsremainhighfordecades.Forinstance,FigureII.1.showstheevolutionoftheBritishpublic
18
CompendiumonDebtSustainabilityandDevelopment
debt,measuredinpercentofGDP.Duringthelast300years,itneverdroppedbelow20percent,reaching
270percentontwooccasionsandaveraging117percent.Thisdebtwasalwayssustainedinthesense
thattheBritishGovernmentneverdefaulted.Wereturntothisexamplebelow.Fornowwejustnotethat
dealing with the issue of debt solvency – and therefore sustainability – requires passing judgment on
eventsthathavenothappenedyet,thatmaycoveraverylonghorizon,measuredindecades,andthat
arelargelyunpredictable.
FigureII.1.TheBritishPublicDebt–1700Ͳ2004
(PercentofGDP)
British public debt
300
250
200
150
100
50
0
1700
1738
1776
1814
1852
Source:BurdaandWyplosz(2005).
1890
1928
1966
2004
Thenextdifficultyisthatthedebtmustbescaledsomehowtocountrysize.Themostpopularapproachis
torelatethedebttotheGDP,asinFigureII.1.,butthechoiceisnotstraightforward.Itdependswhatis
the source of revenues. Public debts are serviced out of government revenues, so what matters is the
taxingabilityoftheGovernment,nowandinthefuture.
Ifthedebtisexternalorpublicbutpartlyowedtotherestoftheworldand/orinforeigncurrency,itwill
beservicedbytheamountofrevenuesinforeigncurrencythattheGovernmentcancollect.Thereislittle
relationshipbetweenGDPandtheadequacyofcollectiblerevenues.Soanotherscalingfactorisrequired
and it is customary to use exports. But this assumes that a constant fraction of exports can be used to
service the debt. The scaling factor – GDP, exports or any other measure – must be forecast over the
relevanthorizonsothatitisnotjustthedebtitselfthatmustbeguessed.Therecanbenopretenseof
precision.
Afurtherdifficultyisthatdebtsarerolledover.EvenlongͲtermbondsarenotlongͲtermenoughtocover
quasiͲpermanent debts.9As the debt is refinanced, borrowing costs change and must therefore be
guessedaswell.Thisrequiresmakingassumptionsonthefuturecourseofdomesticinterestratesforthe
part of the debt that is issued in domestic currency, and assumptions regarding future foreign interest
ratesandcountryriskpremiaforthatpartissuedinforeigncurrency.Interestratescanchangebecauseof
externalconditions–includingsometimescontagionfromfarͲawayevents–whichaffectinunpredictable
waysthesolvencycondition.
9
Thisisnotentirelycorrecthistorically.TheBritishgovernmenthasissuedperpetuitiescalledconsols,bondslackingamaturity.
Oncealargeproportionofthepublicdebt,consolsarenowanoddityunlikelytobefeasiblefordevelopingcountries.
19
DebtSustainabilityAssessment:theIMFApproachandAlternatives
Inextreme,butnotrare,situations,itmayprovetemporarilyimpossibletorefinancethematuringdebt,
muchlesstoissuenewdebt.Thisdoesnotnecessarilymeanthatthedebtisunsustainable;itisacaseof
illiquidity. Illiquidity may none the less force a debt default, even though the debt is sustainable, as
previouslydefined.
DefinitionofSustainability
TwoqualificationsoftheIMFdefinitionimplythatsustainabilityisamoredemandingrequirementthan
solvency.Thefirstqualificationistoruleouta“majorcorrection”intheprimarybalance.Thisprobably
referstosevereexpenditurecutsorlargerevenueincreasesachievedthroughtaxationorpricingofgoods
andservicessuppliedbythepublicsector.Thedefinitionthereforecoversliquidityconstraints–adryingͲ
upoffinancing,eitherdomesticorexternal–thatrequiredrasticadjustments.Thesecondqualification
refers to the “cost of financing”. Financing costs are bound to change over time and are therefore
unpredictable. In particular, they may increase as the debt rises, creating a vicious circle of the type
discussedfurtherbelow.Asaconsequence,adebtmaybesustainabletodayandunsustainabletomorrow,
or conversely. Thus the definition can be unstable. Finally, note that “major” is a matter of judgment,
whichmeansthattheIMFdefinitionisuncomfortablyvague.
TheIMF’sdefinitionisatvariancewiththesustainabilityconceptproposedbyArrowetal.(2004)inavery
different context (the environment). Applied to the debt issue, their definition could be interpreted as
suggestingthatsustainabilityrequiresthatthenetworthofanentity(theGovernmentorthecountry),
defined as the present discounted value of net revenues less the current debt, be on a nonͲdecreasing
trend. This definition differs from the IMF’s in two important ways. First, it does not require solvency.
Solvency is achieved only if net worth is nonͲnegative. The alternative sustainability definition does not
ruleoutthat,initially,networthbenegativeaslongasitisrisingandeventuallybecomesnonͲnegative,
thus meeting the solvency condition.10Second, and importantly for what follows, it does not imply any
specificthresholdforthedebt.
Makingdefinitionsoperational
Thus there are many competing definitions of external or public debt sustainability. The Box below
summarizesandinterpretsthesevariousconcepts.OnetheoreticallyͲpureconceptissolvency.Theother
theoreticallyͲclearconcept,proposedbyArrowetal.(2004),isthatthenetworth(ofthecountryforthe
externaldebtortheGovernmentforpublicdebt)beincreasing,oratanyratenonͲdecreasing.Thesecond
conceptislessstrictthanthefirstonesincesolvencyrequiresthatnetworthbealwayspositive.These
conceptscannotbeimplementedassuchbecausetheyrequireknowledgeofthefutureevolutionofthe
debt.
IMF (2002) adds to solvency the requirement that solvency be always maintained without any major
adjustment. Both because it relies on solvency and because it rests on an unspecified limit to “major
adjustment”,thisdefinitioncannotbeimplementedassuch.Asexplainedbelow,thedefinitionismade
operationalbyrequiringthatthedebtdoesnotexceedathreshold,tobefurtherdiscussed.Itshouldbe
noted that, if the threshold is conservatively set, the resulting definition is more demanding than the
previousone(ifthethresholdisnotbinding,thedefinitionisempty).
TheArrowetal.(2004)conceptcanbemadeoperationalbyignoringtheunobservablepresentvalueof
primarybalancesandrequiringthatthedebtͲtoͲGDPratiobestationary.Sincestationarityisdifficultto
assess in practice, the definition can be implemented by requiring that the debt ratio be on a declining
trend,whichdoesnotruleoutoccasionalbuttemporaryincreases.
10
ThispointisformallystatedintheAppendix.
20
CompendiumonDebtSustainabilityandDevelopment
BoxII.1.TheoreticalandOperationalDefinitionsofDebtSustainability
LetbtbethedebtͲtoͲGDPratioattimet.Simplifyingsomewhat,thevariousdefinitionsinthetextcanbe
summarizedasfollows.
bt d b
,where b isathresholddiscussedinsection3.1below.
x
DSAdefinition:
x
Solvency:thepresentvalueofbtbecomesnegligibleforlonghorizons(limbt/(1+r)t=0astїь),
whereristherealinterestrate.Anequivalentdefinitionisthatthepresentvalueofprimary
balancesшbt.(Seetheappendixforaformalization.)
x
Debtserviceability:solvencyplusnoilliquidity.Illiquidityariseswhenthedebtcannotbeservicedat
aparticularpointintime.
x
IMF(2002)definition:solvencyplusnoneedformajorcorrection.
x
Arrowetal.(2004):networth,i.e.thepresentvalueofprimarybalanceslesscurrentdebt,isnot
decreasingovertime.
x
Debtstationarity:btdoesnotgrowwithoutbounds.Analternativeisthatbtbe(weakly)
declining.
2. AnImpossibilityPrincipleandItsImplications
Because debt sustainability is a forwardͲlooking concept, it cannot be assessed with certainty. In this
rigoroussensedebtsustainabilityassessment (DSA) isimpossible.Atbest,followingproceduressuchas
thosepresentedinSection3below,educatedguessesmaybepossiblebutitisimportanttorecognizeat
theoutsetthatthesearejustguesses,nomatterhowsophisticatedtheymaybe.Theimplicationsofthis
impossibilityprinciplearefarͲreaching.
Giventhelargenumberofguessesthatarerequiredtoreachanyconclusion,thebestthatcanbehoped
for are statements of the type: “there is a probability of x per cent that the debt is sustainable at a
particularhorizon”.Twoaspectsofthisstatementneedtobehighlightedatthisstage.First,DSAcanonly
provideprobabilities.Insomeextremecases,thesemaybe0or100percent,11butgenerallytheywillbe
somewhere between these values but not easily defined. Put differently, DSA is rarely blackͲandͲwhite
andthereforeanimpreciseguidetopolicy.
Second, the probability that a debt is sustainable in the IMF sense is bound to change over time. For
example,ahighlyindebtedGovernmentthatrunsasizeableprimarysurpluswillseeitsprobabilityofdebt
sustainability rise over time. This is in accord with definition of Arrow et al (2004). Conversely a
Governmentthatstartswithalowdebtbutsystematicallyrunslargeprimarydeficitswillhaveadeclining
probabilityofdebtsustainability.
These two possibilities imply that any statement on sustainability is valid only for a particular horizon.
What should that horizon be? In theory, it should be infinite but, in practice, it is determined by the
availabilityofreliableforecasts:ifforecastsofprimarybalances,interestrates,GDP,etc.areextendedto
10 years, the DSA will provide an answer at the 10 year horizon, i.e. a much shorter horizon than the
11
ThecollapseoftheLTCMhedgefundisausefullesson.InSectionC,wewillpointoutthesimilaritybetweenDSAandportfolio
assessment,andwillindeeddiscussvalueatrisk,asophisticatedtechniquedirectlyborrowedfromfundmanagement.Resorting
tothemostadvancedtechniquesavailable,LTCMmanagers–whichincludedNobelPrizeͲwinnerRobertMerton–hadconcluded
thattheirinvestmentwasnear100percentsure.Asitturnedout,anextremelyrareconjunctionofeventsoccurredandLTCM,
arguablythemostprestigiousfund,wentbankrupt.
21
DebtSustainabilityAssessment:theIMFApproachandAlternatives
infiniteonelogicallyrequired.However,sinceeven10Ͳyearforecastsaretotallyunreliable,thehorizonis
boundinpracticetobemuchshorter.Butthisunderminestheconceptualbasisofthisapproach.
As discussed below, a common way of circumventing the horizon problem is to assume “everything
constant”andextendpasttrendstoaninfinitehorizon.Thisisconvenientbuthasunlikelyconsequences,
i.e.theprobabilityoftheassumedpathiscloseto0percent.Suchexercisesdescribepathsthatcannotbe
takenatfacevalue,inparticularforthepurposesofpolicieswithseriousconsequencesforthelivelihoods
ofmanypeople.
Another aspect of the impossibility principle is that sustainability as defined by the IMF requires a
judgmentof whendebtistoolarge.FigureII.1remindsusthat debt canbe verybigand yetsustained.
Recent work has pointed out that “big” is a relative concept.12It is generally considered that the
developingcountriescannotsustainlargedebts.FigureII.2.showsthat,indeed,thepeakinthemidͲ1990s
foremergingmarketswasfollowedbyawaveofcrises.Willtherecentriseforthesecountries,nowabove
the previous peak, usher a new wave of crises? No one knows. Yet, framing the debt sustainability
definitionastheIMFdoesmakesunavoidabletheadditionofanewconcept,namelyadebtceiling.There
isnoprecisewayofdefiningthisceiling.Itmustbebasedonthemaximumamountofresourcesrequired
toservicethedebt,andthusonassumptionsabouteconomiccostsandpoliticalacceptability.Thiswayof
puttingthequestionleadstoanotherimpossibility,thatofassessingadebtceiling.
Finally,risinginterestratesincreasethedebtburdenandreducetheprobabilityofdebtsustainability.A
disturbing aspect of this linkage is that interest rates on public debts, whether in domestic or foreign
currency,includeariskpremium.Theriskitselfisrelatedtotheprobabilityofdefault,i.e.tosustainability.
TheresultisthepossibilityofaviciouscirclethatgoesfromthefearofdebtnonͲsustainabilitytohigher
interestratesandthustoahigherprobabilityofnonͲsustainability.13InotherwordsthemerefearofnonͲ
sustainabilitymakesitmorelikely.DebtdistresscanthusbeselfͲfulfilling.Thismaymeanthatimproperor
incorrectlyinterpretedDSAcanhaveadeleteriouseffectondebtsustainability.
FigureII.2.PublicDebtsinIndustrialandEmergingMarketCountries–1992Ͳ2002
Source:IMF(2003a).
12
13
Forarecentassessment,includingmanyreferences,seeCordellaetal.(2005).
ThisprocessisstudiedinBlanchard(2005).
22
CompendiumonDebtSustainabilityandDevelopment
3. DebtsandInflation
Inflationisa further complication thatisoftenignoredinDSA. Evenforexternaldebt,inflationmatters
becauseinterestandexchangeratesdonotalwaysreflectactualinflation.Forexample,iftheexchange
ratedepreciatesfasterthanprices,foreigncurrencydebtbecomesmoreexpensiveindomesticcurrency.
Thesamehappenswhentheinterestrateondomesticcurrencydebtincreasesbymorethantheinflation
rate. Debt service becomes heavier. Conversely, when interest and exchange rates fail to fully reflect
expectedinflationandthedebtisnotindexedandisdenominatedindomesticcurrency,risinginflation
temporarilyreducesthecostofborrowing.14
DSA should recognize these various possibilities but does not incorporate standard procedures for this
purpose.Onereasonistechnicaldifficulties.Notonlywoulditbenecessarytoforecastinflationbutalso
expectedinflationandnonͲneutralities,i.e.theextenttowhichtheexchangerateandtheinterestrate
failtoreflectexpectedinflation.Whileitispossibletoforecastinflationoverarelativelyshorthorizon,say
two to three years, forecasts beyond that horizon depend on policy actions that are yet to be taken. It
may also be that international institutions, that typically do not condone inflation, are unwilling to
speculateonwhatitcouldbeandhowitcouldbeusedtoalleviatethedebtburden.15
4. LinkwithEarlyWarningIndicators
A large literature has been devoted to early warning indicators which try to identify irregularities that
eventuallyresultinafinancialand/orcurrencycrisis.LikeDSA,earlywarningindicatorsmustbeforward
looking. Crises, and therefore early warning indicators, are beyond the scope of the present paper. The
onlydirectlyrelatedquestioniswhetherahighdebtlevelisacauseoffinancialcrises,amongthemany
potentialones.AccordingtotheextensivesurveyinHemmingetal.(2003),theanswerismaybe.Formal
statistical analyses provide conflicting results on this point. A problem is that they use current fiscal
indicators, the budget balance or the debt level, as potential pointers of impending crisis. So far, DSA
indicatorshavenotbeenused,tothebestofmyknowledge,inearlywarningindicatorestimates.Todoso
wouldprovideagoodgaugeoftheirempiricalrelevance.
C. ApproachestoAssessingDebtSustainability:ACriticalReview
TheimpossibilityprincipledevelopedinSectionB.2representsaformidablehurdle.AllapproachestoDSA
havetorelyonassumptionsaboutthefutureevolutionofbudgetbalances,GDP,interestrates,etc.The
usefulnessoftheconclusionsisdirectlyrelatedtothevalidityoftheseassumptions,whichbydefinition
areneithersafenortestable.Thissectionstartswithacriticaldescriptionoftheapproachchosenbythe
IMF.Itthenpresentsandevaluatessomealternativeapproaches.
1. TheIMFStandardizedApproach
The IMF has decided to systematically attach a standardized DSA to program design and to Article IV
consultations.TheseDSAsexamineboththepublicandexternaldebts.Thestatedintentionistoprovidea
simple, fully transparent and standardized tool that can be readily applied to all countries.16The World
14
Buiter(1985)hasshownthatthegreatreductionoftheBritishpublicdebtover1946Ͳ1970hasmostlybeenachievedthrough
the inflation tax. A full account of this process includes regulated interest rates, i.e. some degree of financial repression. In
countrieswithfullcapitalmobility,thiswillnotbepossible.Awiderdiscussionoftheroleoffinancialrepressionisbeyondthe
scopeofthispaper.AwellͲknowndefenseofsomedegreeoffinancialrepressionisRodrik(1998).
15
AbiadandOstry(2005)provideevidencethatinflationraisestheprimarybudgetsurplus.
16
Infacttherearetwodifferentbutrelatedprocedures,onedesignedforcountrieswithmarketaccessandanotheronedesigned
forlowͲincomecountrieswhichrelymostlyonpublicfinancing.Themaindifferencesarethefollowing.1)Inthecaseofcountries
23
DebtSustainabilityAssessment:theIMFApproachandAlternatives
Bankhasadoptedasimilarprocedure. Unfortunately,theimpossibility principleisin contradictionwith
these intentions. Simplicity is achieved at the cost of improbable assumptions; these assumptions are
transparent,buttheyarelessinnocuousthantheyaremadetoappearbecausetheunderlyingcomplexity
isconcealed.
Focusing here on the external debt part of the exercise, the IMF approach includes the following four
steps:17
(i) AfiveͲyearcentralforecast,orbaseline,ofthevariablesthataffecttheevolutionoftheexternal
debt:theprimarycurrentaccount,GDP,interestandexchangerates,andinflation.
(ii) Theresultingevolutionofthedebt,asashareofGDP,overthenextfiveyears.Thisevolutionis
uncontroversialasitfollowsfromthefollowingaccountingidentity:
bt bt 1
(r g )bt 1 primary balancet WherebisthedebtͲtoͲGDPratio,ristherealinterestrateandgistheGDPgrowthrate.
(iii) Several stress tests that look at the effect on debt of adverse shocks affecting the variables
forecastedinstep((i).Theshocksareasfollows:first,eachofthreevariables(theinterestrate,
realGDPgrowthandtheprimarycurrentaccount)ischangedonebyoneͲhalfstandarddeviation
overthesamefiveͲyearhorizon;thenallthevariablesaresimultaneouslyshockedbyonequarter
standard deviation each over five years; finally the exchange rate is assumed to be depreciated
onceby30percentatthebeginningofthesimulationperiod.
(iv) TheDSAconcludeswithajudgmentonwhetherthedebtlevelsimpliedbyanyorallofthestress
testsaretoohighforthedebttobeconsideredsustainable.
The result is a figure like Figure II.3., which is based on the November 2005 review of the standͲby
agreementwithColombia,seeIMF(2005b).18Thefiguredisplaysvarioussimulatedpathsoftheexternal
debtoverthefiveͲyearperiod2006Ͳ10:thebaselineobtainedinstep((ii))andtheeffectsofthreeofthe
shocksdescribedinstep((iii)).Theseshocksare:aoneͲhalfstandarddeviationcurrentͲaccountshockand
the combined shock, both assumed to last the whole simulation period 2006Ͳ2010; and a 30ͲperͲcent
exchangedepreciationoccurringin2006.
withmarketaccess,theanalysisconcernsboththeexternaldebtandthepublicdebtand,inthecaseoftheexternaldebt,itdeals
with its level while, in the case of lowͲincome countries, it concerns only the external debt, which is measured in net present
discountedvalueterms.2)ForlowͲincomecountries,theDSAusesanexplicitproceduretoestablishdebtthresholds,whilethe
thresholdisleftopentodiscussioninthecaseofthemarketaccesscountries.IMF(2003b)proposestousethesamestandard
proceduretobothgroupsofcountries.Thepresentanalysisignoresthesedifferences.
17
ThisdescriptionfollowstherecentchangesasdescribedinIMF(2005a).
18
Thisseemstobethefirst,andsofaronly,countryreviewthatappliesthechangesasdescribedinIMF(2005a).
24
CompendiumonDebtSustainabilityandDevelopment
FigureII.3.ExampleofDSA:SimulatedPathsoftheDebtͲtoͲGDPRatio
45
40
35
30
25
2004
2005
2006
2007
2008
2009
2010
Baseline
Current account shock
Combined shock
30% depreciation
Source:IMF(2005b).
Whatcanthisinformationbeusedfor?Obviously,thebaselineisnocauseforconcern.Thestresstests,
ontheotherhand,arelessbenign.TheoneͲtimedepreciationraisesthedebtͲtoͲGDPratiobyabout30
percent,presumablybecausetheexternaldebtisinforeigncurrency.Thisisnotathreattosustainability,
however,becausethedebtstartsdeclininginthethirdyear,mostlikelybecausedomesticpricescatchup
withtherateofdepreciation.Moreworrisomearetheeffectsofaworseningofthecurrentaccountand
ofthecombinedshocksinceinbothcasesthedebtkeepsonrising.
The unavoidable question is whether these simulations are sufficient to warrant a policy reaction.
Undoubtedly,werethedebtratiotokeepongrowing,thedebtmusteventuallybecomeunsustainable,
butwhen?Inaddition,sincetheshocksareexpectedtolastforfiveyears,thedebtshouldpresumably
declinebeyondthehorizon.Inordertobeabletodrawanyconclusionfromthisexercise,therefore,one
must be able to conclude that the debt level reached at some point in Figure II.3. is too high to be
sustainable. This, in turn, requires establishing a debt threshold level beyond which danger is looming.
Danger means debt distress, i.e. financing difficulties or, worse, partial or total default. In view of
empirical results that show that the risk of debt distress rises with the size of debt, it seems logical to
establishadebt thresholdbeyondwhich therisks canbedeemedunacceptable.Thisistherationale of
step((iv)).
Should there be a single threshold for all countries? Here again, empirical research shows that the
probability of debt distress depends not just on the debt level itself, but also on a variety of factors,
includingtheprevailingmacroeconomicsituationand,importantly,thequalityofeconomicandpolitical
institutions. A unique common threshold, therefore, is bound either to be too restrictive or too lax,
dependinguponthecountrycharacteristics.ThisiswhytheIMFhasbeguntouseaspartofstep((iv))an
additionalprocedurecalledCountryPolicyandInstitutionalAssessment(CPIA).DevelopedbytheWorld
Bank,CPIAproducesanindexofgovernancequalityforeachcountryproducedbytheWorldBank.19
This index, which ranges from 1 (lowest quality) to 6 (highest quality), is based on 20 indicators. It is
updatedannually,followingaformal andelaborateprocessthatinvolves theBank’s countryteams and
centraldepartments.Foundtoperformwellinstatisticaltests,theindexisusedtoclassifycountriesinto
19
ThisproceduresofaronlyappliestothelowͲincomecountries,presumablybecauseitisinuseatIDA.TheFundisconsidering
applyingittocountrieswithmarketaccess.
25
DebtSustainabilityAssessment:theIMFApproachandAlternatives
threegroups:countrieswithalowCPIAindexareascribedadebtthresholdof30percentofGDP,raised
to45percentfortheintermediategroupofcountries,andto60percentforthecountriesinthehighest
CPIA index group. These thresholds are chosen such that the probability of debt distress is 25per cent
whentheyarereached.Followinganexpertreview,theWorldBankreducedthenumberofcriteriafrom
20to16andhavemadeindividualratingspubliclyavailableforIDAcountriesin2005.
2. DiscussionoftheIMFApproach
Steps ((i))and((ii))aremechanicalimplicationsof theIMF’sforecasts.If theforecastsareaccurate, the
implieddebtlevelisareasonablysafeprevision.TheFundreportsonitsownstudiesthatshowthatthe
forecaststendtoerrontheoptimisticside,withequallyoptimisticdebtpredictions(IMF,2005c).
ProbabilityofWorstCaseScenarios
Thispossibilityofforecastingerrorexplainswhy,“inordertoimposediscipline”onthediscussion,step
((iii))looksatsomeworstͲcasescenarios.Since“worst”canbevirtuallyanything,theprocedureattempts
tobereasonableandtransparent.Tothateffect,theshocksarepreciselycalibrated.Buthowlikelyare0.5
standarddeviationshocks?TheIMFarguesthattheprobabilityofthedebtexceedingtheworstcaseon
thefifthyearisbetween15and30percent,“whichseemsareasonablebalancebetweencapturingthe
mediumͲtermriskstodebtdynamicswithoutbeingsoextremeastobeirrelevantforpolicydiscussions”
(IMF,2005a;p.3).Theemphasisisrightlyputonpolicyimplicationsbuttheargumentraisesthegeneral
questionofwhatcanbelearntfromstresstests.
ThereisnothingwrongwithstressͲtesting.Indeed,itisacommonapproachtoportfoliomanagementand
widely used in the financial industry, as explained in Section C.3. Yet, the implications profoundly differ
betweeneconomicpolicyandportfoliomanagement.Inthefinanceindustry,whenstresstestsreporta
dangerzone,evenahighlyunlikelyone,portfoliomanagersmaydecidetochangetheassetcomposition
oftheirportfolios.Theadjustmentdoesnotcomeforfreesinceitimplieslowerexpectedreturns,butthis
istheusualpricetobepaidforlowerrisk.Itsacceptabilitydependsoninvestorpreference:ifinvestors
areunhappywiththeirportfoliomanagers,theycanchangethem.
Inthe caseofDSA,when stresstestssignalariskysituation, the requiredadjustmentis toimprove the
primary current account. Inevitably, this calls for contractionary macroeconomic policies designed to
compress demand. The costs take the form of falling incomes and rising unemployment. The costs are
bornebythepopulation.AdmittedlycitizenscanvotetheirGovernmentsoutofoffice–whentheregime
is democratic – but only ex post. A Government’s decision to react to events that may occur with a
probabilityof15Ͳ30percentisconsiderablymoresensitivethanthatofportfoliomanagers.
CharacterizationoftheWorstCaseScenarios
Thestresstestsinvolvechangesinonevariableatatimeexceptinthecasewhereallofthemarevaried
togetherina“bad”directionforfiveyears.Howlikelyaresuchchanges?Itisunclearhowthe15Ͳ30per
cent estimate is constructed. Why does it assume that each shock is expected to be maintained over
consecutive 5 years? Does it take into account the fact that some of these shocks may be correlated?
Lettingtheshockslastthewholefiveyearsassumesa100percentautoͲcorrelation.TheoneͲatͲaͲtime
shockassumesazerocorrelation,whilethethreeͲvariableshockassumesacorrelationof100percent.
TheinformationprovidedinIMF(2005a)doesnotshedlightonthisimportantquestion,suggestingthat
correlationsareignored.
What can be done about these problems? Here, as with the threshold question, the proper technical
responseistojackupthelevelofcomplexity,atthecostofreducingtheintendedtransparency.Faced
with the criticism that changes in just one of the variables that drive the debt process have historically
26
CompendiumonDebtSustainabilityandDevelopment
typically affected the others as well, the normal tendency is to acknowledge the point and take up the
challenge. This means using econometric techniques to estimate how, in the past, these variables have
beenrespondingtoeachother’sshocks.20
Theprocedureiswellestablishedbuthasmanydrawbacks.Tostartwith,theestimateswouldhavetobe
conducted country by country, an enormous task that would face serious data availability and
comparability problems. Next, the quality of the estimates is likely to be poor in many cases,21thus
injectingafurtherdoseofuncertaintyregardingthemeaningofthedebtthresholdandthusundermining
its usefulness. In addition, the procedure would turn a simple and transparent procedure into a highly
technicalandcompletelyopaqueexercise,withlittleassurancethatthetestsareplausible.
The challenge is formidable, possibly insurmountable.22It would be a mistake to go in the direction of
addedcomplexity,usingtheabundantparaphernaliaofeconometricinstruments.Resortingtosimplebut
less extreme stress tests (shocks with reasonable overall probability) would be an alternative, but the
resultscouldwellbetoomundanetobeworthconsidering.Thelackofasatisfactorysolutionisnothing
morethananimplicationoftheimpossibilityprinciplepresentedinSectionII.C.
BorrowingandGrowth
MissingfromtheDSAframeworkisthepossiblegrowthͲenhancingeffectofexternalborrowing.Intheory,
acountrywithlowlevelsofhumanandphysicalcapitalstandstobenefitfromexternalborrowing.Ifthe
borrowingiswiselyinvested,thereturnsshouldmorethancoverthecosts.Thebenefitscomeintermsof
acceleratedgrowthandcatchingͲup.23
Thislinkageisexplicitlyignoredinthestresstests.Thismaylooksurprisinggiventhatmultilaterallending
isultimatelyjustifiedbyitsgrowthͲenhancingeffect.Onepossibleexplanationisthatthehorizonistoo
short for the growth effects to materialize. The proper response to this argument is to lengthen the
horizon, not to ignore the link. If debt distress occurs along the way, the expected growth bonus from
externalborrowingwouldbedissipatedbutthisdoesnotjustifyignoringthelink.
Anotherpossibleexplanationisthatborrowedresourcesdonotsystematicallydeliveranygrowthbonus.
There is much evidence that the quality of policies and of political governance matter crucially in this
respect(Cordellaetal.(2005).ThisaspectispartlytakenintoaccountbyCPIAasthequalityofpolicies
andinstitutionsareusedtodeterminedebtthresholds.
Giventheoverwhelmingimportanceofgrowthamorecomprehensiveframeworkisneeded.Bylimiting
theroleofpolicyandinstitutionqualitytothedeterminationdebtthresholds,DSAputsalltheemphasis
ontherisksofoverborrowing.Ignoringtheconditionsunderwhichexternalborrowingcanharmorboost
growthamountstoposingthequestioninadequately.Ifexternalborrowingisgrowthenhancing,therisk
of over borrowing is small, possibly nonͲexistent. If, instead, external borrowing does not exert any
favorable growth effect and possibly stunts growth, the relevance of DSA is moot. Countries in this
20
Some papers have started to explore this issue, see Garcia and Rigobon (2004), Abiad and Ostry (2005) and Celasun et al.
(2005).
21
The degree of precision of such estimates is generally quite limited but in most developing countries data availability and
qualityproblemsarelikelytobemoreserious.
22
IMF(2003b)suggestsusingthetechniquetoderivefancharts,i.e.chartsthatdepicttheevolutionofthedebtfollowingashock
byindicatingthemostlikelypathalongwitharangeofpossibilities.FanchartshavebeenpopularizedbytheBankofEnglandas
partofitsinflationtargetingstrategy.ThisishowtheBankpresentsitsinflationforecasts.Importantly,however,thefancharts
aredesignedbytheBankofEngland’sMonetaryPolicyCommittee.Theyarenottheresultofacomplexeconometricprocedure
but a snapshot representation of what policymakers believe. Fan charts are a great communication tool, which reflect the
considerationsthatgointopolicydecisionsbutnotoutsideexperts’estimationsofwhatislikelytohappen.
23
IntermsoftheformulapresentedinFootnote29,thegapbetweentheinterestcostandthegrowthratedeclines,andthedebt
accumulationbecomeslessdestabilizing,orthegapbecomesnegativeandthedebtisspontaneouslyonadecliningtrend.
27
DebtSustainabilityAssessment:theIMFApproachandAlternatives
situation should only borrow in distress situations and promptly pay back the debt before the burden
becomescrippling.
PolicyResponses
Thestress testsalsoassumethat the Government doesnotreacttotheshocks.Thisisincontradiction
with much evidence that shows that the primary budget reacts to a rising public debt, which should
presumablyalsohaveadampeningimpactontheexternaldebt.24ThustheworstͲcasescenariosmustbe
seenasapredictionthatassumesthatGovernmentsdonotdowhatinfacttheyusuallydo.Thisfurther
reducestheplausibilityofthetests.
CountryPolicyandInstitutionalAssessment(CPIA)
TheinclusioninIMFdefinitionofdebtsustainabilitytheconditionthatdebtlevelsnotbe“toolarge”leads
totheneedtoestablishthresholds.Theobservationthatreasonablethresholdsarelikelytovaryfromone
country to another then requires an explanation of why some countries are more likely to suffer from
debt distress than others. This explanation involves a large number of economic and political
considerationsandrequiresvaluejudgments,averyuncomfortableundertaking.
TheIMFͲIDAsolutionhasbeentolookforstatisticallinksbetweenvariouscausesofdebtdistressandthe
debtlevel.ThetwoacknowledgedbackgroundstudiesareKraayandNehru(2003)attheWorldBankand
anunpublishedIMFpaper.Inlinewiththeliteratureontheroleofgovernance,25theresultingCPIAindex,
whichisreasonablypreciselyestimated,isfoundtoexertasignificanteffectontheprobabilityofexternal
debtdistress.OnthisbasisitwouldbepossibletoassertthatanimprovementintheCPIAindexreduces
the probability of distress and even to compute by how much. This is not how CPIA is used in DSA,
however.
Theprocedureinsteadusestheestimationtoansweradifferentquestion:whatdebtlevelimpliesa25Ͳ
perͲcentprobabilityofdebtdistress?TheanswercannotbebasedonthepartialeffectoftheCPIAindex
only but also involves estimates of the effect of other economic variables. If the resulting overall
estimationdoesagoodjobofexplainingdebtdistressepisodes,itwouldbeagoodcandidatetoestablish
athresholdforeachcountry.Unfortunately,whiletheeffectofeachofthethreevariablesselectedͲdebt,
CPIAindexandrealGDPgrowthͲispreciselyestimated,togethertheyexplainonly23.4percentofthe
probability of debt distress. In a study that seeks to explain 163 episodes of debt distress all over the
world,thisisagoodperformanceͲamongthebestintheliteratureͲandunlikelytobemuchimproved
upon.Yet,thefactthattheanalysisexplainssolittleofthephenomenonofdebtdistressimpliesthatthe
answerishighlyimprecise.SubsequenttestsprovidedbyKraayandNehru(2003)candidlyconfirmthis.
A further problem is that the CPIA index is not applied country by country. Instead, the countries are
classifiedinthreegroupsdependingontheirownCPIAindex.Theeffectofgovernanceisappliedgroupby
group,whichimpliesthattheeffectiseitherexaggeratedorunderestimatedforthecountrieswhoseCPIA
indicesdonotlieinthemiddleoftherange.Thisdistortionriseswiththedistancefromgroupmeans.
This procedure is surprising. On the basis of the estimation, it is possible to compute individual debt
thresholds. Why is it not done? One reason is simplicity. Three thresholds are easier to deal with than
countryͲspecificthresholds.Butthisisaweakjustificationforintroducingseriousdistortionswhichimply
thatthethresholdcannotbetakenseriously.AnotherreasonisthepoliticalsensitivityoftheCPIAindex
forindividualcountries.Thisisunderstandable,buttheresultisthattheDSAthresholdsaretoocoarseto
leadtofirmpolicyconclusions.
24
25
AgoodsurveycanbefoundChapter3ofIMF(2003a).
AgoodreferenceisManasseetal.(2003).
28
CompendiumonDebtSustainabilityandDevelopment
SustainabilityMeasure
It can be argued that the debt level, suitably scaled, is arguably the correct measure for sustainability
analysis.26Whentrackingitsevolutionovertime,however,aproblemarises.Whenevertheinterestrate
exceeds the economy’s growth rate, the debt accumulation process is intrinsically unstable. This is
preciselywhysustainabilityisanimportantissue.27Twodifficultiesfollow.First,relativelysmallchangesin
therealinterestrateandintrendgrowthcantiltthedebtpathfromstabilitytoinstability.Second,when
therealinterestandgrowthratesareclose,smallshockscanhavedramaticallypowerfuleffectsonthe
debtpath.
ThestrengthofthiseffectcanbeseeninFigureII.4.ThefiguredisplaysthebaselinedebttoGDPratioand
theeffectoftheDSAstandardcombinedstressͲtest,bothalreadyshowninFigureII.3.Thethirdcaseadds
tothecombinedtesttheeffectofanexternalinterestrateset3percenthigherthanassumedbytheIMF.
Thusincreasingtheinterestrateproducesasizeableeffect.Comparingthecombinedshockeffectswith
thelowerandhigherinterestrates,weseethatnotonlyisthedebtrisingfasterbut,moreimportantly,
thatthedebtratioisnotstabilized,possiblysuggestingnonͲsustainabilityunderanydefinition.
ThisexampleillustratesthepointthatdebtaccumulationeffectscanbeeyeͲcatchingand,inthisinstance,
mayraiseconsiderablealarmsincethedebtͲaccumulationprocessisunstable.Inreality,primarybalances
willbeadjustedastheresultofpolicymovesandofequilibratingreactionswithintheeconomywiththe
resultthatdebtinstabilityisusuallytakencareof.Ofcoursetherehavebeenepisodesofexplodingdebts,
largely because small slippages can have dramatic effects as the result of the unstable nature of the
process.ThisiswhyputativedebtpathsofthesortproducedbytheIMFaspartofitsDSAprocedurecan
besomisleading.28
26
Asnotedinfootnote16,forlowͲincomecountriesthisprocedureusesthenetpresentvalueofthedebt.While,inprinciple,this
isasuperiormeasure,itscomputationraisesanumberofdelicatequestions,whicharenotconsideredhere.Whenwereferto
debtlevels,wedonotdistinguishbetweenthedebtanditsnetpresentvalue.
27
Whentheinterestrateislowerthanthegrowthrate,thedebttoGDPratioisstableandsustainabilityisassured.InthelongͲ
run,thisisanunrealisticcasebecausegrowthinexcessoftherealinterestrateisacatchͲupphenomenon(acountrythatdisplays
asteadyͲstaterealinterestratelowerthanthegrowthrateisonthe‘wrong’sideofthegoldenruleinthesensethatitsavesand
invests‘toomuch’,suboptimallyrepressingconsumption).Butintheshortrunthisconditionallowscountriestorundownthe
debttoGDPratio.Theinterestratemaybelowerthanthegrowthrateduringaperiodoffastgrowth(asinChinaorIrelandover
thelastdecade)orduringaperiodofacceleratinginflation(asduringsomeperiodsfortheUKshowninFigureII.1).
28
ArelatedconcernappliestotheDSAforlowͲincomecountries.Thechosenmeasure,thenetpresentvalue(NPV)ofthedebt,is
very sensitive to interest changes. This measure is compared to the NPV of the debt ceiling. Should the ceiling itself also be
adjusted?TheIMFdoesnotdothis,arguingthatthereareoffsettingeffectsintermsofexpectedproductivityadjustments.Thisis
likelytobetrue,eventhoughthetimingandsignoftheseeffectsisnotknown.Butitisalsotruethatthesameeffectswillaffect
thepathofthedebtinthesameway.Itisinconsistenttoadjustonemeasureandnottheother.
29
DebtSustainabilityAssessment:theIMFApproachandAlternatives
FigureII.4.EffectontheDebttoGDPRatioofaHigherInterestRateontheCombinedShock
40
35
30
25
2004
2005
2006
2007
2008
2009
2010
Baseline
Combined Shock
Combined Shock with higher interest rate
Source:Author’scalculationbasedonIMF(2005b).
Implementation
TheIMFhasexaminedtheshortexperiencewiththeimplementationofDSA.ItemergesthattheDSAhas
notbeenassuccessfulasitspromotersintended:“withsomeexceptions,sustainabilityassessmentshave
generallynotbeenatthecenterofpolicydiscussionsbetweenstaffandnationalauthorities.Thismaybe
becausethesensitivitytestsareconsideredtooextremetoberealisticor,evenifrealistic,tooextremeto
warrant a policy response. Conversely, there remain concerns that the assumed shocks are too benign.
Finally, from a presentational standpoint, debt sustainability assessments would have greater impact if
theywereintegratedinthebodyofthestaffreportinsteadofbeingrelegatedtoanannex.”(p.15)
ThissituationreflectsreservationsaboutDSAascurrentlypracticed,whichleadstoreluctancetoraisethe
topic with national authorities. The intended transparency of the shocks used for the stress testing is
marred by their low probability of occurrence. Another factor is the “black box” nature of the exercise,
especiallytheassumptionsabouttheeconomy’sresponsetotheshocks.(Infact,itisassumedthatthere
isnoresponse,whichisunrealisticasnotedabove.)Moreimportantly,Staffmaybeembarrassedbythe
question: “so what?”. This question immediately brings to the fore the need to decide whether a
temporarybulgeinthedebtisthreatening.TheanswerismeanttobeprovidedbytheCPIA.TheCPIA,
however, is another “black box” with a large degree of uncertainty. This uncertainty reduces the
usefulness of the CPIA thresholds as a reliable guide for policy. Being unable to answer the “so what?”
question,IMFStaffdownplaytheDSAexercise.
3. OtherApproaches
Until recently, due to the impossibility principle, there have been few other attempts to design
implementableapproachestoDebtSustainability.Theearlyonesacknowledgedtheconcept’ssensitivity
tounavoidableassumptionsbystressingsimplicityandtransparency.Simplicityisjustifiedbytheneedto
make heroic assumptions which imply that the conclusions will always be fragile. Transparency is
necessary to allow users to understand what lies behind the result. More recently, DSA has moved
towards more elaborate procedures, driven by the hope that empirical regularities can generate more
reasonableassumptionsandfacilitateassessmentoftheirplausibility.Acommentaryonotherapproaches
totheproblemsraisedabovefollows.
30
CompendiumonDebtSustainabilityandDevelopment
TheDebtͲStabilizingPrimaryBalance
The classic approach to sustainability asks what is the primary balance required to stabilize the debt
(Blanchardetal.,1991;Buiter,1985).Theobjectivecanbetostabilizethedebteitheratitscurrentlevel
or at any other level deemed more desirable. This approach is simple, transparent and easily
implementablebecauseitrequiresfewassumptions.Initssimplestform,itlooksatthecurrentdebtto
GDP ratio and computes the primary balance which would permanently keep this ratio unchanged. It
requirestwoassumptions:whatwillbetheevolutionoftherealinterestrateandwhatisthepotential
growthrate?Typically,pasttrendsareassumedtoremainstableovertheindefinitefuturebutshockscan
befactoredin,justasintheIMF’sDSA.
UltimatelyIMFdebtͲpathprojectionsandthecomputationofdebtͲstabilizingprimaryaccountsarebased
on the same reasoning and assumptions. Both rest on the debt accumulation identity
bt bt 1
(r g )bt 1 primary balancet
.Yet, there is an important difference. DebtͲpath projections
either indicate that the debt is stable or declining, in which case there should not be any sustainability
issue, or that it is rising, indicating eventual unsustainability. The debtͲstabilizing primary account
approach stops there. The IMF’s DSA goes one step further by exploring adverse shocks. When these
shocksimplyarisingdebt,newquestionsarise.Sincetheshocksare,byconstruction,temporary,arising
debtpathdoesnotmeanunsustainabilityunlessthedebtbecomestoolarge.Thisthenraisesahostof
newissues,whichhavealreadybeendiscussedinSectionB.1.ThevirtueofthedebtͲstabilizingprimary
accountapproachistoavoidtheseissuesor,moreprecisely,topreventthemfrombecomingprominent.
Ofcourse,thequestionofwhatisanappropriatedebtlevelcannotbealtogetheravoided.Lookingatthe
primary account that stabilizes the current debt assumes that the current debt is appropriate. But
alternativetargetsforthedebtlevelcaneasilybelookedat.Thequestionisvexingbecausethetheory
doesnotprovideananswer.TheWorldBank’sCPIAisoneattempttoprovideapracticalanswerbutis
subjecttotheproblemsdescribedinsectionC.2.
HowcouldthealternativeconceptofthedebtͲstabilizingprimaryaccountbeappliedtotheexampleofa
majorshockliketheexchangeͲratedepreciationpresentedinFigureII.3.?Inthebaselinecaseunderthe
IMFDSAthedebt/GDPratioisassumedtodecline.ThusadebtͲstabilizingprimarybalancewouldconsist
of a deficit.29This would provide an answer to the question of what is the balance on current account
needed to keep the debt level unchanged as a percentage of GDP, ignoring capital inflows? The
alternativeapproachcouldalsobeusedtoanswerthequestionofwhatwouldbethebalanceoncurrent
account required to bring the debt level down to a particular level by, say 2010, if the debt level is
perceivedtobetoohigh?Thiswouldbeastraightforwardcalculation.
As noted, the concept of the debtͲstabilizing primary account is formally identical to the IMF DSA, but
interpretationisdifferent,asillustratedinFigureII.5.Herethedebtisassumedinthebaselinescenarioto
decline, and the figures show the effects of the combined shock. The corresponding paths for external
debtandtheprimaryaccountaredenotedas“original”.Theprimaryaccountremainsinsmalldeficitas
the debt must be reduced. If it is assumed that the authorities can control the primary account, the
questionishowtheywouldpursuethepolicygoalofstabilizingthedebt.
OnepossibilitywouldbetofixthedebtatitspreͲshocklevelof2005.Thisrequiresalargeprimarysurplus,
onethatcompletelyoffsetstheeffectoftheshock.Thissurplusisshownas“Stabilized1”intheleftpanel
ofFigureII.5.Therightpanelshowsthatthispolicyrequiresahugeimprovementintheprimaryaccount
in response to the sudden increase in the domestic currency value of the external debt due to the
depreciation.Thepolicyresponseisrelaxedwhenthepriceincreasescatchupwiththedepreciationso
thatthedebtisstabilizedindomesticaswellasforeigncurrency.
29
Theformula,itwillberecalled,isprimarybalance=(interestrate–growthrate)xdebt.
31
DebtSustainabilityAssessment:theIMFApproachandAlternatives
Anotherpossibilityistoletthedebtriseinitiallybuttoaimatreturningittoits2005levelbytheendof
theplanningperiod,here2010.Theleastdisruptivewayofdoingthisistoachieveaprimarysurplusthat
remainsconstantovertheplanningtheperiod.ThisisshowninFigureII.5.as“Stabilized2”.Theprimary
balancenowincreasesmoderately,eventhoughtheshockisunusuallyviolent.Thedownsideisabulgein
thedebtleveluntil2006.
FigureII.5.DebtStabilizingPrimaryBalance
Debt(percentofGDP) Primarybalance(percentofGDP)
45
14
Original
Stabilized 1
Original
Stabilized 2
Stabilized 1
Stabilized 2
12
10
40
8
6
4
35
2
0
30
-2
2005
2006
2007
2008
2009
2010
Source:Author’scalculationbasedonIMF(2005b).
2005
2006
2007
2008
2009
2010
Looking at the debtͲstabilizing primary balance, not just at the effect on the debt path, provides a
different perspective on stress testing. First, it deͲdramatizes the shock effects. It shows that sustained
but moderate primary balance corrections can eventually ensure sustainability – here defined as debt
stabilization –in thefaceofevenunusualshocks. BothexperimentsshowninFigure II.5.quicklyreturn
thedebttoitspreͲshocklevel,immediatelyinthe“Stabilized1”caseandafterfiveyearsinthe“Stabilized
2” case. Given the low probability of the shock, it would make sense to allow the slower return.
Lengthening the horizon would clearly allow for smaller primaryͲ account corrections. It should also be
notedthatadverseshocksarelikelytobecompensatedsoonerorlaterbyfavorableshocks.
Second, this approach also deͲdramatizes the inherent instability of the debt accumulation process
illustratedinFigureII.4.Timepermitting,evenverylargedebtshockscanbedealtwiththroughmoderate
primaryͲaccount corrections. The reason is that a moderate sustained primaryͲaccount correction
eventuallyproducesasalargecumulativeeffectastheshockitself.
Third, this approach brings to the fore important policy implications. Obviously, “Stabilized 2” is more
palatable,economicallyandpoliticallysinceitavoidsamassivespendingcontractionboundtoresultina
severe recession. It is a good general principle that temporary shocks should be met with smoothing
policies,i.e.policiesthatspreadovertimetheadjustmentcosts.30Merelylookingattheoriginaldebtpath
intheleftͲhandsidepanelinFigureII.5.mayconveyasenseofurgencythatisnotnecessarilywarranted.
Ofcourse,the“Stabilized2”pathassumesthattheexternaldebtbulgecanbefinanced,whichmaynotbe
the case for a number of countries. But the answer is that the IMF was created precisely to provide
emergencyfinancinginthefaceofatemporaryshock.
30
Permanentshocks,ontheotherhand,needtobemetbyapermanentoffsettingpolicyassoonaspossible.Thisdoesnotrule
outagradualimplementationiftherequiredpolicyinvolvesadjustmentcosts,whicharebestspreadovertime.
32
CompendiumonDebtSustainabilityandDevelopment
Fourth,theissueofaccesstotemporaryexternalfinancingdrawsattentiontotheallͲimportantcredibility
issue.Thedownsideofthe“Stabilized2”responseisthatthedebtincreasewillbetemporaryonlyifthe
authoritiescancommittoasustainedprimaryaccountsurplus.Absentcredibility,theadjustmentmustbe
frontͲloaded,asinthe“Stabilized1”case.
CredibilityisalsopartoftheIMFDSAsinceitliesattheheartoftheprocedureofthresholddetermination
carriedoutunderCPIA.Yet,FigureII.5pointstoashortcomingofthisapproachwhichfailstodistinguish
betweencaseswhereanincreaseinthedebtlevelismerelythetemporaryconsequenceofatemporary
shock, on the one hand, and those where it results from endemic policy indiscipline.31This is why it is
essentialthatIMFlendingbeaccompaniedbycredibilityͲenhancingconditionality.
Thisexampleillustratesthepointmadeabove:theIMFDSAprocedureimaginesshockswhichmayresult
insizeabledebtbuildͲupsbecausetheauthoritiesareassumed nottoreact. Weseethat,ifgiventime,
relatively moderate primary account improvements can stabilize the debt (this is “Stabilized 2”). There
shouldbenoimplicationthattheshockmustbedealtwithimmediatelyaswith“Stabilized1”.
ValueͲatͲriskstresstests
Financial institutions have developed procedures to explore the risks associated with portfolios in the
formofthevalueͲatͲrisk(VAR)approach.Forfinancialfirms,theobjectiveistheavoidanceofinsolvency.
Attheheartofthisapproacharetwomainideas:thathistoryallowstheevaluationoftheprobabilityof
variouseventsorcombinationsofevents,andthatreactionsshouldtakeintoaccountboththepossible
severityofeacheventanditslikelihood.
The techniques used to measure the plausibility of various risks can also be applied to the debt
sustainability question. The IMF’s approach takes a partial step in this direction when it sets levels for
some variables in stress testing on the basis of their previous behavior. But, as noted above, it ignores
howthesevariablesreacttoeachother.Inprinciple,onecouldgomuchfurtherinthisdirectionbutonly
atthecostofaddingconsiderablecomplexityandopacity.
TheissuehasbeenstudiedbyGarciaandRigobon(2005)andCelasunatal.(2005),sothatitispossibleto
giveasenseofwhatshouldandcanbedone.Ratherthanspecifyingshocksonthebasisofthehistorical
evolutionofindividualvariables,properlyconstructedstresstestsshouldtakeintoaccountthehistorical
interdependenceamongthesevariables.Forexample,inthecasesdisplayedinFigureII.3.thecombined
shockinvolvesasimultaneousdeteriorationinthecurrentaccount,theinterestrateandGDPgrowth.This
combination may be more or less likely than each of its components. For instance, if GDP growth
systematically worsens when the interest rate increases, the combination is as likely as each of its
components.Allowingforsuchcorrelationsenablesbetterappraisaloftheprobabilityoftheshocksthat
areconsidered.
This is the first step of the VAR approach which assumes that historical correlations are likely to be
relevantinthefuture–areasonablebutnotnecessarilycorrectassumption.Thenextstepthenistotake
intoaccountallthepossiblecombinationsofshocksbasedonestimatedcorrelations.Theprocedurecan
beautomatedtorandomlygenerateaverylargenumberofshocksbothsmallandbig,inisolationandin
combination.Eachshockisassociatedwithaprobabilityofoccurrence.32Thelaststepistoassociatewith
eachshockthecorrespondingevolutionofthedebt,muchasintheIMF’sDSA,exceptthateachdebtpath
nowcomeswithaspecifiedprobabilityofoccurrence.
31
ThesamefailuremayhelptoexplainwhytheestimatesofKraayandNehru(2003)explainonlyasmallpartofepisodesofdebt
distress.
32
Technically,thisiscalledMonteͲCarlosimulations.
33
DebtSustainabilityAssessment:theIMFApproachandAlternatives
Onecanthenaskthefollowingquestion:at,forexample,thethreeͲyearhorizonwhatistheprobability
thatthedebtbebetweenahighandalowlevel?Thedifferentcombinationsofhighandlowlevels,with
associated probabilities can then be presented in a “fan chart” like the one shown in Figure II.6. In this
figuredifferentlyshadedrangesidentifyalternativeprobabilitiesforpublicdebtofSouthAfricaoverthe
2005Ͳ2009horizon.33Thedarkestrangecorrespondstoanestimatedprobabilityof80percent,witheach
lighterrangereducingthelikelihoodby20percent.Thefigureshowsthat,thefurtheroutwelook,the
greatertheuncertainty.
This presentation resembles the debt paths shown in Figure II.3. but with important differences. In
contrast to Figure II.3., the shocks are not identified. This is a step forward since the shocks under
considerationinFigureII.3.arearbitraryandthereforeunlikelytobewellsuitedtoanyparticularcountry.
The large number of randomly generated shocks underlying the VAR exercise of Figure II.6. avoid this
criticism.StandardizationoftheIMFDSAistheconsequenceofsimplicity,butitscostisarbitrarinessand
therefore limited credibility. Moreover the VAR approach enables one to judge at a glance how likely
someofthedramaticscenariosare.
FigureII.6.ValueͲatͲRiskAnalysis:TheFanChartforSouthAfrica’sPublicDebt
(ExternaldebtasaproportionofGDP)
Source:IMF(2005d).
HowevertherearecostsassociatedwiththeVARprocedure.Tostartwith,itiscomplex.Fewdeveloping
countries are equipped to carry out such estimations and it would stretch even the staff of any
multinationalinstitutiontodealwithalargenumberofcountries.Inordertoprovidereasonablyreliable
estimates, the procedure requires good data, possibly going far back in the past, and few developing
countrieshavesuchdata.
Complexfanchartexercisesmayalsoprovideanillusoryimpressionthatuncertaintyiswellunderstood.
Whilefanchartsdoprovideusefulinformation,theirprecisionisunknown.Itdependsonthequalityof
thedata,ontheperformanceoftheunderlyingeconometricanalysis,andontherelevanceofhistoryfor
the future. Importantly, perhaps, its “black box” nature goes against the goal of transparency and may
deterpolicyaction.Thustheunavoidablecomplexityandopaquenessmaynotbeworththeeffort.34
33
Thechartdisplaysthepublic,nottheexternaldebt.
It may be ironic that the most advanced countries, which can and often carry out similar exercises, pay limited attention to
thematdecisionͲmakingtime,whileIMF(2003b)andIMFandIDA(2004)callforprioritizingDSAimplicationsinthecaseofthe
lessadvancedcountries.
34
34
CompendiumonDebtSustainabilityandDevelopment
ReactionFunctions
AkeymessagefromFigureII.5.isthatpoliciesdomatter.Adequatepolicyreactionscandealwithshocks,
andthesereactionsneednotbedrastic,giventimeandcommitment.Thisobservationleadstotheidea
thatdebtsustainabilitycanbeachievedthroughtheadequacyofpolicyreactionstoshocks.
Viewed this way, debt sustainability can be assessed by observing how a country’s authorities behave.
Thisleadstotheestimationofpolicyreactionfunctions.Intheareaofmonetarypolicy,suchfunctionsare
knownasTaylorreactionfunctionsandhavebecomeroutine.Theapproachhasstartedtobeappliedto
publicdebt,butapparentlynotyettoexternaldebt.
Thekeyquestionhereiswhethertheprimarysurplusissystematicallyraisedwhenthedebtlevelrises.It
ispossibletoestimatethestrengthofthisreactionandtodetermineathresholdbeyondwhichthedebt
accumulation process is stable, and which thus provides an alternative definition of sustainability. IMF
(2003a)presentsanoverviewoftheemergingresultsonpublicdebtreactionfunctions.Whilethisworkis
stillpreliminary,itseemsthatmanycountriesdopassthesustainabilitytest.Thisisgenerallythecasefor
theadvancedeconomies.Fortheemergingmarketcountriesthereactionisadequateatmoderatedebt
levelsbutprobablynotforhighlyͲindebtedcountries.35
Theadvantageofthisapproachisthatitdoesnotrequireassumptionsaboutlikelyshocksandestimation
oftheirrespectiveprobabilities.Nordoesitrequirepassingjudgmentonwhatisanacceptabledebtlevel,
thusavoidingthecontentiousCPIAprocess.Thelimitoftheapproachis,onceagain,aconsequenceofthe
impossibility principle. Debt sustainability is a forwardͲlooking concept: future Governments are not
boundbypastgovernmentbehavior.Evidenceofpastsustainability,orthelackthereof,isnotguarantee
thatfutureGovernmentswillcontinuetoreactinthesameway.Allthatcanbeconcludediswhetherpast
practicesaredeliveringdebtsustainabilityornot.
Itmightbearguedthatsimplylookingattheexistingdebtlevel,orthehistoryofpastdefaults,provides
the same answer in a much simpler way. But this is not correct. The debt may be high currently either
becauseofundisciplinedpastpoliciesorbecauseofadverseshocks.Inordertoassesswhetherhighdebt
is the result of bad luck or of bad policies we need to disentangle these two assumptions. This is what
reactionfunctionsaredesignedtodo.Forinstance,ifthereactionfunctionindicatesthattheauthorities
havesystematicallyreactedinastabilizingwaytodebtbuildups,wecanconcludethatahighdebtisdue
tobadluck.Inasmuchasbadluckdoesnotstrikeagainandagain,insuchinstancesadebtcanbeassessed
ashighandyetsustainable.ThisisthekeylessonfromFigureII.1.,confirmedbyFigureII.5.
One benefit of this approach is that it focuses attention on the issue of policyͲmaking institutions. The
best guarantee that the authorities will always react to shocks in a debtͲstabilizing way is that their
decisions are embedded in a framework that constrains them to do so. Put differently, sustainability
requiresthatthedebtlevelbesystematicallytreatedasapolicyobjective.Thiscanbedoneinmanyways.
Onesolutionistheadoptionofrules.Thishasbeenthecaseintheareaofmonetarypolicywithrulesfor
moneygrowthorwiththeadoptionofexchangeͲrateanchors.Fiscalruleshavebeenproposedtoensure
publicͲdebtsustainability;theStabilityandGrowthPactoftheEconomicandMonetaryUnioninEuropeis
oneexample.AhugeandinconclusiveliteraturehasexploredthetradeͲoffbetweenrulesanddiscretion.
More recently there has been attention to intermediate solutions in the form of institutions that are
boundbystrictobjectivesbutarealsogivensomeleewaytoexercisediscretion.Muchprogresshasbeen
achieved in the area of monetary policy with the adoption of inflation targeting and independent
monetarypolicycommittees.Similarconsiderationscouldbeappliedtofiscalpolicy(Wyplosz(2005b)).
35
SeealsoWyplosz(2005a)foracomparisonofBrazilandtheOECDcountries.
35
DebtSustainabilityAssessment:theIMFApproachandAlternatives
D. ReviewandConclusions
Alongwithpricestability,lowunemploymentandbalancedgrowth,sustainabilityofexternalandpublic
debt is an essential attribute of good macroeconomic policies.36Along with these other attributes, its
precise definition is elusive and its assessment challenging. This section takes stock of the previous
analysis,developsanumberofprinciples,andadvancessomesuggestions.
1. Assessment
TheoverviewoftheIMF’sDSAandalternativeapproachesyieldsanumberofconclusions.
x
Thevariousapproachestodebtsustainabilitydifferfromoneanotherintwomainrespects:the
definitionofsustainabilityandthewaytheyattempttodealwiththeimpossibilityprinciple.
x
Strict definitions of sustainability start from the solvency condition. These are, sometimes
strengthened,forexample,wheretheIMFaddsanoͲmajorͲadjustmentcondition,andsometimes
relaxed, for example, in the approach of Arrow et al. (2004) to the eventual achievement of
solvency.Weakerdefinitionsfocusonthestationarityofthedebtlevel,usuallyscaledbyGDPor
exports.
x
Implementation of these definitions requires making guesses about the future evolution of key
variables. This gives rise to the impossibility principle: because the future is unknown, any debt
sustainabilityassessmentisonlyvalidwithintheboundsoftheunderlyingguesses.
x
Thereisnowaytoescapeimpossibilityprinciple.Anyapproachisbasedeitheronananalysisof
the past, whose relevance is unknown, or on simulations of what the future might be, which is
unknownbydefinition.Someapproaches–e.g.VARstresstests–combinebothprocedures.
x
TheIMFapproachcombinessimpleandtransparentprocedures(computingdebtpathsbasedon
scenarios) with more elaborate procedures (CPIA) for determining country debt ceilings. The
formerarenecessarilyarbitrary.Thelatterattempttoextractinformationfromthepastthrough
“blackbox”procedures.
x
The impossibility principle does not necessarily provide support for the view that added
complexityallowsformorepreciseassessmentsofsustainability.VARstresstesting,forinstance,
isstateͲofͲtheͲartbut,asfaraspolicyͲmakingisconcerned,thebenefitsareillusory.37
x
Debt sustainability is intimately related to credibility. Credible authorities may adopt a weaker
definition of debt sustainability, eschewing the serious economic and political costs inherent in
strict definitions. Credibility, in turn, emphasizes the role in debt sustainability of policyͲmaking
institutions.
x
PolicyconclusionsdrawnfromDSAexercisesmustbeconsideredwithcare.Sacrificinggrowth–in
theshortandeveninthelongrun–toimpreciselyknownrisksconcerningdebtsustainabilitycan
be very costly. Trading off growth and debt sustainability will always remain more art than
science.
2. Principles
Likeanyguidetopolicymaking,DSAmustbebothsimpleandtransparent.Simplicityisneededtomakeit
possible for every country – especially the less developed countries where growth crucially depends on
36
DebtsustainabilitymaybeseenasapreͲconditionforalltheotherattributes.Itiscertainlynotasufficientcondition.Whether
itisanecessaryconditionremainsopentodebate.
37
ThispointismadebyGoldfajninhisexcellentdiscussionofGarciaandRigobon(2005).
36
CompendiumonDebtSustainabilityandDevelopment
externalborrowing–tobeabletoproduceitsownanalysis.Transparencyisimportantbecausetherange
ofpossiblecausesofdebtdistressisinfinite.Vagueandlowprobabilitythreatsshouldnotinformpolicy
choices.
ThispaperarguesthatDSAoughttorelyonanumberofprinciples.
AccepttheImpossibilityPrinciple
Exceptforobviousbutextremecases,itwill neverbe possible to assert that a debt is unsustainable as
definedbytheIMF.Itsowndefinitionrequirescheckingforsolvency,whichisimpossible.Italsorequires
passing judgment of what is a major adjustment, which involves assessing the willingness and political
abilityoftheGovernmenttocarryoutunpopularpolicies.Thisturncallsforanevaluationoftheimpactof
these policies and of the likely reaction of various segments of society, which depends on the political
regimeand,indemocraticcountries,ontheelectoralcalendar.Finally,exceptforconcessionalloans,DSA
is directly influenced by market sentiment, which can be a source of unpredictable vicious or virtuous
cycles.
ThereisNoTradeͲOffbetweenImpossibilityandSimplicity
Theimpossibilityprinciplerestsontheuncertaintyinherentinpredictingthefuture.Whileeliminatingthis
uncertainty is plainly not an option, a natural temptation is to reduce uncertainty by adopting
sophisticatedtechniquesofassessment.Mostusersareunlikelytograsphowthesetechniquesfunction.
Complexity means opacity. It is an illusion to think that some degree of opacity is worthwhile because
eventhemostsophisticatedinstrumentsdonotavoidtheimpossibilityprinciple.Moreoveropacitymay
also result in the mistaken use of the instruments. By contrast simplicity, which lays bare our lack of
knowledgeofthefuture,isavirtueinitself.
AdoptaWorkableDefinitionofDebtSustainability
Thedifferentdefinitionsandapproachesrevealthatdebtsustainabilityisandwillremainavagueconcept.
Inadditionthereisahugegapbetweentheoryandimplementation.Fromanoperationalviewpoint,two
mainapproachesarepossible:thefirstonerestsondebtthresholdsandthesecondontheevolutionof
debtlevels.Giventheimpossibilityprinciple,ifDSAistoestablishuncontroversialdebtthresholds,atleast
for the time being it should rest on a variant of the second approach which specifies that debt is
sustainableifitisonanonͲincreasingtrend.
EvenBetter,ReplaceDebtSustainabilitywithDebtDistressAvoidance
However,thatadebtlevelbetrendͲdecreasingisneithernecessarynorsufficienttoavoiddebtdistress.
Numerous debt crises have occurred while the debtͲGDP ratio was declining. On the other hand many
countries with long rising debt levels have not run into trouble. In the end, the main reason for paying
attention to the evolution of debts is concern with the possibility of debt distress which, unlike
sustainability,isaclearconcept.
RecognizethatDebtsAreNotNecessarilyBad
Many countries and virtually all Governments are quasiͲpermanently indebted, for both good and bad
reasons.Theviewthatdebtsshouldalwaysbereducedassumesthatalldebtsarebad,whichcannotbe
generally true. Separating good from bad debts is a hopeless undertaking, but it is important to move
awayfromthepresumptionthatdebtaccumulationistobeavoidedunderallcircumstances.38
38
It is puzzling that the IFIs, which routinely emphasize debt reduction, exist mainly to grant loans and are actually the main
creditorstomanydevelopingcountries.
37
DebtSustainabilityAssessment:theIMFApproachandAlternatives
OpenUptheProcessofDeterminingWhetherDebtsAreExcessive
Howshouldoneassesswhetheradebtisexcessive?Whenthedebtistraded,theriskpremiumprovidesa
reasonableguide.ItalsoprovidestherightincentiveforGovernmentstolowertheirdebtlevels.Butwhen
thecountrydoesnothavemarketaccess,thereisnosuchgauge.Yet,thelendersarenaturallyentitledto
have a view. Given that any assessment is bound to be controversial, as the shortcomings of the CPIA
exercisewellillustrate,themultilateralfinancialinstitutionsshouldhaveaproceduretoassessexcessive
indebtednessthatisopenandinvolvesexpertsotherthantheirownstaffs.39
TimeisoftheEssence
Whencurrentdebtlevelsareconsideredexcessive,avoidanceofdebtdistresscallsforadecliningtrend.
Howsteepshouldtherateofdeclinebe?Obviously,bringingthedebtdowntoasafelevelbeforedebt
distressoccursishighlydesirable,butitmaybecostlyintermsofgrowthandemployment.Thereisthusa
tradeͲoffbetweenafastdebtrollͲbackandtheassociatedcosts.ThistradeͲoffmustbecarefullyassessed,
takingdueaccountofeachcountry’sspecificities.
AcceptRisk
Unless current debt levels are considered excessive, keeping them stable is likely to avoid debt distress
under most plausible conditions. To be sure, there will always be exceptional events that will result in
debtdistress.Likealldisastrousevents,thisriskmustbeacceptedasafactoflife.Acompleteguarantee
thatdebtdistresswillneveroccurisillusory,andahighlevelofprotectionisboundtobeverycostly.
3. Suggestions
UsebothApproaches
DSA rests on the debt accumulation process, which is nothing more than an accounting identity:
bt bt 1
(r g )bt 1 primary balancet . At the operational level, one approach is to make
assumptions about the evolution of the primary balance, interest rate r and growth rate g in order to
projectthedebtpath.ThisistheIMF’sDSAapproach.Theotherapproachistoaskwhatshouldhappento
the primary balance to achieve a desirable debt path, given assumptions about the evolution of the
interestraterandgrowthrateg.ThisisthedebtͲstabilizingprimaryaccountapproach.Whichapproachis
moreappropriate?
WhiletheIMFalsocomputesdebtͲstabilizingprimaryaccounts,itspolicyanalysisandgraphicalapparatus
emphasizesdebtpath projections.The presentpaperhasargued thatthe policyinterpretationofdebtͲ
path projections, already subject to the impossibility principle, inevitably leads to a search for debt
thresholds,anothermissionimpossible. Thissuggeststhatitis preferabletorelyonthe debtͲstabilizing
primaryaccountapproach.
Butinfactthereisnoreasontochooseoneovertheother.BothcanbeusedasFigureII.5.shows.Allthat
isneededistheclassicdistinctionbetweentargetsandinstruments.Thedebtpathisatarget.Theprimary
account is the instrument (assuming that the authorities can control it). The policy implications then
follownaturally:DSAbecomesaprocedurethatexplorestheeffectonthedebtpathofvarioussettingsof
the primaryͲaccount instrument. What should this combined approach include and what assumptions
shouldbemade?
39
ThisisinlinewiththerecommendationsoftheexternalpanelthatreviewedtheWorldBank’sCPIA,seeWorldBank(2004).
38
CompendiumonDebtSustainabilityandDevelopment
ParameterSettings
Abaselineprojection,suchascarriedoutbytheIMF,showshowthedebtaccumulationunfoldsonthe
basisofthecurrentlyforeseenprimarybalance,exchange,interestandgrowthrates.Inassociationwith
thisprojectionitisstraightforwardtocomputetheprimarybalancethatwouldstabilizethedebtunder
current conditions, as well as the primary balance required to lower the debtͲtoͲGDP ratio when the
currentlevelisperceivedexcessive.
The baseline is not a forecast, only a statement of where current conditions lead. Currently the IMF
provides two baselines: one that is based on Staff forecasts and one that is based on historical trends.
Neither is adequate. Staff forecasts introduce a degree of arbitrariness. Indeed, IMF (2003b) reports a
tendency for these forecasts to err on the side of optimism. Producing a baseline on the basis of these
forecasts has the merit of consistency, but this comes at the cost of a selfͲinflicted lack of realism. In
addition, the baseline should extend over a horizon which goes beyond the ability to make credible
forecasts.
Historicaltrendshave theadvantageovercurrentvaluesofprovidingsomestability,but thisstabilityis
illusory.TrendprojectionisadequateforGDPgrowth,whichtendstofluctuatearoundareasonablystable
level,withgoodyearsmakingupforbadyears.Butthisistheonlyhistoricaltrendthatshouldbeused.
Theothervariables,theexchangeandinterestratesandtheprimarybalancearepotentiallyvolatileand,
partlyatleast,controlledbytheauthorities.Exploringthedebtimplicationsofthecurrentsettingsismore
informativethanrelyingonhistoricalaveragesthatareoftenoutdated.
Thehorizonshouldbelong,saytenyears.AsarguedinSectionC.3,debtcorrectionsarebestcarriedout
slowly, with small changes in the primary account maintained over a long period. Debt corrections are
inherentlycostly–andmorecostly,thesharpertheyare.Thesamecorrectioncanbeachievedatamuch
lowercostifitistheresultofchangessustainedovermanyyears.
PolicyImplications
Projected over a long horizon, the charts displayed in Figure II.5. provide an adequate framework for
policy discussions. The impossibility principle means that DSA should not lead to automatic policy
conclusions, a fact well recognized in IMF (2003b). For this reason, the more transparent are the
parametersettings,thesmalleristheriskoftheirbeinggivenanoverlyprominentrole.
Thebaselinedebtprojectionimmediatelyindicateswherethedebtisheading.Theprimaryaccountwhich
isdebtͲstabilizing–ordebtͲreducingwhenthedebtisdeemedexcessiveandalowerlongͲruntargetcan
beagreedupon–providesareasonableevaluationofwhatisrequiredtoachievedebtsustainability.
Itisthenpossible,indeeddesirable,toask“whatif?”questions.Inpolicydiscussions,manyquestionscan
beaskedandeasilyanswered. Itisstraightforward toproduce chartssimilar toFigureII.5. showing the
mechanical effects on debt and the stabilizing primary account of changes in interest or growth rates,
whether they are permanent or temporary. But it is important to keep in mind that such an exercise is
purelymechanical,becauseitignoresthelinkagesamongthevariablesthatdrivedebtaccumulation.For
example,changesintheprimarycurrentaccountmayrequireactingontheexchangerate,whichinturn
willnotonlyaffectthedebtͲGDPratiointermsoflocalcurrencybutmayalsoleadtodifferentinterestͲ
ratelevels.
ThishassomeresemblancetotheIMFapproachtoDSAbutdiffersfromitintwoimportantways:1)there
isnopretenseofprovidingforecastsandofassessingtheirlikelihood;and2)theresultsareanswersto
questions asked by policymakers and not simply readyͲmade suggestions that the debt might be in a
dangerzone.
39
DebtSustainabilityAssessment:theIMFApproachandAlternatives
InstitutionsMatter
Credibility is an essential component of DSA, yet it is largely hidden. Credibility affects exchange and
interest rates, and can trigger virtuous or vicious circles. The CPIA is one way of recognizing the
importance of credibility but suffers from its assumption that institutions are given. Policymaking is not
justaboutsettingmacroeconomicvariables.ItshouldalsogiveaprominentroletoshapingpolicyͲmaking
institutions.
A number of countries have taken steps to improve their policyͲmaking institutions in the area of fiscal
policy,withmuchsuccess.BrazilandChile,forinstance,haveadoptedformalproceduresthatbindpolicy
actionswithinaframeworkthatputsdebtsustainabilityattheforefront.MostEastAsiancountrieshave
informallydonethesame,relyingonnormsinsteadoflegalarrangements.MechanicalDSAimplicitly(and,
as just noted, CPIA explicitly) takes institutions as given. It would seem important to downplay the
mechanical part of DSA and, in contrast, to emphasize forcefully the contributions that adequate
institutionscanmaketoavoidingdebtdistress.
40
CompendiumonDebtSustainabilityandDevelopment
References
Abiad A and Ostry JD (2005). Primary Surpluses and Sustainable Debt Levels in Emerging Market
Countries.IMFPolicyDicusssionPaperPDP/05/6.
ArrowKetal.(2004).AreWeConsumingTooMuch?JournalofEconomicPerspectives,18(3):147–172.
Blanchard OJ (2005). Fiscal Dominance and Inflation Targeting: Lessons from Brazil. In: F. Giavazzi, I.
GoldfajnandS.Herrera(2005)InflationTargeting,DebtandtheBrazilianExperience.MITPress:
49–80.
Blanchard OJ et al. (1990). The Sustainability of Fiscal Policy: New Answers to an Old Question. OECD
EconomicStudies15:7–36.
Borensztein E et al. (2004). Sovereign Debt Structure for Crisis Prevention. Occasional Paper, 237,
InternationalMonetaryFund.
BuiterW(1985).AGuidetoPublicDebtsandDeficits.EconomicPolicy,1:13–62.
Bulow J and Rogoff K (1989). A Constant Recontracting Model of Sovereign Debt. Journal of Political
Economy,97(1):155–78.
BurdaandWyplosz(2005).Macroeconomics,AEuropeanText.FourthEditions,OxfordUniversityPress.
Celasun O, Debrun X and Ostry JD (2005). Primary Surplus Behavior and Risks to Fiscal Sustainability in
EmergingMarketCountries:A“FanͲChart”Approach.IMFPolicyDiscussionPapers,PDP/05/6.
Chang R and Velasco A (1999). Liquidity Crises in Emerging Markets: Theory and Policy. In: B. Bernanke
andJ.Rotemberg,eds.NBERMacroeconomicsAnnual,theMITPress.
Cordella T, Ricci LA and RuizͲArranz M (2005). Debt Overhang or Debt Irrelevance? Revisiting the DebtͲ
GrowthLink.IMFWorkingPaper05/223.
Garcia M and Rigobon R (2005). A Risk Management Approach to Emerging Market’s Sovereign Debt
Sustainability with an Application to Brazilian Data. In: F. Giavazzi, I. Goldfajn and S. Herrera
(2005).InflationTargeting,DebtandtheBrazilianExperience,MITPress:163–188.
HemmingR,KellMandSchimmelpfenningA(2003).FiscalVulnerabilityandFinancialCrisesinEmerging
MarketEconomies.OccasionalPaper218,InternationalMonetaryFund.
Hostland D and Karam P (2005). Assessing Debt Sustainability in Emerging Market Economies Using
StochasticSimulationMethods.IMFWorkingPapersWP/05/226.
IMF(2002).AssessingSustainability.PolicyDevelopmentandReviewDepartment,SM/02/166.Available
at:www.imf.org/external/np/pdr/sus/2002/eng/052802.htm.
IMF(2003a).PublicDebtsinEmergingMarkets:IsItTooHigh?WorldEconomicOutlook,September:113–
152.
IMF (2003b). Sustainability Assessments—Review of Application and Methodological Refinements.
Availableat:www.imf.org/external/np/pdr/sustain/2003/061003.pdf.
IMF and IDA (2004). Debt Sustainability in LowͲIncome Countries: Further Considerations on an
Operational
Framework
and
Policy
Implications.
Available
at:
www.imf.org/external/np/pdr/sustain/2004/091004.pdf.
IMF(2005a).InformationNoteonModificationstotheFund’sDebtSustainabilityAssessmentFramework
forMarketAccessCountries.Availableat:www.imf.org/external/np/pp/eng/2005/070105.htm.
41
DebtSustainabilityAssessment:theIMFApproachandAlternatives
IMF(2005b).Colombia:FirstReviewUndertheStandͲByArrangement.CountryReport,05/392.Available
at:www.imf.org/external/pubs/ft/scr/2005/cr05392.pdf.
IMF(2005c).AreFundStaffProjectionsofDebtMoreOptimisticUnderProgramContexts?Availableat:
www.imf.org/external/np/pp/eng/2005/110905.pdf.
IMF(2005d).SouthAfrica:SelectedIssues.CountryReport,05/345.
Jeanne O and Zettelmeyer J (2001). International Bailouts, Moral Hazard and Conditionality. Economic
Policy,33:407–32.
Kraay A and Nehru V (2003). When Is External Debt Sustainable? Available at: www.imf.org/external/
np/res/seminars/2003/lic/pdf/kn.pdf.
Manasse P, Roubini N and Schimmelpfenning A (2003) Predicting Sovereign Debt Crises. Working Paper
03/221,InternationalMonetaryFund.
MissaleA(2000).PublicDebtManagement.OxfordUniversityPress.
RodrikD(1998).WhoNeedsCapitalͲAccountConvertibility?In:S.Fischerandothers(1998).Shouldthe
IMF Pursue CapitalͲAccount Convertibility? Essays in International Finance, 207, International
FinanceSection,DepartmentofEconomics,PrincetonUniversity,May.
World Bank (2004). Country Policy and Institutional Assessments: An External Panel Review Panel
Recommendations and Management FollowͲup. Available at: www.siteresources.worldbank.org/
IDA/Resources/CPIAExpPanRepSecM2004Ͳ0304.pdf.
WyploszC(2005a).InstitutionsforDebtStabilityinBrazil.In:F.Giavazzi,I.GoldfajnandS.Herrera(2005).
InflationTargeting,DebtandtheBrazilianExperience.MITPress:193–222.
WyploszC(2005b).FiscalPolicy:InstitutionsVersusRules.NationalInstituteEconomicReview,191:70–
84.
42
CompendiumonDebtSustainabilityandDevelopment
Annex:AnAlternativeDebtSustainabilityCondition
Section B.1. suggests an alternative definition of debt sustainability. This appendix briefly characterizes
thelinkofthisdefinitionwiththesolvencycondition.
LetBtbethedebtoutstandingatbeginningofperiodt,Rt,t+ithediscountfactorbetweenperiodstand
t+i,andSttheprimarybudgetbalance.Thedebtaccumulationprocessimplies:
n
Bt ¦ Rt ,t i St i
Rt ,t n Bt n
i 0
Solvencyisdefinedbytheusualtransversalitycondition:
lim Rt ,t n Bt n d 0
nof
whichcanberewrittenas:
f
Bt d ¦ Rt ,t i S t i
i 0
Thenetworthoftheentity(Government,country)is:
f
Vt
¦R
t ,t i
S t i Bt
andthereforesolvencysimplyrequiresVtш0.
i 0
Thealternativesustainabilitycondition(basedonArrowetal.(2004))isthatVtbetrendͲincreasing,i.e.:
Vt+n–Vtш0formostn.
Thus,itmaybethatinitiallyVtч0butthesustainabilityconditionimpliesthatthereexistsahorizonN
suchthat,foralln>N,Vt+nш0.
43
CompendiumonDebtSustainabilityandDevelopment
CHAPTERIII
THEMECHANICSOF
DEBTSUSTAINABILITYANALYSIS
AnhͲNgaTranͲNguyenandAlbiTola
(UNCTAD)
A. Introduction
Debtsustainabilityhasbecomeakeyissueinthediscussionondebtindifferentfora.Theconceptofdebt
sustainabilityisdifficulttodefineinpracticeanddependsonanswerstoanumberofquestions.Whenisa
certainlevelofdebttoohighandunsustainable?Howimportantisdebtsustainabilityfordevelopment?
Shoulditbeamajorobjectiveandshouldeconomicpoliciesbeadjustedaccordingly?Alternativelyshould
developmentbeanobjectivewhichcanoverridedebtsustainability?Ifso,whatdoesthismeaninterms
of policies? Or on the contrary should there be acknowledgement that debt sustainability and
developmentaretoointerconnectedforsuchaseparationofobjectivestobefeasible?
Answers to these questions are not straightforward. The literature on debt sustainability offers a wide
rangeofanalysesoftheissuefromdifferentperspectives.Thebasicelementsofthedifferentapproaches
are highlighted in this paper, with particular reference to their practicality and relevance. Four main
approachesarecovered:
x
x
x
x
Presentvalueanalysis;
Financinggapsanalysis;
Theindicatorsofdebtcrisis;and
AdevelopmentpolicyͲbasedframework.
The analytical underpinnings of these approaches are linked, but the focus on particular aspects of the
debt problems is typically different. The last approach attempts to address various concerns within a
comprehensivemacroeconomicandpolicyͲbasedmodel.
45
TheMechanicsofDebtSustainabilityAnalysis
B. DebtIndicatorsandEarlyWarningofCrisis
In the aftermath of recent debt crises economists have attempted to identify early warning indicators
whichwouldsignalinadvancetheprobabilitythatacurrencyordebtcrisiswilloccur.Suchindicatorsare
intendedtofacilitatecorrectivemeasures.Onesuchindicatorisanunsustainablecurrentaccountdeficit
associatedwithlowGDPgrowth.Thattypicallycharacterizeddebtcrisesinthe1980s.
Accordingtothemodelsofdebtdynamicsdiscussedbelow,acountry’sdebtaccumulationissustainable
as long as the growth of GDP is greater than the real interest rate. However, access to credit can be
abruptlystoppedifcreditorsbecomeworriedaboutincreasingdebt.Inaddition,highgrowthratesofGDP
accompaniedbycapitalinflows,cancausetherealexchangeratetoappreciate.Thiscanleadtolossof
competitivenessandfurtherdeteriorationofthecurrentaccount.
Anotherimportantindicatorofdebtsustainabilityisthesizeofexports,whichenhancethedebtor’sability
to generate the foreignͲcurrency revenues needed to service debt. However, a large export sector can
makeacountryvulnerabletotermsͲofͲtradeandforeigndemandshocks.Forexample,asitisshownby
Corsettietal.(1998),theexportsofSouthEastAsiancountriesweresubjecttonegativetermsͲofͲtrade
shocksin1996priortothefinancialcrisisof1997followingthefallinthepricesofsemiconductorsand
other exports, the increasing competition of cheaper goods from China, and decreasing demand from
Japanduetolongstagnationofitseconomy.
Radelet and Sachs (1998) point out that so long as investments financed by external borrowing are
channeled to productive activities, they can contribute to growth. However, overͲreliance on external
financingcanbethesourceofmacroeconomicinstabilityifitinducesanappreciationoftherealexchange
rate.PriortotheAsianfinancialcrisisasignificantpartofcapitalinflowswasdirectedtosectorsproducing
nonͲtradedgoodsandtorealestate,neitherofwhichgenerateforeignexchangerevenue.
RadeletandSachs(1998)alsoidentifytheincreasingfragilityoffinancialsectorasasignofanupcoming
debtcrisis.AscredittotheprivatesectorgrewrapidlypriortotheAsianfinancialcrisis,banksincreased
theirrecoursetoshortͲtermforeignborrowing.Bankswerethusexposedintwoways:(1)toexchangeͲ
rate risk since they were borrowing in foreign currency for onͲlending in domestic currency; and (2) to
maturitymismatchessincetheyborrowedshortandlentlong.
Traditionalindicatorssuchasaslowingofexportgrowth,adeterioratingcurrentͲaccount deficitandan
overvaluedexchangeratearenotnecessarilyareliablesourceofwarnings.Somerecentresearchonearly
warningsystems(EWS),describedinBergetal(2004),hasbeenmodelͲbased.Theaimofthisresearch,of
which examples are Kaminsky, Lizondo and Rienhart (KLR) (1998) and work of the IMF Developing
CountriesStudiesDivision(DCSD),hasbeentodevelop“theexchangemarketpressureindex”.Ifthevalue
ofthisindexexceedsitsmeanbymorethanthreestandarddeviations,thenthecurrencyisexposedtoa
serious risk of devaluation. Although the models developed for this purpose use different econometric
techniques,theyaredesignedtoestimatetheprobabilityofan“event”onthebasisofvariousindicators.
The indicators of the DCSD model include realͲexchange devaluation, changes in foreign reserves, the
ratioofshortͲtermdebttoforeignexchangereserves.TheKLRmodelusesalsoincludesindicatorssuchas
thedomesticcreditgrowth,changeinmoneymultiplier,andtheratioofforeignͲexchangereservestoM2.
TheIMFalsoexaminedtheratiooftheshortͲtermdebttoreservesasawarningindicatorbutfoundthatit
performedbetterasanindicatorofliquiditythanofexternalsolvencyproblems.Forexample,priortoits
debt crisis of January 2002 Argentina had a better indicator for this ratio than Turkey in 1999Ͳ2000.
However, its ratio of debtͲservice to current receipts was twice as large as that of Turkey and more
successfullyanticipateditssubsequentsolvencyproblems.
46
CompendiumonDebtSustainabilityandDevelopment
TheIMFhasdrawnattentiontothepossibilityofincludingmodelsofcountryriskmodelsandsovereignͲ
risk indices in the EWS, pointing to an econometric model of Eichengreen and Moody (2000) in this
context. This model is used to estimate the determinants of emergingͲmarket debt spreads and to
forecast currency problems on this basis. Inter alia the authors found that the debtͲserviceͲtoͲexports
ratiowashighlycorrelatedwiththelevelofspreads.
OtherstudiesofEWSmodels(reviewedinBergetal.,2004)pointtoamixedperformanceinpredicting
crises. While they tend to perform better than nonͲmodelͲbased indicators such as sovereign spreads,
sovereign ratings or market surveys, they have often signaled false alarms. On the other hand, these
modelshavecoveredalargenumberofmacroeconomicandfinancialvariablesformanycountriesandfor
long periods. They have highlighted invaluable information about the relative vulnerability of different
countries(BergandPatillo,2000).
Aliquidityproblemarisesbecauseofabunchingataparticulartimeofdebtobligationswhichcannotbe
fullyservicedonthebasisofexistingrevenues.However,thedebtcanberepaidifrecoursetoexternal
fundingisavailableonatemporarybasisorifthedebtisrestructuredsothatdebtobligationsarebetter
matchedtothedebtor’srevenues.Insolvency,ontheotherhand,denotestheinabilityofthedebtorto
pay in full his debt obligations owing to more structural problems which cannot be solved simply by a
rearrangementofpaymentsdue.
In theory a country is solvent even if it runs a huge current account deficit as long as it is capable of
producingcurrentaccountsurplusesinthefutureand,asdescribedbelow,GDPincreasesatarateabove
the rate of interest. In practice this condition has been shown to be unrealistic. According to Roubini
(2001)themainproblemwithsuchdebtdynamicslieswiththefactthataGovernmentcannotcredibly
committoruntherequiredfiscalandbalanceͲofͲpaymentssurplusesinthefuture.Thisauthorprefersthe
simpleratioofforeigndebttoGDPratioasanindicatorofbothsolvencyandsustainability.Ifthisratiois
increasing,thenalargertradesurpluswillberequiredtoachievesolvencyandthismustbeincreasedstill
furtheriftherealinterestrateisbiggerthanGDPgrowth.
Other studies have attempted to determine empirically the thresholds beyond which a “debt crisis” (or
solvency problems) will develop. The debt crisis is defined as an event in which there are arrears of
principal or interest on external obligations, or in which the country reschedules or restructures its
external debt. These studies (IMF, 2002a) have found that a debt crisis occurs typically at debt to GDP
ratios below 50Ͳ60 per cent. The “transfer problem” implies that a country must also generate foreignͲ
exchange receipts through an export surplus sufficiently large to service its debt. This implies that the
exportsͲtoͲGDP ratio (or another indicator for the same purpose such as the debtͲ serviceͲtoͲexports
ratio)isrelevant.ThesurveyofIMF(2002a)showedthatfordebtcrisesatexportͲtoͲGDPratiosbelow20
percentthreeͲquartersoccurredatdebtͲtoͲGDPratiosoflessthan60percentofGDP,whileforcrisesat
exportͲGDPratiosofbetween20and40percentthecorrespondingdebtͲtoͲGDPratiowasbetween60
and80percent.
Themainindicatorsusedtomeasureliquidityaretheratioofforeignexchangereservestoimports,the
ratio of foreign exchange reserves to shortͲterm debt, the share of shortͲterm debt in total debt, and
interestpaymentstoforeignexchangereserves.Althoughacountrycanbesolvent,itrisksadebtdefault
orcrisisifitdoesnothaveenoughliquiditytoserviceitsshortͲtermdebt,perhapsowingtoabunchingof
loansmaturingataparticulartime.ThismighthappenduringafinancialpanicwhereshortͲtermcreditors
decide not to renew their loans or ask for a repayment. Thus debtor countries should monitor their
maturity structure to make sure that they have enough shortͲterm assets to cover their shortͲterm
liabilities.
Assessingthesolvencyandtheliquidityofacountryisadynamicprocesswhichshouldtakeaccountof
differentshocksthatmightreduceitscapacitytoserviceitsdebt.Theseshockscouldbedropsinexport
47
TheMechanicsofDebtSustainabilityAnalysis
earnings,unfavorablemovementsinthetermsoftrade,orincreasesinoilprices.Roubini(2001)draws
theattentiontowhathecalls“selfͲfulfillingsolvencytraps”.Thisiswhereahighexternalorpublicdebt–
but not necessarily at a level entailing insolvency Ͳ will be considered by the markets as a situation
sufficientlyriskyforthespreadsoninterestratestoincrease.Asaresult,thedebtorcountrywillpaymore
interestandwillaccumulatedebtmorerapidly.Thisisanothercasewherealiquiditycrisismayturnintoa
solvencycrisis.
Liquidityandsolvencyconceptsareusuallyinterlinkedinpracticesothatitisverydifficulttodistinguish
between the two. It is generally difficult to determine whether incapacity to pay is temporary or
permanent.Moreoverproblemsofilliquiditycanturnintoinsolvencyiftheyarenottackledintime.
C. ThePresentValueofFutureIncome
How should debt sustainability be defined? Since debt is generally incurred to finance investment, one
approachinvolvesananalogywiththemicroͲlevelanalysisofinvestmentprojects.Heredebtsustainability
isdeterminedbytheconditionthatthepresentvalueofthefutureincomestream(netofexpenditure)
derivedfrominvestmentprojectsshouldbeatleastequaltothenominalvalueofdebtusedtofinance
them.Therearetwowaystoobtainthepresentvalueoffutureincomestreams.Thepresentvaluecanbe
computed by discounting the streams of income by the interest rate of the debt. Another way is to
compute the expected internal rate of return, defined as the rate of discount applied to future income
whichwouldmakestheexpectedpresentvalueofincomeequaltothenominalvalueofdebt.Inthiscase
debtsustainabilityisassurediftheexpectedinternalrateofreturnisatleastequaltotheinterestrateof
thedebt.
Application of the present value approach to debt sustainability to the debt of a country nonetheless
involvesaddressingmanychallengesatonce:
x
Theinvestmentprojectsfinancedbydebtarethesumofprivateandpublicprojectswithdifferent
ratesofreturn.Thelattermayincludeprojectsaddressingsocialratherthaneconomicneeds.
x
The income generated by investment can be denominated in foreign currency (in the case of
exports)orinlocalcurrency.
x
External debt sometimes is used to smooth out cyclical changes in consumption (which do not
generateanyincome), debtbeingincurredduring timesofdepressedgrowth andrepaidduring
highͲgrowthperiods.
x
Debt can carry concessional or market interest rates and have different maturities and grace
periods,whichmakethechoiceoftheappropriatediscountrateacomplicatedissue.
Notwithstanding these difficulties, the present value approach has been widely applied to developing
countries’ debt. The IMF implicitly uses this concept of the present value in their approach to debt
sustainability (see IMF, 2002a). It defines debt sustainability as “a situation in which a borrower is
expectedtobeabletocontinueservicingitsdebtswithoutanunrealisticallylargefuturecorrectiontothe
balanceofincomeandexpenditure”.40Underlyingthisdefinitionofsustainabilityareconceptsofsolvency
andliquidity,whicharedefinedonthebasisofthepresentvalueapproach.
40
Sustainability, according to the IMF, rules out: (1) debt restructuring; (2) Ponzi games where the borrower indefinitely
accumulates debt faster than its capacity to service debt is growing); (3) moral hazard whereby the borrower lives beyond its
meansbyaccumulatingdebtintheknowledgethatamajordebtservicereductionwilleventuallybeneeded.
48
CompendiumonDebtSustainabilityandDevelopment
Solvencyissecuredwhenthepresentvalueofadebtor’scurrentandfutureprimaryexpenditure(E)isno
greaterthanthepresentvalueofitscurrentandfuturepathofincome(Y),netofanyinitialindebtedness
(D):
f
¦
i 0
f
Et i
d ¦(
i
– (1 r
t j
)
j 1
i 0
Yt i
(1 rt ) Dt 1 )
i
– (1 r
t j
)
j 1
.
A liquidity crisis occurs when a debtor’s liquid assets and available financing are insufficient to meet or
rollͲover its maturing liabilities, regardless of whether the solvency condition is met. Sustainability is
reachedwhenadebtor’sliabilitypositionsatisfiesthepresentvaluebudgetconstraintwithouttheneed
foramajorcorrectioninthebalanceofincomeandexpenditure,givenitscostsoffinancing.
Thepracticalityofsuchanotionofsolvencyisquestionable.First,thetimehorizonisinfinitelylong,which
makes planning or time framing of Government’s budget impossible. Furthermore, the appropriate
discountratemuststillbechosen.Differentchoiceswillresultindifferentestimatesofpresentvalue.
The present value approach is also applied to the analysis of the primary balance of the Government
budget or the balance of payments on current account excluding payments. Based on the national
accounting identities, the balance, Pt, that is the difference between revenue and expenditure or the
currentͲaccountbalance,inbothcasesexcludinginterestpayments,isequaltothechangeindebtDtplus
interestpayments:
Dt
(1 r ) Dt 1 Pt 1 Assumingaconstantinterestrate,recursiveapplicationofthisformulagives:
i
Dt ¦
j 1
D
Pt j
(1 r )
j 1
t i 1
i 1
(1 r )
Asitendstoinfinity,thetermontherightͲhandsidetendstozero.Thismeansthatthepresentvalueof
debt in the indefinite future converges to zero and reflects the unwillingness of lenders to allow the
debtor perpetually to pay its interest obligations by borrowing more. Hence, for i sufficiently
f
Dt
large,
ensue.
¦
j 0
Pt j
(1 r ) j 1 ,whichimpliesthatalldebtmusteventuallybepaidbackforsustainabilityto
Thisobservationleadstotheinterestingconclusionthattheabilitytogenerateasurplusisaprecondition
forlongͲtermdebtsustainability.Inthecaseofthefiscalbalancethisisaprimarysurplus.Inthecaseof
thebalanceofpaymentsoncurrentaccountthedebtorcountrymusteventuallyexportandearnenough
foreignexchangethroughatradesurplustorepaydebt(thesocalledtransferproblem).
AdifferentversionofthepresentvalueapproachistheIMFNetPresentValueofDebt.InthecaseoflowͲ
income countries thatreceiveloansonconcessionalterms, creditorsapplythe conceptof “Netpresent
Value”41(NPV)ofloanstocalculatethe“realburden”ofdebt,usingaformulaforwhichthedebtservice
onconcessionalloansisdiscountedatthemarketrateofinterestinordertoreflectthe“trueopportunity
41
Theterm«NetPresentValue»(NPV)isusedhereinawaythatdivergesfromstandardcorporatefinancejargonforwhichNPV
isthedifferencebetweenthepresentvaluesoftheincomeandcostofaninvestmentproject.Infact,amorecorrecttermhere
wouldbethe“presentvalue”(PV).
49
TheMechanicsofDebtSustainabilityAnalysis
cost” of the loans. This NPV concept was first applied in the context of Paris Club rescheduling on
concessionaltermsandtheninthecontextoftheInitiativefortheHeavilyIndebtedPoorCountries(HIPC).
The IMF applies it to the lowͲincome countries in its debt sustainability analysis framework (IMF, 2003,
May).
TherationalebehindtheuseofthisversionofNPVisthatdiscountingthestreamoffuturedebtͲservice
payments by an appropriate market interest rate provides an aggregate measure of the effective debtͲ
serviceburdenimpliedbyagivendebtstock.However,asseenatthebeginningofthissection,inorderto
gaugewhetherthedebtorcanserviceitsdebt,thedebtor’sfuturestreamofincomehastobediscounted
either at the rate of interest on the loan which has been contracted or at the rate of return on the
investmentproject.
ThisNPVapproachdoesnot,therefore,reflectthedebtor’scapacitytopay.Ifatall,thiscapacityistaken
intoaccountwhenconcessionaltermsaregrantedtolowͲincomecountries.TheversionofNPVisrathera
conceptofinteresttocreditors,usedtomeasurethegrantelementofaid.Itcanalsobeusefulasameans
to determine comparable opportunity costs for donors for burden sharing purposes in debtͲrelief
operations.
D. TheFinancingGap
Thefinancinggapanalyticalapproachesarebasedonthethreenationalaccountidentitiesrelatedtothe
balanceofpayments,domesticinvestmentandsavings,andgovernmentbudget,whichlinkexternaldebt
withdifferentfinancinggaps:thecurrentaccountdeficit,ortheshortageofdomesticsavingscompared
todomesticinvestments,orthegovernmentbudgetdeficit(partofwhichcannotbefinancedbydomestic
publicdebt).
Assuch,thefinancinggapsarejustaccountingidentities.Economistshaveexpandedtheseidentitiesand
added behavioral equations in order to project financing gaps or to analyse the dynamic stability
conditions of external debt. The use of gap models to project financing gaps and aid requirements has
been widely accepted by international financial institutions. However this approach has been much
criticized, notably because of exͲpost inaccuracies of these projections and because of the instability of
the key variable, the incremental capitalͲoutput ratio (ICOR),42used to determine the growth path of
income.
Despite these shortcomings, financing gaps as determined by accounting identities remain useful
indicators for policymakers in the short run for the analysis of the origins of external debt and of the
burdenofdebtservicingongovernmentbudget.
1. DebtandNationalAccountingIdentities
Threenationalaccountingidentitiesrelatedtothebalanceofpayments,domesticinvestmentandsavings,
andthegovernmentbudgetshowthelinksofexternaldebttodifferentfinancinggaps.
(1)ThebalanceͲofͲpaymentsidentityisstatedasfollows:
Et ( DtE DtE1 )
Et Pt* M t Pt X t Et > i * DtE1 FDI t ( Rt Rt 1 )
@
42
SeeEasterly(1997).
50
CompendiumonDebtSustainabilityandDevelopment
where
DtE istheexternaldebtstockindomesticcurrency, Et thenominalexchangerate(domesticper
unit of foreign currency),
Pt * foreign prices, M t import volume, Pt domestic prices, X t export volume,
it* theforeigninterestrate, FDI t thenetflowofforeigndirectinvestment(includinglongtermportfolio
R
equityinvestment),and t thestockofforeignexchangereserves.
This identity shows that new debtͲcreating capital inflows fill that part of the currentͲ account deficit
(
Et Pt * M t Pt X t Et i * DtE1 )notfinancedbyFDIandthechangeinreserves.
(2)TheinvestmentͲsavingsidentityisderivedfromthenationalincomeidentity:
Yt
*
Ct S t
C t I t ( Pt X t Et Pt M t )
Where C t isconsumption, S t issavings,and I t isinvestment.
sothat
St I t
Pt X t Et Pt* M t CombiningthebalanceͲofͲpaymentsandinvestmentͲsavingsidentitiesgivesanotherequationforexternal
debt:
E (D E D E )
(I S ) E
> i* D E
FDI 'R
@
t
t
t 1
t
t
t
t
t 1
t
t
ThisidentityshowsthatnewdebtͲcreatingcapitalinflowsfillthegapbetweendomesticinvestmentplus
InterestspaymentsanddomesticsavingswhichnotfinancedbyFDIandthechangeinreserves.
' DG t ' D tD E t ' DG tE it D tD1 E t i t* DG tE1 Pt (3)TheGovernmentbudgetidentityisstatedas:
DD
where t is Government’s domestic debt, Gt 1 is government consumption expenditure and Pt is the
primarysurplus.
Pt Tt ETt (Gt IGt DTt ) T
ET
G
Where t isgovernmentrevenue, t externaltransferstoGovernment, t government consumption,
IGt governmentcapitalformation,and DTt domestictransfersandsubsidies.
This identity above shows that new debtͲcreating capital inflows fill the gap between the total budget
deficitandtheamountofthedeficitthatcanbefinancedbyissuingdomesticdebt.
2. TheStabilityofExternalDebtDynamics
TheroleofforeigncapitalinthedevelopmentprocesshasfirstbeenanalysedinthecontextofthetwoͲ
gap models, whereby debt or capital inflow helps in filling the resource gap resulting from of shortage
foreignexchangeearnings(derivedfromidentity(1))orofsavings(derivedfromidentity(2)).Asdebtis
assumedtocontributetogrowth,overtimetheresourcegapisgraduallynarrowedandtowardstheend
ofthecycle,thedebtorcountrywillhaveenoughsurplusesinresourcestorepayitsdebt.
51
TheMechanicsofDebtSustainabilityAnalysis
This virtuous cycle of debt is characterized by a stable dynamic path of debt accumulation, which
decreases over time. Assuming unchanged exchange rate and expressing the variables in the same
currency(withE=1),thedynamicinteractionsbetweendebtononehandandincomegrowthandexport
growthontheotherhand,woulddeterminetheconditionsunderwhichthedebtaccumulationprocess
canbecontrolled.Thebasicmodelshownbelowillustratessuchdebtdynamics.43
Thefirstequationdescribesdebtaccumulation(D)tofinancetheresourcegap,whichisdefinedhereas
thedifferencebetweenimportsandexports(MͲX)andtheinterestpaymentsonearlierdebt(iD).
dD
(1) dt
M X iD
Theincreaseinoutputissimplytheproductofinvestment(I)andtheinverseoftheincrementalcapitalͲ
outputratio(1/k):
dY
(2) dt
1
I
k Thesavingsgap(thedifferencebetweeninvestmentIanddomesticsavingsS)andtheforeignexchange
gap(thedifferencebetweenimportsMandexportsX)areequalexpost:
I S
MX
dD
iD
dt
(3)
Other specifications can be added to this basic model. For example, imports can be broken down into
importsofcapitalgoodsandimportsofconsumptiongoods.Exportscanbespecifiedtogrowexogenously
ortodependontheinvestmentrate.
Considerthesavingsgapfirst.
Thesavingsfunctionsimplystatesthatsavingsaretheproductofthemarginalpropensitytosave(s)and
domesticproduct:
(4) S=sY
Usingthesavingsgap(IͲS)andrearrangingtheterms,debtaccumulation(expressedasaratioofdebtto
output)canbesetinfunctionoftheoutputgrowthrate(g),investment(gk)andsavings(s)rates:
d (D / Y )
dt
(5)
D
(i g ) ( gk s )
Y
Equation5statesthatthechangeintheratioofdebttodomesticproductdependsonthecurrentlevelof
this ratio, as well as the difference between the interest rate and the growth rate, and the difference
betweentheinvestmentrateandsavingsrate.Thesolutiontothisdifferentialequationwilldeterminethe
conditionofstabilityofthedebtaccumulationprocess:
(6)
D
Y
gk s § D0 gk s · (i g ) t
¸e
¨
g i ¨© Y0
g i ¸¹
whereD0/Y0isthevalueofD/Yattimet=0.
43
Notethatthemodelisinrealterms,i.eallvariablesaredividedbytheGDPdeflator.
52
CompendiumonDebtSustainabilityandDevelopment
Fromequation6itcanbeseenthatthedynamicpathwillbestableandconvergetoequilibriumonlyif
the rate of GDP growth is higher than the interest rate. If the interest rate is higher than the rate of
growth,debtwillgrowataneverͲincreasingrateandwillreachanexplodinglevel.
Withg>i,theasymptoticvalueofthedebtͲincomeratiois:
LimD/Yї[(gk–s)/(g–i)]
tїь
The maximum amount of D/Y (or the maximum level of sustainable debtͲtoͲGDP) is determined by the
differencebetweenthemarginalinvestmentrate(gk)andsavingsrate(s).
Alternatively, using the trade gap (M Ͳ X) and rearranging the terms, debt accumulation can be set in
functionoftheinterestrate(i)andtherateofgrowthofexports(x).
d (D / X )
dt
(7)
D
MX
(i x) X
X whereD/XisthedebtͲtoͲexportratio.
Equation7statesthatthechangeintheratioofdebttoexportisthesumofthecurrentlevelofthisratio
multipliedbythedifferencebetweentheinterestrateandtherateofgrowthofexports,andthecurrent
trade gap divided by the current level of exports. The solution to this differential equation gives the
followingtimepathofthedebtͲexportratio:
(8)
D
X
1 §M X
¨
xi© X
1 §M X
· ª D0
¸«
¨
¹ ¬X0 x i © X
·º ( i x ) t
¸» e
¹¼
whereD0/X0=debtͲexportratioattimet=o.
Here,thestabilityconditionisthattheinterestrateshouldbelowerthantherateofgrowthofexports(i<
x).ThetrajectoryofD/Xcanbeexponentiallyascendingordescending,dependingonwhetherxisgreater
orsmallerthani.
Withi<x,D/Xwillconvergetoanequilibriumwiththefollowinglimit:
LimD/Xї[1/(xͲi)][(M–X)/X]
tїь
AsthesizeofthedebtͲtoͲexportratiodependsonthetradegap,thislimittosustainabilitycanbevery
highifthetradegapisverylarge.
Finally, taking into consideration the fiscal gap, as derived from the Government budget identity (and
assumingthattheGovernmentincursonlyexternaldebtanddoesnothavedomesticallycontracteddebt),
applyingthesameprocedureasaboveyieldsthefollowingequation:
d (D / Y )
dt
(9)
D
PB
(i g ) Y
Y 53
TheMechanicsofDebtSustainabilityAnalysis
PBistheprimarydeficitoftheGovernmentbudgetandgistherateofgrowthoftheeconomy.Thistime
thestabilityconditionisthatgexceedsi.
The debt dynamics depicted above imply that as long as GDP grows faster than real interest rate, a
countryissolventeveniftheratioofforeigndebttoGDPkeepsgrowing.Thesameappliesforfiscaldebt.
Likewise, as long as export grows faster than real interest rate, a country is solvent even if the ratio of
foreigndebttoexportsisincreasing.
In practice, these concepts have shown to be not very realistic. In particular, even though the stability
conditionscanbesatisfied,theasymptoticvaluesofdebt,aswellasthetradegapandfiscaldeficitwhich
determinethesevalues,canbeveryhigh.Nothinginthemodelsignalsthecapacityofdebtorstorepay
thefullamountofthisdebt.Unlesslendersarewillingtolendoveraprolongedperiodwhileknowingthat
debtisnotgoingtoberepaid,debtorsneedtorunfiscalsurplusortradesurplusinordertoreducetheir
debt.
3. TheIMFandWorldBankDSAFramework
ThethreefinancingͲgapidentitiesarealsousedasthebasisoftheframeworkemployedbytheIMFand
theWorldBankfordebtsustainabilityassessment(DSA).Theframeworkconsistsoftwotemplates,one
relatedtotheexternaldebtsustainabilityandtheothertothefiscalsustainabilityofthepublicsector.
The external debt sustainability template analyses the debt incurred externally by domestic residents
(boththepublicandtheprivatesector).UsingthebalanceͲofͲpaymentsidentityasthepointofdeparture
and rearranging the terms derived from the algebraic transformations of this identity, the following
equationcanbeusedtodecomposechangesinexternaldebt:
d t 1 d t
1
(r g U (1 g ) HD (1 r ))d t tbt 1
(1 g U gU )
where d isthedebtͲtoͲGDPratio, D istheshareofdomesticͲcurrencydebtintotalexternaldebt, H is
the change in the exchange rate expressed in US$ per local currency unit, tb is the currentͲaccount
balance excluding interest payments in per cent of GDP, U is the change in the domestic GDP deflator
expressedinUS$, g istherealGDPgrowthrate,and r istheinterestrate.
Thisequationallowsthefollowingdecompositionoftheseparatechannelsaffectingtheevolutionofthe
debtͲGDPratio:
x
thenonͲinterestcurrentaccountdeficit,tb;
x
r
dt
therealinterestͲrate, (1 g U gU ; g
dt
therealgrowthrate, (1 g U gU ) ;
x
U (1 g ) HD (1 r )
dt
.
priceandexchangeͲratechanges, (1 g U gU )
x
Thethreelasteffectscanbecharacterizedasthecontributionsofendogenousdebtdynamics.
54
CompendiumonDebtSustainabilityandDevelopment
The fiscal sustainability template analyses the behavior of the debtͲtoͲGDP ratio with all variables
expressed in domestic currency. Using the Government budget identity as the point of departure and
rearrangingthetermsderivedfromalgebraictransformationsofthisidentity,thefollowingequationcan
beusedtodecomposechangesinpublicdebt:
dt 1 dt
1
(rˆ S (1 g ) g HD (1 rˆ))dt pbt 1
(1 g S gS )
where d is the debtͲtoͲGDP ratio, pb is the primary balance, r̂ is a weighted average of domestic and
foreign interest rates, D the share of foreignͲcurrency denominated public debt, S the change in the
domesticGDPdeflator,andgtherealGDPgrowthrate.Changesintheexchangerate(localcurrencyper
U.S.dollar)aredenotedby H ,with H ! 0 indicatingadepreciationofthelocalcurrency.
ThisequationallowsthefollowingdecompositionofthechannelsaffectingtheevolutionofthedebtͲGDP
ratio:
x
theprimarydeficit pb ,
x
r S (1 g )
dt
g
g
(
1
S
S
)
;
therealinterestrate,
g
dt
g
g
(
1
S
S
)
;
therealgrowthrate,
HD (1 r )
dt
g
g
(
1
S
S
.
exchangeͲratedepreciation,
x
x
Again,thethreelasteffectscanbecharacterizedasthecontributionsofendogenousdebtdynamics.
Byidentifyingdifferentfactorscontributingtothegrowthofthedebtratios,thetemplatesindicatethe
channelsthroughwhichdebtcanbereduced,ifthelevelistoohigh.Notethatinthetemplatesthereal
GDP growth rate and export growth rate are exogenously given so that there is only a oneͲway
relationshipfromGDPandexportgrowthtodebtaccumulation.Theshortcoming,therefore,isthelackof
areverserelationshipfromGDPanddebtgrowthtogrowthofexports,whichlimitsuseofthetemplates
forexaminingcertainscenariosfordebt,GDPandexportgrowth.
Furthermore,theIMFandWorldBankDSAframeworkleavesopenthequestionoftheappropriatelevel
of debt around which debt should be stabilized. An example of ad hoc character of the application of
IMF/WorldBankDSAframeworkindeterminingasustainablelevelofdebtisthesettingoftheindicative
thresholdsfortheratiosofNPVofdebttoGDP,toexportsandtogovernmentrevenueatdifferentlevels
forlowͲincomecountriesaccordingtothepolicyperformanceofborrowingcountriesasmeasuredbythe
CountryPolicyandInstitutionalAssessment(CPIA)Index.Thefollowingtableillustratesthesethresholds.
55
TheMechanicsofDebtSustainabilityAnalysis
TableIII.1.IndicativeExternalDebtBurdenIndicators1/
NPVofdebtinpercentof:
Exports
GDP
Revenue3/
Debtserviceinpercentof:
Exports
Revenue3/
QualityofPoliciesandInstitutions2/
Poor
Medium
100
30
200
15
25
150
40
250
20
30
Strong
200
50
300
25
35
1/SeeIDAandIMF“OperationalFrameworkforDebtSustainabilityAssessmentsinLowͲIncomeCountriesͲFurther
Considerations”,2005.
2/Country’swithaCPIAbeloworequalto3.25aredefinedtohaveapoorqualityofpoliciesandinstitutions,while
aCPIAabove3.75indicatesagoodquality.
3/Revenueisdefinedexclusiveofgrants.
There are several problems with using the CPIA as the sole criterion for determining debt thresholds.
Historical series for the CPIA index are not publicly disclosed (only data for IDA countries starting from
2005 are disclosed). As a consequence, all analyses that link debt sustainability to the CPIA have been
conductedbyWorldBank/IMFstaffandnoexternalresearcherhasbeenallowedtotesttherobustnessof
thelinksbetweenthesetwovariables.ItisalsoquestionablewhetherthequantitativeimpactoftheCPIA
ontheprobabilityofdebtdistressislargeenoughtoformulatedebtthresholdsonlybasedontheCPIA.
Moreover,itisnotclearwhethertheCPIAisindeedameasureofpoliciesorjustaleadingindicatorofa
debtcrisis.
4. Debt,TradeandGrowth
Interlinkagesbetweentrade,growthanddebtcanbeshownmoredirectlybyrearrangingthebalanceof
payments identity of section B.1 in accordance with the algebraic transformations in section I of the
Appendix.Thesegivethefollowingequation:
d t d t 1
m x (i f p f )d t 1 (ctot g )d t 1
wheredistheratioofexternaldebttoGDP,mtheratioofimportstoGDP,xtheratioofexportstoGDP,
iftheforeigninterestrate,pftherateofchangeofforeignpriceindex,phthedomesticpriceindex,ctot(=
e + pf – ph) the rate of change of the terms of trade (e being the rate of change of the exchange rate
expressedasdomesticperunitofforeigncurrency),variablesbeingexpressedindomesticcurrency.
IgnoringtheeffectofrealforeigninterestrateandincludingatermrepresentingtheeffectofFDI(fdi,the
ratioofnetflowsofFDItoGDP),thisequationhasbeentestedempiricallythroughapanelregressionof
data for seven countries (Argentina, Bangladesh, Bolivia, Kenya, the Republic of Korea, Malaysia, and
Uganda)overtheperiod1981Ͳ2004.Theresultsareasfollows(tͲstatisticsinparentheses):
dt–dtͲ1=0.27Ͳ0.36(xt–mt)Ͳ0.37fdiͲ0.14gͲ0.18ctot
(6.07)(5.54)(2.30)(1.68)(0.66)
R2=0.1858
56
CompendiumonDebtSustainabilityandDevelopment
ThevariablesontherightͲhandsideoftheequationexplain18Ͳ19percentofthevariationsindebtratios.
Thecoefficientsaresignificantatthe5Ͳor10ͲperͲcentlevelexceptforthatofthetermsͲofͲtradechange.
Allthecoefficientshavetheexpectedsigns.Understandably,thetradebalanceplaysahighlysignificant
roleindebtaccumulation:atradedeficitaddstodebt,whileatradesurplusreducesit.FDIandgrowth
reducedebtaccumulation.
5. ContributionofDebttoGDPGrowth
Theprecedingsuggeststhatgrowthreducestheexpansionofthedebtratio.Butwhatisthecontribution
of debt to growth? In order to assess this relationship, the balanceͲofͲpayments identity is rearranged
again to show the relationship between growth as a dependent variable and other variables including
debtflows(seeappendixforthealgebraictransformations).Thisrelationshipisreflectedinthefollowing
equation(derivedinsectionIIoftheAppendix).44
*
*
e
e
w
w
dY (1 OK \ )(dP/ P dE/ E dP / P ) (1 O)(d('D ) / 'D ) OH dY / Y
Y
S
where O istheinitialratioofexportstoimports,ʗthepriceelasticityofimports,ʋtheincomeelasticity
ofimports,ɻthepriceelasticityofexports,ɸtheincomeelasticityofexports,ȴDethenetexternaldebt
flows, P the export price index, P* the import price index, E the exchange rate (domestic per unit of
foreigncurrency),andYwworldincome.
Panelregression,coveringthesamecountriesandthesameperiodasintheanalysisabove,wasusedto
estimatealternativerelationships((i)and(ii))basedonthisequation.
(i)g=0.04Ͳ0.109ȴd+0.049ctot+0.054ȴx
(10.75)(5.25)(2.00)(1.91)
R2=0.25
where g is real GDP growth, ȴd the change in the ratio of external debt stock to GDP, ctot the rate of
changeofthetermsoftrade(UNCTADindexoftermsoftrade),andȴxthegrowthrateofconstantUS$
exports
Allcoefficientsaresignificant.Thesignofthecoefficientofdebtisnegative,signifyingthatanincreasein
thedebttoGDPratioreducesgrowth.
(ii)g=0.338+0.127ȴf+0.094ctot+0.045ȴx
(8.05)(2.22)(3.23)(1.46)
R2=0.11
inwhichthedebtstockisreplacedbyavariable(ȴf)measuringtheratioofnetexternalresourceflowto
GDP:ѐf=ft ͲftͲ1andf=nd+fdi +oda,ndbeing theratioof net transfersondebt (debtflowsnet of
amortizationminusinterestpayments)toGDP,fditheratioofnetflowsofFDItoGDP,andodatheratio
ofgrantstoGDP.
44
Thirwall and Hussain (1986) derived a similar equation, expressing growth in terms of the volume effect of relative price
changes,thetermsoftrade,andthegrowthoftheworldeconomyandcapitalflows.
57
TheMechanicsofDebtSustainabilityAnalysis
The variables on the rightͲhand side of (ii) explain only 11 per cent of GDP growth However, the
coefficients except that of export growth are statistically significant. This time external debt is not
specifiedintheequationassuchbutisincludedasoneofthecomponentsofthevariablerepresenting
netexternalresourceflowswhichhaveastatisticallysignificantpositiveimpactongrowth.
E. DevelopmentPolicyͲBasedApproachtoDebtSustainability
TheabovediscussionshowsthatthereisadiversityofapproachestoDSA.Eachapproachhasaparticular
focus and serves a different purpose, whether it be debt management, crisis prevention, or debt relief.
Technical indicators should be supplemented by policy considerations and other kinds of analysis if
countriesaretomanagetheirexternaldebtinasustainableway.
Bearing in mind all external and domestic factors contributing to debt sustainability, under a
developmentͲpolicy approach, debt sustainability is not viewed only from the narrow perspective of
reducinganunsustainablelevelofdebtbutisalsointegratedintotheoveralldevelopmentstrategyofa
country.Underthisapproach,debtshouldbemanagedinsuchawayastomaximizeitscontributionto
sustainabledevelopment.
Suchanapproachincorporatestheviewthatexternalindebtednesscannotbesustainableinthelongrun
ifthedevelopmentstrategyadopteddoesnotleadtoanincreaseinforeignexchangeearningstorepay
thedebtonlyaftertheotherdomesticresourcerequirementsofthedevelopmentstrategyhavebeenmet.
Thusthepointofdepartureofasustainabledebtstrategyisaclearvisionofthecountry’sdevelopment
trajectory. Debt should be integrated into this development trajectory by encouraging efficient use of
externaldebtwhichbalancesitscostsandbenefitsinthecontextofthetrajectory.
Thepanoplyofpoliciesintegratingdebtintoacountry’sdevelopmentstrategywouldaimataddressing
differentsituations:
x
x
x
policiestoenhanceanefficientuseofdebtinlinewithdevelopmentobjectives;
policiestoadjusttoshocksinordertoavoiddebtcrises;
policiestodealwithdebtcrisesandtorestoregrowth.
Integrationofdebtanddevelopmentstrategydoesnotexcludepoliciestoreduceexcessivelevelsofdebt
but emphasizes the context of a growthͲoriented approach to debt sustainability. Furthermore such
integration is based on acknowledgement that in an interdependent world prevention of a debt crisis
often also requires actions at the international level, based on international cooperation to ensure
adequatetransferofresourcesfordevelopmentaswellastradingopportunitiesfordebtorcountries.
The establishment of an effective institutional framework up for debt management is essential for the
implementation of a sustainable debt strategy. Within this framework specific roles and responsibilities
shouldbeassignedtotheministryoffinance,thecentralbankandthedebtmanagementagency,i.e.the
differentgovernmententities,andtheframeworkitselfshouldbeadaptedtotheadministrativecapacity
ofeachdebtorcountry.
Development is not a smooth process, and no country can be sheltered from the threat of a debt or
financialcrisis.Countriescanmoreeffectivelyadjusttodebtandcurrencycrisesiftheymanagetoforesee
theeventslikelytotriggerthem.Takingearlyadjustmentmeasures,wherepossible,couldinsomecases
help to mitigate the gravity of the crisis and shorten its duration. The indicators reviewed in this paper
should help in this respect. They should also help countries to assess the costs and benefits of debt
renegotiations.
58
CompendiumonDebtSustainabilityandDevelopment
References
BergAandPattilloC(2000).TheChallengesofPredictionEconomicCrises.IMFEconomicIssues.22,July.
BergA,BorenszteinEandPattilloC(2004).AssessingEarlyWarningSystems:HowHaveTheyWorkedin
Practice?IMFWorkingPaper(WP/04/52),March.
Corsetti G, Pesenti P and Roubini N (1998). What Caused the Asian Currency and Financial Crisis. NBER
WorkingPapers,6833,December.
EichengreenBandModyA(2000).WouldCollectiveActionClausesRaiseBorrowingCosts?NBERWorking
Papers,7458,January.
IMF(2002a).AssessingSustainability.PreparedbythePolicyDevelopmentandReviewDepartment,May.
IMF(2002b).EarlyWarningSystemModels:TheNextStepForward.In:GlobalFinancialStabilityReport.
Washington,DC,InternationalMonetaryFund,March.
IMF(2003).DebtSustainabilityinLowͲIncomeCountries:TowardsaForwardͲLookingStrategy.Prepared
bythePolicyDevelopmentandReviewDepartment,May.
InternationalDevelopmentAssociationandInternationalMonetaryFund(2005).OperationalFramework
forDebtSustainabilityAssessmentsinLowͲIncomeCountries:FurtherConsideration(IDA/R2005Ͳ
0056).
KaminskyG,LizondoSandRienhartC(1998).LeadingindicatorsofCurrencyCrises.IMFStaffPapers,47,
0:62–98.
RadeletSandSachsJ(1998).TheOnsetoftheEastAsianFinancialCrisis.Mimeo,February.
Roubini N (2001). Debt Sustainability: How to Assess Whether a Country is Insolvent. New York, Stern
SchoolofBusiness,NewYorkUniversity,20December.
ThirwallAPandNureldinͲHussainM(1982).Thebalanceofpaymentsconstraint,capitalflowsandgrowth
ratedifferencesbetweendevelopingcountries.OxfordEconomicPapers,November.
59
TheMechanicsofDebtSustainabilityAnalysis
ANNEX
1. Debt,TradeandGrowth
FromtheBalanceofPaymentsidentity:
1)
Ph X EPf M i f ED EF
0
X arerealexports(orvolumeofexports)
M arerealimports(orvolumeofimports)
Ph isdomesticpriceindex
Pf
isforeignpriceindex
D isexternaldebtexpressedinforeigncurrency
F areexternalflows,includingnewdebtandnetFDI,expressedinforeigncurrency
E isnominalexchangerate(domesticcurrencyperunitofforeigncurrency)
if
istheforeigninterestrate
Weassumethattherearenochangesininternationalreserves.
DividingallvariablesbynominalGDP,
x m if
2)
ED EF
PhY PhY
m
where x = PhX PhY ,
F
relatedsothat
d
Define
PhY ,weget:
0
EPf
Ph
*
M
Y and for simplicity it is assumed that all external flows are debt
dD
dt ED
f
PhY and
EF
PhY Differentiatetheexpressionfor d :
( EdD DdE ) Ph Y ED ( dPh Y Ph dY )
dd
dt
( Ph Y ) 2
EdD DdE ED(dPhY Ph dY )
PhY
( PhY ) 2
EdD DdE EDdPh 1
EDdY 1
PhY
Ph PhY
Y PhY EdD DdE EDp h EDg
PhY
PhY
PhY 60
CompendiumonDebtSustainabilityandDevelopment
ED>dD / D dE / E p h g @
PhY
Wecanrewritetheaboveexpressionas:
dd
dt
d (e p h g ) ED F
PhY D = d (e p h g ) f e rateofchangeofexchangerate
p h rateofchangeofdomesticprices
grealGDPgrowthrate(dY/Y)
Finally,wereplace f withtheaboveexpressioninequation2:
x m i f d (e p h g ) d dd
dt
or
dd
3) dt
0
m x (i f e p h g ) d
Further,wecanrewriteequation3as:
dd
dt
where
or
pf
m x i f d (e p f p h g p f ) d
representstherateofchangeofforeignprices
dd
dt
m x (i f p f ) d (ctot g ) d
where ctot (= ph Ͳ pf Ͳ e) is the rate of change of
thetermsoftrade.
2. ContributionofDebtandCapitalFlowstoGrowth
First,thebalanceofpaymentsidentity:
(D e D e )
E P * M P X i * Dte1 'FDI t ( Rt Rt 1 ) t 1
t t
t
t
t
4) t
Deisexpressedindomesticcurrency.
De
Forsimplicity,wewillassumethatallcapitalinflowsaredebtͲrelatedandsoincludedinto t ,thatthese
inflowsareestimatedonanetbasis(i.e.afterallowingfori*DtͲ1),andthattheforeignexchangereserves
areunchanged(
'Rt
'Dte Pt * X t
0 ).Then:
Et * Pt* * M t 61
TheMechanicsofDebtSustainabilityAnalysis
Takinglogsonbothsides:
ln('Dte Pt * X t )
Differentiating:
d ('Dte P * X )
'Dte P * X
ln( Pt* * Et * M t ) *
dP
dE dM
*
E
M P
'D e P * X
E * P* * M
t
t
t
t
t
t wecanrewritethisexpressionas:
Since d ( P * X ) dP * X P * dX and
{P*X/P**E*M}{dP/P+dX/X}+{ѐDet/P**E*M}{d(ѐDet)/ѐDet}=dP*/P*+dE/E+dM/M
'Dte
P* X
*
*
Replacing P * E * M by O and P * E * M by 1 O weget.
§ d ('D e ) ·
¸ O §¨ dP dX ·¸
(1 O )¨
¨ 'D e ¸
X ¹
© P
¹
©
5)
dE dP * dM
* E
M
P
where O representstheinitialratioofexportstoimports.
Substitutingtheexpressionforrealimports,
dM / M
\ (dP * / P * dE / E dP / P) SdY R / Y R
andrealexports,
dX / X
K (dP / P dE / E dP * / P * ) H dY w / Y w
,
whereɻisthepriceelasticityofexports,ʗthepriceelasticityofimports,ʋtheelasticityofdemandfor
imports,ɸtheelasticityofdemandforexports,YrtheGDPofthedebtorcountry,andYwforeignincome,
andsolvingfordYr/Yr,weget:
(1 OK \ )(dP/ P dE/ E dP* / P* ) (1 O)(d('De ) / 'De ) OH dYw / Y w
S
6)dYr/Yr=
OntheRHSofthisexpression,weseethatthebalanceͲofͲpaymentͲconstrainedgrowthrateofrealGDP
dependson:
*
*
therateofchangeoftermsoftrade (dP / P dE / E dP / P ) ;
the combined effect of price elasticity of imports and export and relative price changes/changes in the
*
*
termsoftrade (K \ )(dP / P dE / E dP / P ) ;
the combined effect of rate of change of debtͲrelated foreign capital inflows and the trade deficit as a
e
e
proportionofimports,(1Ͳʄ) d ('D ) / 'D ;
thecombinedeffectofgrowthrateoftheratioofexportstoimports,theelasticityofdemandforexports,
andforeignincome,ʄɸYw.
62
CompendiumonDebtSustainabilityandDevelopment
CHAPTERIV
ANANALYTICALFRAMEWORKFORDEBT
SUSTAINABILITYANDDEVELOPMENT
ValpyFitzGerald
(UniversityofOxford)
A. Introduction
1. ExternalDebtandDevelopment
This paper is intended to contribute to the further development of debt sustainability analysis for
developing countries, within a framework that does not take the narrow view of debt sustainability as
beingreachedsolelybyreducingexcessivecurrentlevelsofdebt.Ratheritviewsdebtsustainability,asan
integralpartofasuccessfuldevelopmentstrategy,closelylinkedtoexportgrowth.
There are at least three good reasons for developing country Governments to borrow abroad: (i) the
economicreturnonpublicinvestmentindevelopingcountriesissuperiortothecostofborrowedcapital
so that growth can be accelerated by prudent use of debt without excessively reducing current
consumptionlevels;(ii)domesticfirms(particularlysmallandmediumenterprises)cannoteasilyborrow
abroadandtermsarebetterforsovereignborrowerssothatitisefficientfortheGovernmenttousedebt
foronͲlendingtoproductivesectors,particularlyexports;and(iii)theexternalitiesfrompublicinvestment
ininfrastructure,health,education,etc.arelargeandpositivebutcannotgenerallybecapturedinreturns
toforeigndirectinvestors.
ForeignprivateinvestorscanbenefitfrominvestingindevelopingͲcountrysovereigndebtastheratesof
return are higher than those obtainable on OECD government bonds, while the risk due to possible
defaultcanbemitigatedbyappropriatediversificationofportfolios.However,financefromthissourceis
available only for “emerging markets” – that is middleͲincome countries and a few large lowͲincome
countries.MostlowͲincomecountries,ontheotherhand,donothaveaccesstoexternalprivatecapital
except for foreign direct investment in natural resource sectors owing to problems associated with
contractenforcement,informationasymmetry,andeconomic externalities.Inconsequencebilateralaid
63
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
donors and multilateral development banks act as financial intermediaries to provide loans on suitable
termsonthebasisoftheirownabilitytoraisefundsinglobalcapitalmarkets.
External debt has to be repaid in foreign exchange, so that trade plays a critical role. The relationship
between external borrowing and trade is the key to a successful external debt strategy, as external
indebtedness cannot be sustainable in the long run if the development strategy does not lead to an
increaseinforeignexchangeearningsaboveimportrequirementssufficienttorepaythedebt.Thepoint
ofdepartureofasustainabledebtstrategyis,therefore,aclearvisionbytheGovernmentofthecountry’s
developmenttrajectoryanditsrelationtoitstradepotential.
2. TheCurrentEmpiricalContextforExternalDebtAnalysis
It is essential to take into account the situation of different types of debtor countries (middleͲincome
countries, lowͲincome countries, HIPCs, etc.) both because lenders are diverse (with distinct objectives
andleverages)andbecausecountries’economicstructuresandvulnerabilitytoexogenousshocksdiffer.
InthissubͲsectionwetakeabrieflookataggregatedataorganizedbyregionalgroups,incomelevels,and
debt difficulties. This disguises many problems at the country level but gives a good idea of the overall
issues.
AsTableIV.1.indicates,thedebtburdeninrelationtoexportsisthreetimeshigherinlowͲincomethanin
middleͲincomecountries.HowevermiddleͲincomecountriesowethreeͲquartersofalldevelopingͲcountry
debt.Thusthe“debtproblem”isanissueofintegrationintotheworldeconomyifconsideredfromthe
point of view of a middleͲincome country, but an issue of economic development if viewed from the
perspectiveofalowͲincomecountry.
TableIV.1.ExternalDebtandExportsbyIncomeLevel,2003
Exports
(US$bn)
ExternalDebt Debt/Exports
(US$bn)
(Percent)
176
523
297
MiddleͲincomecountries
1813
1815
100
Totaldevelopingcountries
1999
2339
117
LowͲincomecountries
Source:WorldBank(2005b).
Thetotalvalueofexternaldebtanddebtservicevarieswidelybyregionandbydebtorstatus,atTable
IV.2.indicates.By2003netexternalborrowingwasquitelowcomparedtooutstandingdebtinallthree
regionsidentifiedhere,45butonlyinDevelopingAsiaarereservessufficientlylarge(particularlysincethe
midͲ1990sfinancialcrises)tocoverexternaldebtliabilities.InLatinAmerica(andbyextensionin“UDC”
countries with recent debt difficulties) reserves barely cover debt service, leading to serious liquidity
difficulties.InAfrica(andbyextensiontheHIPCgroup)reservesareatleastfourtimesdebtservicebut
this has no practical significance because the debt is not traded. It is worth noting that if the overseas
assets of the private sector were recorded and entered here, the net asset position of developing
countries would be positive – and in this sense there is no “developing country debt issue” as such but
ratheraserioussovereigndebtproblem.
45
TheCISandEasternEuropewereinfactthemainnetborrowersin2003.
64
CompendiumonDebtSustainabilityandDevelopment
TableIV.2.ExternalDebtofDevelopingCountriesbyRegionandDebtorStatus,2003(US$bn)
Total
Externaldebt
OfficialReserves
Debtservice
Netexternalborrowing
Exceptionalfinancing
2724.3
1412.6
437.8
91.5
32.4
of
which
Developing
Asia
695.7
670.1
105.5
18.7
6.2
Latin
America
and
Caribbean
759.0
196.2
174.3
0.6
14.4
Africa
UDC
HIPC
278.0
90.9
25.2
3.8
6.7
804.2
168.7
112.3
3.8
13.0
145.8
19.7
3.6
2.0
5.1
Source:IMF(May2005).
Notes:“UDC”areUnsustainableDebtCountries(author’sdefinition)witharrearsand/orreschedulingduring1997Ͳ2001;
HIPCare“highlyindebtedpoorcountries”underconsiderationbytheWorldBankandIMFfordebtcancellation;
“debt service” is actual payments of interest on total debt plus amortization payments on longͲterm debt,
incorporating exceptional financing; “exceptional financing” is arrears on debt service, rescheduling of debt
serviceanddebtforgiveness.
Theexternaldebtstructurevariesintwodimensions–maturityandcreditor.AstableIV.3.shows,most
debtis“longͲterm”(thatiswithamaturityofoneyearormore)andhasanaveragematurityoftheorder
oftenyears.AfricaandtheHIPCcountriesrelymostlyonofficialcreditors,whileAsiaandLatinAmerica
rely more on private lenders. Within this latter category bonds predominate over bank credit, although
thedifferenceisnotgreatinpracticefromthepointofviewoftheborrower.
TableIV.3.StructureofExternalDebt,byMaturityandCreditor,2003(US$bn)
ShortͲterm
LongͲterm
Totaldebt
Officialcreditors
Privatecreditors:
bankcredit
bonds
Total
377.9
2344.9
2724.3
1021.9
722.1
960.0
of
Developing
which
Asia
106.3
589.4
695.7
292.6
161.9
241.3
LatinAmerica
andCaribbean
89.7
669.3
759.0
204.5
185.8
368.7
Africa
UDC
HIPC
19.4
34.8
3.3
258.6 769.4 142.5
278.0 804.2 145.8
213.1 491.3 132.0
42.5 183.2 10.8
22.4 129.7
3.0
Source:IMF(May2005).
Notes:ForUDCseenotestotableIV.2.
In relation to exports it is clear from Table IV.4. that the major “debt overhang” difficulties are
encounteredinLatinAmericaandAfrica,wheremostoftheUDCandHIPCaretobefound.Inrelationto
debtserviceLatinAmerica(andbyextensiontheUDCs)hasthemostseriousproblem.Africancountries
(andthustheHIPC) benefitfromsofter,aidͲrelateddebtterms –and,indeed,donotfullyservice their
debt.ThedifferencesininterestratespaidreflectthedifferencesbetweenmiddleͲincomecountrieswith
accesstoprivatelendingandlowͲincomecountrieswhichrelyonofficiallenders,ontheonehand,and
thehigherdefaultriskinLatinAmericacomparedtoAsia,ontheother.
65
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
TableIV.4.IndicatorsofDebtinRelationtoTrade,2003
of
which:
Total
Percentofexports
ExternalDebt
InterestPayments
Amortization
Percentofdebt
Interest
Amortization
111.3
4.3
13.6
3.9
12.2
Latin
America
and
Caribbean
Developing
Asia
73.2
2.6
8.5
3.6
11.6
199.3
10.8
34.9
5.4
17.5
Africa
UDC
HIPC
144.6
4.1
9.0
2.8
6.2
208.9
6.8
22.3
3.3
10.7
321.1
3.6
6.6
1.1
2.1
Source:IMF(May2005).
3. CoverageofthePaper
This paper focuses on external debt. It does not examine domestic debt, even though with currency
convertibilitypublicdebtdenominatedindomesticcurrencycanrepresentacontingentclaimonforeign
exchangereserves–albeitatanundefinedexchangerate.Nordoesthispaperaddressdebtcrisesassuch
and the subsequent renegotiations and restructurings. None the less, many of the analytical results for
prudentdebtmanagementdiscussedbelowarerelevanttodebtsustainabilityatthedomesticaswellas
theexternallevelbecauselattersustainabilitycannotbesensiblyplannedforeitherexceptinthecontext
ofasustainabledebttrajectory.
The structure of the paper is as follows. Section B provides the appropriate national accounting
framework for debt analysis, and then sets out the traditional “gap models” (savings, trade and fiscal)
along with a summary of the modern critique of this approach. Section C outlines the modern
intertemporalapproachtodebtanalysis,derivestheoptimaldebtlevelinrelationtooutputandexports
foranopen developingeconomy,andproposestwo“goldenrules”forexternaldebtmanagement. The
macroeconomic consequences of external debt are addressed in Section D, which starts with the key
impact on real exchange rates and follows this with the framework for analysing the effects on fiscal
balancesandincomedistribution.SectionEexplainshowcreditrationinginglobalcapitalmarketsmeans
thatdebtlevelsarenotdeterminedbyborrowers,andgoesontoanalysetheimpactofinterestͲrateand
trade shocks under these circumstances. Finally, Section F derives policy conclusions for both domestic
Governmentsandtheinternationalcommunity.
B. Debtandthe“FinanceGap”Model
1. NationalAccountingsandDebt
Debtaccountingisquitecomplexbecausedebtflows–inflowsoffreshdebtcapitalandoutflowsofdebt
service–enterintotheprocessofsavingsandinvestment,thebalanceofpayments(onbothcurrentand
capital account) and the fiscal framework. In the savings and investment balance, net debt flows
constitute“externalsaving”.Inthecurrentaccountofthebalanceofpaymentsinterestpaymentsondebt
areanoutflowoffactorincome.Inthecapitalaccountnetdebtflowscreatechangesinthenetexternal
assetposition.Inthefiscalaccountsgrossdebtinflowsareanexternalresource,whileamortizationand
66
CompendiumonDebtSustainabilityandDevelopment
interest payments are major expenditure items. Furthermore net debt flows in a particular year, in
combinationwithinheriteddebt,determinedebtfornextyear,thusintroducingadynamicelementinto
debtaccounting.
Theseaccountingidentitiestellusnothingaboutthebehaviorofthevariouscomponents:inotherwords
theyarenotamodel.However,theydoclarifythecomplexrelationshipbetweendebtandthedomestic
economy,andalsounderlinethefactthatthecomponentsmustbereconciledͲinotherwords,“addup”.
Weusethefollowingnomenclature:
Yaggregateoutput(i.e.GDP)
Caggregateconsumption
Xexportsofgoodsandservices
M importsofgoodsandservices
Sdomesticsaving
I investment(grossfixedcapitalformation)
i interestrateonexternaldebt
ɷamortizationrateonexternaldebt
Dexternaldebt
Ggovernmentexpenditure
Tgovernmentrevenue
Rofficialforeignexchangereserves
WestartwiththeaggregatedemandͲsupplybalance
Y M { C I X [B.1]
which,whenthesavingsͲinvestmentidentityisinserted,yieldsthe“accumulationbalance”
[B.2]
Thebalanceofpaymentsoncurrentaccountincludesnotonlyexportsandimportsofgoodsandservices,
but also factor income (income from capital and workers’ remittances). To simplify the exposition we
include here only the interest payments on (public) external debt at this stage. Note that, if these are
includedin[B.2],thenthedefinitionofsavings(S)isnationalsavingsandthatofoutput(Y)isGNP.
Thecurrentaccount(CAB)andthecapitalaccount(KAB)areoppositeandequal.Thus
CAB { X M iD [B.3]
andthecapitalaccountis
KAB { 'D 'R [B.4]
sothat
CAB KAB { 0 [B.5]
67
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
Note that private capital flows (and the corresponding factor payments) can be inserted into this
accounting framework very simply. Foreign private assets (A) include portfolio holdings overseas
(sometimes misleadingly called “capital flight”) and FDI abroad by domestic companies. Foreign private
liabilities(L)includebothforeignborrowingbydomesticfirmsandinwardFDI.Thefullcapitalaccountcan
thusbewritten46
KAB { 'D 'L {'R 'A} [B.4a]
However,intherestofthispaperweshallassumethattheGovernmentistheonlyexternaldebtor–not
leastbecausethe“externaldebt”statisticsreflectpublicsectorandpubliclyguaranteedexternaldebt.
Inthisframeworkweassumethatthefiscalbalanceisclosedonlybyforeignborrowing–thusexcluding
monetaryissue(“seignorage”)anddomesticborrowingfromconsideration:
T 'D { G iD [B.6]
Last,butfarfromleast,wehavethelawofmotionfortheexternaldebtitself(D)intermsofitsprevious
periodvalue(DͲ1),andnewborrowing(ȴD)andthedepreciationrate(ɷ)inthefollowingtwoalternative
forms:
D { D1 GD1 'D
'D { D D1 GD1 [B.7]
2. “FinancingGap”ModelsofDebtandGrowth
In the context of work on development strategies from the 1960s through the 1980s “financing gap”
modelsprovidedthebasicanalyticalframeworkforbothlendersandborrowers.47Inthesemodels,48the
objectiveoftheplanneristomaximizetherateofoutputgrowth(y)subjecttotheconstraintsimposedby
domestic savings (i.e. the capacity to invest), the external sector (i.e. the capacity to import) or fiscal
balances(i.e.thecapacitytospend).
The savings constraint exists because available funds are determined by the domestic economy’s
propensitytosave(s)andtheinflowofexternalfinance(F),whichinturndeterminesthemaximumlevel
ofinvestment(I)thatcanbeundertakenandthustherateofgrowth(y).
The external constraint exists because the level of imports (M) cannot exceed the foreign exchange
available from exports (X) and capital inflows (F). Exports are assumed fixed in the short term, due to
capacityconstraintsand/orlimitedexternalmarkets.Theavailabilityofimportsdeterminesthemaximum
levelofoutput(Y)foragivenimportpropensity(m).
The fiscal constraint exists because growth depends on public investment (either because it constitutes
the bulk of investment, as in poor countries, or because it is essential in order to promote private
investment,asinmiddleͲincomecountries).Publicinvestmentisassumedtobeaconstantproportion(p)
oftotalinvestment.Publicinvestment,andthusgrowth,isconstrainedbybudgetarybalance(Z).
46
AcompletecurrentͲaccountidentitywouldincludeprivateinflowsandoutflowsoffactorincome
Avramovicandothers(1964)isagoodsurveyofthetraditionalmethodologyforanalysingtherelationshipbetweendebtand
growth.
48
Thereisalargeliteratureonthesemodels,whichoriginateswiththeHarrodͲDomargrowthmodelofthesavingsconstraint,
andcontinueswithChenery&Strout(1955),whomodeledtheexternalconstraint.Thismodelwasthenextendedtoincludealso
thefiscalconstraint.Goodformalexpositionsofallthree‘gapmodels’aregivenbyBacha(1990)andTaylor(1994).
47
68
CompendiumonDebtSustainabilityandDevelopment
Inconsequenceoftheseassumptionsexternalfinance(F)actsasasourceof“externalsaving”,tofillthe
gapbetweendomesticsavingsandtotalinvestment,actingasaformof“importsupport”andasasource
offiscalrevenue.ThroughthesechannelsFaffectsboththelevelofinvestmentandtherateofgrowthof
GDP.Thismodelstillinformsmostoftheempiricalpolicydebateaboutaid,debtandforeigninvestment.
Theplanningproblemisthustomaximizeywhere
Yt = k Kt
K t = K t -1 + I t
S t = s Y t (t g )Yt
M t = m1 Y t + m2 I t
Dt Dt 1 (i G ) Dt 1
Ft
Zt
( g t )Yt pI t
[B.9]
[B.8]
subjecttothethreeconstraints
I t d St + F t
M t d X t + Ft
Z t d Ft
(wheretistaxrevenueandggovernmentexpenditureasaproportionGDP).
Theoutcomedependsonwhichofthethreeconstraintsactuallybindsatanyonepointintime,whichis
anempiricalissue.
ThesavingsͲconstrainedmaximumgrowthrate(
y s*
y s*
)canbederivedas:
Fº
ª
k « s (t g ) »
Y ¼ [B.10]
¬
The main concern of aidͲrelated policy modeling in most developing countries is the externallyͲ
constrainedmaximumrateofgrowth(
y e* =
y e*
),whichcanbederivedas:
k ªX F
º
m1 »
«
Y
¼ [B.11]
m2 ¬
Finally,thefiscallyconstrainedrateofgrowth(
y *f =
y *f
)canbederivedas
k ªF
º
(t g ) »
«
p ¬Y
¼ [B.12]
Allthreegrowthratesareofcourseincreaseinresponsetonetdebtinflows(i.e. wy / wF ! 0 ),butwith
differentderivatives.Whichbindsdependsonthecharacteroftheeconomy.Generallyitisreasonableto
assumethatinthepooresteconomiesthesavingsconstraintisbinding,andthatexternalandfinallyfiscal
69
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
become binding as economic development advances. The effect of net debt flows is likely to be
progressivelygreaterineachofthesethreestagesbecausegenerally
p m2 1
wy *f
wF
!
wy e* wy s*
!
wF
wF [B.13]
3. TheLimitationsof“FinancingGap”Models
The “financing gap” model continues to form the basis for the trade, aid and growth linkages in the
mediumͲtermmacroeconomicprogrammingmodelusedbytheWorldBank:theRevisedMacroeconomic
StandardModel(RMSM).49Italsoinformstheshorttermmonetaryprogrammingframeworkusedbythe
IMF.50Thesetwomodelsstillformtheessentialanalyticalunderpinningforthemissionreportsofthetwo
BrettonWoodsinstitutionsonstabilizationandadjustmentprograms.51TheUNDPmakesestimatesofthe
externalfinancingrequirementsofpoorcountriesonasimilarbasiswhenpreparingformeetingsofdonor
consortia.
However,thelastdecadehaswitnessedgrowingawarenessofthelimitationsofthesemodels,whichno
longercorrespondeithertomodernmacroeconomictheoryortomacroeconomicpolicypracticeinopen
economies. Indeed from a neoclassical viewpoint this analytical tradition is regarded as invalidating the
proposals from the Bank and the Fund on additional lending and debt forgiveness.52However, their
persistenceisdoubtlessdueinlargeparttotheiranalyticalsimplicityandthefactthattheparameterscan
beestimatedeasilyandquicklyfromavailablemacroeconomicdataindevelopingcountries.
Withoutgoingsofarastoreject“financinggap”models,itispossibletoidentifyfourareasofweakness
whichneedtoberemediedinordertoproduceasounderconceptualframeworkandanalyticalmodelfor
quantifyingdebtsustainability.Theseare:
x
First,thecoefficientsinthebehavioralequations(particularlytheconstraints)areassumedtobe
stable and exogenous, rather than endogenously determined. In the case of savings, empirical
evidence and Keynesian theory suggest that domestic saving (and thus consumption) in fact
adjuststotheleveloffixedinvestmentandforeigninflowsofcapital.53Again,thefiscalbalance
canalwaysbeadjustedbyvaryinggovernmentexpenditure.
x
Second,intheexternalbalanceoftrade,exportsareassumedtobegivenandimportstodepend
onlyonthelevelofeconomicactivity.Thisignorestheeffectoftherealexchangerateonboth
import and export volumes, and thus the possibility of adjusting to foreign exchange shortages
without having to reduce growth.54It also underplays the role of exchange rates in determining
thefiscalbalance.
x
Third,“financinggap”modelsassumethatextraexternalfinancealwayscontributestogrowth,by
simply and directly adding to savings, import capacity or fiscal resources and thus allowing
investment–andthusgrowth–torise.However,itisestablishedthatexternalresourcesoftenin
49
SeetheAddison(1989)andhttp://www.worldbank.org/data/rmsm/index.htmforanupdatedversionofthispaperplusother
RMSMdocuments.
50
See IMF (1987), which in turn derives from Polack (1957). See also Baquir and others (2003) for the growth linkages in IMF
models.
51
See Aghenor and Montiel (2003) for a recent survey, and Khan and others (1990) for a formal statement of the relationship
betweenthetwomodels.
52
SeeEasterly(1999).
53
SeeFitzGerald(2003a).
54
SeeDornbusch&Helmers(1988).
70
CompendiumonDebtSustainabilityandDevelopment
practiceleadtoincreasedconsumption.55Moreovertheinvestmentundertakenmaynotleadto
increasedexportsandthusdebtͲrepaymentcapacity.
x
Fourth,andmostseriouslyfromananalyticalviewpoint,the“financinggap”modeldoesnotallow
for intertemporal optimization by economic agents: that is, the fact that households, firms and
Governmentstakeinvestment,savingandborrowingdecisionslookingforwardovermanyyears.
Theassumptionofintertemporaloptimizationisthebasisofmodernmacroeconomicsingeneral
and for small open economies in particular; and allows resource allocation behavior to be
endogenized.56
C. SustainableDebtLevels
1. TheOptimalDebtLevel,ExportCapacityandIntertemporalMaximization
The contemporary approach to debt sustainability starts from the same foundation as the modern
macroeconomictheoryofopeneconomies,whereapparentbalanceofpaymentsdisequilibriaintheshort
run can be seen as part of an intertemporal equilibrium based upon expectations by economic actors
about the future. The small open economy is composed of overlapping generations of households
optimizingconsumptionandsavingovertimeandoffirmsmakinginvestmentdecisionsbasedonprofit
maximization.57Currentaccountsurpluses(ordeficits)generatenetasset(orliability)positionswiththe
restoftheworld,whichinturnaffectthefuturebehavioroffirmsandhouseholds.
Ifthereisfreeaccesstointernationalfinancialmarketsatagiveninterestrate(i)andnoissuessuchas
debtdefault,thenthecountryobeystheFisherianmaximandseparatesthedecisiontoinvestfromthe
decisiontoconsume.58Focusinghereonthedecisiontoinvest,firmschoosetheirinvestmentstrategyso
as to maximize the wealth of their shareholders when measured at world interest rates. The
intertemporalequilibriumstrategy59amountstoselectinganinvestmentrate(k*)thatisasolutionto
f
max ³ exp( it )(Qt J t )dt
k
Qt
K
t
0
Q( K t )
I t GK t
k { I t / Qt
[C.1]
whereQisthelevelofnetoutput,Jthecostofinstallingnewcapital,Kthecapitalstockandɷtherateof
depreciation.
In order to find a tractable solution to this general problem, we have to specify the relevant functional
forms.Westartoffbydefiningnationalincome(W)asoutputminusdebtinterestcosts,wheredebtalso
playsaroleincapitalformation,suchthat
55
AtleastsinceGriffin(1970).
SeeObstfeld&Rogoff(1995)andSen(1994).
57
Thisisnowastandardformulation:seeforinstanceObstfeld&Rogoff(1995).
58
Thesavingsratedependsonthesocialratediscountfactorandtheintertemporalelasticityofsubstitutionofconsumption,on
theonehand,andthe(world)interestrate,ontheother.SeeSen(1994).
59
SeeCohen(1994)forthederivation.
56
71
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
W Y iD
W ' ( D) ! 0
W ' ' ( D ) 0 [C.2]
inorderthatamaximumshouldexist.60Thisformulationisalsoconvenientbecausethetwoconstraints
reflectdecliningabsorptioncapacityanddebtoverhangeffectsrespectively.Undertheseconditions,the
optimaldebtlevelwillbedefinedbytheconditionformaximizingWwithrespecttoD:
wW
wD
wY
wi
(i D
)
wD
wD
0
[C.3]
Inotherwords,debtshouldbecontracteduptothepointwherethemarginaladditiontooutputequals
themarginaladditiontointerestcosts.Ceterisparibus,thehighertheinterestrate,thelowertheresulting
optimaldebtlevel;andthelargerthepositiveimpactofthatdebtonoutput,thehighertheoptimaldebt
level.
Tofindtheoptimaldebtlevel,westartwithastandard61endogenousͲgrowthproductionfunctionofthe
form
Y
aK [C.4]
Leavingasidethelasttermin[C.3]andthusassumingthattheinterestrateisunaffectedbythedebtlevel,
wehavethefollowingmaximizationcondition:
wW
wD
wK
wD
wY
i
wD
i
a
wY wK
i
wK wD
0
[C.5]
Thekeyissueisthusshowntobetheeffectofdebtoninvestment.Weshallexaminetheparticularcase
where the domestically funded capital (K1) is already installed, there is no previous debt, and the
authoritiescontemplatemovinginoneperiodtotheoptimaldebtlevel,byprovidingextracapitalstock
(K2)fundedbyexternaldebt
K
K1 K 2 [C.6]
Externaldebt(D)isthencontracted.Afixedproportion(ʄ)ofthisisusedtofundtheinstallationofnew
productivecapital(directlyasinruralinfrastructure,orindirectlyasloanstoexporters),whiletherestis
allocatedtootheractivitiessuchassocialinvestments(healthetc),coverageofcurrentͲaccountdeficits,
ornonͲeconomicinfrastructure.Thecostofthisproductivepublicinvestment(J)isaquadraticfunctionof
theinvestmentrate:62
60
Otherwisetheoptimaldebtlevelwouldbeinfinite,ofcourse.
SeeRebelo(1991)forthebasisofthe‘AK’modelusedhere,andAghion&Howitt(1999)foracomprehensivesurveyofmodern
endogenousgrowththeory.
62
SeeHeijdraandvanderPloeg(2002:40)andalsoCohen(1994:490).
61
72
CompendiumonDebtSustainabilityandDevelopment
D
J
O
§ I K2 ·
¸
K 2 ¨¨1 2 K1 ¸¹
©
[C.7]
WithI>I
J
Sowecannowspecifytheobjectivefunction[C.2]as
W
a ( K1 K 2 ) § I K2 ·
¸
K 2 ¨¨1 O © 2 K1 ¸¹
[C.8]
i
anddifferentiatingwithrespecttoK2yieldstheoptimalsolutionintermsoftheratio(ɶ)betweendebtͲ
fundedcapitaland“domesticallyfunded”capital:
wW
wK 2
Kˆ 2
K1
a
K ·
i §
¨¨1 I 2 ¸¸
K1 ¹
O©
1 § aO
·
1¸
¨
I© i
¹
0
J
[C.9]
Notethattheoptimalcapitalstructurecoefficient(ɶ)canbenegative–whichwouldimplyaccumulation
offoreignassetsinsteadofborrowingabroad.
WefindtheoptimaldebtͲtoͲoutputratio(ɽ),bysubstituting[C.9]into[C.7]and[C.4]:
Dˆ
Yˆ
T
Kˆ 2 §
Kˆ ·
J
¨1 I 2 ¸ K1 1 JI ¨
¸
K1 ¹
O ©
O
a ( K1 Kˆ 2 ) aK1 (1 J )
Dˆ
Yˆ
J (1 JI )
aO (1 J )
[C.10]
Clearlyɽisincreasinginɶ,andthusby[C.9]theoptimaldebtͲoutputratiowillbeloweredbyanincrease
intheinterestrate(i)(asweshouldexpect),butwillbeloweredbyanincreaseintheproportionofdebt
fundsallocatedtoproductiveinvestment(ʄ)orintheoverallproductivityofcapital(a).Thisresultcanbe
generalizedtoasteadyͲstategrowthsituationbecauseinsuchasituationY/Kisconstant(andthusboth
components of capital grow at the output growth rate), and thus D/Y must be constant. If the optimal
debtlevel(ɽ)ishigher,thendebtcanbesafelyraised.
Overall capital productivity requires some further comment in the context of this study. We assume a
simplifiedformoftheexternallyconstrainedeconomydiscussedintheprevioussectionsuchthat:
M mY
MdX
X DEK [C.11]
73
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
whereexportsareafunctionoftheproportion(ɴ)ofthecapitalstocklocatedintheexportsectorwith
knownproductivity(ɲ).Ifitisassumedthatthesecondconstraintof[C.11],i.e.thatM=X,substitutionof
thefirstexpressionof[C.11]into[C.4]andtheresultintothethirdexpressionof[C.11]gives
a
Kˆ 2
K1
DE
m
1 § DE O
·
1¸
¨
I© m i
¹
J
[C.9a]
Inotherwords,theoptimaldebtlevelriseswiththeproportionɴofdebtͲfundedcapitalstockallocatedto
theexportsector.However,itfallswithanincreaseininterestratesortheimportcoefficient.Toputthis
anotherway:longͲrundebtsolvency–andthustheavoidanceofdebtcrisesarisingfromtradeorcapital
marketshocks–requirestheallocationofahigherproportionofthefundsraisednotonlytoproductive
investmentbutalsotoinvestmentintheexportsector.
Finally,wecanalsodefinetheoptimaldebtserviceratio(ʘ)fromthisresult,where
Z
(i G ) Dˆ
Xˆ
[C.12]
bysubstituting[C.10]and[C.11]into[C.12]toyield:
Z
(i G )
Dˆ
Yˆ
Yˆ
Xˆ
(i G )T
a
DE [C.13]
The optimal debt service ratio (ʘ) will decrease with a higher interest rate (i) because its negative
influenceontheoptimaldebtͲoutputratio(ɽ)outweighsthatoftheexplicititermin[C.13].
2. The“GoldenRules”forDebtSustainability
The “law of motion” for external debt from the previous section (equation [B.7]) can be expressed in
termsoftheprimary63currentaccountbalance(P)ontheassumptionthatthisdebtistheonlyformof
externalfinance:
Dt
(1 i ) Dt 1 Pt 1 [C.14]
whichthroughrepeatedsubstitutionyields
Dt
(1 i ) t
t 1
Pt
¦ (1 i)
t
t 0
+D0 [C.15].
Whenngoestoinfinity,thepresentvalueofdebt(i.e.thelefthandsideoftheequation)goestozeroand
weretrievetheintertemporalbalanceofpaymentsconstraint
63
Thatis,excludinginterestpayments.
74
CompendiumonDebtSustainabilityandDevelopment
f
D0
Pt
¦ (1 i )
t
t 0
[C.16]
Inotherwords,alldebtmusteventuallybepaidback.
However, in practice, developing country financial authorities and debt managers have to work on a
shortertimescaleandwithouttheluxuryofsearchingforoptimalsolutions.Theceilingon“prudent”debt
is conventionally expressed as a share of output or as a ratio of debt service to exports, the former
reflectinglongerͲtermsolvencyconsiderationsandthelattershorterͲtermliquidityones.
Oncetheprudentialceiling(d)onthedebtoutputratiohasbeenreached,debtmanagementstrategyis
logicallynottoexceedit.Thusthe“goldenrule”isthat
Dt Dt 1
d
Yt
Yt 1
d
[C.17]
Foragivenrateofoutputgrowth(y)andexpressingtheprimarydeficitasaratio(p)ofoutputwehave
(1 i ) Dt 1 Pt
(1 y )Yt 1
dt
Dt
Yt
dt
1 i
dp
1 y
(1 i ) Dt 1 Pt
(1 y )Yt 1 Yt
[C.18]
sothatthe“goldenrule”forthedebtͲoutputratiois
§ 1 i
·
p t d ¨¨
1¸¸ | (i y )d
©1 y
¹
[C.19]
In other words, a primary deficit (p<0) can only be safely incurred if the growth rate is higher than the
interestrate(y>i).
Ifweexpresstheruleintermsofthecurrentaccountbalanceproper,asaproportion(c)ofoutputthen
thegoldenrulebecomes
c p id
c t yd [C.20]
In other words, the maximum current account deficit as a proportion of GDP is the rate of growth
multipliedbytheprudentdebtͲGDPratio.
We can now turn to the second “golden rule” related to the ratio of debt service to exports. The
derivationisverysimilartothatforthefirstrule,butexpressedintermsofthesecondceiling(ʍ):
Oncetheprudentialceiling(d)ondebtinrelationtooutputratiohasbeenreached,thenextrequirement
of debtͲmanagement strategy is that service payments on the resulting Dt in relation to exports should
notexceedʍ.Thusthe“goldenrule”isthat
75
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
(i G ) Dt (i G ) Dt 1
d
Xt
X t 1
V
[C.21]
Foragivenrateofexportgrowth(x)andexpressingtheprimarydeficitasaratio(p’)ofexportswehave
(i G ) Dt
Xt
Vt
(i G ){(1 i ) Dt 1 Pt }
(1 x) X t 1
1 i (i G ) Dt 1 (i G ) Pt
1 x
X t 1
Xt
1 i
V (i G ) p
1 x
Vt
sothatthe“goldenrule”forthedebtserviceratiois
pc t
[C.22]
V § 1 i
· ix
V
1¸ |
¨
i G ©1 x
¹ i G [C.23]
Inotherwords,andmoregenerallythataprimarydeficit(p’<0)canonlybesafelyincurrediftheexport
growthrateishigherthantheinterestrate(x>i).IfweexpressthesecondruleintermsofthecurrentͲ
accountbalanceproper,asaproportion(c’)ofexports
cc
p c iV
x
V
cc t i G [C.24]
3. ConvergenceandExpectations
PolicymakerswithresponsibilityfordebtattempttoadoptatleastamediumͲtermview,andwhentheir
debtlevelsareabovetheprudentlimits,thena“convergence”strategymustbeadoptedinordertoreach
theselimitswithinareasonablenumberofyears.Supposethatwewishtoreachtotheprudentiallimit(d)
of the debtͲGDP level over a number of years from the present level (d’) by reducing the debt by a
proportionueachyearovernyears,then
u
§ dc d ·
¨
¸
© dc ¹
1/ n
[C.25].
ThefirstgoldenruleisthenreͲexpressedas:
~
p t (i u y ) d
~
c t (u y )d
[C.26]
Whethernewdebtissustainabledepends,therefore,onexpectationsaboutthefuturegrowthofoutput
andofthedeterminantsofthebalanceoftradeandofthecurrentaccount,namelyexportgrowthand
future interest rates and terms of trade. As we have seen, the debt level itself will affect growth in an
optimalsolutionsothattheusetowhichthedebtistobeput,andthusfutureproductivity,arealsopart
of the solution. Debtors and creditors should have agreed on these forecasts before signing a debt
76
CompendiumonDebtSustainabilityandDevelopment
contract.Forinstance,ifweexpecttherateofchangeofthetermsoftrade(h)tohaveaprojectedvalue
inthefuture,thenthisisdistinguishedfromexportvolumegrowth( x )sothat[C.24]isrewrittenas
cc t hx
V
i G [C.24a] For a debt contract to be agreed upon by debtor and creditor, both must agree on projections of key
parameters; or if they disagree, at least the overall outcome must be anticipated as profitable to both
sides..Butasexemplifiedbythedebtcrisesoftheearly1980sandofthemiddle1990sandbythepresent
plightoftheHIPCs,eventsdonotalwaysturnoutasexpected.Expectationsonbothsidesarethuscrucial
tothelending/borrowingdecision–therecanbeno“overͲborrowing”without“overͲlending”.
D. FiscalConsequencesofExternalDebt
Analytical frameworks such as those developed in sections B and C can be used explore the effects of
policy towards external debt. One important issue under this heading is the relation between debt and
fiscalpolicy.Heretheanalysisstartsfromanadjustedversionofequation(B.6)inwhichtaxrevenueand
foreignfinancingconstrainfeasiblelevelsofgovernmentexpenditureandservicespayments(interestand
amortization)onexternaldebtasfollows:
G (i G ) D d T F c [D.9]
Thusanincreasedgrossdebtflow(Fʚ)allowsafiscalexpansion(i.e.Gtorise).However,accumulateddebt
itselfgenerateslargebudgetaryitemswhichinsomecasesbecomethelargestsingleitemofgovernment
expenditure,crowdingoutotherexpenditurecategories.64
Thisconstraintcanbeelaboratedandsimplifiedtotakeaccountofadditionalassumptions:
x
x
x
x
x
SinceexternaldebtisdenominatedinforeigncurrencyandtherestoftheGovernment’sbudget
indomesticcurrency,thedebttermismultipliedbytherealexchangerate(e=(Epf/)/pd);
Debtamortizationflowsarenettedout:,
A strict budgetary rule is observed that only allows a maximum fiscal deficit (q) in domestic
currencytobefinancedfromseignorageand/ordomesticborrowing;
TheprudentdebtͲoutputratio(d)ismaintained;
Taxrevenueisagivenshare(t)ofnationalincome
(D.9)isthenrewrittenas:
G ieD d (t q )Y de'Y [D.10] Diving through by Y and rearranging, we then obtain the constraint on the share (g) of government
expenditureinnationalincomeintermsofthefamiliardebtparameters(d,i)andtherateofgrowth(y)of
output:
g d (W q ) e( y i )d [D.11]
Absent a serious tax reform (ȴʏ) or a more relaxed monetary stance (ȴq), the governmentͲexpenditure
share (g) in national income is highly dependent on the debt parameters, on the one hand, and the
64
Foralldevelopingcountriesin2003,theaverageratioofexternaldebtservicetoGDPwas6percent.Theaverageratiooftax
revenuetoGDP(t)was15percentandofpublichealthexpendituretoGDPwas3percent(WorldBank,2005a).
77
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
growth rate (y) and the real exchange rate (e), on the other, Ͳ both of which variables are themselves
stronglyaffectedbythedebtstrategy.Inparticular,anincreaseineconsequentupondevaluationwhen
therateofoutputgrowthislow(i.e.y<i)–acommonoccurrenceduringdebtcrisesͲwillhaveastrongly
negativeimpactonthefiscalconstraintandthusongovernmentexpenditure.
This subordination of government expenditure as illustrated by the fiscal constraint [D.11] to debt
managementhasatleastthreemajorconsequences:
(a)
It is difficult to give priority to increasing social provision in general (and poverty reduction in
particular)byexpandingrealhealthandeducationexpenditurefasterthanpopulationgrowth;
(b)
ItisnotpossibletoengageinanactivecounterͲcyclicalfiscalpolicyinordertoreducetheimpact
of exogenous shocks on investment and growth, for example, by expanding infrastructure
expendituretomaintaincapacityutilization;
(c)
LongͲtermplanningofpublicexpenditureisrenderedmeaningless,withnegativeeffectsforthe
efficiencyofpublicservices,infrastructureprovisionandtheutilizationofscarceadministrative
skills.
E. DebtVulnerabilityandExternalShocks
1. DeterminantsofDebtFlows
Sofarwehavebeenworkingontheassumptionthatdevelopingcountriescanchoosethelevel(D)ofdebt
thattheycontractatagiveninterestrate(i).Thisistheconventionalassumptionineconomicanalysisas
well as in policy debates when reference is made to “overͲborrowing”. In fact, however, lenders
determinethevolumeofchangesindebtandtheinterestrateisnotgiven.Internationaldebtflowsare
subjecttoaformofcreditrationing.
Official lending (that is by bilateral donors or multilateral organizations) is always determined by the
lenderonitsowncriteria,althoughtheseshouldinprinciplesupportsustainabledevelopmentandthus
coincidewiththoseoftheborrower.However,theborrowerdoesnotdecidethedebtlevel.Ratherthe
overallvolumeofofficiallendingisdeterminedbytheinstitutionalstrategyofthelender.Withinthetotal
regional and country allocations lending depends upon both the technical appreciation of development
prospectsͲandthusthesustainabilityofdebtͲ;ontheonehand,andthegeopoliticalpressuresofdonor
Governments,ontheother.
Givenaceilingofofficiallendingfromdonorsinanyoneperiod,developingͲcountryGovernmentstendto
contractdebtuptothislimit.Itisinthissensethat“creditrationing”existsforthiscategoryoflending.It
is extremely rare for developing countries – particularly small lowͲincome countries without access to
privatecapitalmarkets–toturndownofficiallendingproposals.Theinterestrateandmaturityofofficial
debtisalsosetbythelender,usuallyonasubsidizedbasis.Eligibilityisdecidedbythedonor.
IninternationaldebtmarketsforbothbondsandbankloanstothegreatmajorityofdevelopingͲcountry
Governments (“sovereigns”) another form of credit rationing obtains. This reflects the influence of
uncertainty in the loan market creates which causes adverse selection, as the two sides have different
perceptions ofriskandlenders cannotdistinguish between borrowersastotheirabilitytorepayinthe
future.ItalsoreflectsthelowertoleranceofriskonthepartofOECDinvestorsinforeignthanintheirown
markets,asituationwhichleadstoaninefficientallocationoftheirportfoliosknownas“homebias”.65
65
SeeFitzGerald&Babilis(2005).
78
CompendiumonDebtSustainabilityandDevelopment
Consider the initially upwardͲsloping supply schedule of bank loans or bond purchases in Figure IV.1.
below.Thisshowsthereturnspread,r(theexcessoftheriskyovertherisklessrate,thelatterbeingthe
rateofinterestongovernmentbonds,alongtheordinateandthevolumeoflendingalongtheabscissa.
The competitive international banking market is made up of many nonͲcollusive bank lenders and
borrowers. Banks are priceͲtakers in deposit markets but set lending rates (i.e. spreads) to maximize
expectedreturns.Higherlendingrateshaveanadverseselectioneffectsonborrowersbyincreasingthe
perceivedrisksoflending.Theseinturnthusincreaseactualdefaultriskowingtotheincreasedburdenof
interestpaymentsandtheenhancedincentivetodefaultduetoriseswithdebtandinterestrates.Beyond
a certain point the debt supply schedule will be backwardͲsloping. Banks’ unwillingness to differentiate
betweendifferentriskreflectstheirimperfectinformationonfundamentals(e.g.defaultrisk)aswellas
theirfearofcovariantriskbetweenborrowers(contagion).
Thedemandschedule(Dd)infigureIV.1.fordebtisthebackwardͲslopingcurveforthesupplyofcapital.
Competitivelendersmaximizetheirdebtholdingsatthepoint(Dd,r*):atthisprice(i.e.returnspread)
the potential supply of capital or demand for debt assets from developing countries (Ds) is in excess of
demand for capital or the supply of assets (Dd)– in other words, the willingness to borrow exceeds the
willingnesstolend.
D
Dd ( r*) Ds ( r*) [E.1]
The market interest rate in foreign currency (if) to emerging market borrowers is determined by two
elements, the riskless world rate (iw) and the risk premium (r). The risk premium is the product of the
perceived66probability of default (ʋ) and an appropriate of investors’ degree of risk aversion (A).67
PerceiveddefaultriskdependsuponindicatorsdiscussedinsectionC.BsuchasthedebtͲGDPratio(d)and
thedebtserviceratio(ʍ).68
FigureIV.1.CreditRationinginGlobalDebtMarkets
66
Theperceptionisthatoflenderstypicallyinfluencedbycreditratingsagencies.
Thus the risk premium is only equal to the underlying default risk if the financial market is strictly riskͲneutral and there is
perfectinformation;sothatyieldspreadsshouldnotgenerallybeinterpretedasmeasuresof‘countryrisk’–seeCunninghamand
others(2001).
68
Aswellascruderliquiditymeasuressuchasthe‘quickratio’mentionedinSectionF.1.
67
79
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
Thuswehave:
if
iw r
SA
S S (d , V ) [E.2]
r
Clearly an increase in d or
default risk (which is why the loan supply curve eventually
becomesbackwardͲsloping)andthusnotonlyraisesinterestcostsbutalsoreducesloanavailability.Note
that the risk premium (r) depends on forecasts of debt default probability, and thus on expectations of
exportandoutputgrowth,ontheonehand,andontherisktoleranceofinvestors,ontheother.
2. CapitalͲMarketShocks
Oneshockisanincreaseinratesofinterestinacountrywithamajorfinancialmarket.Thisshockaffects
interestratespaidbydevelopingcountrysovereignborrowers(if)intwowaysasEquation[V.2]indicates:
firstly,bysimplyraising)theriskfreerate(iw),and,secondly,byraisingtheriskpremium(r)owingtothe
increaseinthedebtͲserviceratio(ʍ).
However, another more commonly observed market shock results from shifts in the demand for
emergingͲmarketdebtduetochangeswithindevelopedcountrymarkets–suchaschangesinregulations,
fluctuationsinriskaversionamongstlendersandinvestors,orcontagionfromotherdebtors.Theselead
to“horizontal”downwardshiftsintheassetdemandcurveinFigureIV.1.
The macroeconomic and distributional consequences for emerging markets can be disproportionately
large.69This results from a fundamental asymmetry in international capital markets: while capital flows
arerelativelysmallinrelationtothehomeeconomiesoflendersandinvestors,theyaremuchlargerin
relation to host markets. The effect of the shocks is exacerbated by hysteresis: 70 owing to the
irreversibilityofinvestmentandwageͲpricestickiness,adownswingdoesnotleadtheeconomybackto
whereitwasbeforetheupswing.FluctuationsintherealexchangerateassociatedwithshortͲtermcapital
flowsalsoleadfirmstomisallocateinvestmentbetweenthetradedandnonͲtradedsectors,withnegative
consequencesforgrowth.71
Another potentially serious negative effect of debt shocks on growth is not felt directly through the
balanceofpaymentsbutratherthroughtheeffectoninvestoruncertaintyaboutfuturemacroeconomic
conditions72andpolicychanges73whenthedebtlevelexceedstheprudentiallimit,apositioncommonly
knownas“debtoverhang”.Butthisriskcanbereducedbygovernmentaction.EveniftheGovernment
cannotcrediblypreͲcommittorepaydebt,investingingrowthbeforeborrowingcanmakeforeignlenders
aswellasdomesticinvestorsmoreoptimisticaboutgrowthprospects.
3. GlobalTradeShocks
Globaltradeshockscantakevariousformswhichinclude:
69
See FitzGerald (2001). Interestingly, this was the position taken by the IMF in the 1998 World Economic Outlook (‘Financial
Crises:CharacteristicsandIndicatorsofVulnerability’).However,by2005theWorldEconomicOutlookhadbecomemuchmore
sanguine,attributingmostofemergingͲmarketvolatilitytodomesticfundamentals.
70
A model of this process is set out in Chapter 6 of FitzGerald (2003). On the macroeconomic theory of hysterisis and pathͲ
dependencyseeHeijdra&vanderPloeg(2002),Chapter2.2andAppendixA.6.4.
71
SeeFitzGerald&Perosino(1999).
72
SeeFitzGerald,Jansen&Vos(1994).
73
SeeRodrick(1991).
80
CompendiumonDebtSustainabilityandDevelopment
x
x
x
Suddenmovementsinexportprices,particularlyforprimarycommodities,duetodemandshifts
indevelopedcountriesorsupplychangesbyotherproducers;
Unexpectedshiftsinimportprices,particularlythoseforessentialcommoditiessuchasoil;and
Thelossofaccessforexporterstoparticulardevelopedcountrymarketsduetochangesintrade
barriersordomestic(e.g.health)regulations.
TheseshocksobviouslyhaveaneffectonthedebtserviceͲexportratiobychangingthedenominator:for
example,asuddenfallinprimarycommoditypriceswillraisethisratio,eventhoughdebtserviceitselfhas
not changed, and can render a previously sustainable debt unsustainable. SecondͲorder effects depend
uponwhatpolicyresponsetheauthoritiestake.Insummarytheyhavefouroptions:
x
Incurringmoredebtinordertosustainimportlevelsandmaintainthelevelofeconomicactivity.
Thisislikelytoappreciatetherealexchangerate(oratleastpreventitfromdepreciating)andto
preventanincreaseinexports.TheresultisafurtherriseinthedebtͲserviceratiothroughlower
exportsinadditiontothehigherdebt.
x
MaintainingthedebtlevelandallowingthecurrencytodepreciateinordertoimprovethecurrentͲ
accountbalanceandstimulateexports.InthiscaseexportsdonotfallandthedebtͲserviceratio
does not increase. However, owing to the increased burden of servicing the external debt in
domesticcurrencythefiscalbalanceisworsenedwithconsequentcutsinsocialexpenditurecuts.
Moreovertheincomedistributionworsenswithdecliningrealwagesandinflation.
x
Maintaining the debt level and stabilizing the real exchange rate. This is likely to be associated
with cutting the level of economic activity in order to depress imports, prevent inflation and
balancethecurrentaccount.
x
Any one of the above policies combined with a reallocation of debt funds to exports with good
marketssoastomaintainexportgrowthandthusreducethedebtͲserviceratio.
Which policy option is adopted determines the impact of a trade shock on debt sustainability. The
domestic policy choice between exchangeͲ rate shifts and demand management depends on local
economic structures and political processes, as well as pressures from creditors or international
institutions. The “golden rule” in this context is well known: “treat negative shocks as permanent and
positiveshocksastemporary”.Itisclearlybettertoreducedebtinresponsetoimprovedtradeconditions
thantoincreaseitwhentheydeteriorate.Nonetheless,developingͲcountryGovernmentsfrequentlydo
theexactopposite:increasingdebtduringdownswingsandnotreducingitagaininupswings.Moreover
the tendency to apply public external debt to nonͲtraded sectors (which is often encouraged by the
internationalinstitutions)reducestheabilitytocopewithtradeshocks.
F. Conclusions:PrinciplesforDebtManagementinDevelopmentStrategies
1. TheParametersofDebtPolicy
ThemainindicatorsunderlyingprudentdebtmanagementareshowninTableIV.5.Itisclearthatthose
economies with unsustainable debt (i.e. the UDCs which were in arrears and/or undertook debt
reschedulingduringthe1997Ͳ2001period)stillhaveveryhighdebtͲGDPratiosclosetotheconventional
upperboundof60percent.Thisceilingisderivedfromexperienceofcountrieswhichgetintomajordebt
81
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
difficultiesandisinfactconsiderablylowerthanthatsetbytheWorldBank74inthecontextoftheHIPC
initiative.
ThoseUDCswhichdependonprivatecreditorsaremainlyinLatinAmerica:theirdebtͲserviceratiosand
interest payments as a proportion of debt are higher and the average maturity shorter than for other
countriesintheregion.TheLatinAmericanUDCsaretypicallysufferingfromaliquidityproblem,reflected
in the fact that the ratio of reserves to shortͲterm obligations (or “quick ratio” as it is known by debt
traders)islessthanunity,makingthemsusceptibletospeculativeattack.TheAfricanHIPCsincontrast,
appear to be insolvent rather than illiquid: their inability to repay principle results in very long implicit
maturities(i.e.yearsrequiredtopayoffdebtatpresentratesofamortization).Inmarkedcontrast,Asia
appearstobebothsolventandliquid.
TableIV.5.IndicatorsofDebtVulnerability,2003
Total
ExternalDebt(percentofGDP)
38.1
DebtService(percentofexports) 17.9
InterestPayments(percent
3.9
ofdebt)
Implicitmaturity(years)
8.2
LiquidityRatio
1.73
of
Developing
which:
Asia
25.4
11.1
3.6
8.6
3.16
LatinAmerica
Africa UDC HIPC
andCaribbean
43.9
49.9 63.2 86.9
45.8
13.1 29.2 10.2
5.4
2.8 3.3
1.1
5.7
0.74
16.1 9.4
2.04 1.14
48.7
2.86
Source:Author’scalculationsbasedonIMF(May2005).
Notes: “UDC” are Unsustainable Debt Countries’ with arrears and/or rescheduling during 1997Ͳ2001; HIPC are “highly
indebtedpoorcountries”underconsiderationbytheWorldBankandIMFfordebtcancellation;interestpayments
divided by debt outstanding should be compared with longͲterm average interest rates in advanced economies
averaging 5 percent in this period; “implicit maturity” is outstanding debt divided by amortization payments;
“liquidity ratio” is the ratio of reserves to payments in the form of interest and principle on shortͲ term debt
principleplusservicepaymentsonlongͲtermdebt.
Nonetheless,asTableIV.6.shows,sustainabilityas measuredbyallindicatorsandforalldebtorclasses
clearlyimprovedbetween1996and2003.Thisappearstobeduetoexportgrowthandcontrolofimports
(whichallowedcurrentͲaccountbalancestomoveintosurplusinmanycountries,especiallyinAsiaand
LatinAmerica)ratherthantosignificantreductionsindebtlevels.Indeed,allregionsappeartoberunning
current account deficits that are less or surpluses which are larger than indicated by the prudent rules
showninthetable.
Theseindicatorspointtotheeffectsofcreditrationingonthepartofcreditorsandstabilizationeffortson
thepartofdebtors.Theysuggestthatthereisroomtoinitiateanewcycleofincreaseddebtlevelssolong
asitisaccompaniedbyprudentmacroeconomicpolicy.
74
See World Bank (2004). ‘Moderately indebted’ countries are those with a ratio of the present value of contracted debt
payments (PV) to GNP of over 132 percent and of PV to exports of goods and services (XGS) of over 48 percent, while ‘highly
indebted’countrieshavePV/GNPofover220percentandPV/XGSofover80percent.Noexplanationisgivenforhowtheseexact
figures are derived. These ratios are also difficult to compare with the IMF data used in this paper because the ratio of PV to
nominaldebtdependsonthetermsofthedebtitself.
82
CompendiumonDebtSustainabilityandDevelopment
TableIV.6.ChangesinDebtSustainability1996Ͳ2003
ExternalDebt(percentofGDP)
1996
2003
GDPgrowth(percent)
1996Ͳ2003
CurrentͲaccount balance (per
centofGCESR(2005para.117).DP)
1996
2003
“prudentvalue”(ʍ)
DebtService(percentofexports)
1996
2003
Exportgrowth
1996Ͳ2005
CurrentͲaccount balance (per
centofexports)
1996
2003
“prudentvalue”(ʍ)
Total
of
which:
Developing
Asia
37.8
38.1
5.1
31.2
25.4
6.6
Latin
America
and
Caribbean
35.0
43.9
2.6
Ͳ
Ͳ
Ͳ1.9
21.5
17.9
10.8
Ͳ
Ͳ
Ͳ2.1
Africa
UDC
HIPC
69.0
49.9
3.9
51.0
63.2
3.5
126.9
86.9
4.8
Ͳ1.9
3.1
Ͳ1.9
Ͳ2.2
0.3
Ͳ1.0
Ͳ1.1
Ͳ0.1
Ͳ2.3
Ͳ
Ͳ
Ͳ2.0
Ͳ
Ͳ
Ͳ5.1
13.9
11.1
12.0
46.7
45.8
7.0
20.3
13.1
8.1
29.2
29.2
8.1
22.6
10.2
7.1
Ͳ7.3
8.9
Ͳ1.5
Ͳ14.7
1.4
Ͳ3.2
Ͳ3.6
Ͳ0.29
Ͳ1.4
Ͳ
Ͳ
Ͳ2.4
Ͳ
Ͳ
Ͳ1.24
Source:author’scalculationsfromIMF(May2005).
Note:fordefinitions,seeSectionIIIabove..Thedebt/GDPanddebtͲservice/exportlevelsusedinthecalculationofprudential
CABand“goldenrule”arethesimpleaveragesof1996and2003.
2. PolicyImplicationsforDevelopingCountries
Debt levels must clearly be kept within prudent limits and Governments should make credible
commitments to keep within these constraints, employing appropriate legislation if necessary. Such a
policyisessentialtoreduceuncertaintyfordomesticfirmswhicharethemainvehiclesfortheinvestment
on which growth depends. A debt overhang and the prospect of deflationary stabilization policies and
debtrestructuring(orevenmoratoria)implyfuturelossesofsales,profitsandassetvalues.
Debt should be contracted on the longest terms possible. The cost of servicing should be kept at a
minimum subject to appropriate control over the vulnerability to future capitalͲmarket or worldͲtrade
shocks.Suchcontrolmayimplythathigherinterestratesareareasonablepricetopayforloansoflonger
maturityifvulnerabilitycanbereducedthereby.75
The use of funds generated by external debt should be geared to ensuring repayment capacity. This
meansthatasubstantialproportionofthesefundsshouldbeallocatedtothesupportofexportgrowth.
ThisdoesnotimplythattheGovernmentshouldbedirectlyengagedinexportproductionbutratherthat
75
Missale(1999)demonstrateshowthisprinciplehasbeenappliedinOECDcountries.
83
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
funds should be used to support appropriate infrastructure provision, the supply of longͲterm credit to
exporters,andtrainingfortheworkforce.
The support of export growth also involves maintaining a competitive real exchange rate, which has
implications bothfor nominalexchangeratemanagementand forwagebargainingpolicy.ThepoliticalͲ
economy constraints on excessive reduction of real wages are best countered by appropriate
commitments to output and employment growth. Low real interest rates and an expansionary credit
policyareneededtosupporttheinvestmentratelikelytoberequiredforthetargetrateofgrowth.Thisin
turn means that the domestic financial system should to some extent be shielded from international
capitalmarkets.
The recent popularity of inflation targeting as the core of stabilization policy in emergingͲmarket
economiesdoesnothelpreducedebtvulnerabilitybecauseithastheeffectofincreasingvulnerabilityto
cyclicalcapitalflows.Openingofthecapitalaccountandafloatingexchangeratehasbeenaccompanied
by reliance on a single monetary policy instrument (the interest rate) and rigid fiscal rules in emergingͲ
marketeconomiesinpursuitofpricestability.Thisprecludesnotonlycountercyclicalmonetaryandfiscal
policybutalsotheuseoftheexchangeratetomaintainexportcompetitivenesswhichisakeyelementof
prudent debt management. There is a strong argument for emergingͲmarket Governments to adopt a
counterͲcyclical monetary stance in response to capital flows. This would need to be supported by real
exchangeͲratetargeting,bankcreditregulationandamoreactivefiscalstance.76
IfsuchapolicyistobesuccessfulinmiddleͲincomecountrieswithsubstantialshortͲtermprivatecapital
flows, there is thus a strong case for intervention through various controls to reduce the volatility of
capital flows.77These controls now usually take the form of taxes, regulatory measures (such as setting
specialreserveordepositlevelsforinflows),andtargetedmoneyͲmarketoperations,whilequantitative
controlshavebecomelesscommon.
3. PolicyImplicationsfortheInternationalCommunity
Thereareanumberofotherpolicyareasthatcanonlybeaddressedbytheinternationalcommunity.
Substantial debt reduction has not yet been forthcoming, even for HIPC countries, due to difficulties in
budgetaryallocationsforthecorrespondingassetwriteͲdowns.Thisisaninternalaccountingmatterfor
OECDcountriesandrequiresurgentsolution.Furtherdebtrestructuringcanreducetheliquidityproblem
ofdebtͲservicepressureonthecurrentaccount.However,itdoesnotreducetheinvestmentdisincentives
fromdebtoverhangandmayevenmakethemworsebyincreasinguncertainty.78
Giventhatexportgrowthisakeycomponentofprudentdebtmanagement,accesstoOECDmarketsfor
developing country exporters is crucial to their ability to contract debt prudently, while accelerating
economicgrowthandpovertyreduction.Thesameistrueofmeasurestoreducespeculativefluctuations
inprimarycommodityprices.Ideally,thesewouldbecombinedwithlinkageofdebtrepaymentstoexport
levels–atleastinthecaseofpaymentstoofficialcreditors.79
Since capital shocks to developing countries usually originate within OECD financial markets, policy
towardsthemshouldbebasedonrecognitionoftheirexternalcharacter.Onesteptoreducetheimpact
of these exogenous shocks would be for the IMF to provide temporary finance on a larger scale, more
quicklyandwithlessconditionalityinordertofacilitatesmoothdebtmanagement.Inthelongerrun,itis
76
SeeFitzGerald(2005b).
SeeFitzGerald(2005a).
78
Thiseffectfaroutweighsanypotentialmoralhazardimplicitin‘frontloading’debtforgiveness.
79
In practice markets are very unlikely to accept sovereign bonds with yields linked to commodity exports. However, certain
primarycommoditiescanbeusedascollateralforborrowing.
77
84
CompendiumonDebtSustainabilityandDevelopment
essentialtodeepenthemarketfordevelopingͲcountrydebtinOECDcountriesby:lengtheningthetenor
of instruments, taking measures to increase their liquidity, and encouraging their inclusion in the
investmentsofpensionandinsurancefunds.80
80
Suchmeasureswouldalsoreducetheriskinessofsovereigndebt.
85
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
References
AghionPandHowittP(1998).EndogenousGrowthTheoryCambridge.MA,MITPress.
Addison D (1989). The World Bank Revised Minimum Standard Model: Concepts and Issues. Working
PaperSeries231.WashingtonDC,WorldBank.Anupdated(1999)versionofthispaperplusother
RMSMdocumentsareavailableat:www.worldbank.org/data/rmsm/index.htm.
Agenor PR and Montiel PJ (1999). Development Macroeconomics. 2nd edn Princeton, NJ, Princeton
UniversityPress.
AvramovicDetal.(1964).EconomicGrowthandExternalDebt.Baltimore,MD,JohnsHopkinsUniversity
Press.
Bacha EL (1990). A ThreeͲGap Model of Foreign Transfers and the GDP Growth Rate in Developing
Countries.JournalofDevelopmentEconomics,32.
BaqirR,RamcharanRandSahayR(2003).IMFProgramDesignandGrowth:WhatistheLink?Research
WorkshoponMacroeconomicChallengesinLowIncomeCountries,Washington,DC,International
MonetaryFund.
BuiterWH(1990).PrinciplesofBudgetaryandFinancialPolicy.NewYork,HarvesterWheatsheaf.
Chenery HB and Strout AM (1955). Foreign Assistance and Economic Development. American Economic
Review,65.
Cohen D (1994). Growth and External Debt. In Van der Ploeg F ed. Handbook of International
Macroeconomics,Oxford,Blackwell
Cunningham A, Dixon L, Hayes S (2001). Analysing Yield Spreads on Emerging Market Sovereign Bonds.
BankofEnglandFinancialStabilityReview,175,December.
Dornbusch RandHelmersF(1988).TheOpenEconomy:ToolsforPolicymakersinDevelopingCountries.
NewYork,OUP.
EasterlyW(1999).TheGhostoftheFinancingGap:TheGrowthModelUsedintheInternationalFinancial
Institutions.JournalofDevelopmentEconomics,60.
FitzGerald V, Jansen K and Vos R (1994). External Constraints on Private Investment Decisions in
Developing Countries. In: Gunning JW, Kox H, Tims W and de Wit Y eds. Trade, Aid and
DevelopmentBasingstoke.Macmillan.
FitzGeraldV(2001).ShortͲTermCapitalFlows,theRealEconomyandIncomeDistributioninDeveloping
Countries. In: GriffithͲJones S, Montes M and Nasution A eds. ShortͲterm Capital Flows and
EconomicCrises.Oxford,OxfordUniversityPressforUNU/WIDER.
FitzGeraldV(2003a).GlobalMarketsandtheDevelopingEconomy.Basingstoke,Palgrave.
FitzGerald V (2003b). The Instability of International Demand for Emerging Market Assets. In: FfrenchͲ
DavisRandGriffithͲJonesSeds.FromCapitalSurgestoDrought:SeekingStabilityforDeveloping
Economies.Basingstoke,PalgraveforUNU/WIDER.
FitzGerald V (2005a). Policy Issues in Market Based and NonͲMarket Based Measures to Control the
Volatility of Portfolio Investment. In: Green CJ, Kirkpatrick C and Murinde V eds. Finance and
Development:SurveysofTheory,EvidenceandPolicy.Aldershot,EdwardElgar.
FitzGeraldV(2005b).MonetaryModelsandInflationTargetinginEmergingMarketEconomies.In:Arestis
P, McCombie J and Baddeley M (2005). The New Monetary Policy. Cheltenham, UK and
Northampton,MA,EdwardElgar.
86
CompendiumonDebtSustainabilityandDevelopment
FitzGerald V (forthcoming). International Risk Tolerance, Capital Market Failure and Capital Flows to
Emerging Markets. In: A. Shorrocks ed. Thinking Ahead: the Future of Development Economics.
Basingstoke,PalgraveforUNU/WIDER.
FitzGeraldVandBabilisS(2005).RiskAppetite,HomeBiasandtheUnstableDemandforEmergingMarket
Assets.InternationalReviewofAppliedEconomics,19(4).
FitzGeraldVandPerosinoG(1999).TradeLiberalization,EmploymentandWages:ACriticalapproach.In:
BarbaͲNavarretti G, Faini R and Zanalada G eds. Labour Markets, Poverty and Development,
Oxford,Clarendon.
Griffin KB (1970). Foreign Capital, Domestic Savings, and Economic Development. Oxford Bulletin of
EconomicsandStatistics,32(2).
Heijdra BJ and van der Ploeg F (2002). The Foundations of Modern Macroeconomics. Oxford, Oxford
UniversityPress.
IMF (1987). Theoretical Aspects of the Design of FundͲSupported Adjustment Programs. IMF Occasional
Paper55.WashingtonDC;InternationalMonetaryFund.
IMF(2003).PublicDebtinEmergingMarkets–IsittooHigh?WorldEconomicOutlook2003,ChapterIII.
IMF,WorldEconomicOutlook.
KhanMS,MontielPJandHaqueNU(1990).AdjustmentwithGrowth:RelatingtheAnalyticalApproaches
oftheWorldBankandtheIMF.JournalofDevelopmentEconomics,32.
McCombie J and Thirlwall AP (1994). Economic Growth and the Balance of Payments Constraint.
Basingstoke,Macmillan.
MissaleA(1999).PublicDebtManagement.Oxford,OxfordUniversityPress.
ObstfeldMandRogoffK(1995).TheIntertemporalApproachtotheCurrentAccount.In:GrossmanGM
andRogoffKeds.HandbookofInternationalEconomics,3.Oxford,Blackwell.
PolakJJ(1957).MonetaryAnalysisofIncomeFormationandPaymentsProblems.IMFStaffPapers,6(1).
RebelloS(1991).LongRunAnalysisandLongRunGrowth.JournalofPoliticalEconomy,94.
Rodrik D (1991). Policy Uncertainty and Private Investment in Developing Countries. Journal of
DevelopmentEconomics,36.
RomerPM(1986).IncreasingReturnsandLongRunGrowth.JournalofPoliticalEconomy,94.
SachsJD(1989).DevelopingCountryDebtandtheWorldEconomy.Chicago,IL,UniversityofChicagoPress
forNBER.
Sen P (1994). Savings, Investment, and the Current Account. In: van der Ploeg F ed. The Handbook of
InternationalMacroeconomics.Oxford,Blackwell.
TaylorL(1994).GapModels.JournalofDevelopmentEconomics,45.
Taylor L (ed.) (1993). The Rocky Road to Reform: Adjustment, Income Distribution, and Growth in the
DevelopingWorld.Boston,MA,MITPressforUNU/WIDER.
TranͲNguyen AN (2004). Debt ServicingͲCapacity of Developing Countries: Conceptual Approaches and
MeasurementProblems.Geneva,UNCTAD(processed).
UNCTAD(2005).WorldInvestmentReport2005.Geneva,UNCTAD.
WorldBank(2004).GlobalDevelopmentFinance2004.Washington,DC,WorldBank.
87
AnAnalyticalFrameworkforDebtSustainabilityandDevelopment
WorldBank(2005a).WorldDevelopmentIndicators.
WorldBank(2005b).WorldDevelopmentReport.
88
CompendiumonDebtSustainabilityandDevelopment
CHAPTERV
THEDEBTEXPERIENCESOF
UGANDA,KENYAANDBOLIVIA
DamoniKitabire,PeterMichaelOumo,FrancisM.Mwegaand
PaulBeckerman81
A. Introduction
This chapter reviews the debt experiences of three of the world’s poorest economies, namely Uganda,
KenyaandBolivia.Thechapterhighlightstheconditionsanddebtproblemsthatunderpinnedthefailure
ofsuccessivedebtinitiativestorendertheirdebtpositionsustainable.
Amajorcontributoryfactortothisfailureisthatthethreecountries’exportsremainconcentratedona
handful of commodities, all of which suffered significant deteriorations in the terms of trade since the
1980s. Moreover, the three countries also experienced severe climatic shocks, such as severe droughts
(Kenya)andElNino(Bolivia).Tothesefactorsmustbeaddedpoliticalturmoil,instabilityandwars.
Uganda and Bolivia have records of having been exemplary pupils of Washington Consensus policies.
Kenya followed similarly orthodox approaches to macroͲeconomic management, albeit against a
background of turbulent relations with creditors. All three countries went through a succession of
programs. The reform efforts revived growth at the outset but sustained per capita gains failed to
materialize.
ThecountryreviewsinsectionsII,III,andIVhighlighttheroleoffactorsaffectingsustainabilitythatshould
havebeenincorporatedinpastdebtreliefanalyseswithspecialemphasisonexportdiversification,fiscal
81
Section B is based on a paper by Damoni Kitabire and Peter Michael Oumo (Ministry of Finance, Planning and Economic
Development of Uganda) (Kitabire and Oumo (2005)), Section C is based on a paper by Francis M. Mwega (Department of
Economics, University of Nairobi) (Mwega (2005)), Section D isbased on a paper by PaulBeckerman (Independent Consultant)
(Beckerman(2006)).
89
TheDebtExperiencesofUganda,KenyaandBolivia
positions, and new financing. The three experiences are then compared and contrasted in Section E.
SectionFsummarizestheprincipalfindings.
B. Uganda
1. Introduction
Despite three decades of attempts to reduce the external debt burden, debt sustainability still eludes
Uganda.Thecountry’sprincipaldebtproblemhasbeenitsheavydebtserviceburden.Despiteremarkable
GDPgrowthsincethe1990sandimprovementsinexportearnings,theeconomyremainsdependenton
rainͲfedagricultureandvulnerabletoshocksdeliveredbyworldcommoditymarkets.Thecountryisalso
stillheavilydependentondonoraidwhichcurrentlyfinancesabout40percentofthebudget.
Since the 1991 debt crisis, Uganda has developed a fairly coherent debt strategy. However, its debt
burden remained high until the HIPC Initiative put Uganda on a sustainable debt path momentarily.
UnfortunatelytheHIPCInitiativedidnotleadtoapermanentexitfromdebtproblems.Thecountryhas
borrowedheavilypostͲHIPCtoachievetheMDGsanditsdebtindicatorsareunsustainableagain.
2. EconomicPerformanceandPolicies
Uganda entered the 1980s with a degree of political stability that allowed GDP growth to recover to a
positive 1.7 per cent in 1980Ͳ1983. Thereafter, industrial production declined due to foreign exchange
shortageandthepoorstateofinfrastructure,whileagriculturalproductionalsolagged.In1983/84fiscal
there was fiscal slippage on an IMF stabilization program which was cancelled in late 1984. Political
instabilityandaprotractedguerrillawarledtoanewGovernmenttakingpowerinJanuary1986.
InMay1987,thenewGovernmentembarkedonanEconomicRecoveryProgrammewiththesupportof
IMF,WorldBankandothers.Thiswasfollowedin1989byaStructuralAdjustmentProgramme(SAP).Its
focus was on limiting the involvement of the state in economic activities, the liberalization of trade,
financialͲsectorandmarketingactivities,theprivatizationanddivestitureofpublicenterprises,andmore
generally the promotion of privateͲsector participation in production. The program resulted in an
acceleration of GDP growth to an average rate of 6.9 per cent per annum between 1991/92 and
1999/2000(seeFigureV.1.).
By 2000, the structure of the Ugandan economy had changed dramatically. In 1982/83, agriculture
accountedfor53.6percentofGDP,butitssharedeclinedto36.3percentin2004/05.Atthesametime,
thesharesofindustryandservicessteadilyincreased,thatofservicesrisingfrom35.2percent1982/83to
36.6percentin1990/91andbecomingthelargestin2001/02.However,Ugandaremainsvulnerableto
weatherchangesasthecountry’sagriculturalsystemreliesheavilyonrainͲfedsmallholderagriculture.
90
CompendiumonDebtSustainabilityandDevelopment
FigureV.1.Uganda’sGDPGrowth1982/83–2004/05
(Percent)
25%
20%
15%
10%
5%
0%
3
/8
-5%
82
19
/03
/99
/01
/05
/97
/93
/95
/87
/89
/91
/85
00
02
98
04
94
96
88
90
92
84
86
20
19
20
20
19
19
19
19
19
19
19
-10%
Agriculture
Industry
Services
GDP at Market Prices
Source:UgandabureauofStatistics.
Largelydependentonprimarycommodities,Uganda’sexportgrowthhasbeenerratic.Followingreform
efforts,growthreboundedintheearly1990s.Thiswasreinforcedbythecoffeepriceboomof1993/94Ͳ
1996/97. Following efforts to diversify away from coffee, the share of coffee in Uganda’s exports has
declined from 70 per cent in the 1990s to about 20 per cent since 2000/01. Fish has become Uganda’s
leadingexport,followedbycotton,tea,tobacco,andflowers.
As shown in Figure V.2., Uganda’s terms of trade (TOT) have been erratic but with an overall secular
decliningtrend,largelydeterminedbytheinternationalpriceofcoffee.TheTOThaverecentlyimproved
andchangeshavebeenpositivesince2002/03.Deterioratingtermsoftradehaveadirectimpactondebt
sustainability.Coffeeexportpricesin2003/04were49percentlowerthanenvisagedatthetimeofHIPC
IIcompletion.64percentofthedeteriorationintheratiooftheNPVofdebttoexportsbetween2002
and2004wasduetofallingcoffeeexportprices.
FigureV.2.ChangesinUganda’sTermsofTrade1989/90–2004/05
(Percent)
100%
80%
60%
40%
20%
19
89
/9
19 0
90
/9
19 1
91
/9
19 2
92
/9
19 3
93
/9
19 4
94
/9
19 5
95
/9
19 6
96
/9
19 7
97
/9
19 8
98
/9
19 9
99
/0
20 0
00
/0
20 1
01
/0
20 2
02
/0
20 3
03
/0
20 4
04
/0
5
0%
-20%
-40%
Source:BankofUganda.
91
TheDebtExperiencesofUganda,KenyaandBolivia
Under the Economic Recovery program initiated in 1987, reforms in trade policy gradually eased
quantitative restrictions and were geared towards export promotion. Trade licensing schemes were
abandonedandcoffeemarketingwasliberalizedinthelate1980s.In1992,thetaxoncoffeeexportswas
abolished.Itwasbrieflyreintroducedin1994tolimittheappreciationoftheexchangerateasaresultof
thecoffeeboom,andabolishedagainin1996.Importdutieswererationalizedin1992toarangeof10Ͳ60
percent,andwerefurtherreducedtoarangeof10Ͳ50percentin1994.
Initially, the exchangeͲrate policy involved repeated devaluations and rationing of the available foreign
exchange under various schemes. A foreignͲcurrency retention scheme was introduced in 1988 and
extended in 1989. In 1990, the exchange market was liberalized with the legalization of the parallel
(kibanda) market. In 1992, an exchange rate auction market was created. The foreign exchange market
was fully liberalized and the exchange rate was floated in 1993. In 1997, the capital account was
liberalized.
Topromoteforeigninvestment,UgandaenactedanInvestmentCodein1991.ThisreversedlongͲstanding
antipathy towards foreign investment and introduced standard provisions regarding investment
incentives, profit repatriation and protection against expropriation. FDI rose from US$43.2 million in
1992/93toUS$670millionasofend2004/05.
Successive reforms have enabled Uganda to manage its fiscal balances more prudently but have not
reduced the country’s dependency on donor aid for financing its budget. In the 1990s, over half of
Uganda’sbudgetwasfundedbydonoraidandthisratiowasstill40percentin2005/06.Upto1996over
half of the aid received was in the form of loans, though grants became more important subsequently
(AtingiͲEgo2005).
DealingwiththeDutchDiseaseeffectsoftheseflowshasbeenthesourceofasignificantriseindomestic
debtservicing.DutchDiseaseeffectsputappreciationpressuresontheexchangerate,withinterestrate
rises owing to the attempt to contain the inflationary effects of the inflows on liquidity. According to
AtingiͲEgo (2005), since 1998 Dutch Disease effects in Uganda have adversely affected investment and
imports.
3. ExternalDebt
Uganda’s debt problems date back to the 1980s. Debt continued to accumulate (despite the
Government’s increasing inability to service it) due to continuing foreign exchange shortages. By 1986,
Uganda’s debt stock had grown to US$1.4 billion, up from US$680 million in 1980. Between 1986 and
1990,becauseofthereconstructionandrecoveryprogramandofalackofaneffectivedebtmanagement
strategy, both the debt stock and debt service went out of control. Large sums were borrowed on
unfavorabletermsandarrearsaccumulated,theburdenbeingexacerbatedbydelinquentprivateͲsector
loansguaranteedbytheGovernment.
By the late 1980s, Uganda faced a debt crisis. In 1990, the Government ran out of foreign exchange
following a sharp decline in terms of trade due largely to a decline in the price of coffee. Debt service
obligations amounted to over 60 percent of export earnings. Drastic action was therefore necessary to
reversethecollapseinthebalanceofpayments,promptingthedevelopmentofUganda’sfirstintegrated
debtmanagementstrategyin1991.
AsshownintableV.1.,mostofUganda’sdebt(63percentin1991to88percentasof2004)isowedto
multilateral institutions and is therefore longͲterm. Owed mainly to IDA and ADF the debt is also on
concessionalterms,i.e.has10yearsofgrace,andarepaymentperiodof30yearsforIDAandof40years
forADF.ThedebttoGDPratiohasdeclineddrasticallyfromapeakof98percentin1992andhasrecently
stabilized in a range of 60 Ͳ70 per cent. The ratio of debt service to exports has also undergone sharp
92
CompendiumonDebtSustainabilityandDevelopment
fluctuationsbutsincetheendofthe1990shasstabilizedataround20percentlargelyduetoHIPCdebt
reliefinitiativeandthepromotionofnonͲtraditionalexports.
TableV.1.Uganda’sDebtStructureandIndicators1980–2004
US$ Million
Total Debt Stock
o/w arrears
Multilateral
Bilateral
o/w Paris Club
Non Paris Club
Other
Mulitilateral (% Debt Stock)
Bilateral (% Debt Stock)
Other (% Debt Stock)
Debt Service
Debt/GDP
Debt Service/Exports
Debt /Exports Ratio
1980
689.0
101.0
n.a
n.a
n.a
n.a
n.a
1986
1,422.0
92.0
n.a
n.a
n.a
n.a
n.a
1987
1,945.0
99.0
n.a
n.a
n.a
n.a
n.a
1988
1,923.0
136.0
n.a
n.a
n.a
n.a
n.a
1989
2,177.0
190.0
n.a
n.a
n.a
n.a
n.a
1990
2,583.0
298.0
n.a
n.a
n.a
n.a
n.a
1991
2,591.6
370.9
1,643.6
811.8
285.5
526.3
136.2
1992
2,647.4
583.1
1,755.9
651.4
273.2
378.2
240.1
1993
2,637.1
301.8
1,815.9
697.3
281.7
415.6
123.9
1994
2,999.4
251.0
2,156.1
730.4
332.0
398.4
112.9
1995
3,386.9
233.4
2,487.9
787.9
380.1
407.9
111.1
1996
3,515.8
250.3
2,655.1
755.1
350.6
404.5
105.6
1997
3,660.2
316.6
2,763.0
796.0
339.1
456.9
101.2
1998
3,631.6
275.5
2,826.8
748.6
325.0
423.6
56.2
1999
3,499.6
241.5
2,782.6
649.9
288.2
361.7
67.1
2000
3,579.9
232.0
2,936.3
593.2
260.6
332.6
50.5
2001
3,397.5
286.8
2,893.3
476.1
131.5
344.7
28.1
2002
3,825.2
301.7
3,318.1
488.5
111.4
370.1
18.5
2003
4,296.3
318.7
3,720.4
555.5
122.8
432.7
20.4
2004
4,310.0
342.9
3,782.8
510.3
66.1
444.2
16.9
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
n.a
63.4%
31.3%
5.3%
66.3%
24.6%
9.1%
68.9%
26.4%
4.7%
71.9%
24.4%
3.8%
73.5%
23.3%
3.3%
75.5%
21.5%
3.0%
75.5%
21.7%
2.8%
77.8%
20.6%
1.5%
79.5%
18.6%
1.9%
82.0%
16.6%
1.4%
85.2%
14.0%
0.8%
86.7%
12.8%
0.5%
86.6%
12.9%
0.5%
87.8%
11.8%
0.4%
57.0
54.6%
17.2%
208.2%
172.0
32.7%
43.2%
357.3%
160.0
66.4%
43.8%
532.9%
202.0
54.2%
62.3%
593.5%
186.0
147.0
148.0
131.0
140.6
167.8
59.5%
86.0%
83.4%
98.2%
87.4%
81.3%
61.2%
59.8%
66.1%
65.2%
83.2%
66.1%
716.1% 1050.0% 1157.0% 1317.1% 1560.4% 1180.9%
150.7
64.3%
22.6%
507.8%
142.2
64.0%
19.7%
486.3%
155.9
64.0%
18.9%
443.8%
154.6
59.5%
24.4%
573.1%
162.9
62.0%
22.4%
481.8%
133.4
65.9%
20.1%
540.0%
146.1
65.0%
21.6%
501.9%
133.6
71.3%
19.1%
545.7%
172.0
74.4%
22.2%
555.1%
179.7
68.8%
19.4%
464.4%
Source:MinistryofFinance,Planning&EconomicDevelopmentandBankofUganda.
Firstattemptsatdevelopingadebtmanagementsystemcamein1983withtheformationoftheExternal
Debt Management Office (EDMO) within the Bank of Uganda (BoU). In 1986, two other offices; the Aid
CoordinationUnit(ACU)intheMinistryofFinancePlanningandEconomicDevelopment(MFPED),andthe
TreasuryOfficeofAccounts (TOA)were mandated tomanageand disburse external debttogetherwith
the EDMO. The ACU, now called Aid Liaison Department (ALD), was responsible for seeking and
negotiatingnewloansinlinewithGovernment’sfinancingrequirements.
The key features of Uganda’s debt adopted in the 1995 strategy include seeking grant funding before
contracting any loan and ensuring that all loans are strictly on IDAͲcomparable terms. Loans must be
approved by the beneficiary sector and the development committee before being contracted, and they
mustbeinlinewithsectoralandpovertyreductiontargets.LoansarethenscrutinizedbytheMinistryin
chargeandcheckedagainstbudgetarytargets,afterwhichcabinetandparliamentaryapprovalaresought.
TechnicalcapacityfordebtmanagementinUgandaiswelldeveloped.Particularlysince1995,Ugandahas
madesustained,tangibleprogressincapacitybuildinginallaspectsofdebtmanagement.Moreover,BoU
has now a good debt recording capacity and a complete upͲtoͲdate computerized database which uses
UNCTAD’sDebtMonitoringandFinancialAnalysisSystem(DMFAS).
4. DomesticPublicDebt
In Uganda the issuance of Government debt has not only had the normal function of meeting revenue
shortfalls, but also that of financing the sterilization of foreign aid inflows. As sterilization efforts
intensifiedattheendofthe1990s,Treasurybillsalesrosefrom23percentto32percentofcommercial
bankholdingsbetween1998and2004despitethefactthatdomesticdebtwastypicallyintherangeof1Ͳ
2percentofGDP(seeTableV.2.).Interestpaymentsondomesticdebt,however,doubledin1995Ͳ2000
owingtothepolicyofhighinterestratesassociatedwiththeattempttomanagetheconsequencesofhigh
aidinflows.ByaddingtopressuresonthefiscalbalancetheseinterestpaymentscontributedtopostͲHIPC
difficulties.
93
TheDebtExperiencesofUganda,KenyaandBolivia
TableV.2.DomesticDebtinUgandaandKenya,1980Ͳ2000
(Percent)
Domesticdebt
(inpercentofGDP)
Totaldebt
(inpercentofGDP)
Domestic/Totaldebt
Domesticinterest
payments/totaldebt
Source: Christensen(2004)
1980Ͳ89
1990Ͳ94
1995Ͳ00
1980Ͳ89
1990Ͳ94
1995Ͳ00
1980Ͳ89
1990Ͳ94
1995Ͳ00
1990Ͳ94
1995Ͳ00
Kenya
Uganda
Average
HIPC
21
23
22
81
100
74
25
23
29
71.7
74.5
2
1
2
2
74
59
100
1
4
14.4
29.4
11
12
15
62
102
118
25
19
22
49.7
51.9
9
6
8
69
138
169
22
6
6
42.3
42.5
Decision NonͲHIPC
Point2/
3/
10
14
18
7
8
23
73
53
143
59
164
59
25
30
35
7
7
40
43.6
60.8
43.2
65.3
Notes:
1/BothdomesticdebtsincludeTreasuryBillsandGovernmentStocks,withKenyaalsoissuingbonds
2/IncludesUganda
3/IncludesKenya
5. TheDebtStrategyfrom1991to1995
In early 1991, the Government of Uganda embarked on a comprehensive debt strategy, including a full
debt audit. The 1991 Debt Strategy focused on overcoming the immediate debt payment crisis and
developing mechanisms to ensure that it did not reͲoccur. The first objective of this strategy was to
provideasolutiontothecashflowproblemthroughdebtrestructuring.Thisnecessitatedclearingarrears
andreducingdebtservicetolevelsconsistentwithUganda’sabilitytopay.Thesecondobjectivewasto
improve debt management structures. This resulted in the strengthening of debt management by
requiringministriestoworkwiththeAidCoordinationUnit(ACU)intheMinistryofFinance.Inaddition,
strictlimitsonborrowingwereputinplace,witharequirementtoexhaustallsourcesofgrantfinancing
beforeconsideringnewloans,whichhadtocomefromhighlyconcessionalsources.
By1991,UgandahadalreadyundertakenfourrestructuringoperationswithintheframeworkoftheParis
Clubin1981,1982,1987,and1989.Unfortunately,theserestructuringoperationswerenotsufficientto
easethedebtoverhangfortworeasons.First,in theParisClubs1to3,negotiationscoveredonlydebt
fallingdueduringashortconsolidationperiod(12Ͳ18months).Second,untilParis8,onlypreͲcutoffdebt
(accountingfor4percentofthetotaldebtstock)waseligiblefordebtrelief.Moreover,thedeminimis
clauseexcludedloansoflessthanSDR500,000fromrescheduling.
The 1991 strategy also addressed the country’s commercial debt. Although this debt accounted for just
over9percentoftotaldebtstockin1992,mostofitwasinarrears.UgandaembarkeduponadebtbuyͲ
backstrategy,financedbytheWorldBank.Theofferpricewasfixedat12centsperdollarinDecember
1992,andtheclosingdatewasinFebruary1993.Overall,thebuybackwasverysuccessful.
The1991debtstrategywassuccessfulinmanyways.Itestablishedclearproceduresfornegotiatingnew
loans and strengthened debt management. It helped to increase the proportion of payments made on
time.Itledtolargereductionsincommercialdebtanddebtservice.Consequently,thedebttoGDPratio
fellfrom83percentin1991to64percentin1995.Thestockofarrearsfellfrom15percentin1991to7
percentin1993,whilemultilateraldebtincreasedfrom61percentto75percentoftotalexternaldebt
94
CompendiumonDebtSustainabilityandDevelopment
overthesameperiod.However,thisputthecountryinadifficultpositionbecausemultilateraldebtcould
notberestructured.
Therewerethreemainweaknessesinthe1991debtstrategy.Firstwastheinsufficientreductioninlong
term multilateral debt. Secondly, the country continued to require large amounts of new financing to
supportthereformprogramwiththedangerofincreaseddebtservicingobligationsifthenewfinancing
wasnotconcessionalenough.Thirdly,therewerestillsomeproblemswithdebtmanagementstructures.
A comprehensive review of the debt strategy was carried out in 1995 with help from the Swedish
Government.Thenewstrategywhichemergedfocusedonfourobjectives:
Reductionofthemultilateraldebtserviceburdenthroughbilateralgrants;
Increasingtheconcessionalityofnewborrowingandthequalityofloanfinancedinvestment;
Improvingdebtandreservemanagement;
Improvingcoordinationwithdonors,andlobbyingforlongtermmultilateraldebtreduction.In
fact, in November 1995, a Multilateral Debt Fund (MDF) was established, with contributions
usedtoservicedebt.Thestrategyalsointroducedtherequirementofparliamentaryapprovalof
newloans.
(a)
(b)
(c)
(d)
6. TheHIPCDebtReliefInitiative
InApril1998,UgandabecamethefirstcountrytobenefitfromHIPCDebtReliefInitiative.PriortoHIPC
debtrelief,thenominalvalueofUganda’sexternaldebtstockwasUS$3.5billion,andtheNPVofdebtto
exportsratiowas294percent.UnderHIPCI,UgandareceiveddebtreliefofUS$347millioninNPVterms.
Ofthisamount,79percentwasduetomultilateralcreditorssothatforthefirsttime,debtreliefhada
largemultilateralcomponent.Uganda’sNPVofdebttoexportsratiowassupposedtofall196percent,i.e.
belowthethresholdratioof202percent.
However, Uganda’s debt swiftly returned to unsustainable levels, mainly on account of the El Nino
weatherphenomenon,whichseverelyaffectedexportperformancein1999.Hence,inMay2000,Uganda
received further relief under Enhanced HIPC. Prior to this, in June 1999, Uganda’s external debt stock
reachedUS$3.6billion.TotalreliefunderHIPCIIwasexpectedtoamounttoanadditionalUS$656million,
withmultilateralcreditorscontributing83percent.ThetotalreliefundertheHIPCasawholewasUS$1
billioninNPVterms,orunderonethirdofthepreͲHIPCnominaldebtstock.
7. PostͲHIPCDevelopments
Since HIPC II completion, Uganda’s external debt sustainability as measured by NPV of debt to exports
ratio has deteriorated. Uganda’s NPV of debt to exports ratio had reached 280 per cent according to a
June2004analysis.
Anumberoffactorshavecontributedtothedeteriorationindebtindicators.
x
First,istheimpactoffallingcoffeepricesonexportearnings,whichwere57percentand36per
centlowerin2002/03and2004/05thaninitiallyenvisaged.
x
Secondly,risinginterestratesreducedtheconcessionalityofthecountry’sdebt.
x
Thirdly,attheEnhancedHIPCdecisionpoint,estimatesfornewfinancinginthemacroeconomic
framework and balance of payments projections were not fully incorporated in the Debt
SustainabilityAnalysis.
x
Fourth,theinitiativewasweakenedbytherefusalofsomecreditorstoparticipate.
95
TheDebtExperiencesofUganda,KenyaandBolivia
x
Fifth, and most importantly, Uganda has borrowed more than US$1.6 billion since HIPC II
completion, 85 per cent of which is owed to IDA and ADF, primarily to finance the Poverty
EradicationActionPlan(PEAP),Uganda’soverͲarchingpolicyframeworktoeradicatepoverty.This
heavyrelianceonborrowedfundsreflectslimitedimprovementsindomesticrevenues,whichhas
leftthecountryhighlydependantonexternalassistance.
Inthe2004budgetspeechitwasannouncedthatceilingswouldbeputonannualtoachieveagradual
declineintheNPVofdebttoexportsratiotosustainablelevels.
Its tumultuous history aside, Uganda’s experience serves to underscore that without a comprehensive
debtstrategyitisimpossibletousedebtfordevelopment.Inaddition,thefailuretodiversifytheexport
basehasleftthecountryatthemercyofprimarycommodityprices
Uganda needs to consolidate the gains of the debt strategy it has pursued since 1991. The institutional
arrangementsforexternalborrowingshouldclearlyoutlinetheroles,responsibilities,andobligationsof
all stakeholders. Uganda is currently attempting to ensure that borrowing is strictly for enhancing
productivity and competitiveness. Moreover the quality of infrastructure built with past borrowing has
fallen into a dilapidated state even before the loans are repaid so that there is a serious risk of
accumulatingfurtherdebtforitsrepair.
C. Kenya’sDebtExperience
1. Introduction
Kenya did not experience one big default. Rather, it has had serious recurrent debt servicing problems,
withadebtcrisispeakingin1991.Theseproblemsoccurredagainstabackgroundofnegativeexogenous
financial and trade shocks arising from the vulnerability of the Kenyan economy and the prices of key
primarycommoditiestoweatherconditions.
2. TheEconomicEnvironment
(a)OverallEconomicPerformance
The 1980Ͳ84 period was characterized by various adverse external and internal shocks (including two
severedroughts),globalrecessionandreducedcapitalinflowsfollowingthe1982debtcrisis.Itwasalso
characterized by inability to satisfy the IMF credit ceilings and Government borrowing conditionalities,
leadingtothecancellationofanumberofprograms.In1985Ͳ90,economicgrowthwasrelativelyrapid,
partly due to an increase in coffee and tea prices and a decline in petroleum prices. The Government
adoptedaproͲcyclicalpolicyandincreasedpublicexpenditure(bothcapitalandcurrent)morethanthe
increaseinrevenue.
Inthefirsthalfof1990s,theeconomyreceivedmoreshocks:adroughtin1991/1992,oilpriceincreases
due to the Gulf War, an aid embargo in 1991Ͳ93, and ethnic clashes in 1992. These shocks were
accompaniedbyanincreaseinthebudgetdeficit,risinginflation,andlargeexchangeratedepreciations,
astheforeignexchangemarketwasliberalized.Inthesecondhalfofthe1990s,economicgrowthdeclined
furthertoanaverageof1.9percent,assimilarinstabilitiescontinued.
As shown in table V.3., the performance of Kenya’s export sector has been lacklustre and exports have
grownlessthanGDPsinceindependence.TheshareofexportsinGDPdecreasedfrom21.8percentin
1980 to 12.5 per cent in 2004. Tea, horticulture and coffee are by far the most important exports,
96
CompendiumonDebtSustainabilityandDevelopment
accounting for 54 per cent in 2000Ͳ2004. Kenya’s terms of trade have also declined substantially.
Dependenceonprimarycommodityhasalsomeantthatthetermsoftradeareveryvolatile.
TableV.3.Kenya:ofExports,TermsofTrade(TOT)andForeignDirectInvestment(FDI)
Year
Exports
(K£million)
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Average
487.64
513.86
545.74
633.08
754.81
785.10
957.97
753.41
917.74
999.83
1232.38
1533.83
1708.08
3625.21
4170.72
4656.18
5696.30
5722.95
5722.25
5770.3
5988.2
6071.7
6569.7
6835.45
7953.05
GDP
Exports
Exports
(K£million) (inpercentof (US$million)
GDP)
2235.37
21.8
1318.0
2597.23
19.8
1388.8
2944.62
18.5
992.2
3316.63
19.1
994.8
3851.78
19.6
1041.2
4374.62
17.9
957.4
5083.98
18.8
1182.6
13.3
913.2
5648.23
6480.62
14.2
986.8
7451.34
13.4
925.8
8377.78
14.7
1022.8
9540.33
16.1
1091.6
11402.53
15.0
943.6
14185.41
25.6
1063.2
16903.24
24.7
1162.0
19205.79
24.2
1674.8
21865.55
26.1
2071.2
31161.76
18.4
1944.9
34701.44
16.5
1738.3
37173.95
15.5
1528.9
39817.15
15.0
1529.2
48391.90
12.5
1560.1
51938.20
12.6
1715.2
57089.00
12.0
1781.4
63685.80
12.5
2056.5
Growthof
exports
(percent)
5.37
Ͳ28.56
Ͳ4.78
10.20
Ͳ8.05
23.52
Ͳ22.78
8.06
Ͳ6.18
10.47
6.73
Ͳ13.55
12.67
75.13
Ͳ10.05
23.67
Ͳ6.10
Ͳ10.63
Ͳ12.05
0.02
2.02
9.94
3.86
15.44
0.10
TOT
1982=100
FDI/GNI
(percent)
123
108
100
94
110
92
103
85
88
79
69
82
79
90
101
95
93
102
100
86
84
79
78
81
77.4
1.1
0.2
0.2
0.4
0.2
0.5
0.5
0.5
0.0
0.8
0.7
0.2
0.1
0.0
0.1
0.4
0.1
0.4
0.4
0.4
1.3
0.5
0.4
Source:EconomicSurvey,VariousIssues.
CollierandGunning (1999)attributemuchofKenya’sweak growth performance to geographyandrisk.
MuchofthecountryisalsosemiͲaridsothatagriculturalproductionintrinsicallyrisky.Kenya’sgeography
also means that transport costs are high, quite aside from deficiencies in infrastructure. But they also
argue that trade shocks caused an economic decline because of overͲregulation and the Government’s
lossofcontroloverpublicexpenditure.Azam(1997)shows thatinsufficient privateinvestmentandthe
failure to increase human capital accumulation contributed to the slowing of growth in the 1980s and
1990s.
(b)LiberalizationStrategies
Inthe1980s,Kenyahada“managedͲfloat”exchangerateregime.Theperiodwitnessedacuteshortages
of imported inputs due to nonͲavailability of foreign exchange. This resulted not only in frequent
interruptionsinproductionbutalsoinchronicunderͲutilizationofinstalledcapacity.Kenyatookaseries
of measures that gradually removed foreign exchange controls and liberalized the exchange rate,
including a large devaluation of the shilling. In 1992Ͳ1993 the official exchange rate and the interͲbank
foreignexchangerateweremerged,controlsoncurrentandcapitalaccounttransactionswereremoved.
Furtherliberalizationfollowedin1995.
97
TheDebtExperiencesofUganda,KenyaandBolivia
Similarly in the 1980s and 1990s Kenya implemented trade reforms, which eliminated most nonͲtariff
barriers and lowered tariffs, substantially opening the economy. The maximum tariff rate was reduced
from 170 per cent to 70 per cent over 1987Ͳ1993. Recently, under obligations of the East African
Communitycustomsunion,tariffbandswerereducedtothreewithamaximumexternaltariffof25per
centsinceJanuary2005.
Kenyaliberalizeditscapitalaccountoverthesameperiod.Reformsalsoincludedtheeasingofrestrainton
foreign ownership and the establishment in 1990 of the Capital Markets Authority (CMA). The Nairobi
StockExchangemarketopenedtoforeigninvestorsinJanuary1995.Toinsureagainstthepotentialriskof
liquidity crises delivered by exogenous shocks or speculative activities, Kenya has followed other
developingcountriesinaccumulatingforeignreserves.
Kenya also embarked on financial sector reforms. Positive real interest rates, the target of the market
reforms,aimedatenhancingefficiency.InstitutionalreformsfocusedonstrengtheningtheCentralBank,
particularlyinitssupervisoryandregulatoryroles.Inmonetarypolicytherewasashifttomoreindirect
instrumentslikeopenͲmarketoperations.Therewasafinancialcrisisin1998whichledtotheliquidation
ofseveralbanks.Muchofthefinancialdeepeningwhichhasresultedisduetotheconversionofdeposits
ofnonͲbankingfinancialinstitutionstocommercialbanksdeposits.
The Central Bank has maintained a high interestͲrate regime to stabilize the exchange rate and has
pursuedagenerallytightmonetarypolicyinthefaceofinflationarypressures.Oneoftheconsequences
hasbeenwidespreaddistressedborrowingsothatbanks’portfolioshaveincludedmanynonͲperforming
loans.Thedeclineincredithasbeenassociatedwithdeclininginvestment.
Kenya’sfiscalpolicyislinkedtoitsexternalindebtedness.Kenyaisheavilydependentonaidinflowsforits
governmentfinances,withaidaccountingfor45percentofthebudgetatthepeakin1991(O’Brienand
Ryan,1999).Throughoutthe1990s,foreignaidaveragedabout9percentofGDP,accountingforabout
20 per cent of the annual government budget and financing slightly over 80 per cent of development
expenditures(Njeru2004).
3. ExternalDebt
Kenyaisasamoderatelyindebtedcountry.Thecountry’sexternaldebtincreasedfromUS$4.2billionin
themidͲ1980stoapeakofUS$7.5billionin1991,decliningtoUS$6billionin2002.AsaproportionofGNI
itincreasedfrom70.8percentofGNIin1985toapeakof156percentin1993butthendeclinedto49.2
percentin2002.(SeeTableV.4.)Externaldebtserviceincreasedtoapeakof39percentofexportsin
1988butthendeclinedto13percentin2002.
Almost all of Kenya’s external debt is either public or publicly guaranteed and owed primarily to
Governments and multilateral organizations. For the period 1985Ͳ2002, private nonͲguaranteed debt
generallyaccountedforlessthat15percentofthetotal.ShortͲtermdebtaccountedforbetween54and
69percentofoutstandingstocks.Theaveragegraceperiodisabout6.9years,theaveragegrantelement
about50.9percent,andtheaveragematurityperiodabout26.5years.Bilateralaidhasbeenmainlyin
theformofgrants(72percentofthetotal),whereasmultilateralaidhasmainlybeenintheformofloans
(86percent),mostlyfromtheWorldBankgroup.
While Kenya’s external debt to GNI ratios are currently less unfavorable than at the beginning of the
1990sandareevensustainableaccordingtoHIPCcriteria(IMF2003),thestockofexternaldebtandits
servicing nevertheless poses a major problem for two reasons. First, debt servicing is still a large
proportion of export earnings and government expenditures. Second, a large external debt creates
uncertaintiesforinvestmentsandunderminesthecredibilityofdomesticpolicies(Elbadawietal.,1997).
Pattilloetal.(2002),usingapaneldatasetof93developingcountriesover1969Ͳ98,findthattheaverage
98
CompendiumonDebtSustainabilityandDevelopment
impactofexternaldebtongrowthbecomesnegativeforadebttoGDPratioof35Ͳ40percent.Kenya’s
externaldebtsignificantlyexceedsthisthreshold.
TableV.4.DebtIndicatorsofKenya
Year
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Totaldebt
External ExternalDebt Principal
service
stocks
debt
arrears
(US$million) (aspercent (aspercentof
(US$
Exportsof
ofGNIa)
million)
goodsand
services)
4180.6
70.8
38.7
4.1
4602.8
65.9
35.6
6.2
5782.9
75.4
39.8
12.6
5808.9
71.2
39.0
25.5
5889.6
73.7
36.6
49.4
7057.6
87.2
35.4
71.8
7457.8
98.3
32.6
155
6902.6
90.7
31.1
263.3
7115.4
156.0
27.1
409.8
32.9
9.2
7128.9
105.5
7313.4
84.2
30.4
6.1
6811.4
75.4
27.8
14.8
6455.6
62.2
22.1
56.6
6808.1
60.9
23.2
105.7
6450.2
64.5
25.7
177.9
6159.2
61.1
18.7
133.6
5561.6
49.9
15.8
149.9
6031.2
49.2
13.6
236.5
Domestic Foreign
Interest
Budget
Debt
financing
arrears
deficit
(US$
((asper (aspercent (aspercent
million) centofGDP ofGDPb) ofbudget
b
)
deficitb)
10.9
17.3
28.3
40.3
64.6
94.7
140.9
188.8
241.7
81.2
31.4
9.9
27.5
58.2
68.8
42.1
34.4
56.8
3.7
4.8
7.5
3.7
3.8
4.3
5.0
1.3
4.5
5.8
1.3
1.2
2.2
0.8
0.7
0.9
1.6
12
16
27.8
27.1
26
24.6
26.9
25.9
33.5
27.2
24.9
21.5
22.4
21.1
20
19.7
18.1
42.7
40.6
21.9
17.9
21.3
54.3
40.4
44
42.6
6.2
175.1
12.7
48.8
Ͳ135.8
168.5
268.7
84.9
aSource:WorldBank,GlobalDevelopmentFinance,2004.
bSource:KenyaEconomicSurvey,VariousIssues.
Kenya has yet to develop a coherent strategy for managing aid flows. Aid design, process and
implementation have been ad hoc through issues of circulars from the Ministry of Finance (MOF). The
defaultpolicyistoaccommodateasmuchforeignaidasismadeavailable.TheexternalLoansandCredits
Act specifies limits on borrowing to a principal amount outstanding to no more than 650 million Kenya
poundsattheprevailingexchangerate,or“suchhighersumastheNationalAssemblymaybyresolution
approve”.Thelatterisaloopholeroutinelyusedbyministersinparliament.Themanagementofforeign
aid and external debt are the responsibility of several government ministries and agencies. The main
governmentsdepartmentsdealingwithdonorsandloans(theExternalResourcesDepartment(ERD)and
Loans Division and the External Debt Management (DMD)) are highly constrained in human resource
capacityintermsofnumbersandskills.Thecountryalsolacksdebtmanagementobjectives.
4. PublicDomesticDebt
Tofinanceitsbudgetdeficits,Kenyahasborrowedondomesticmarketsaswellasabroad.AsTableV.4.
shows, Kenya’s domestic debt accounted for 25 per cent of GDP for much of the 1990s, and foreign
financinginsomeyearscoveredahighproportionofthebudgetdeficit.Attheirpeakin1993/94interest
payments on domestic debt amounted to 47.6 per cent of government revenues and 24.8 per cent of
governmentexpenditurerespectively.Thesepercentagesexceedwidelyusedbenchmarksforsustainable
rates of interest on domestic debt. Sterilization of the inflows associated with the foreign financing of
99
TheDebtExperiencesofUganda,KenyaandBolivia
budget deficits contributed to tight credit markets and recession, thereby undermining growth and
contributingtodebtproblems.
5. TheEvolutionofKenya’sExternalDebt
Kenya’s first debt problems followed the drought and trade shocks of the early 1980s, with external
debt/GNIratioexceeding70percentby1985.Afterabriefimprovement,followingmoreexternaltrade
shocks and the ethnic clashes of 1992, debt levels rose again, and Kenya accumulated arrears on both
interestandprinciple.ArrearspeakedinmidͲ1993afterthecuttingofaidandofrelationswithdonors.By
1994KenyahadrescheduleddebtsworthUS$500millionitowedtotheParisClub(WorldBank2003).In
1998itbegannegotiationstorescheduledebtowedtoprivatelendersattheLondonClub.82
Debt indicators improved as a result of the 1994 debt rescheduling. When HIPC was launched in 1996,
KenyawasdeclaredcapableofachievingsustainabilitywithanNPVofdebttoexportsoflessthan150per
cent(148percent).However,furthertrade,climateandpoliticalshocks(surroundingthe1997elections
and another suspension of foreign aid) worsened the situation again in the late 1990s (see Table V.4.).
Following the approval of an IMF program in 2000, Kenya rescheduled with the Paris Club under
“Houston”termswithan agreementcovering US$300millionofarrears.However,arrearscontinued to
accumulate and, following a third IMF program in 2003, a new Paris Club deal was secured in January
2004 covering US$353 million of arrears. External debt stocks were not significantly reduced by the
agreementsof2000and2004.Indeed,thestockincreasedfromUS$5.5billionin1999toUS$5.7in2004.
Inrecentyears,therehasbeensomeimprovementineconomicperformance.Afiscalstrategyhasbeen
establishedtocontrolexpenditureoverthemediumtermandtherehasbeenareversalofthedeclining
trendindomesticrevenuesaswellasasmallrepaymentofpublicdebtin2004/05.However,Kenyahas
been demanding moredebtrelieffollowingthe MDRIinitiative,particularly sinceitsserviceburdenhas
beenexceedingMDGspendingforyears.
As with Uganda, Kenya’s debt accumulation has been closely related to its fiscal needs. Kenya’s case
shows that in the absence of a debt strategy, external debt is unlikely to serve a development agenda.
Kenya’s development and public investment expenditure suffered from both fluctuations in external
financingandtheburdenofdebtservice
Kenya’s debt woes are also related to its continued dependency on agriculture and on primary
commodities.SevereclimaticandtermsͲofͲtradeshockshaveunderminedgrowthanddeepenedpoverty.
Despite diversification in the production and export base the economy remains vulnerable to adverse
exogenousshocks.
Kenya has failed to develop significant debtͲ management capacity or clear aid strategies. Arrears have
tended to accumulate even when external debt stocks were not growing. Owing to poor relations with
donorsthecountryhasalsomissedoutonmajordebtreductioninitiatives.Mostdebtrestructuringshave
concentratedonliquidityproblems,i.e.arrears.Weaknessindebtmanagementandaidstrategyarelikely
toresultinacontinuationofthecountry’shistoricalpatternofdebtproblems.
82
TheformeragreementledtothecancellationofUS$21millionofarrearsandmaturities,whilethelatterdealeventuallyledto
thereschedulingUS$45millionofdebt.
100
CompendiumonDebtSustainabilityandDevelopment
D. Bolivia’sDebtExperience
1. Introduction
Bolivia is topologically rugged country with arid agricultural conditions and low population density. The
naturalͲ resourcesͲbased economy has generated only limited employment, and exports earnings have
contributedlittletorelievingpoverty.BoliviawasoneofthefirstbeneficiariesofHIPCInitiativesbecause
itstrackrecordasaliberalizingreformer.However,sincethelate1990sBolivia’sperͲcapitarealGDPhas
stoppedgrowing.Politicaloppositiontoreformhasintensified.ThedebtstockreturnedtopreͲHIPClevels.
Bolivia’sexperienceraisesquestionsconcerningcurrentdebtreliefarrangements.
2. OverallEconomicPerformance
In the early 1980s, like many other Latin American economies, Bolivia’s economy slid into recession
associatedwithsurgingworldinterestratesandthe1982debtcrisis(seeFigureV.3.).Between1981and
1988perͲcapitarealGDPdeclinedby15percent.Overthesubsequent10yearsperͲcapitarealGDPgrew
atanannualaveragerateof2percent.In1998Ͳ2003itstagnated.
In the 1990s economic growth was revived by rising investment associated with the “capitalization”
process (see below) and with the export of natural gas. FDI inflows became more important than debt
from the midͲ1990s, and remittances were also an important source of external financing. Inflows of
financing from private foreign creditors have been adversely affected by past experience of losses.
However,recentlymultilaterallendersliketheAndeanDevelopmentCorporation(CAF)andtheIDAhave
beenasignificantsourceofcredit.
As an exporter of naturalͲresource products Bolivia has been vulnerable to adverse price shocks and
termsͲofͲtrade movements. The prices of its key exports collapsed spectacularly in the early 1980s.
Commodity prices did not recover significantly in the 1990s. Figure V.4. shows the decline in Bolivia’s
termsoftradesince1991.Exportpriceslostaquarteroftheirvaluebetween1991and2000,whileimport
pricesdriftedupwardswithworldinflation.Decliningtermsoftradeunderminedebtexportratios.
Additionally Bolivia was affected by a series of shocks as of 1997: El Niño; the East Asian crisis in
September1997;theRussiancrisisofAugust1998;andtheBrazilianandArgentinecrisesof1999Ͳ2001.
More generally countryͲspecific factors have hindered growth though to an extent difficult to measure.
Thesefactorsincludeharshtopographyandclimate,ethnicandlinguisticdiversity,regionaldivisions,and
ahistoryofpoliticalinstability.
101
TheDebtExperiencesofUganda,KenyaandBolivia
FigureV.3.Bolivia:PerͲcapitarealGDP,PrivateConsumption,andPublicExternalDebt,
1970Ͳ2004
$1,800
$1,600
$1,400
$1,200
$1,000
$800
$600
$400
$200
$0
1970
1975
1980
1985
1990
1995
2000
Per-capita real GDP
Per-capita real non-government consumption
Per-capita real public external debt (incl. debt to IMF and interest arrears)
Source:InternationalFinancialStatistics(InternationalMonetaryFund).
FigureV.4.Bolivia:TermsofTrade,1991Ͳ2004
(June;1991=100)
140.0
120.0
Import prices -->
100.0
Export prices -->
80.0
Terms of trade -->
60.0
40.0
20.0
0.0
1991
1992 1993
1994
1995 1996
Terms of trade (1991 = 100)
1997 1998
1999 2000
Export prices (1991 = 100)
2001
2002 2003
2004
Import prices (1991 = 100)
Source:CentralBankofBolivia(website).
3. Bolivia’sLiberalizationandStructuralͲReformPolicies
Since 1985 successive Bolivian Governments have carried out some of Latin America’s most ambitious
liberalizationandreformprograms.Thereformprocessbeganwiththe1985stabilizationprogram,which
102
CompendiumonDebtSustainabilityandDevelopment
vanquished hyperinflation. The Bolivian Government introduced a “New Economic Policy” of: fiscal
discipline;priceandinterestͲrateliberalization;liftingofcontrolsoncrossͲborderfinancialflows;aunified
marketͲbased exchange rate; and trade liberalization. Price liberalization, which had ended the control
and subsidies of prices, was none the less replaced by price capping in 2000 after export prices rose
sharply.
ExchangeͲrate management has been at the centre of Bolivia’s stabilization efforts since 1985. The
authoritiesallowedthepesotofloatandendedmultipleexchangeͲratepractices.InJanuary1987,anew
currency, the “boliviano”, was introduced, at a rate of one peso per million. In early 1988 the new
currencystabilizedatabout2.3bolivianosperdollar.TheCentralBankhasmanagedtheexchangerateas
acrawlingpeg,movingitinlinewiththedifferencebetweenBolivia’sandworldinflationrates.Thispolicy
hasledtothemaintenanceofrelativelyhighforeignexchangereserves(seeFigureV.5.).
Persisting dollarization has complicated exchangeͲrate policy. Despite compulsory conversion of dollar
bank deposits into Bolivian pesos in the early 1980s, there has remained a large amount of informal
dollarization, which has contributed to inflationary pressure. Since 1985 dollarͲdenominated accounts
haveaccountedfor85Ͳ90percentofdepositsandloans.Bolivia’scocatradehasalsocontinuedtobringa
largeinflowofdollars,contributingtodollarization.
Tight monetary control has been fundamental to the maintenance of price and exchangeͲrate stability.
Between1987and2004,theaverageannualrateofincreaseinconsumerpriceswasonly8.7percent;and
the average annual rate of increase in the price of the U.S.dollar in bolivanos was 8.2percent. On the
whole,Bolivia’sexchangeͲratepolicyhascontinuedtosupportstabilizationsince1985.
FigureV.5.Bolivia:YearͲendForeignͲExchangeReserves,1980Ͳ2003
Months
Per cent of GDP
7.0
16.0
6.0
14.0
12.0
5.0
10.0
4.0
8.0
3.0
6.0
2.0
4.0
1.0
2.0
0.0
0.0
1980
1985
1990
1995
Months of imports of goods and non-factor services
2000
Per cent of GDP
Source:CentralBankofBolivia
The midͲ1990s witnessed a “second generation” of reforms, which centered on three elements:
restructuring and capitalization of key sectors; pensions’ reform; and significant decentralization. The
“capitalization” program of 1995Ͳ1996 was an alternative to politically unfeasible privatization. The
Governmentauctionedtherightto50ͲperͲcenttemporaryownershipstakeandmanagementcontrolin
selected enterprises accompanied by a commitment to carry out specified capital expenditures. The
program was successful in the sense that the enterprises which were capitalized exceeded agreed
103
TheDebtExperiencesofUganda,KenyaandBolivia
investment targets, and services improved (IMF2005). A closely associated reform was the 1996
HydrocarbonsLawdesignedtoenhanceforeigninvestment,particularlyinthedevelopmentofnewfields.
The reform did succeed in attracting substantial investments, and led to discovery and exploitation of
large gas reserves. However, government revenue from the sector was disappointing, and deepening
foreignparticipationhasbeensourceofpopularresentment.
In 1997 the Government undertook reform of the troubled pensions system along the lines of Chile’s
pension reform, i.e. shifted to a contribution system. As for decentralization, representative governing
bodiesweresetupfordepartmentsandprovinces.Theseinstitutionsweregivensignificantfiscalroles,
including shares of government revenue. However, the transfer of revenue and responsibilities proved
politicallycontentious,andcontributedtoBolivia’sfiscaldifficulties.
Throughoutthesechanges,theGovernmentlackedfirmpoliticalsupport.Ambitiousasthereformswere,
theydidlittleforordinaryBolivians.Inresponse,manyBolivianstriedtoescapepovertyby“rentͲseeking”
strategies involving publicͲsector employment, smuggling activities, or participation in the illicit cocaͲ
derivatives trade. Thus the political process became closely linked to persistent pressures for public
employmentandsubsidization;smugglingbecameubiquitous;andsuppressionofthecocatradehasbeen
impossible. Since securing public positions has become a basic function of political parties, it is hardly
surprisingthattheadministrationhasbeenpronetoinefficiency,overstaffingandcorruption.
Thereformsofthe1990shavebeguntounderminefiscalbalancesdespitetheexistenceofpolicyrules
such as forbidding the printing of money and mechanisms to control government expenditure and to
ensureagoodflowofforeigntraderevenues.Taxrevenuehasstabilizedsince1998atabout12Ͳ13per
cent of GDP with customs revenues steady at about 1per cent of GDP. Hydrocarbons reform had an
unexpectedly large upͲfront fiscal cost, especially when royalties were cut in 1997. Earnings from
hydrocarbonshadbeenaround10percentofGDPbutby2004theyhadslidto6.4percent.Receiptsfrom
fuel excises initially rose after 1997 but stagnated in 2000 when fuel prices were frozen. On the
expendituresidepersonnelcostsare10percentofGDP.Thecostsofdecentralizationandofthepension
reformturnedoutwellabovewhatwasanticipated.Lastly,domesticinterestpaymentshavebeenrising
inlinewithdomesticborrowing.
4. Bolivia’sExternalDebt:StructureandMainFeatures
Externaldebtgrewindollartermsfromthe1970suntilHIPCdebtreliefin2001.Inthe1970sthebulkof
thedebtwasbilateraldebtowedtocommercialsourcesandborrowedmostlyfordevelopmentpurposes,
notablyinfrastructure(communications,roads,airports).Between1980and1987thegrowthofBolivia’s
totalexternaldebtaccelerated,increasingfromUS$2.7billiontoUS$5.8billionͲfromjustunder60tojust
over 140per cent of GDP. The prime reason for this surge was increases in world interest rates. In
addition, international recession drove down Bolivian export prices. As a result Bolivia could no longer
meetitsdebtͲserviceobligationstocommercialbanks,andwentintoarrearsanddefault.
Between 1989 and 1992, Bolivia’s overall external debt stock stabilized at about US$4billion, rising to
US$5billionafter1996.Meanwhile,highereconomicgrowthduringthemidͲ1990sreducedthedebtͲGDP
ratio somewhat. In 1998 and 2001 Bolivia received about US$1billion in HIPC debt relief, reducing the
debtͲGDPratio.However,thisreductionprovedtransitoryandwithintwoyearsslowgrowthandheavy
borrowing from multilateral sources raised the debtͲGDP ratio to where it had been before HIPC debt
reduction.
Since the second half of the 1980s Bolivia has cut its dependence on commercial bank finance.
Multilateralagenciesincreasedtheirlendinginthelate1980stoassiststabilization,andinthe1990sto
support liberalization and structural reform. Thus, of the endͲ2004 total external debt of US$4.6billion
104
CompendiumonDebtSustainabilityandDevelopment
US$4.3billionwasowedtomultilateralentitieswiththeIDAaccountingforUS$1.7billion(seeFiguresV.6.
andV.7.).
FigureV.6.Bolivia:YearͲEndPublicandPubliclyͲGuaranteedExternalDebt,1970Ͳ2004
(US$billion)
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
Multilateral
Bilateral
Private sources:
Source:GlobalDevelopmentFinance(WorldBank).
Debtserviceremainedwithinarangeof4to5percentuntiltheendofthe1990s.Itthensurgedbrieflyin
2000and2001duetorelativelyhighrepaymentflows.Bolivia’sexternaldebtͲserviceͲtoͲexportsratiowas
generallybeenabove20percentuntilitfellbelow20percentafterHIPCdebtreduction.Bolivia’sinterest
burdenwaskeptdownbytheconcessionalnatureofmuchofitsdebtsincethesecondhalfofthe1980s.
FigureV.7.Bolivia:YearͲEndPublicandPubliclyͲGuaranteedExternalDebt,1970Ͳ2004
(PercentofGDP)
200.0
180.0
160.0
140.0
120.0
100.0
80.0
60.0
40.0
20.0
0.0
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
Multilateral
Bilateral
Private sources:
Source:GlobalDevelopmentFinance(WorldBank).
105
TheDebtExperiencesofUganda,KenyaandBolivia
Bolivia debt management has improved significantly since the early 1980s. At that time, Bolivia had no
administrative system of debt management and governance as such, though the Government formed
committeestodealwithcommercialbanks.In1985debtwasconsolidatedinthenationalTreasury,anda
ministerialͲlevelcommitteewasformedtoworkoutastrategy.In1987commercialbanksdecidedtooffer
reliefthroughdebtbuybacksbuttheoperationswerecarriedoutadhocbyexpertswithoutthehelpof
sophisticateddebtͲmanagementsystems.
Under the basic institutional arrangement eventually adopted the Government assigned the bulk of
managing and monitoring of external debt to the Central Bank because of its institutional depth and
analyticalcapacities.Sincethe1980sBolivia’stechnicaldebtͲmanagementcapacityhasimprovedsteadily
andBolivia’sdebtpoliciesarenowhighlytransparent.TheConstitutionrequiresparliamenttoapproveall
newborrowing.
5. DomesticPublicDebt
Alongside of its external borrowing to finance government expenditures Bolivia has also borrowed
domestically particularly after the 1985 reforms. Thus domestic public debt rose steadily from 1991 to
overUS$1bnin2000,thendoublingtojustunderUS$2billionby2004,i.e.from13to21percentofGDP.
The issuance of domestic obligations can help Governments to deepen their financial sector and widen
theirrevenuebase.ButaswithotherHIPCs,Boliviahadtopayhighinterestratesonitsdomesticdebt,
whichisnotcontractedonconcessionalterms.Interestpaymentsoninternaldebtrodefrom0.4to1.8
per cent of GDP between 1998 and 2004, while interest payments on external debt remained stable
aroundonepercentofGDP.Domesticdebtreached21percentofGDPin2004,andtotalpublicdebt95
per cent of GDP. Initial fiscal sustainability targets under HIPC programs overlooked this source of
indebtednessandtheresultingpressureonfiscalbalances.
106
CompendiumonDebtSustainabilityandDevelopment
TableV.5.Bolivia:DomesticPublicDebt1991Ͳ2004
Year
DomesticDebt
(US$million)
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
385.6
396.2
296.8
97.5
217.0
438.8
508.4
827.9
983.6
1100.1
1508.6
1504.3
1724.4
1992.9
Source:Cowanetal2006.
TotalPublicdebt
(US$million)
4258.4
4429.8
4300.1
4576.5
4999.6
5080.8
5040.0
5487.3
5557.4
5560.1
5920.6
5804.7
6768.7
6944.0
Domesticdebt
(aspercentof
GDP)
7.2
7.0
5.2
1.6
3.2
6.0
6.4
9.8
11.9
13.2
18.7
18.3
20.1
21.3
TotalPublic
debt(aspercent
ofGDP)
86.9
85.5
80.2
78.3
77.8
74.8
70.1
74.5
79.3
79.7
92.2
89.1
98.8
95.5
6. PastandPresentExternalDebtPracticesandStrategy
Sincetheearly1980sthreebroadphasesofBolivia’sdebtstrategycanbedistinguished.Thefirstbegan
withtheonsetofthedebtcrisisin1982.AtthattimeBoliviareliedheavilyonexternaldebttocoverits
fiscaldeficit,andwhendebtflowswerecutoff,theGovernmentshiftedtomonetaryfinancing,generating
hyperinflation.In1984Boliviadeclaredamoratoriumondebtservice.ThemarketvalueofBolivia’sdebt
to commercial banks plunged to 10Ͳ15per cent of its face value by the midͲ1980s. The second phase
lasted from the August1985 stabilization program until 2000 and consisted largely of reducing debt
throughvariousinitiativesandincreasingrecoursetoconcessionalflows.Since2001newmultilateraldebt
inflowshaveoffsetHIPCdebtreduction,politicalturmoilhasintensified,andGDPgrowthhasstagnated.
OncethestabilizationprogrambeganinAugust1985,theauthoritiesrestoredrelationswiththeIMFand
other creditors. The 1986 IMF program opened the way to new financing. Bolivia took a pioneering
approachtoitscommercialͲbankdebt(aboutUS$650millionin1986).Usingfundsprovidedbydonors,it
retired the bulk of its debt by purchasing it at deeply discounted values.83Bolivia’s debt to commercial
bankswasmostlyeliminatedbytheearly1990s.
Thereafter, Bolivia sought relief on its bilateral debt through the Paris Club. It went through six
reschedulingsbetween1986and1995.Between1986and2003,BoliviahadthreeIMFprogramsinvolving
SDR515million, including one of the first Poverty Reduction and Growth (PRG) Facilitiesin 1998. In
April2003, Bolivia secured an IMF agreement for SDR 129 million amidst economic and political crises.
Fearing that the collapse of the agreement would aggravate Bolivia’s problems, the IMF waived some
conditions,andBoliviadrewSDR102millionbyMarch2005(IMF2005).
In September1998, multilateral and bilateral creditors provided Bolivia debt relief amounting to
US$449million in netͲpresentͲvalue (NPV) terms, at the “completion point” of its HIPC process. Of this
total,bilateralcreditorsandtheIADBeachaccountedforabout35percent,theWorldBankforabout12
per cent, CAF for 9 per cent, and the IMF for about 6percent. The conditions Bolivia satisfied for
83
TheBolivianbuybackoperationof1987,whereUS$253millionwererepurchasedatabout11centsperdollar,wasoneofthe
firstlargeͲscalebuybackcarriedoutwiththespecificpurposeofreducingacountry’sexternaldebtsincethe1930s.
107
TheDebtExperiencesofUganda,KenyaandBolivia
completion were its history of IMFͲ led programs. The estimated equivalent stock reduction was
US$760million(7.8percentof1998GDP)(IMF&IDA1998).
In May2001, Bolivia became the second country to reach its completion point under “enhanced” HIPC.
This time, official creditors provided debt relief of US$834million in NPV terms. Conditions satisfied for
completionincludedasuccessfulreviewofperformanceunderthe1998PRGFacility,theformulationofa
PovertyReductionStrategy,andcontinuedimplementationofsocialͲsectorprograms.
Subsequently, however, Bolivia’s economic performance has deteriorated. A widening public deficit
meant that Bolivia had to secure increased financing. Over 2001Ͳ2003, the Government used external
sourcestocoverabout60percentofitsfinancingrequirements.Ofthetotalaboutonehalfcamefrom
theIADBandtheWorldBankatconcessionalratesandtheremainderfromtheCAFatnonͲconcessional
rates.
The premise of Bolivia’s debt strategy was that stabilization, liberalization, structuralͲreform and debtͲ
reduction would create a dynamic economy and high real GDP growth. But the results have been
disappointing.TheIMF2005staff’sreportnotesthatBolivia’spostͲ1998slowdownmaybeunderstoodas
theconsequenceof“aprotractedsequenceofexternalanddomesticshocks…”(IMF2005,p.12).Reforms
had not achieved their goals, and the political and social roots of Bolivia’s problems were deeper than
recognized.
ForthemajorityofBolivians,theoverridingobjectiveofstructuraleconomic policyought tohavebeen
theidentificationofways tobringabout theirparticipationintheeconomy. However,liberalizedprices
discouraged enterprise in areas capable of generating employment, and reforms led to cuts in public
sector employment. Through their political protests Bolivians are now demanding that the Government
make these problems its highest priority. Negotiated debt relief is of little consequence in this context,
since as in much of the rest of Latin America, repayments are viewed as illegitimate since debt was
accumulatedinthefirstplacewithoutpopularconsent.
It is not clear how much Bolivia benefited from the HIPC initiative. By the time Bolivia negotiated HIPC
relief it had already reduced burdensome commercial and bilateral debts. When debt reduction is not
accompaniedbymeasuressuccessfulinreducingbothinternalandexternalborrowingrequirements,debt
indicators are likely to grow again. The IMF failed to grasp the deeper problems behind Bolivia’s debt
accumulation, hiding instead behind customary calls to deepen reforms along lines already tried
unsuccessfully.
E. DebtExperiencesCompared
Despite nearly three decades of adjustment, stabilization, extensive reforms and liberalization, Uganda,
KenyaandBoliviahavenotescapedtheirpredicamentashighlyindebtedpooreconomies.Comparisonof
thethreeeconomiespointstothreecommonremainingfeatures:
x
x
x
Failuretoreducepovertysignificantly;
Failuretogeneratesustainedgrowth;and
Failure to reduce their vulnerability to external shocks and the continued concentration on
vulnerablesourcesofincome.
TheassumptionofIMFandWorldBankprogramshadbeenthatreformandliberalizationwoulddeliver
growth by instituting prudent macroeconomic management and eliminating price distortions. In many
cases, the policy measures not only caused momentary pain but also had longerͲterm adverse effects.
Higher interest rates did not automatically result in financial deepening, but did increase defaults of
108
CompendiumonDebtSustainabilityandDevelopment
privateenterprisesandthefiscalcostsofdomesticborrowing(orboth).ExchangeͲratedevaluationfailed
to make primary commodity exports more competitive, and trade liberalization undermined the
government revenue base. Cuts in government spending undermined public investment, thereby
weakeningprivateinvestmentandhumancapitalformation.
InallthreecasesthereformsadoptedwerenotsufficienttoovercomemanydeepͲrootedproblemsand
structuralweaknesses.Fiscaldisciplinehasbeenbeneficial,butitdoesnotexpandataxandrevenuebase
curtailedbymeasuressuchastariffremovals.Hence,40Ͳ50percentofgovernmentactivitiesinUganda
and Kenya continue to be financed by foreign aid. Likewise, orthodox reforms have not reduced high
productionandtransportationcostsinherenttothethreecountries’difficultgeographyandtopography.
Neither the reforms nor the debt initiatives have adequately recognized or produced solutions to the
extreme vulnerability of the three countries to strong external shocks. All three countries have a heavy
concentration of economic activity and exports in a few primary and unprocessed commodities, whose
prices have been highly volatile and subject to sharp declines. The effects of this concentration are
exacerbatedbythedependenceoflargepartsofthepopulationonrainͲfedagriculture.
Itisagainstthiscontextthatthethreedebtorshavehadtomanagetheirexternaldebtburdens.Whilethe
countries were catapulted into debt traps at different times and with different intensities, debt
sustainabilitycontinuestoeludeall three.Althoughinheriteddebtstockshavebeenreducedand there
have been shifts to concessional financing and grants, the three debtors continue to experience the
pressuresofhighdebtburdens.
Recent debt crises in all three cases have originated in the government sector, i.e. the inability of the
Government to service foreign loans. However, the defaults and arrears were caused less by the
ballooningofdebtstocksthanbysuddenandunexpectedsharpshortfallsinrevenuesduetoexogenous
shocks,namelyrisinginterestratesorcollapsingexportearningswhichleddebtratiostosoar.
AllthreeGovernmentshaveincreasinglyresortedtodomesticborrowing,albeittodifferentdegrees,to
financegovernmentbudgets.Domesticdebtsanddebtburdenshaveonlyrecentlybeenincludedindebt
sustainability analyses (World Economic and Social Survey 2005). Domestic debt tends to be more
expensive than external finance, so that its costs worsen fiscal difficulties or widen fiscal deficits. This
underminesthebeneficialeffectsofoperationsreducingexternaldebt,andisoneofthereasonsforthe
failureofHIPCinitiatives.InUgandaandKenyadomesticdebthasalsobeenissuedtosterilizeofficialaid
inflows.InUgandasuchsterilizationhashadtheconsequencethattherewouldbenoimprovementinits
debtserviceafterHIPCIIafterallowanceforthecostoftheTreasuryBillsissuetosterilizeaidflows.
Atthetimeoftheirfirstcrisesnoneofthethreecountrieshadinplaceameaningfuldebtstrategyoreven
good management systems to monitor or analyse debt. This has changed substantially in the case of
UgandaandBolivia,whichbothnowhaveadequatetechnicalcapacitytomanagetheirdebt.OnlyKenya
stilllacksanadequatedebtmanagementsystem.
Since the 1990s, debt strategies have been determined by official creditors. Rescue from defaults and
fresh finance depended on the IMF and multilaterals, which initiated the countries’ adjustment and
reformprogramsaspreͲconditionsfordebtrestructuringswiththeLondonandParisClubs.Thedetailsof
theagreementsreachedexplainwhytherewasaneedforcontinuousandrepeatedreschedulingefforts.
These resulted from early cutͲoff points, the exclusion of too many types of debts and creditors, debt
reliefinadequatetoeaserepaymentdifficulties.
ThelaunchoftheHIPCinitiativewasarecognitionofthefollowing:
x
x
Debtproblemsparticularlyforpoorercountriesreflectinsolvencyratherthanilliquidity;
PartialandprotractedreͲschedulinghasnotprovidedapermanentexitfromrestructuring;and
109
TheDebtExperiencesofUganda,KenyaandBolivia
x Reachingsustainabledebtpathsrequiresdebtreduction.
The fact that HIPC had to be enhanced almost as soon as it was born highlighted its similarity to the
reluctant and partial approach to debt problems that characterized previous initiatives. It also reflected
weak analytical bases, which, for example, overlooked fiscal criteria for sustainability. More seriously,
debt sustainability analyses were not based on realistic and comprehensive scenarios, and
underestimated thevulnerabilitytoand theextent ofexogenousshocks (see forexample Nissankeand
Ferranini (2006). Even the most compliant countries included in the initiative had been consistently
thrownoffcoursebysuchshocks,includingduringandafterHIPC.
ThepostͲHIPCproblemsofUgandaandBoliviarevealotherflaws:
x
x
x
x
Thelimitsofanarrowfocuswhichdefinestheattainmentofdebtsustainabilityintermsofdebt
ratiosbelowthresholdsatonepointintime;
FailuretotakeintoaccountthattheriseinpostͲHIPCborrowingcouldquicklyreversegains(the
resultinUganda’scaseofafailuretoplacealimitonthestockofnewborrowing);
Theproblemofusingloansinsteadofgrantstofinancepovertyalleviationprograms;
Themoregeneraldifficultyforpoorereconomiesofachievingthereformsoffiscalpolicywhich
makepossibleobservanceofdomesticdebtthresholds.
In 2007 Bolivia had unsustainable debt based on fiscal criteria and Uganda’s debt sustainability has
deteriorated since HIPC II completion. Kenya’s debt indicators remain unsatisfactory and its second
PovertyandGrowthFacilityisunderreview.
F. ConcludingRemarks
All three countries covered by this study have extensively liberalized their trade and foreignͲexchange
regimes and their financial sectors. Part of this liberalization was undertaken at the countries’ own
initiative.Butwithrespecttomanyofthemeasuresthecountrieshadlittlechoice,sincereceivingdebt
relief and aid from multilaterals depended on implementation of conditions in agreed programs. These
conditionslimitedthespaceofpolicymakersandfailedtodelivereitherbroadͲbasedorsustainedgrowth.
These experiences show that the programs on which debt relief was conditional were based on an
inadequateapproach.Byfailingtodelivergrowthortostabilizerevenuesandexportearnings,theyalso
failed to provide a sufficient improvement in the ability to repay or service debt. Debt initiatives were
blinkeredandpartialintheircoverage,draggingeachcountryintoanunendingseriesofnegotiationsand
reschedulings.
Onelessonoftheseexperiencesistheneedforamuchdeeperandmorecomprehensiveunderstanding
of debt sustainability and of solvency which goes beyond thresholds and liquidity ratios, however
rigorously derived. Another lesson is that it is impossible to use debt to spur economic growth and
development without a coherent debt management strategy. A third lesson is that new external
borrowing by poorer countries will contribute to growth only if directed at expenditures that enhance
productivityandcompetitiveness.
110
CompendiumonDebtSustainabilityandDevelopment
References
AfricanDevelopmentBankandOECD(2005/2006).AfricanEconomicOutlook,2005/06,Kenya.Available
at:www.oecd.org/dataoecd/34/31/36740590.pdf.
AjayiSI(1991).MacroeconomicApproachtoExternalDebt:TheCaseofNigeria.AERCResearchPaper,8,
December.
AllenMandNankaniG(2004).DebtSustainabilityinLowͲIncomeCountries–ProposalforanOperational
Framework and Policy Implications. Washington, DC, International Monetary Fund and
InternationalDevelopmentAssociation,February.
AndersenLE and Osvaldo N (2000). TheHIPCInitiativeinBolivia.LaPaz, UniversidadCatólicaBoliviana,
November. Available at: www.bcb.gov.bo/deudaexterna/documentos/general/the_hipc_
initative_in_bolivia.pdf.
AtingiͲEgo M (2005). Budget Support, Aid Dependency, and Dutch Disease: the Case of Uganda. Cape
Town,ConferenceonPractioners’ForumonBudgetSupport,May2005.
Aiyar S, Berg A and Hussain M (2005). The Macroeconomic Challenge of More Aid. Finance and
Development, 42 (3), September. Available at: www.imf.org/external/pubs/ft/fandd/2005/09/
aiyar.htm.
AzamJP(1997).PublicDebtandtheExchangeRateintheCFAFrancZone.JournalofAfricanEconomies,6
(1),March.
BeckermanP(2006).Bolivia’sEconomicPerformanceandExternaldebtSincethe1980s.UNCTADproject
onCapacityBuildingforDebtSustainabilityinDevelopingCountries.Geneva(January).
BeckermanPandSolimanoAeds.(2002).Ecuador’sCrisisandDollarization,Washington.DC,WorldBank.
Bleaney et al. (1995). Revenues Instability with Particular Reference to SubͲSaharan Africa. Journal of
DevelopmentStudies,31(6),August.
BurnsideCandDollarD(2000).AidPoliciesandGrowth.AmericanEconomicReview,90:847–868.
CollierPandGunningJ(1999).ExplainingAfrica’sEconomicPerformance.JournalofEconomicLiterature,
March.
Collier P and Reinikka R (2001). Post Reconstruction and Liberalization: An Overview. In: Collier and
Reinikkaeds.Uganda’sRecovery.TheRoleofFarms,FirmsandGovernment.Washington,DC,The
WorldBank.
CowanK,LevyͲYeyatiE,PanizzaUandSturzeneggerF(2006).SovereignDebtintheAmericas:NewData
andStylizedFacts.ResearchDepartmentWorkingPaper,577.InterͲAmericanDevelopmentBank,
October.
Christensen J (2004). Domestic Debt Markets in SubͲSaharan Africa. IMF Working Paper, 04 (46),
Washington, DC. Available at: http://www.internationalmonetaryfund.com/external/pubs/ft/wp
/2004/wp0446.pdf.
EasterlyW(2001).TheElusiveQuestforEconomicGrowth.Cambridge,Massachussets,MITPress.
EasterlyWandSchmidtͲHebbelK(1993).FiscalDeficitsandMacroeconomicPerformanceinDeveloping
Countries.TheWorldBankResearchObserver,8(2).
ElbadawiI,NduluBandNdung’uN(1997).DebtOverhangandEconomicinSubͲSaharanAfrica.In:Iqbal
andKanbur.ExternalFinancingforLowIncomeCountries.WashingtonDC,InternationalMonetary
Fund.
111
TheDebtExperiencesofUganda,KenyaandBolivia
Geithner T (2003). Sustainability Assessments Ͳ Review of Application and Methodological Refinements.
IMFPolicyDevelopmentandReviewDepartment,Washington,DC,June.
Geithner T (2002). Assessing Sustainability. Washington, DC, IMF Policy Development and Review
Department,May.
Government of Uganda (1998). Vision 2025: A Strategic Framework for National Development.
BackgroundPapers,2,Kampala.
GreeneJandVillanuevaD(1990).PrivateInvestmentinDevelopingCountries:AnEmpiricalAnalysis.IMF
StaffWorkingPaper(WP/90/40).Washington,DC.
Henstridge M and Kasekende L (2001). Exchange Reforms, Stabilisation and Fiscal Management. In:
Reinikka and Collier P eds. Uganda’s Recovery. The Role of Farms, Firms and Government. The
WorldBank.Washington,DC.
IMF(2003).Kenya:DebtSustainabilityAnalysis.IMFCountryReport,03(400).Washington,DC.
IMF (2004). SubͲSaharan Africa Regional Economic Outlook. Washington, DC. Available at:
www.internationalmonetaryfund.com/external/pubs/ft/afr/reo/2004/eng/02/pdf/reo1004.pdf.
IMF (2005). Bolivia: Ex Post Assessment of Longer Term Program Engagement. Washington, DC, IMF
Country Report, 05 (139),April.
IMFandInternationalDevelopment Association (1997).Bolivia: FinalDocumentontheInitiativeforthe
HeavilyIndebtedPoorCountries(HIPC).27August.
IMF and International Development Association (1998). Bolivia: Initiative for the Heavily Indebted Poor
Countries(HIPC).September.
IMF and International Development Association (2000). Bolivia: Decision Point Document for the
EnhancedHeavilyIndebtedPoorCountriesHIPC)Initiative.January.
IMFandInternationalDevelopmentAssociation(2001).TheChallengeofMaintainingLongͲTermExternal
DebtSustainability.Washington,DC,April.
IMF and International Development Association (2001). Bolivia: Completion Point Document for the
EnhancedHeavilyIndebtedPoorCountries(HIPC)Initiative.Washington,DC,May.
Independent Evaluation Group (2006). Debt Relief for the Poorest: An Evaluation of the HIPC Initiative,
WorldBank.Washington,DC.Availableat:www.worldbank.org/ieg.
Johnson A (2001). Key Issues for Analysing Domestic Debt Sustainability. Publication, 5. Debt Relief
International.
Kenya (2003). Economic Recovery Strategy for Wealth and Employment Creation 2003–2007. Nairobi,
GovernmentPrinter.
KitabireDandOumoPM(2005).Countrystudy:Uganda.UNCTADprojectonCapacityBuildingforDebt
SustainabilityinDevelopingCountries.Geneva(December).
M’Amanja D and Morrissey O (2005). Fiscal Policy and Economic Growth in Kenya. Working Paper,
CP05/06. Centre for Research in Economic Development and International Trade, University of
Nottingham.
MatyamaF(2001).TheNationalDebtandtheHIPCDebtReliefInitiative–Uganda’sExperience.Kampala,
PaperPresentedattheILIWorkshoponDebtManagementandNationalBudgeting,September.
MbireBandAtingiM(1997).GrowthandForeignDebt:theUgandaExperience.AERCResearchPaper,66.
112
CompendiumonDebtSustainabilityandDevelopment
MoralesJA(1988).InflationStabilizationinBolivia.In:BrunoM,diTellaG,DornbuschR,andFischerSeds.
InflationStabilization:TheExperienceofIsrael,Argentina,Brazil,BoliviaandMexico.Cambridge,
Massachusetts,MITPress:307Ͳ347.
MussaM(2000).MeetingtheChallengesofGlobalisation.Nairobi,PaperpreparedforthePlenarySession
oftheAERCBiͲAnnualResearchWorkshop,May.
MuwangaͲZake ESK (2000). External Debt Management and Policy. Kampala, Paper Presented to the
MasterofArtsEconomicPolicyManagementClassatMakerereUniversity,July.
MuwangaͲZakeESKandStephenN(2001).TheHIPCDebtReliefInitiative–Uganda’sExperience.Helsinki,
PaperPresentedattheUnitedNations(UNI/WIDER)ConferenceonDebtRelief,August.
MwegaFM(2005).ExternalDebtManagementinKenya.UNCTADprojectonCapacityBuildingforDebt
SustainabilityinDevelopingCountries.Geneva(July).
MwegaFM,MwangiNandOleweͲOchiloF(1994).MacroeconomicConstraintsandMediumͲTermGrowth
inKenya:AThreeͲGapAnalysis.AERCResearchPaper,23.
Mwega FM and Ndung’u NS (1996). Macroeconomic Policies and Exchange Rate Management Kenya.
Report of an AERC/ICEG Collaborative Project on Macroeconomic Policies and Exchange Rate
ManagementinAfricanCountries.
MwegaFMandNdung’uNS(2004).ExplainingAfrica’sGrowthPerformance:ACaseStudyofKenya.Paper
preparedforanAERCCollaborativeResearchProject.
NjeruJ(2004).TheImpactofForeignAidonPublicExpenditure:TheCaseofKenya.AERCResearchPaper,
135,March.
Nissanke M and Ferrarini B (2006). Assessing the Aid Allocation and Debt Sustainability Framework:
Working Towards Incentive Compatible Aid Contracts. Helsinki, Wider Development Conference
onAid:Principles,PoliciesandPerformance,June.
O’BrienFSandRyanTCI(1999).AidandReforminAfrica:KenyaCaseStudy.Mimeo,WorldBank.
Oumo PM (2005). Debt Management Process: Uganda’s Experience, Current Situation & Future Plans.
Tokyo,aPaperpresentedattheExternalDebtManagementWorkshop,January.
PattilloCA,PoirsonHandRicciL(2002).ExternalDebtandGrowth.IMFWorkingPaper02(69).
PsacharopoulosGetal.(1992).PovertyandIncomeDistributioninLatinAmerica:TheStoryofthe1980s.
Report,27,Washington,DC,LatinAmericaandCaribbeanTechnicalDepartment,WorldBank.
PratiAandTresselT(2006).WhatistheMostEffectiveMonetaryPolicyforAidͲReceivingCountries?IMF,
EconomicandSocialAffairs,DESAWorkingPaper,12(ST/ESA/2006/DWP/12).February.
Rajan R (2004). Sovereign Debt Structure for Crisis Prevention. IMF Occasional Paper, 237, Washington,
DC.
TheRepublicofUganda(2004).PovertyEradicationActionPlan2004/05–2007/08.Kampala,Ministryof
Finance.
The Republic of Uganda (2002). Uganda PostͲHIPC Debt Strategy and New Financing Workshop Report.
Kampala,November.
The Republic of Uganda (2004). Uganda National Debt Strategy and New Financing Workshop. Draft
Report,Kampala,July.
RoeARandGriggsJ(1990).InternalDebtManagementinAfrica.AERCSpecialPaper,4.
RongeE(2001).ExportBoomRegimesandtheDebtAccumulationProcess:TheKenyanExperience.
Helsinki,WiderDevelopmentConferenceonDebtRelief,17–18August.
113
TheDebtExperiencesofUganda,KenyaandBolivia
Rudaheranwa N (1999). Transport Cost and Export Trade of Developing Countries: A Case Study of
Uganda.PhDThesis,UniversityofNottingham.
SachsJ(1986).TheBolivianHyperinflationandStabilization.Cambridge,Massachusetts,NationalBureau
ofEconomicResearchWorkingPaper,2073,November.
Serven L (1997). Irreversibility, Uncertainty and Private Investment: Analytical Issues and some Lessons
fromAfrica.JournalofAfricanEconomies,Supplementtovol.6(3).
StiglitzJ(2002).GlobalizationanditsDiscontents.England,PenguinBooks.
UNCTAD(2000).CapitalFlowsandGrowthinAfrica.Geneva.
Unidad de Análisis de Política Economica, Ministerio de Finanzas, Bolivia(2000).Bolivia: Economic
Performance and Social Issues: From the 80s Up to Date. March. Available at: www.nssd.net
/pdf/boecon1.pdf.
United Nations (2005). World Economic and Social Survey. New York. Available at: www.un.org/esa/
policy/wess/wess2005files/wess2005web.pdf.
United Nations (2007). World Economic Situation and Prospects. New York. Available at:
www.un.org/esa/policy/wess/wesp2007files/wesp2007.pdf.
vanTrotsenburgAandMacArthurA(1997).TheHIPCInitiative:DeliveringDebtRelieftoPoorCountries.
Washington,DC,InternationalMonetaryFundandWorldBank,February.
World Bank (1992). Kenya Financial Sector Adjustment Project. Project Completion Report. Washington,
DC,December.
WorldBank(2003).Kenya:PolicyAgendatoRestoreGrowth.Washington,DC.
WereM(2001).TheImpactofExternalDebtonEconomicGrowthandPrivateInvestmentsinKenya:An
EmpiricalAssessment.Helsinki,WiderDevelopmentConferenceonDebtRelief17–18August.
114
CompendiumonDebtSustainabilityandDevelopment
CHAPTERVI
CASESTUDIES:ARGENTINAAND
THEREPUBLICOFKOREA
MarioDamill,RobertoFrenkel,MartínͲRapettiandYungChulPark84
A. Introduction
This paper examines recent crises that shook the economies of Argentina and the Republic of Korea as
well as the international financial system. Both were capitalͲaccount crises in apparently successful
middleͲincomedevelopingeconomies.Whilebothcountrieshadexperienceddebtcrisesbeginninginthe
late 1970s, Argentina’s default of 2000Ͳ2001 and the Republic of Korea meltͲdown of 1998 were
exceptional in their severity. International rescue packages led by the IMF were organized in both
instancesandwereasourceofpoliticalcontroversy.
TheArgentinecrisisanddefault,thelargestinrecentyears,isstillsubjecttodisagreementastoitscauses.
SectionIIascribescentralimportancetoawrongdiagnosisofthecrisisbytheIMF,whichconcentratedon
addressing a fiscal disequilibrium during a liquidity crunch. The country’s political leadership shared the
IMF’s belief, as is evident from the various fiscal adjustment programs undertaken. Several factors Ͳ
externalaswellasinternalͲdidpushpublicdebttowardsunsustainablelevels,particularlyinthecontext
of a recession. However, structural features of the economy such as the convertibility regime and the
dollarization of the banking system were of critical importance to the default, which led to a historical
fallingͲoutbetweenArgentinaandtheIMFandanacrimoniousdebtrestructuring.
84
Section B is based on a paper by Mario Damill, Roberto Frenkel and Martín Rapetti (Researchers at CEDES, Buenos Aires)
(Damill, Frenkel and Rapetti (2005)), Section C is based on the paper by Yung Chul Park (Graduated School of International
Studies,SeoulNationalUniversity)(Park(2005)).
115
CaseStudies:ArgentinaandtheRepublicofKorea
TheexperienceoftheRepublicofKoreadescribedinSectionIIIalsosuggeststhatIMFpolicyprescriptions
worsened the crisis by helping to push the economy into default. Section III also highlights the role of
banking crises, collapsing financial markets, and irresponsible foreign borrowing by chaebols. Yet, as in
Argentina,theimmediatetriggersofthecrisiswereadverseexternalshocks,namelytheweakeningofthe
Yenandregionalcontagion.SectionIVcomparesmajorfeaturesofthetwocrises.
B. LessonsfromtheArgentineCrisisandDefault
1. Introduction
This section challenges the leading explanations of the latest Argentinean debt crisis, whereby
uncontrolledpublicspendingisperceivedasthemaincauseofdebtaccumulation,crisisanddefault.
Firstly, it is shown that the effects of rises in interest rates rises were the main driver of public debt
dynamicsattheendofthe1990s.Evenifallowanceismadefortheeffectofuncertaintiesaboutpublic
debt sustainability on investors’ assessment of the country’s position, the main source of the
deteriorationwasnotfiscalpolicybutfinancialfragilityandcontagion.
Secondly, the role of macroeconomic policies – particularly exchange rate policy Ͳ in generating an
unsustainabledebtpathisemphasized.InthisregardtheArgentinecaseisanextremeexampleofbadly
managed financial integration leading to high interest rates, low growth, and vulnerability to financial
contagionandvolatilityofcapitalflows(Frenkel,2003b).
Thirdly,thepaperchallengesacommonlyheldopinionthatthedefaultdecisionwasmainlyresponsible
for the deep crisis in Argentina. It shows on the contrary that the abrupt contraction in activity and
employment occurred before the default as the Government tried to keep debt service on track. The
defaultprovedtobeoneoftriggersthatsubsequentlyallowedrecovery.
Fourthly, the section examines how debt restructuring took place in the context of a confrontational
relationshipbetweentheIMFandArgentina.Themostunusual–indeedunprecedentedͲfeatureofthis
processwasthattheIMFdidnotparticipateinthedesignoftherestructuring.
2. MacroeconomicPerformanceinthe1990s
Between 1977 and 1982 Argentina went through a phase of financial opening and accelerated
indebtednessthatendedinmassivecapitalflight,exchangeͲratecrisis,anddefault.Thiswasfollowedbya
longperiodofinternationalcreditrationingbetween1982and1990.The1991Ͳ2001periodalsoendedin
crisisanddefault.Adistinguishingfeatureofthesetwoperiodsistheroleplayedbytheprivatesectorin
thegenerationofexternalfinancialobligations.Despitethestrongriseintotalexternaldebtinthe1990s
theshareofpublicexternalintotaldebtdeclinedbyover20percentagepointswhichsuggeststhatfiscal
disequilibriumwasnotthemaincauseofthecrises.85
Argentinaenteredbothphasesofacceleratedindebtednessinthecontextofstabilizationprogramsbased
on a fixed nominal exchange rate. These programs set in motion proͲcyclical macroeconomic processes
which left the economy vulnerable to negative external financial shocks. (Frenkel, 1983; Taylor, 1998;
Frenkel,2003a).
In1981thestabilizationpolicybasedontheexchangeͲrateanchorwasabandoned.Anewphasefollowed,
characterized by massive devaluations of the peso. These devaluations were accompanied by higher
85
SelectedindicatorsforArgentina’seconomyduring1977Ͳ2006aregivenintable1.
116
CompendiumonDebtSustainabilityandDevelopment
internationalinterestratesandaneventualpeakoftheratioofforeigndebttoGDPofnearly60percent
in 1982. The public sector’s share in external debt also rose in this period because the Government
assumed a considerable proportion of the private foreign debt. During the second phase of the 1990s
totalexternaldebtincreasedbutmostofthatrisewasgeneratedbytheprivatesector.
Thefiscalbalancewentthroughthreeperiodsinthe1990s.(seetableVI.2.)During1991Ͳ94theaverage
deficit,whichinthe1980swasabout7percentofGDP,decreasedtolessthan1percentofGDP.This
wasmainlyduetoanimprovementintheoverallpublicsectorbalance.Nonetheless,publicdebtrosein
the early 1990s because the Government assumed debts that were not registered in the fiscal balance,
especiallydebtsofpublicͲsectorpurveyorsandofthesocialsecuritysystem.
In1994newnegativepressuresemergedonpublicfinancesduetothreefactors.Firstly,asocialsecurity
reform that created the Private Pension Funds led to a significant fall in contributions. Secondly, the
regionalboomwasfollowedbytheconsequencesoftheTequilaeffectin1995,whichwasmanifestedina
sharp rise in the countryͲrisk premium of Argentina’s interest rates (see table VI.3.). Thirdly, the
Governmentattemptedtocountertheseconsequencesbyloweringthetaxburdenontradables.Between
1995and1997thepublicdebt/GDPratioroseslightlybeforestabilizing.
The Russian and Brazilian crises in 1998 resulted in a new jump in the countryͲrisk premium. This was
accompanied by a recession and increased financial vulnerability of debtors. A sharp rise in interest
payments had already begun in 1996. By 2000 these payments amounted to nearly 19 per cent of
governmentrevenues.Recessionandhigherinterestrateslargelyexplaintheexplosivepathstakenbythe
publicdebtanddeficit,whichhadtheconsequencethatthepublicdebt/GDPratioincreasedbyalmost20
percentagepointsbetween1997and2001.
3. MacroeconomicPerformanceBeforeandAftertheDefault
Themacroeconomicstoryofthelate1990scanbedescribedasaswingfromeuphoriatodepression.The
negative turnaround in the external environment experienced in 1997Ͳ1998 left the economy with a
significantandgrowingcurrentͲaccountdeficit,anappreciatedrealexchangerate,andavisiblelackof
policyinstrumentstodealwiththeproblem.Hence,restrictivefiscalpolicieshadtobearthemainburden
of attempts at adjustment. The expectation was that fiscal discipline would trigger greater confidence,
leading to a recovery in domestic expenditure which would push the economy out of recession. De la
Rua’sadministrationacceptedthisargument,andtheIMFgaveitssealofapproval.
However, the result was failure. Fiscal policy alone was impotent to counter large macroeconomic
imbalances,whichweremostlyrootedintheexternalsectoroftheeconomy.Theeconomysufferedthe
longestrecessionsincetheFirstWorldWar.
Capital inflows contracted sharply in response to the contagion caused by the Mexican crisis at the
beginningof1995(seetableVI.4.).ForeignͲexchangereservesalsofell.However,therecessionwasshortͲ
livedthankstotheeffectsoftheIMFͲledpackageoffinancialsupport.AfterabriefrecoverythecountryͲ
risk premium began to increase again after the devaluation in Thailand in 1997. As noted above, a
sustainedcontractionstartedaftertheRussiandefaultin1998.
Duringtheearly1990stherewerelargeprivatecapitalinflows,followedbyacontractionin1995.Capital
inflowstothepublicsectorweremorestable,beingsustainedintherecessionof1995andduringthat
which began in 1998. Private capital inflows recovered in 1996 but were accompanied by outflows ofa
similarmagnitude.From1998onwardsthenetinflowturnedintoalargenetoutflow.
117
CaseStudies:ArgentinaandtheRepublicofKorea
Theincreaseintheforeignpublicdebtoftheentireperiodfrom1991onwardsexceededUS$35billion.
ThisamountisquiteclosetotheincreaseintheforeignfinancialobligationsofthenonͲfinancialprivate
sectorwhich,however,weremorethanoffsetbytheriseofthesector’sexternalassets.
In December 1999, a new Government took office. As previously mentioned, this Government believed
thatthemaincauseoftheeconomicdepressionwasfiscalmismanagement.Successivepackagesoftight
fiscalmeasureswereapplied.EffortstopreventdefaultincludedaFiscalResponsibilityLawin1999that
setamandatorydecliningtrendforthepublicdeficitdesignedtobringittozeroinafewyears.BymidͲ
2001themeasuresbecamedesperateandincludedanunprecedented13ͲperͲcentacrossͲtheͲboardcut
in public wages and pension benefits. Coming after years of severe recession, these cuts did not
contributetosocialpeace.
Theexpected“confidenceshock”nevermaterialized.Indeed,theroundsofcontractionaryfiscalpolicies
onlyreinforcedthedeflationarytrend.During2000and2001theGovernmentattemptedtocomplement
fiscal measures with some financial initiatives. It also implemented important debt swaps aiming at
convincingthepublicthattherewasnoriskofdefault.Bytheendof2000,apackageoflocalandexternal
support of about US$40 billion was announced (the “blindaje” or financial shield). The IMF led the
operationwithaUS$13.7billionextensionofthestandͲbycreditinforcesinceMarch2000.However,two
monthslater,acrisisinTurkeyledtoasharpriseinthecountryͲriskpremium.
Asareactionavoluntarydebtswap(the“megacanje”)ofbondsofUS$30billionwaslaunchedinJune.
However, because the newly issued bonds carried interest rates of about 15 per cent, they fuelled the
perception that debt had become unsustainable. Another voluntary swap directed at domestic
bondholdersinvolvingUS$42billionofpublicbonds,waslaunchedinNovember2001.Allthesemeasured
failed to halt the withdrawal of bank deposits and the fall of international reserves which began in
October2000.
From the beginning of December 2001 the Government established tough restrictions on capital
movements and on cash withdrawals from banks. It was hoped these measures would hold back the
demandforforeigncurrency,preservethestockofreserves,andmakeitpossibletoavoiddevaluation.In
fact,theyusheredintheendoftheregime.TheDecembermeasuresthrewthecountryintosocialand
politicalunrest.Inthefirstdaysof2002,thecurrencyͲboardregimewasofficiallyabandoned,andwithit
theoneͲtoͲoneparityofthepesototheUS$.
AfterthreeyearsofrecessioneconomicactivitysufferedaparticularlyabruptfallasofmidͲ2001.Social
indicators such us the unemployment rates and poverty indexes, which had worsened in the 1990s,
deteriorated further, adding to social tensions and to the political crisis (Damill, Frenkel and Maurizio,
2003).
Thecatastrophicfallinoutputandemploymentcontinuedforawhileaftertheendoftheconvertibility
regime. However, contrary to mainstream beliefs and quite extraordinarily, a recovery started only one
quarterafterthedevaluationanddefault.Itwastriggeredbythesuddenchangeinrelativepricesinfavor
ofsectorsproducingtradables.
The turnaround was associated with a set of policies aimed at recovering basic macroeconomic
equilibrium.Thepoliciesincludedthefollowing:
(a)
The imposition of restrictions on capital outflows and exchange controls, including under the
lattertherequirementthatexporterssellapartofforeigncurrencyearnings;
(b)
The establishment of taxes on exports, which allowed the authorities to capture some of the
benefitsofthedevaluationforexporters’incomes;
(c)
Aflexiblemonetarypolicyaimedtoassisttherecoveryofbanks;
118
CompendiumonDebtSustainabilityandDevelopment
(d)
AnexchangeͲratepolicyaimedatavoidingtheappreciationofthepeso.
The IMF had insisted on the immediate free flotation of the peso. For a short period the Government
adoptedthisregime.Oncetheexchangeratewasfreetofloat,theexchangeratemovedabruptlytolevels
of close to 4 pesos per US$. The reintroduction of exchange controls was designed to contain further
movement. Soon afterwards the demand for pesos started to recover with US$ in excess supply. This
resultingstabilizationhelpedtohalttheriseindomesticprices,asdidthefreezingofpublicutilityrates.
GDP recovery of the first half of 2002 had a short first phase in which aggregate demand barely rose.
Whatstoppedtherecessionwasarecoveryindomesticproductionwhichwasnowmeetinganincreased
proportion of domestic demand as imports contracted sharply. Investment rose by nearly 40 per cent
between2002and2004,beingfollowedcloselybyprivateconsumption.
Economicrecoverytookplaceinacontextofseverecreditrationing.Investmentwasfinancedbyretained
profits.A“wealtheffect”fromtheexternalassetsholdingsoftheprivatesector,alsohelped.Theseassets
– now estimated at over US$100 billionͲ rose in value as result of exchange rate depreciation, and in
relationtothepricesofdomesticassetssuchasrealestate.
Improvementinthecurrentaccountstartedin1998.Theabruptcontractionofimportsaftertheendof
convertibilityhelpedtotransformadeficitofalmostUS$3billionin1998intoasurplusofUS$17billionin
2002.
On the fiscal front between 2001 and 2004 there was an improvement in the overall balance of the
ConsolidatedPublicSectorfromadeficitof5.6percentofGDPin2001toasurplusof3.5percentin2004
(see table VI.5.). This reflected improvements in the three major components, the primary balance,
interestpayments,andtheaggregatebalanceoftheprovinces.
The most important factor in the improvement of the primary balance was an improvement in tax
revenues due mainly to those on exports and income. In table VI.7. interest payments are shown as
decliningby2.5percentofGDP.However,thisdoesnotindicatetheeffectofthesuspensionofpayments
onexternalpublicdebt,whichatthe2004exchangeratewouldhaveamountedtoabout10percentof
GDP:
4. DefaultonExternalDebtandtheRestructuringProposals
Thesuspensionofservicepaymentsonpartofpublicdebtwasdeclaredon24December2001.Outofa
total of US$144.5 billion US$61.8 billion in public bonds and some US$8 billion in other liabilities were
affected.Thedevaluationofthepesohadamajorimpactontheeconomy’scontractualobligations,given
thepervasivedollarizationofcontracts.Afewdaysafterthedevaluation,aspartofpoliciestoattenuate
theshock,theauthoritiesissuednewdebt.
Themainsourceofthenewindebtednesscamefrominterventionsinthefinancialsystem,andledtoa
US$14.4billionriseinpublicdebt.InFebruary2002theGovernmentdecidedtoundertakeacompulsory
conversion of foreignͲcurrency bank deposits at a rate of 1.4 pesos per dollar.86The withdrawal of
depositswasrestrictedto1,500pesosperpersonperweek.Bankcreditsinforeigncurrencyweresubject
to conversion at a rate of one peso per dollar. This “asymmetric pesoification” of credits and deposits
causedasignificantlossinbanks’networththatwascompensatedbytheGovernment.Newdebtissued
forthispurposeamountedUS$5.9billion.
86
Whenthemeasurewassanctioned,thedollarwasataround2.15pesos.Fourmonthslater,thedollarexchangeratereached4
pesos,decliningsmoothlythereafter.FromMarch2003,theparitystabilizedatbetween2.8Ͳ3pesosperdollar.
119
CaseStudies:ArgentinaandtheRepublicofKorea
The pesoification of private deposits and the forced resetting of their maturities triggered legal claims.
Many of the courts’ rulings were favorable to these claims. In response the Government of President
Duhalde launched three offers for the voluntary swap of deposits for new public bonds, which were
widely accepted by savers. This measure alleviated the financial system’s liquidity problems, but at the
expenseofincreasingpublicdebtbyafurtherUS$6.1billion.
Another source of increases in public debt was the transfer of the bank liabilities of provincial
GovernmentstothecentralGovernment,whichamountedtoUS$9.7billion.ThecentralGovernmentalso
assumed the loss due to the assumption of obligations on provincial Governments’ bonds that had
performedascurrencybetween2001and2003.87Bothtransactionswereguaranteedbyaproportionof
thefutureflowofnationaltaxresources.Lastly,in2002Ͳ2003,publicdebtalsorosebyUS$2billiondueto
obligationstoemployees,pensionersandpurveyors.ThisfollowedaSupremeCourtrulingstatingthatthe
13ͲperͲcentcutofpublicwagesandpensionsinJuly2001wasunconstitutional.
In February 2002, the Government decided to convert into pesos all debts issued in foreign currency
undertheArgentineanlegislation.ThemeasurewouldaffectUS$57.5billionofmostly“guaranteedloans”
issuedaftertheNovember2001swap.88Itwasalsodecidedtoapplyfixedinterestratesof2Ͳ5.5percent
tothe“new”debt.Thepesoificationofthe“guaranteedloans”reducedthe dollarvalueofthe debtby
aboutUS$22.1billion.However,duetotheindexationofthisdebt,bytheendof2003,thevalueofthese
obligationshadrisenbysomeUS$7.3billion.
Insummary,variousmeasuresrelatedtomanagingtheconvertibilitycollapseandthedefaultledpublic
debtstockstoincreasebyUS$28.2billionbetweenDecember2001andDecember2003.
Inthesecondhalfof2003,thefirstofficialstepstowardsrestructuringthedefaulteddebtweretaken.In
September, after an agreement with the IMF, the Government announced the main guidelines of a
restructuring proposal at the annual meeting of the IMF and the World Bank in Dubai. The “Dubai
proposal”established thattheofferwould be directed toholdersofbondsissued untilDecember2001
withauniformtreatment,therestofthedebtbeingserviced.8990
TheGovernmentacknowledgedadefaulteddebtstockofaboutUS$87billionexcludingunpaidinterest.A
75ͲperͲcent haircut was imposed on this amount with new bonds to be issued up to a maximum of
US$21.8 billion. The issue of three new bonds (Par, QuasiͲPar and Discount) was announced for this
purpose.Thefirsttwowouldreceivemoderatehaircuts,butthatontheDiscountbondwouldbehigher.
TheproposalwasconsistentwiththeprimarysurplustargetthathadbeenrecentlyagreedwiththeIMF
(2.4percentofGDPforthecentralGovernmentand3percentfortheconsolidatedpublicsector).
Argentina’screditorsexpressedstrongdisapproval,arguingthatthecountrywasinapositiontomakea
betterofferbycommittingitselftoahigherfiscaleffort.TheIMFexertedpressureontheGovernmentfor
signs of “goodͲfaith”. Similar pronouncements were made in June 2004 by GͲ7 finance ministers. In
response the Government announced a new proposal in Buenos Aires, aimed at getting closer to the
creditors’ positions. The eligible debt was the same as that defined in Dubai.91But in exchange for
defaulted debt new bonds would be issued up to a total of US$38.5 billion, depending on the level of
87
Asofmid2001,someprovincialgovernmentsissuedbondsthatperformedasmoney.WhentherescueprocessstartedinMay
2003,thestockof‘quasiͲmonedas’wasat7.5billionpesos(2percentofGDP).
88
Themeasureaffectedpublicbonds,bilateralloans,debtwithcommercialbanksandotherobligations,whichalladdedtoabout
US$15billion.
89
Thissetofobligationswastermed‘eligibledebt’.Itconsistedof158instruments,issuedin7differentcurrenciesandunder8
jurisdictions.
90
Someobligations(bilateraldebt,debtwithcommercialbanksandothercreditors)remainedwithoutdefinition.InDecember
2003,thesedebtsamountedtoUS$7.5billion.
91
TheBuenosAiresofferspecifiedsomedetailsomittedinDubai.Itwasclarifiedthatthe‘eligibleamount’comprisedthestockof
bondsatDecember31,2001,plusaccruedinterestsuptothatdate.
120
CompendiumonDebtSustainabilityandDevelopment
acceptance.92Later,itwasmadeclearthattheswapwouldcomprisebothcapitalandinterestarrearsand
theamountofthenewbondswasincreasedtoamaximumofUS$38.5Ͳ41.8billion.Thethreedifferent
bondsweremaintainedinthenewproposal.
TheBuenosAiresproposalimpliedahigherfuturefiscaleffort.TheGovernmentwasineffectcommitted
toaprimarysurplustargetof2.7percentofGDPduringthefirstfiveyears,thistargeteasingtoaround
2.3percentofGDPasof2014.Undertheassumptionof3.3percentannualaveragegrowth,projections
indicated that the fiscal effort would finance most interest payments. However, even if the multilateral
organizations agreed to refinance debt due to them, the Government would still have to obtain annual
fundingofabout2percentofGDPfortenyearsaftertheswap.
EvidencethatArgentinawouldfaceaheavydebtburdenaftertheswapdidnoteasecreditors’demands.
ImmediatelyaftertheannouncementinJunebondholders’organizationsrejectedtheproposal.Financial
analysesshowed thatasubstantialhaircutofabout73Ͳ80percentwasimplied.Thesize ofthehaircut
depended crucially on the discount rate used in the calculation. That used was the yield of assets of
emergingͲmarketcountriesratedasofsimilarrisk,i.e.12Ͳ14percent.
By late 2004 developments on international capital markets unexpectedly started to play in favor of
Argentina.GreaterworldliquiditystimulatedtheappetiteforriskandforemergingͲmarketsdebt,andled
to a reduction of developing countries’ risk premium.93In this new context estimates of the haircut
impliedbyArgentina’sproposalwerereducedandtheswaplookedmoreattractive.Thepresentvalueof
offeredbondscalculatedatthenewdiscountratewas30Ͳ35centsonthedollar.Thiswassimilartothe
marketpriceofthedefaultedbonds.
The improvement in the financial environment paved the way for the Government finally to launch the
swapwithoutintroducinganychangetotheJune2004proposal.94TheswapstartedonJanuary14,2005.
On May 3, 2005, the Government announced that acceptance had reached 76.15 per cent. This meant
thatUS$62.3billionoftheoldbondswouldbeexchangedforaboutUS$35.3billionofnewinstruments
and GDP growthͲlinked coupons. The operation reduced public external debt by US$67.3 billion,95and
attenuatedthepublicfinances’exposuretoforeignexchangerisk,sincearound44per centof thenew
bondsweredenominatedinlocalcurrency.
5. Argentina,theIMFandtheInternationalFinancialSystem
At first glance it may seem striking that the crisis and the massive default took place in a country
considered an example of the success of Washington Consensus policies. From the IMF’s perspective
Argentina’s currency board had been a prime example of a feasible corner solution for exchangeͲrate
policyinanemergingmarket(Fischer,2001).Yetatthesametimeitwaswidelybelievedthatthedebt
andtheconvertibilityregimewerenotsustainable,astheprogramdidnotinvolveanysubstantialchanges
tomacroeconomicpolicy.
Argentina’s program aimed at reͲestablishing confidence through commitments to fiscal austerity.
However, the recession and the liquidity crunch meant that it was implausible that the issuing of fiscal
92
In the lower acceptance scenario the recognition of interest arrears would include the period until December 31, 2003 for
aboutUS$18.1billion,whereasinthehigheracceptancescenarioitwouldincludeinterestsarrearstillJune30,2004,forUS$1.4
billion.
93
The JP Morgan EMBI+ index decreased to an average of 375 basis points in the last quarter of 2004, whereas the Brazilian
countryriskͲpremiumfellto417basispoints.
94
Torelieveitselffromcreditors’pressures,thegovernmentgaveuptherighttochangetheguidelinesbysendingabilltothe
Congresspreventingtheadministrationfromdoingso.Congressquicklyapproved.
95
According to minister Lavagna, at the end of 2004, the haircut would reduce debt stocks from US$191.2 billion to US$123.9
billion.Thepublicdebt/GDPratiowouldhavefallenfrom113to72percent.
121
CaseStudies:ArgentinaandtheRepublicofKorea
signalswouldbesufficienttostopthecrisis.Bythetimeofthereductionsingovernmentexpenditurein
midͲ2001, there were good reasons to think that multilateral resources would end up financing private
capitalflightwithoutpreventingadefault.
After the changes at the head of the IMF in 2001, the Fund’s relationship with Argentina became
increasinglystrained.IMFrecommendationsplayedanegativeroleinstabilizationandrecovery.Aprime
examplewasexchangeͲratepolicy.InFebruary2002theIMFdemandedtheimmediateflotationofthe
exchange rate, threatening not to reestablish negotiations in its absence. The implementation of this
measure predictably led to an abrupt rise in the price of the dollar and an acceleration of inflation.
Similarly, there was a clash over the management of the crisis in the banking sector. The Lavagna
Government wanted gradual action and voluntary options, while the IMF promoted heroic “solutions”
suchasbankliquidations.
TheseexamplesshowthattheFundoperatedonthebasisofthediagnosisthat(1)theexchangemarket
couldnotbestabilized,(2)ahyperinflationaryprocesswasunavoidable,and(3)reestablishmentofsome
degreeoffinancialintermediationindomesticcurrencysoonwouldbeimpossible.Theimplementationof
themeasurespromoted bythe IMF wouldhavetransformeditsdiagnosisintoaselfͲfulfilling prophecy.
The IMF maintained its policy line until May 2003 when the Deputy Manager Director recognized the
deficiencyoftheFund’sdiagnosis.
The 2002 and 2003 agreements were signed in the context of a highly confrontational relationship
between Argentina and the IMF. In September 2003, a threeͲyear agreement to refinance debts to the
IMFwasagreed.Thetermsofconditionalitywereonlyestablishedforthefirstyear,astheGovernment
refusedtocommittohighertargetsforsubsequentones.Targetsincludednewregulationsofprivatized
public utilities, measures to strengthen financial system, and a new law about the distribution of fiscal
revenues between the national and provincial Governments. The conditionality also included a clause
underwhichthecountrywastodisplay“goodfaith”inthetreatmentofexternalcreditors.Theambiguity
ofthetermlefttotheIMFagreatmarginofdiscretioninitsevaluation.
AyearlaterArgentinahadcomfortablyfulfilledthequantitativetargetsbutnotthequalitativeones.The
most significant one under the latter heading was probably the finalization of the renegotiation of
contractsandtheestablishmentofanewregulatoryframeworkforprivatizedpublicutilities.Whilethe
IMF was conducting its evaluation, Argentina was presenting the debt restructuring proposal and
organizing the swap. The relationship between Argentina and the Fund reached an impasse. The IMF
couldhaveterminatedtheagreementonthebasisofthefailuretofulfillqualitativetargets.Thatwould
have signified a serious negative shock for a country in the middle of the debt restructuring process. It
couldalsohaveledtofinancialdifficultiesfortheIMFsinceArgentinawasalargeborrower.
Theimpassewasovercomebythesuspensionoftheprogramuntilthebeginningof2005atArgentina’s
request.Thereafter,ArgentinarepaidtotheIMFallprincipalandinterestthatcouldnotbepostponed.In
theperiod2002Ͳ2004itmadenetprincipalpaymentsofmorethanUS$2.1billion,andinterestpayments
of US$1.9 billion. As these figures compared with net receipts of US$23 billion in 1994Ͳ2001, the
ArgentineanMinisteroftheEconomydescribedtheIMFasmovingfrombeinga“lastͲresortlender”toa
“privilegeddebtpaymentscollector”.
A crucial element in the process was the Government’s view that international financial crises and
defaults are the result of excessive debts attributable to the irresponsible behavior of borrowers and
lenders. This irresponsible behavior is encouraged by the implicit guarantee given by the IMF’s rescue
packages. Hence, there should be less intervention by the IMF both under normal conditions and in
default situations. Argentina’s Government requested nonͲintervention of the IMF, arguing further that
the restructuring proposal did not involve additional multilateral funding. The high haircut was seen as
122
CompendiumonDebtSustainabilityandDevelopment
proportionaltotheirresponsibilityshownbythemarket.Indeed,Argentina’sstrategyillustratedboththe
flawsoftheinternationalfinancialsystemandtheviabilityofalternativewaystosolveproblems.
By2006ArgentinahadrestartednegotiationswiththeIMFfromapositionstrengthenedbythehighlevel
ofacceptanceoftheswap.Thenegotiationsgavegreaterlegitimacytotheoperation.Toorigidposition
by the IMF risked being politically uncomfortable for some GͲ7 Governments, and would have
contradictedtheacceptanceofthehaircutbyprivatecreditors.
Moreover, with the high acceptance of the swap, the IMF faced a fait accompli in that the outcome
indicatedanassumptionbythemarketthatArgentina’smultilateraldebtwouldberefinanced.Stillmore
uncomfortably for the IMF the Fund had not participated in the design of the proposal. This clearly
clashedwithIMF’sinstitutionallogicinthattherefinancingofacountry’sdebtwassupposedtorequire
itsapprovalofnewloans.Therefore,byacceptingArgentina’sdemandstheIMFappearedtobeaccepting
achangeinitsrole.
Thesetensionswereexacerbatedbythespecialcircumstancesthattheinstitutionwasgoingthrough.The
IMF had actively participated in the restructurings of sovereign debts with the private sector since the
1980s.TherecentSDRMinitiativewasintendedtobeanextensionofthattradition,andwasanattempt
todefine,formalizeandstrengthentheIMF’sroleincasesofsovereigndebtdefault.AfterWallStreetand
theUnitedStatesrejectedtheSDRMinitiative,thisroleoftheIMFremainsilldefined.Thisisnotthefirst
time that the Governments of developed countries –particularly the United StatesͲ have redefined the
functionsoftheIMFduringtheprocessofdealingwithimmediateandspecificproblems.Forexample,the
1995 Mexican crisis led to IMF rescue packages for capitalͲ as opposed to currentͲaccount crises.
Argentina’s case may eventually contribute to a redefinition of the functions of the IMF in the
internationalfinancialsystem.
C. ExternalDebtManagementoftheRepublicofKoreaduringtheCrisesof1979Ͳ
1980and1997Ͳ1998
1. Introduction
During the past four decades, the Republic of Korea has experienced a number of periods of financial
stress.Themostseriouswasthe1997Ͳ98crisisthatbroughtthecountrytothebrinkofdefault.Theother
periodsofstress,includingthecrisisof1979Ͳ80,werelessdamaging(Park,1986;Cooperetal.,1994).In
many respects the causes were similar: they included investment booms in the periods leading to the
crises,largeandgrowingcurrentͲaccountdeficits,andappreciationsoftherealexchangerate.However,
the 1997Ͳ98 financial meltͲdown was a capitalͲaccount crisis, of which the Republic of Korea had no
previousexperience.96
TheRepublicofKoreaengineeredaquickrecoveryfrombothcrises.Intermsofeconomicfundamentals
there was no reason to believe the Republic of Korea was any more vulnerable to a crisis during the
second half of the 1990s than it had been two decades earlier. Nevertheless, the cost of resolving the
secondcrisiswasfargreater,andthetwocrisesfolloweddifferentadjustmenttrajectories.
SectionBdiscussesthebuildͲupandresolutionofthe1979Ͳ80debtcrisis.ThisisfollowedinSectionsC
and D by an examination of macroeconomic developments prior to and in the aftermath of the second
crisis.SectionEexploresthelessonsandSectionFcontainsasummaryofthemainpoints.
96
SelectedeconomicindicatorsforRepublicofKoreafor1975Ͳ1985and1995Ͳ2004aregivenintables6and7.
123
CaseStudies:ArgentinaandtheRepublicofKorea
2. The1979Ͳ1980DebtCrisis
The Republic of Korea economy slowed in 1979 after three years of strong growth, while the current
accountslidintodeeperimbalance,risingtoadeficitof6.6percentofGDPin1979andof8.3percentin
1980fromoneof2percentin1978.In1980outputcontractedby1.5percentandtheconsumerprice
index (CPI) soared to 29 percent. The economy was thus experiencing stagflation with a large currentͲ
accountimbalance.AtthesametimetotalexternaldebtasaproportionofGDPswelledto42.6percent.
In these circumstances a traditional IMFͲsupported prescription would have included a strong dose of
stabilizationmeasurestogetherwithacurrencydevaluation.ButRepublicofKoreapolicymakersoptedfor
adifferentgrowthͲfirstpolicy.TothesurpriseoftheIMFandtheinternationalfinancialcommunity,the
economyreboundedin1981,growing6.2percent.
At the centre of Republic of Korea economic policy in the midͲ1970s was the plan for the heavy and
chemicalindustries.Thispolicyentailedtaxincentives,lowͲcostbankcredit,andothersubsidiesmostlyto
large firms belonging to the Republic of Korea’s industrial groups or chaebols. The result was an
investmentboomleadingtoariseintheratioofgrossinvestmenttoGDPfrom28.7percentin1977to36
percentin1979.Atthesametimetheeconomyoverheated,withannualincreasesinrealwagesin1976Ͳ
78averagingover18percent.Asteephikeinagriculturalpricescausedbyapoorharvestin1978further
aggravatedinflationarypressures.
Despite this, the Republic of Korea Government was determined to maintain a dollarͲpegged exchange
rate.Thisledtoanappreciationoftherealexchangerate,whichinturnunderminedexportearnings.At
thesametime,theRepublicofKoreasufferedadverseexternalshocks.Itwashitbythesecondoilcrisisin
1979,sufferinga15ͲperͲcentdeteriorationinitstermsoftradein1979Ͳ1980.
Furthermore, the Republic of Korea was thrown into political turmoil by the assassination of President
Park in 1979. The new military Government of May 1980 was hardly in a position to adopt a strong
stabilization program. Political uncertainties worsened Republic of Korea economic prospects. Not
surprisingly, businesses adjusted by cutting investment, fixed investment falling by 11 per cent in 1980.
Theeconomysankintoadeeprecessionin1980,whichwasaggravatedbyacrisisintheinformalcredit
market. However, surprisingly the currentͲaccount deficit did not shrink as expected. This was because
consumption remained strong: consumers considered the fall in output transitory and cut their savings
rather than their consumption. As a result the share of saving in GDP dropped more than that of
investment.
Lackingsupportforastabilizationprogram,thecaretakerGovernmentfocuseditspolicyresponsetothe
deteriorating current account on the exchange rate. The won was devalued visͲàͲvis the US$ by 27
percentin1980,andthereaftertheRepublicofKoreamovedtoamanagedfloattiedtoabasketofmajor
international currencies. On the macroeconomic front the Government gave priority to stopping the
economicdownturn.
Hereitscommitmentwastobroadlyconceivedstabilizationtogetherwithfinancialreformandcorporate
restructuring. In its view inflation was at the root of the deterioration in income distribution, of labor
unrest, and of the weakening of the country’s export competitiveness. A growthͲfirst strategy would
succeedonlyifthedeficitonthecurrentaccountwasbroughtundercontrolandfinancedexternally.The
prospectforsuchapolicywasuncertainastheRepublicofKoreahadoneoflargestexternaldebtsamong
developingcountries.Nevertheless,debtservicelevelsremainedwithinasustainablerange.Thegamble
paid off. The Government maintained an expansionary policy until 1983 when it began restraining
domesticdemand.By1981inflationwasalreadysubsidingandtheeconomyrecoveredfullyonlyayear
aftertherecessionof1980.
124
CompendiumonDebtSustainabilityandDevelopment
What were the factors responsible for the dramatic turnaround? Haggard and Collins (1994) single out
threedevelopments:
x
An improvement in the external environment due to (1) falling prices of oil and raw materials
leading to a better terms of trade and lower inflation, (2) declining international interest rates,
and(3)anappreciationoftheyenagainsttheUS$;
x
Decliningrealwagesinboth1980and1981,partlyduetomoreflexiblelabormarkets;
x
The depreciation of the real exchange rate, which improved the Republic of Korea’s export
competitiveness.
However,otherimportantfactorsalsohelpedpulltheRepublicofKoreaeconomyoutofcrisis.Onewas
theclosedcapitalaccount.ThisallowedflexibilityaswellaseffectivenessformonetarypolicyinafixedͲ
exchange rate regime. Despite the economic crisis and political turmoil, the Republic of Korea did not
experienceanycapitalflightoranywithdrawalofforeignloans.TheothercrucialfactorwastheRepublic
ofKorea’sabilitytofinanceitscurrentͲaccountdeficitexternally.Thecountrywasneverdeniedaccessto
internationalfinancialmarkets,althoughitsborrowingcostswentup.
By 1983, stability returned alongside of the resurgence in growth. By now both domestic demand and
exportearningswerestrong.Inthesecircumstancesacontinuationofloosemonetaryandfiscalpolicies
couldhaverekindledinflation.Furthermore,totalexternaldebtremainedatover47percentofGDP.To
reduce the debt burden the current account had to move in the direction of surplus. This explains the
Government’sshifttoastabilizationpolicywhichwassustaineduntil1988.
3. The1997Ͳ1998Crisis
(a)InvestmentBoomFueledbyForeignBorrowing
TheRepublicofKoreaeconomyreboundedstronglyfromaslowͲdownin1992and1993.Thisgrowthwas
ledbyexportsandinvestment(39percentofGDPin1996).Inthatyear,thedeficitoncurrentaccount
was a little over 4 per cent of GDP and apparently manageable yet a major financial crisis followed in
1997Ͳ1998.
Expansion of investment on this scale in an economy with still small financial markets led to higher
external borrowing. Two major developments can help to explain this debtͲfinanced investment surge.
Thefirstwasthestrengtheningoftheyenfromthesecondhalfof1992tothefirsthalfto1995.Thisrise
endedinthespringof1995whentheyenhitthelevelof79.5yentothedollar.Theyen’sappreciation
broughtaboutasharpincreaseinRepublicofKoreaexportearningsbecausemanyofitsindustrieswere
indirectcompetitionwiththoseofJapan.
The second development was increased financial openness, which increased the availability of lowͲcost
foreign credit. In the period 1996Ͳ1998 external debt rose from 28 to 47 percent of GDP. Much of the
inflows during 1995Ͳ1997 consisted of shortͲterm borrowings by domestic financial institutions, which
usedtheproceedstofinanceinvestmentsbychaebols.Theconsequencesincludedseriouscurrencyand
maturitymismatchesinthebalancesheetsoffinancialinstitutions(Park,1998,andParkandSong,2002).
Atthesametime,RepublicofKoreaindustrialgroupswereincreasingtheirinvestmentsabroad.Muchof
this investment was financed with foreign credits. This helps to explain a rise in the foreign debts of
domesticfirmsfromUS$35.6billionin1996to US$43.2billion ayearlater. Theliabilitiesoftheforeign
subsidiariesand branches ofRepublicofKoreafirmswereestimatedto have exceeded US$51billionat
theendofJune1997.
125
CaseStudies:ArgentinaandtheRepublicofKorea
(b)TheBurstingoftheInvestmentBubbles
Attheendof1995,theJapaneseyenbegantodepreciate.Atthesametimethetermsoftrademoved
against the Republic of Korea, and continued to deteriorate for the next two years. This termsͲofͲtrade
shock, which reflected falling demand for the Republic of Korea’s major exports, worsened the current
accountandsetinmotionaweakeningoftheeconomy.Therealexchangeratecontinuedtoappreciate
throughout 1996. However, policymakers did not react owing to their preoccupation with industrial
restructuring and the belief that a strong won would help facilitate the shifting of resources from light
manufacturingtomoreskillͲandknowledgeͲintensiveindustries.
Despiteaslowdownoftheeconomyasofthesecondhalfof1996largeindustrialgroupsdidnotadjust
theirproductionandinvestment.Astheirinventoriespiledup,sodidtheirlosses.However,commercial
banks became less accommodating. As a result chaebols tuned to highͲcost, shortͲterm loans from any
foreignfinancialinstitutionswillingtolendtothem.
WhyweretheRepublicofKorea’sindustrialgroupssoinflexibleinadjusting?Theanswerlayinsomeof
theircharacteristics.Oneofthesewastheirpredatorybehaviorofcompetingformarketsharemorethan
for profit, a feature nurtured by an industrial policy geared to obtaining economies of scale in major
export industries. Thus, when profits fell, they tried to protect their share by investing more, and by
diversifyingatthefirstopportunityintonewindustries.
Furthermore,rigidandbureaucraticmanagementsystemswithdecisionͲmakingconcentratedatthetop
made it difficult for chaebols to adjust quickly to changing market conditions. Reluctance to reduce
investmentledtohighlyleverage.Asurveyconductedin1997showsthattheaveragedebtͲequityratioof
the30largestchaebolswasabove380percentin1996,fourtimesashighasthatofTaiwanProvinceof
China (Korea Institute of Economics and Technology, 1997). The high leverage and the balance sheet
mismatches of the corporate sector proved to be the Republic of Korea economy’s most damaging
structuralweaknessesduringthecrisis.
By1994thenewGovernmentwascommittedtomarketopeningasaresultofitsWTOagreement.This
agreementreducedtheroomforindustrialpolicy,whilereductionincontroloverfinancialactivitytook
away much of the rest of the Government’s ability to coordinate investments. Moreover reform of
corporategovernanceorfinancialregulationfailedtokeeppacewithmarketopening.
(c)TheFinancialMarketCollapseof1997
Theexternallyfinancedinvestmentboomcouldonlylastsolong.Oncetheinvestmentbubbleburst,the
number of corporate bankruptcies also soared as did the volume of nonͲperforming loans at financial
institutions.BetweenDecember1996andJune1997,nonͲperformingloansasaproportionoftotalloans
doubled(Park1998).Thefirstmajorcorporatecasualtyinthesecondhalfof1996wasHanbo,the14th
largestchaebol.
TheinvestigationintotheHanbocollapserevealedthatmanyloanstothisgrouphadbeenmadeunder
politicalpressure.Theextentofunholytiesbetweenpoliticiansandindustryandthescaleofcorruption
shocked both the Republic of Korea people and foreign investors. Moreover, the pervasiveness of
corruptionunderminedtheconfidenceofforeigninvestorsintheGovernmentandtheeconomy,thereby
helpingtobringaboutthecrisis.
MorehighͲprofilebankruptciesfollowed.TheKiagroupwasputintoliquidationinOctober1997,andwas
followedbyfivemorechaebols.Bytheendof1997,theRepublicofKoreahadalameͲduckGovernment
whichwasunabletorestorestabilitytotheRepublicofKoreafinancialmarkets.Foreigninvestorsknew
this,andbeganwithdrawingfundsfromtheRepublicofKoreastockmarketduringtheautumn.
126
CompendiumonDebtSustainabilityandDevelopment
ReflectingtheineffectivenessoftheGovernment,exchangeͲratepolicyinthelastthreemonthsbefore
the crisis drifted into inconsistency and unpredictability. The won had been under strong pressure of
depreciation since early 1997. Throughout the year, the Government stated that it would defend the
exchangerate.Whenthewon/US$exchangerateapproachedthepsychologicallyimportantlevelof1000,
theGovernmentintervenedheavilyinthemarket,onlysuddenlytowithdrawafewdayslater.
Between June and November 1997, the Bank of Korea’s reserve holdings fell by US$10 billion. The
Government further strained investors’ credulity by failing to divulge the true level of foreign exchange
reserves.ItassertedthattheBankofKoreastillheldaboutUS$30billioninreserves,whentheactuallevel
ofusablereserveshadalreadydroppedbelowUS$22billioninMarch.BytheendofNovemberthefigure
hadfallentoUS$7billiondollars.
The dire financial situation was further compounded by changes in sovereign credit ratings. Between
JanuaryandNovember1997,Moody’sadjustedtheratingdownwardtwice,andS&Pthreetimes.Bythe
sametokenthepremiumonRepublicofKoreasecuritiesrose.Foreignbanksbegantorefusetorollover
shortͲtermloanstotheRepublicofKorea.Theactionsofthecreditratingsagenciesgeneratedavicious
cycleofdecliningratingsandmarketsentiment.
4. ManagementofandRecoveryfromthe1997Ͳ98Crisis
BytheendofOctober1997thefinancialsituationwasoutofcontrol.Foreigninvestorsmovedoutofthe
stockmarketindroves,andRepublicofKoreabankswereincreasinglyunabletorollovertheirshortͲterm
foreignloans.Toavoiddefault,theywereforcedtoturntotheBankofKoreaforliquidityortoresortto
foreignovernightloans.
No action was taken until the announcement on 19 November of a reform package, which included
measures for the disposal of nonͲperforming loans and a widening of the band for exchangeͲrate
movements. In the prevailing panic, the market hardly noticed. Three days later, unable to control the
situation,theGovernmentpubliclyapproachedtheIMFforassistance.NegotiationsbetweentheRepublic
ofKoreaGovernmentandtheIMFwerecompletedinarecordtimeof10days.TheIMFagreedtoprovide
atotalofUS$21billiontobedisbursedoverathreeͲyearperiod.Italsosecuredfinancialcommitments
totalingUS$36billionfromtheWorldBank,theAsianDevelopmentBank,theUnitedStates,Japan,and
othersasasecondlineofdefense.
IMF conditionality required tight monetary policy, a fiscal surplus, sweeping financialͲsector reform
includingfurtherliberalization,greaterflexibilityinthelabormarket,andrestructuringthechaebols.By
the end of December, a 25ͲperͲcent interest rate ceiling and most capital controls were abolished. The
limit on aggregate stock ownership by foreigners was raised to 55 percent, and the shortͲterm money
market was also to be deregulated. However, the swift conclusion of negotiations did little to change
market sentiment which was also affected by the political uncertainties concerning the outcome of the
presidential elections due in December 18. The won/dollar exchange rate continued to depreciate;
interestratessoared;andstockpriceswentintoanoseͲdive.
The squeeze on the money supply together with banks’ efforts to meet the 8Ͳ percent Basel capital
adequacyratiobyApril1998reducedtheavailabilityofbankcredit.InDecemberthepercentagerateof
loan defaults jumped to 1.49 from 0.14 a year earlier, while business failures were five times higher.
ExternallenderssawthattheIMFfinancingwhichhadbeenagreedwasshortoftheamountofforeign
debt repayment due. There were also concerns that tight monetary and fiscal policies would depress
economicactivitysomuchthattheRepublicofKorea’sabilitytoserviceitsdebtwouldbeundermined.
Interestratesshotuptothedizzyingheightof40percent,andthewon/depreciatedtoalevelof1,995per
dollar.
127
CaseStudies:ArgentinaandtheRepublicofKorea
Thefinancialsituationwasclearlyunsustainable,andrumorsbegantocirculatethattheRepublicofKorea
mighthavetodeclareadebtmoratorium.OnChristmasEve,theIMFandtheGͲ7countriescameupwith
anotheremergencyfinancingprogramofUS$10billion,drawingonthesecondlineofdefense.Thenew
package succeeded in turning around market sentiment as it demonstrated the resolve to rescue the
RepublicofKoreafromfinancialcollapse.Foreignlenderswantedtobeassuredofpaymentsofprinciple
and interest. They asked for and received government guarantees on private debt on the basis of the
argument that this would facilitate debt restructuring and new credit extension. By January 1998,
international creditor banks agreed to convert most of the shortͲterm debt of Republic of Korea banks
(US$24 billion) into longͲterm loans, with government guarantees that mature over one to three years,
andinterestratesof2.25Ͳ2.75pointsaboveLibor.
In1998thegrowthrateofGDPplungedtoͲ6.9percentfrom+4.7percentayearbefore.Pricesleaped
by7.5percent,thewondepreciatedby27percentvisͲàͲvisthedollar,andtheunemploymentreached8
per cent, the highest since the 1960s. Surprisingly, the crisis was short lived. The rebound was no less
drastic than preceding fall. The Republic of Korea economy grew by 9.5 per cent in 1999, and recovery
continuedthereafter.
TheinitialGDPcontractionin1998waslargelycausedbythecollapseofinvestment.TheconsumptionͲ
GDPratioremainedfairlystable,whiletheinvestmentͲGDPratiodroppedsharplyto25percent.In1998,
therewasahugecurrentͲaccountsurplusofalmost12percentofGDP.Thiswasbecauseimportdemand
declinedby22percentin1998,whileexportsfellbyunder3percent,movementswhichreflectedthe
influenceofboththerecessionandthedepreciationofthewon.
An empirical examination by Park and Lee (2002) of worldwide patterns of adjustment in 160 currency
crisis episodes from 1970 to 1995 shows a widespread tendency for countries to undergo a VͲtype
recovery of real GDP growth similar to that experienced by the Republic of Korea after the 1997Ͳ1998
crisis.Thestudyalsoshowsthatalargerealdepreciation,expansionarymonetaryandfiscalpolicy,andan
improvement in the global economic environment are usually responsible for the upturns. All of these
developmentswerepresentduringthesecondRepublicofKoreacrisis.WhatdistinguishestheRepublicof
Koreaexperiencefromothersarethedegreeoftheinitialcontractionandsubsequentrecovery.Thiswas
duetothefollowingfactors:
x
ExchangeRateDepreciationandOpenness:inviewoftheRepublicofKorea’srelativelyhighlevel
ofopennessandrelativelylargetradesector,adepreciationoftherealexchangeratewasgoing
tohaveanespeciallylargeimpact.
x
Favorable External Environment: the Republic of Korea economy was the beneficiary of an
improvement in the external trading environment. The global economy was strong in 1999.
Moreover,theRepublicofKoreaexportsalsobenefitedfromhigherpricesofsemiconductors,and
fromanappreciationoftheyenwhichimproveditsindustries’competitiveness.
x
MacroeconomicPolicyAdjustments:realizingthedepthoftheslowdown,theIMFagreedtorelax
monetary and fiscal policies as early as April 1998. The ensuing expansion of money supply
preventedafurthercontractionofdomesticdemand.
The positive role of expansionary macroeconomic policies in the postͲcrisis recovery has raised the
questionofwhethertheinitialtighteningwastooharsh,maintainedfortoolong,andasaconsequence
deepened the crisis. In order to deal with the crisis, the IMF chose a traditional policy prescription
designedformanagingacurrentͲaccountcrisis,whichcomprisedtightmonetarypolicyandfiscalausterity.
However, the Republic of Korea crisis involved principally the capital account. In these circumstances
increasedinterestratesresultedinwidespreadbankruptcieswhichdidlittletorestorefinancialstability
andtheconfidenceofforeignlendersandinvestors.
128
CompendiumonDebtSustainabilityandDevelopment
TheIMFandsupportersofthecontractionarymonetarypolicyarguethatintheabsenceofsuchapolicy
capitaloutflowsandthebankrunwouldhavecontinued.ThosewhodisputetheIMFviewsuchasRadelet
andSachs(1998)andFeldstein(1998),ontheotherhand,maintainthattheRepublicofKoreaproblem
wasoneofliquidity.Therefore,thetraditionalIMFstrategywaslikelytohavedonemoreharmthangood
asitdrovemanyhighlyleveragedbutviablefirmsoutofbusiness,therebydeepeningeconomicrecession.
5. LessonsoftheTwoCrises
Bothdebtcriseswereinpartprecipitatedbyinvestmentboomsfinancedbyforeignborrowing.Theratios
ofexternaldebttoGDPweresimilar,andtheRepublicofKoreaexaggeratedthecrisesbyadheringfortoo
longtorigidexchangeͲrateregimes.Inbothcases,TheRepublicofKoreaeconomyreboundedswiftlyin
bothcases,butthescarsofthe1997Ͳ98crisisweremoreextensiveanddeeper.
Themostsignificantdifferencebetweenthetwocrisesinvolvedthepolicyresponses.Inthe1979Ͳ80crisis
the Republic of Korea policymakers took advantage of the country’s continuing access to international
financialmarketstofinancethedeficitoncurrentaccountinthebeliefthateconomicfundamentalswere
strongandthattheeconomywasafflictedbyatransitoryimbalance.InthesecondcrisistheGovernment
hadtoseekIMFfinancingthatsubjectedtheeconomytoawiderangingarrayofpolicychanges.Itpaida
highpriceintermsoflostoutputandofthecostofresolvingbankruptfinancialinstitutionsandbailingout
insolvent corporations,whichamounted16 percentof GDPin1998.Ithad nochanceofreplicatingthe
strategyofrelianceonexternalborrowingfollowedafterthe1979Ͳ1980owingtoitsincreasedfinancial
opennessandthemorelimitedpossibilityofrecoursetocapitalcontrols.
GreaterfinancialopennesswastheresultofthepolicyofeconomicliberalizationpursuedsincethemidͲ
1980s,whichhadalsoresultedinamoreopentraderegime.TheGovernmenthadopenedthefinancial
sectorandderegulatedcapitalͲaccounttransactionaheadofthebidtojointheOECDintheearly1990s.
By the time of the 1997Ͳ98 crisis broke out the Government had been reforming institutions and
restructuringitsfinancial,corporate,andpublicsectorsformorethanadecade.
A financially open economy with a relatively inflexible exchange rate lacks an effective buffer against
external financial shocks. Moreover orderly financial opening requires an efficient financial regulatory
system to monitor risks. The reform of the regulatory system lagged in the Republic of Korea at a time
whenfinancialinstitutionsweretakingonnewrisks,especiallyintheiroperationsabroad.
AccordingtoEichengreen,WyploszandRose(1996)therearethreetypesofdistortionthatcangiverise
toafinancialcrisis.Thefirstisasymmetricinformationwhereborrowersorissuersofdebtorequitytake
advantageofsuperiorinformationascomparedwiththatoflendersandinvestorsabouttherisksoftheir
business.Asymmetricinformation,isassociatedwiththedangerofherdbehavioronthepartofforeign
investors and financial institutions. Second is moral hazard in both domestic and international financial
markets. This denotes the danger that those who expect protection against loss through bailͲouts by
publicauthoritieswilltakegreaterrisksthantheywouldotherwise.Thethirdisanydistortionthatcould
leadtotheinstabilityintheexchangerateassociatedwithmultipleequilibriainforeignexchangemarkets.
AllofthesedistortionswerepresentintheRepublicofKoreaintherunͲuptothe1997Ͳ98crisis.
Before and during the early years of market liberalization foreign lenders and investors did not care to
learnaboutthestructuralweaknessesofRepublicofKoreabanksandcorporategovernancebecauseof
governmentguarantees.OnlywiththegrowingexposureoftheKoreaneconomytointernationalfinancial
marketsdidtheirawarenessincreaseofbalanceͲsheetmismatchesatbanksandchaebols.Bythetimethe
Thai crisis spread to other parts of East Asia in September 1997, the Republic of Korea began losing
reserves. Lacking confidence concerning the adequacy of Republic of Korea reserves, lenders and
investorsbegantoreducetheirexposuretothecountry,refusingeventorenewshortͲtermloans.Both
129
CaseStudies:ArgentinaandtheRepublicofKorea
borrowersandlendersweretoblameforbringingonthecrisisͲborrowersowingtotheirdisregardfor
prudenceandriskmanagementandlendersowingtotheirshortͲtermismandherdmentality.
InternationalfinancialmarketsandRepublicofKoreapolicymakersshareresponsibilityforfailingtocarry
outreformswhichwouldhavereducedmoralhazard.Commercialandmerchantbankhadlongoperated
with implicit government guarantees. Together with inadequate supervision these guarantees provided
incentivestobankstoborrowlargeramountsoffundsabroad,andtoinvestinriskierprojectsthanthey
wouldotherwise.
Moral hazard also appears to have affected the lending behavior of foreign financial institutions. These
expected to receive national treatment. Assuming that they too would benefit from government
guarantees, foreign banks did not conduct careful credit analyses of Republic of Korea borrowers.
Moreover, when the crisis broke out, few foreign banks attempted to reschedule loans to troubled
Republic of Korea banks in sharp contrast to their behavior towards delinquent borrowers in their
domesticmarkets.
Finally, creditors believed that, as a group, they could pressurize the Republic of Korea Government if
there was a crisis. In the event this assumption was to prove justified since their pressure played an
importantroleinthedecisionoftheRepublicofKoreaGovernmenttoseekIMFfinancing.Thebankswere
aware that a debt moratorium was not a realistic option owing to the large number of lenders and
borrowers involved. Banks’ recourse to this pressure also reflected their knowledge that IMF programs
favorcreditorsoverdebtors(Soros1998).
Thecrisisof1997Ͳ98wasacapitalͲaccountcrisisinwhichtheinitialcurrentaccountͲimbalancedidnot
playaprimaryrole.Massivecapitaloutflowsprovokedaliquidityandcreditcrisis.Intheseconditionsthe
traditionalIMFstabilizationprogramdidnotwork,andaninfusionoffreshcapitalwasrequiredtostop
thebleedingoftheeconomy.
It is natural to ask whether the Government could have followed the same policy as that pursued in
response to the crisis of 1979Ͳ1980, i.e. combining a growthͲoriented macroeconomic policy with
continuedrelianceonexternalborrowing.Itishardtobelievethatfinancialmarketstodaywouldsupport
anythingbutamacroeconomicstabilizationprogram,evenifthereweregoodgroundsforthinkingthat
thecrisiswouldbetransitory.InsuchanenvironmenttheaccumulationoflargereservesthroughcurrentͲ
account surpluses by major emergingͲmarket countries as insurance against the imposition of
inappropriatestabilizationprogramsbecomesfullyunderstandable.
D. ConcludingRemarks
ThesecondcrisesofboththeRepublicofKoreaandArgentinawerecapitalͲaccountcrisesthattookplace
in economies that had liberalized capital transactions and that were thus integrated into international
financialmarkets.CapitalinflowswhichfuelledthegrowthprecedingthecrisesandwhichintheRepublic
ofKoreacasebecameaninvestmentboomweretransformedintooutflowswhichledtomeltͲdowns.In
bothcasesIMFpolicyprescriptionsworsenedthecrises.
During much of the 1990s Argentina experienced strong growth. However, as early as 1995 adverse
developmentsintheexternalenvironmentbegantotriggereconomicdifficultiesandthecountrysuffered
aminiͲcrisistogetherwithasharpdeteriorationinitsfiscalbalancefollowingtheMexicancrisisof1995.
Afterarecoveryin1996Ͳ1997Argentina’sriskpremiumbegantoriseagainandforeignborrowingbecame
more costly. External debt was increasing, while the ability to pay was being undermined. A series of
rescue packages failed to restore confidence, and were unable to stop eventual bank runs and the
bleedingofforeignexchangereserves.Thedollarizationofbankcreditsandofthecontractualstructureof
130
CompendiumonDebtSustainabilityandDevelopment
theeconomymadethecollapsemoresevere.PublicͲsectordebtroseafterthedefaultowingtomeasures
takenbytheGovernmentaspartofitsinterventioninthefinancialsystem.
The Republic of Korea crisis also came after a period of high growth, and was triggered by the
depreciationoftheyenandadverseshockstoitsexports.Thesechangestriggeredgreaterattentionon
the part of foreign lenders to the scale of foreign borrowing by chaebols and to deteriorations in their
balance sheets. The rises in bankruptcies and nonͲperforming loans that followed heralded a financial
market crisis with foreign borrowers refusing to rollͲover major bank loans. Given the context of the
earliercrisisinThailandthewoncameundermassiveattackasbankrunsandcapitaloutflowscontinued.
ThesedevelopmentsunderlinedtheimportanceofbetterfinancialregulationandcorporategovernanceͲ
andnotjustgoodmacroͲeconomicmanagementͲasessentialelementsofsuccessfuldebtmanagement.
In both countries the meltͲdowns led to sharp falls in GDP growth. Resolution of the debt problems
followed different courses. In the Republic of Korea case, bank lending to private borrowers was more
important and resolution involved the conversion of shortͲterm bank loans into longerͲterm loans with
government guarantees. In Argentina debt securities were more important and restructuring involved
theirconversionintoalternativesecuritieswithlowercouponsorvaluesandlongermaturities.
InbothcasestheIMFprogramsincludedillͲconceivedpolicymeasuresduetomistakesindiagnosis,which
worsenedtheimpactofthecrises.InArgentina,theausteritymeasuresdeepenedtherecession,thereby
undermining payment capacity and accelerating default. In the Republic of Korea, earlier relaxation of
monetaryandfiscalpolicycouldhavemeantthatbankruptciesandlostoutputwouldhavereachedless
than16percentofGDP.
In both cases, the recovery was aided by favorable external developments such as the improvement in
appetiteamongstlendersandinvestorsfordevelopingcountryrisk,easingofinterestratesandimproved
export markets. In Argentina default also provided a respite to the fiscal balance and the domestic
economy.
Inbothcases,devaluationcompressedimportsaswellashelpingexports(whoseincreasewasparticularly
notable for the Republic of Korea). The turnarounds were surprisingly quick: the worst of the crisis in
Argentina was in December 2001, and signs of recovery were evident in the first half of 2002; and the
Republic of Korea crisis collapse of 1998 was followed by a spectacular recovery as early as 1999. This
followsapatternidentifiedbyLevyͲYeyatiandPanizza(2006)accordingtowhich,bythetimeadefault
occurs,thelossesintermsofoutputandgrowthhavealreadytakenplacesothatitsoccurrencecoincides
withthebeginningofeconomicrecovery.Animplicationofthispatternisthat,oncefirmexpectationsof
defaulttakeholdandthemeltͲdownstarts,measurestopostponethedefaultmaywellbemorecostly
thanthedefaultitself.
Beyondacertainpoint,neithercountrycouldhavedoneanythingtostopexternaldebtfromfollowingan
exploding path. Herd behavior delivered the final blows, As Park (2005) notes, international financial
markets are not a good source of shortͲterm liquidity for emerging economies, when they are
experiencing financial instability. The lesson drawn by several emergingͲmarket countries has been to
accumulate reserves as a form of insurance. If these economies felt assured of adequate liquidity
assistance from international financial institutions or regional financial cooperative arrangements, they
wouldbelessinclinedtofollowthispolicy.
IMFpolicyfailuresandtheperceptionthatitsidedwithcreditorsinthesetwocriseshavecontributedto
underminingofitsauthorityamongstdevelopingcountries.Argentina’sdebtrestructuringproposalsand
independent recovery program have set a precedent for crisis resolution not mediated by the IMF.
However, the faith of the United States and private creditors in individual debt workͲouts under rules
subjecttoonlyminormodificationsincomparisonwiththepresentregimeisunlikelytoconstituteafullyͲ
131
CaseStudies:ArgentinaandtheRepublicofKorea
fledged, unquestioned alternative. Moreover impetus from these quarters in favor of further capitalͲ
accountliberalizationhasnowbeenlost.
132
CompendiumonDebtSustainabilityandDevelopment
References
CetrángoloOetal.(2000).LasostenibilidaddelapolíticafiscalenAméricaLatina.Elcasoargentino.In:
Talvi E and Végh C eds. ¿Cómo armar el rompecabezas fiscal? Nuevos indicadores de
sostenibilidad.IDB,Washington,DC.
Cetrángolo O and Jiménez JP (2003). Política fiscal en Argentina durante el régimen de convertibilidad.
SerieGestiónPública,108.SantiagodeChile,CEPAL.
Cooper RN (1994). Korea’s Balance of International Payments. In: Haggard S, Cooper R, Collins S, Kim C
andRoST.
Damill M (2000). El balance de pagos y la deuda externa pública bajo la convertibilidad. Boletín
InformativoTechint,303.BuenosAires.
Damill M and Frenkel R (2006). Argentina: Macroeconomic Performance and Crisis. In: FfrenchͲDavis R,
NayyarD,andStiglitzJEcomps.StabilizationPoliciesforGrowthandDevelopment,CapitalMarket
LiberalizationProgram,InitiativeforPolicyDialogue.NewYork,November.
Damill M, Frenkel R and Juvenal L (2003). Las cuentas públicas y la crisis de la Convertibilidad en
Argentina.DesarrolloEconómicoͲRevistadeCienciasSociales,170.BuenosAires.
Damill M, Frenkel R and Maurizio R (2003). Políticas macroeconómicas y vulnerabilidad social. La
Argentinaenlosañosnoventa.SerieFinanciamientodelDesarrollo,135.SantiagodeChile,CEPAL.
DamillM,FrenkelRandMaurizioR(2002).Argentina:Adecadeofcurrencyboard.Ananalysisofgrowth,
employmentandincomedistribution.EmploymentPaper,2002/42.Geneva,InternationalLabour
Office.
Damill M, Frenkel R and Rapetti M (2005). Lesson from the Argentine case of debt accumulation and
default. UNCTAD project on Capacity Building for Debt Sustainability in Developing Countries.
Geneva(July).
Eichengreen B, Wyplosz C and Rose A (1996). Contagious Currency Crises. National Bureau of Economic
ResearchWorkingPaper5681,ScandinavianJournalofEconomics,98.
FeldsteinM(1998).RefocusingtheIMF.ForeignAffairs,77:20Ͳ33.
FischerS(2001).Exchangeratesregimes:isthebipolarviewcorrect?DistinguishedLectureonEconomics
inGovernment.NewOrleans,AmericanEconomicAssociation,January6.
Frenkel R (1983). Mercado financiero, Expectativas Cambiarias y Movimientos de Capital. El Trimestre
Económico,L(4)200.México.
FrenkelR(2003a).GlobalizationandFinancialCrisesinLatinAmerica.CEPALReview,80.SantiagodeChile.
Frenkel R (2003b). From the Boom in Capital Inflows to Financial Traps. Barcelona, Initiative for Policy
Dialogue(IPD),CapitalMarketLiberalizationTaskForce.2–3June.
Haggard S et al., eds. (1994). Macroeconomic Policy and Adjustment in Korea, 1970–1990. Harvard
InstituteforInternationalDevelopment,KoreaDevelopmentInstitute.
Haggard S (1994). From the Heavy Industry Plan to Stabilization: Macroeconomic Policy, 1976–1980. In:
HaggardS,CooperR,CollinsS,KimCandRoST.
HaggardSandCollinsS(1994).ThepoliticalEconomyofAdjustmentinthe1980s.In:HaggardS,CooperR,
CollinsS,KimCandRoST.
Hausman R and Velasco A (2002). Hard Money’s Soft Underbelly: Understanding the Argentine Crisis.
Harvard,BrookingTradeForum.
133
CaseStudies:ArgentinaandtheRepublicofKorea
Availableat:www.ksghome.harvard.edu/~rhausma/paper/btf02_hard_money.pdf.
IMFIEO(2004).ReportontheevaluationoftheroleoftheIMFinArgentina,1991–2001.Washington,DC,
IndependentEvaluationOffice,InternationalMonetaryFund.
IMF(2001).Argentina:CountryReport2001,01/90.Washington,DC.
IMF(2004).Argentina:SecondReviewUnderStandͲByArrangement.IMFCountryReport,04(195),July.
IMF (2005). Public Information Notice (PIN), 05 (83). Washington, DC, June. Available at:
www.imf.org/external.np/sec/pn/2005/pn0583.htm.
KoreaInstituteofEconomicsandTechnology(1997).EconomicReview,29,December.
LevyͲYeyati E and Panizza U (2006). The elusive costs of sovereign defaults. Working Paper, 581. InterͲ
American Development Bank, November. Available at: www.iadb.org/res/pub_desc.cfm?pub_id
=WPͲ581.
MussaM(2002).ArgentinaandTheFund:FromTriumphtoTragedy.WorkingPapers.Washington,DC.,
InstituteforInternationalEconomics.
OECD,ExternalDebtStatisticsSeries2(1995–1998).Electronicdatabase.
OECD,ExternalDebtStatisticsSeries3(1998–2002).Electronicdatabase.
Park YC (1986). Foreign Debt, Balance of Payment, and Growth Prospects: The Case of the Republic of
Korea,1965–88.WorldDevelopment,14:1019–1058.
Park YC (1998) Financial Crisis and Macroeconomic Adjustments in Korea, 1997–98. In: Financial
LiberalizationandOpeninginEastAsia:IssuesandPolicyChallenges.KoreaInstituteofFinance.
Park YC and Song CY (2001). Financial Contagion in the East Asia Crisis –with Special Reference to the
RepublicofKorea.InternationalFinancialContagion,KluwerAcademicPublishers,May.
ParkYCandLeeJW(2002).FinancialCrisisandRecovery:PatternsofAdjustmentinEastAsia,1996–99.
ADBInstituteResearchPaper,45,October.
Park YC (2005). Economic Liberalization and Integration in East Asia: a Post Crisis Paradigm. Oxford
UniversityPress,December.
Park YC (2005). A tale of two crises: Korea’s experience with external debt management 1979Ͳ80 and
1997Ͳ98. UNCTAD project on Capacity Building for Debt Sustainability in Developing Countries.
Geneva(September).
Simpson L (2006). The role of IMF in debt restructuring: Lending into arrears, moral hazard, and
sustainabilityconcerns.G24DiscussionPaperSeries,40.UNCTAD,May.
RadeletSandRadeletJ(1998).OnsetoftheEastAsianFinancialCrisis.PresentedatNBERCurrencyCrisis
ConferenceandtoUSAgencyforInternationalDevelopment.January.
SorosG(1998).TheCrisisofGlobalCapitalism:OpenSocietyEndangered.NewYork,PublicAffairs.
WorldBank(2006).GlobalDevelopmentFinance,electronicdatabase.
134
135
(Percent)
RealGDPannualgrowth
InflationCPIchange
NominalExchangeRate(PesoperUS$)
TermsofTrade(goods)
TermsofTrade(goodsandservices)
(inUSDmillionsunlessotherwisestated)
Totalexports,fob
Totalimports,fob
CurrentAccountBalance
Currentaccount(aspercentageofGDP)
GrossReserves
Publicpubliclyguaranteeddebt(aspercentageofGDP)
Privatenonguaranteeddebt(aspercentageofGDP)
Publicfinances
(PercentofGDP)
Central Government, total revenue and grants
Central Government, total expenditure and net lending
Central Government balance
Outputandtrade
5651
Ͳ3799
1126
2.0
3331.7
8.9
5.7
6.4
176.0
191.2
1977
6401
Ͳ3488
1856
3.2
5161.6
11.5
5.3
Ͳ3.2
175.5
95.2
1978
7810
Ͳ6028
Ͳ513
Ͳ0.8
9589.6
12.6
8.0
7.0
159.5
65.5
1979
13.3
16.7
Ͳ3.4
8021
Ͳ9394
Ͳ4774
Ͳ6.3
6914.6
13.5
8.7
1.5
100.8
39.5
1980
12.6
21.3
Ͳ8.7
9143
Ͳ8431
Ͳ4712
Ͳ6.0
3446.5
13.5
15.6
Ͳ5.4
104.5
139.6
18.5
Ͳ18.3
1981
10.5
19.2
Ͳ8.7
7623
Ͳ4859
Ͳ2353
Ͳ2.9
2675.2
19.8
14.0
Ͳ3.2
164.8
488.8
Ͳ17.0
Ͳ8.9
1982
TableVI.1.Argentina:SelectedEconomicIndicators,1977Ͳ2006
CompendiumonDebtSustainabilityandDevelopment
11.9
18.4
Ͳ6.5
7835
Ͳ4119
Ͳ2436
Ͳ2.8
1332.6
29.3
12.0
4.1
343.8
306.2
Ͳ3.5
6.1
1983
12.2
18.2
Ͳ6.0
8100
Ͳ4118
Ͳ2495
Ͳ2.7
1392.6
29.0
11.2
2.0
626.7
542.4
7.0
Ͳ0.2
1984
15.8
19.7
Ͳ3.9
8396
Ͳ3518
Ͳ952
Ͳ1.1
3441.1
42.3
5.2
Ͳ7.0
672.2
789.6
Ͳ13.9
12.4
1985
16.5
18.3
Ͳ1.8
6852
Ͳ4406
Ͳ2859
Ͳ2.7
2905.2
38.7
4.1
7.1
90.1
56.7
Ͳ14.8
Ͳ3.0
1986
136
Central Government, total revenue and grants
Central Government, total expenditure and net lending
Central Government balance
(Percentunlessotherwisestated)
RealGDPannualgrowth
InflationCPIchange
Nominalexchangerate(PesoperUS$)
TermsofTrade(goods)
TermsofTrade(goodsandservices)
(inUS$millionsunlessotherwisestated)
Totalexports,fob
Totalimports,fob
CurrentAccountBalance
Currentaccount(inpercentofGDP)
GrossReserves
Publicpubliclyguaranteeddebt(aspercentageof
GDP)
Privatenonguaranteeddebt(aspercentageofGDP)
Publicfinances
(PercentofGDP)
Outputandtrade
15.6
19.6
Ͳ4.1
13.5
18.0
Ͳ4.4
37.5
1.4
45.2
1.7
Ͳ2.0
343.0
308.2
6.9
Ͳ8.7
9134
Ͳ4892
Ͳ1572
Ͳ1.2
3569.4
1988
6360
Ͳ5343
Ͳ4235
Ͳ3.9
1834.1
2.5
131.3
127.4
Ͳ14.2
Ͳ10.0
1987
14.1
26.2
Ͳ12.1
67.6
2.3
9573
Ͳ3864
Ͳ1305
Ͳ1.7
1664.4
Ͳ7.0
3079.8
4736.7
1.4
Ͳ7.9
1989
14.3
16.0
Ͳ1.7
33.2
1.3
12354
Ͳ3726
4552
3.2
4803.1
Ͳ1.3
2314.0
1051.8
Ͳ12.2
Ͳ8.0
1990
TableVI.1.(continued)
16.7
18.0
Ͳ1.2
25.1
0.9
11978
Ͳ7559
Ͳ647
Ͳ0.3
6211.1
10.5
171.7
95.6
Ͳ2.0
44.9
1991
CaseStudies:ArgentinaandtheRepublicofKorea
17.1
17.4
Ͳ0.2
18.9
18.0
0.9
19.4
2.7
20.8
1.0
6.3
10.6
0.8
Ͳ1.4
3.0
13268.9
Ͳ15632.5
Ͳ8205.89
Ͳ3.5
14001.4
1993
12398.9
Ͳ13794.8
Ͳ5547.75
Ͳ2.4
10200.2
10.3
24.9
3.9
Ͳ4.5
1.3
1992
18.9
19.4
Ͳ0.5
19.5
5.1
16023.3
Ͳ20162.2
Ͳ10979.5
Ͳ4.3
14550.4
5.8
4.2
0.0
17.1
1.2
1994
18.6
19.6
Ͳ0.9
21.3
6.2
21161.7
Ͳ18804.3
Ͳ5117.96
Ͳ2.0
14515.4
Ͳ2.8
3.4
0.1
Ͳ0.2
Ͳ5.8
1995
17.6
20.1
Ͳ2.5
22.9
7.0
24042.7
Ͳ22283.2
Ͳ6769.98
Ͳ2.5
18324.1
5.5
0.2
0.0
7.7
9.2
1996
137
19.0
20.3
Ͳ1.3
18.5
20.1
Ͳ1.6
9.3
8.0
25.8
22.8
3.9
0.9
0.0
Ͳ6.0
Ͳ4.8
26433.7
Ͳ29530.9
Ͳ14482
Ͳ4.8
24769.9
1998
26430.8
Ͳ28553.5
Ͳ12138.1
Ͳ4.1
22336.8
8.1
0.5
0.0
Ͳ2.0
0.0
1997
21.9
Ͳ2.5
19.4
9.6
28.6
23308.6
Ͳ24103.2
Ͳ11942.8
Ͳ4.2
26268.3
Ͳ3.4
Ͳ1.2
0.0
Ͳ6.1
Ͳ4.4
1999
22.0
Ͳ2.4
19.5
9.1
29.8
26341
Ͳ23889.1
Ͳ8980.62
Ͳ3.2
25147.7
Ͳ0.8
Ͳ0.9
0.0
10.1
9.9
2000
22.6
Ͳ3.7
18.8
11.8
32.0
26542.7
Ͳ19157.8
Ͳ3780.42
Ͳ1.4
14553.4
Ͳ4.4
Ͳ1.1
0.0
Ͳ0.5
Ͳ0.7
2001
33.4
Ͳ15.2
18.2
27.9
88.3
25650.6
Ͳ8473.1
8719.69
8.5
10489.8
Ͳ10.9
25.9
206.5
Ͳ0.4
Ͳ0.6
2002
25.8
Ͳ5.2
20.7
21.9
74.5
29938.8
Ͳ13134.2
8092.6
6.2
14153.9
8.8
13.4
Ͳ5.3
9.8
8.6
2003
27.8
Ͳ4.3
23.4
15.5
65.6
34575.7
Ͳ21311.1
3218.98
2.1
18980.6
9.0
4.4
0.8
0.4
1.8
2004
26.2
Ͳ2.5
23.7
14.4
33.1
40386.8
Ͳ27300.1
5690.52
3.1
27266.9
9.2
9.6
Ͳ0.7
Ͳ3.0
Ͳ2.1
2005
25.9
Ͳ1.7
24.2
10.4
29.9
46456.4
Ͳ32584.8
7998.22
3.7
30996.2
8.0
10.9
5.2
6.5
5.7
2006
Source:UNNationalAccountStatistics;IMFBalanceofPaymentsStatistics,InternationalFinancialStatisticsandWorldEconomicOutlookdatabases;WorldBankGlobalDevelopmentFinance
database.
(Percentunlessotherwisestated)
RealGDPannualgrowth
InflationCPIchange
Nominalexchangerate(PesoperUS$)
TermsofTrade(goods)
TermsofTrade(goodsandservices)
(inUS$millionsunlessotherwisestated)
Totalexports,fob
Totalimports,fob
CurrentAccountBalance
Currentaccount(inpercentofGDP)
GrossReserves
Publicpubliclyguaranteeddebt(as
percentageofGDP)
Privatenonguaranteeddebt(as
percentageofGDP)
Publicfinances
(PercentofGDP)
Central Government, total revenue and
grants
Central Government, total expenditure
and net lending
Central Government balance
Outputandtrade
TableVI.1.(continued)
CompendiumonDebtSustainabilityandDevelopment
CaseStudies:ArgentinaandtheRepublicofKorea
TableVI.2.Consolidatedfiscalbalance(NationalAdministrationandProvinces)
(asapercentageofGDP,annualaverage)
National Adm inistration
Prim ary Surplus
Total
Period
Prim ary
Interest
without Social
Balance
Surplus
paym ents
Security
(2)
(1)
Av erage
1981-90
nd
-4.4
1.9
-6.2
Av erage
1991-94
2.1
1.3
1.2
0.1
Av erage
1995-97
1.7
-0.3
1.7
-2.0
Av erage 1998-01
3.1
0.5
3.1
-2.7
Av erage 1991-01
2.3
0.6
2.0
Consolidated
Public Sector
Balance
(3)
-1.5
-7.0
-0.6
-2.6
-4.1
-2.4
Source:Authors´calculationsbasedonMinistryofEconomy,CetrángoloandJiménez(2003)andGaggero(2003).
(1)Primarybalanceexcludingreceiptsandexpendituresofnationalsecuritysystem.
(3)=(2)+ProvincesandBuenosAiresCitybalances.
TableVI.3.Totalpublicinterestpayments,TaxcollectionͲGDPratioandsovereignrisk
premium
(inpercent)
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Tax
Average
Interest
Sovereing
collection as interest rate payments /
risk premium
percentage
on public tax collection
(annual
of GDP
debt
ratio
average)
(1)
(2)
(3)
18.8
20.8
21.3
21.1
20.9
19.6
21.0
21.4
21.4
21.9
21.0
19.2
23.1
s.d
6.6
5.0
5.5
6.1
5.8
6.7
7.6
8.3
8.9
9.4
5.2
1.9
5.5
8.3
6.0
6.9
9.2
9.7
10.9
12.2
15.9
18.5
23.4
13.3
9.6
9.6
6.9
4.9
5.9
12.4
6.5
3.3
5.8
7.2
11.5
14.8
-.-.-
Source:Authors´calculationsbasedonMinistryofEconomy.
(1)IncludesSecuritySystemreceipts.
(2)CalculatedasaratiobetweeninterestpaymentinperiodtanddebtattheendoftͲ1.
(3)Taxreceiptsincludethosefromsocialsecuritysystem.
138
CompendiumonDebtSustainabilityandDevelopment
TableVI.4.Changeinforeigndebtandforeignassetsbysectorandperiod
(US$million)
Changes in
Net external
External debt of
External assets of
debt
of private
Public
Financial
Private
Financial Private Sector
sector (2)-(3)
Total
Sector (1)
Sector
Sector (2)
Sector
(3)
Period
1991:4
to 1994:4
8,529
5,726
10,321
24,575
1,728
566
9,755
1994:4
to 1995:4
5,924
2,952
4,361
13,238
821
11,174
-6,813
1995:4
to 1998:2
9,222
11,579
15,607
36,407
15,307
15,050
557
1998:2 to 2000:4
8,523
-555
3,139
11,107
-4,274
11,876
-8,737
2000:4 to 2001:4
2,975
-8,053
-688
-5,766
-10,665
12,865
-13,553
Total
35,173
11,649
32,740
79,561
2,917
51,531
-18,791
Source:Authors´estimationsonthebasisofdatafromtheMinistryofEconomy.
(1)IncludingtheCentralBank.
TableVI.5.Fiscaladjustment:ResultsoftheConsolidatedPublicSector(CPS)
(aspercentofGDP)
Concept
2001
National Public Sector
Tax receipts
Taxes on exports
Financial tax (*)
VAT
Income tax
Other taxes (**)
Other receipts
Total receipts
Total expenditures
Primary expenditures
Interest services
Primary result
Total result of the NPS
Provinces (***)
Total result of the CPS
13.8
0.0
1.1
3.1
2.5
7.2
4.9
18.8
22.0
18.2
3.8
0.5
-3.2
-2.4
-5.6
2004
18.7
2.3
1.5
3.4
3.4
8.1
4.8
23.5
20.9
19.6
1.3
3.9
2.6
0.9
3.5
Variation
(2004-2001)
4.9
2.3
0.4
0.4
0.9
0.9
-0.1
4.7
-1.1
1.4
-2.5
3.3
5.9
3.3
9.2
Source:Authors´calculationsbasedonMinistryofEconomy.
(*)Taxonbankdebitsandcredits.
(**)Includestaxessharedwithprovinces,whichareincludedas.
expendituresinPrimaryexpendituresastransferstoprovinces.
(***)IncludingtheCityofBuenosAires.
139
140
87.0
TermsofTrade(1985=100)
93.4
39.5
28.2
12.7
1.6
Realeffectiveexchangerate(1993=100)
TotalexternaldebtaspercentageofGDP
ShortͲtermDebtaspercentageofexternaldebt
DebtServiceRatio(longtermdebtonly)
ForeignExchangeReserves(US$billion)
404.5
13.9
Realexportgrowth
NominalExchangeRate(Won/USD)
Ͳ8.8
CurrentAccountaspercentageofGDP
(percentunlessotherwisestated)
ExternalSectorIndicators
1975
3.0
10.4
28.6
35.5
108.6
484
99.3
51.8
Ͳ1.1
1976
19.8
17.6
Realwagesannualgrowth
InflationCPIchange
28.7
10.2
26.7
15.3
28.7
25.3
28.4
InvestmentaspercentageofGDP
25.0
30.2
19.8
20.7
8.9
Investment
10
1977
SavingaspercentageofGDP
10.6
1976
5.9
1975
RealGDPannualgrowth
(Percent)
MacroeconomicIndicators
4.9
11.3
4.3
26.2
28.1
107.6
484
111.3
26.5
Ͳ2.0
1978
8.5
36.1
18.3
30.0
10.0
6.8
1979
29.4
33.2
113.1
484
106.2
30.2
0.0
1977
18.2
33.1
14.7
30.3
34.4
9.3
1978
5.7
13.7
27.1
32.1
120.1
484
108.9
18.4
Ͳ6.6
6.6
14.0
34.6
42.6
107.6
484
94.4
16.3
Ͳ8.3
1980
92.5
21.4
Ͳ6.4
1981
6.9
14.7
31.5
45.4
108.4
7.0
16.1
33.4
48.7
106.3
681
96.5
2.8
Ͳ3.3
1982
7.9
28.7
7.2
26.3
11.1
7.3
1982
607.4
Ͳ0.5
Ͳ4.0
1979
29.6
21.4
31.8
28.7
25.4
Ͳ3.1
Ͳ10.7
25.0
6.2
1981
Ͳ1.5
1980
TableVI.6.RepublicofKorea:SelectedEconomicIndicators,1975Ͳ1985
CaseStudies:ArgentinaandtheRepublicofKorea
6.9
16.3
30.0
47.8
101.9
731
97.4
11.9
Ͳ1.8
1983
7.2
29.0
3.5
29.5
17.4
10.8
1983
7.6
16.3
36.5
46.2
99.3
776
99.5
19.6
Ͳ1.4
1984
6.4
30.3
2.2
31.8
10.9
8.1
1984
7.7
21.3
22.9
48.4
93.4
806
100.0
3.6
Ͳ0.8
1985
6.6
30.0
2.4
32.2
5.3
6.8
1985
141
37.7
InvestmentaspercentageofGDP
Source:BankofKorea.
20.4
52.0
24.2
33.2
36.6
47.3
32.7
48.2
33.7
73.7
1,401
74.1
28.2
34.4
84.3
1,189
114.1
8.6
Ͳ2.8
116.7
5.5
1999
2.2
5.6
31.0
33.9
12.2
8.5
2000
11.7
1998
0.8
11.2
29.1
ForeignExchangeReserves(US$billion)
45.8
ShortͲtermDebtaspercentageofexternaldebt
28.2
100.5
950
122.2
5.0
Ͳ1.6
1997
7.5
Ͳ9.3
25.0
35.8
23.2
TotalexternaldebtaspercentageofGDP
108.1
804
125.4
3.7
Ͳ4.1
1996
4.4
2.5
36.0
37.9
Ͳ22.9
Ͳ2.3
35.8
9.5
Ͳ6.9
4.7
8.3
1999
1998
1997
DebtServiceRatio(longtermdebtonly)
105.0
771
138.5
Realeffectiveexchangerate(1993=100)
NominalExchangeRate
TermsofTrade(2000=100)
30.3
Realexportgrowth
1995
Ͳ1.7
5.0
6.7
38.9
35.7
8.4
7.0
1996
CurrentAccountaspercentageofGDP
(percentunlessotherwisestated)
ExternalSectorIndicators
Realwagesannualgrowth
4.4
6.4
36.5
SavingaspercentageofGDP
Inflation
13.1
9.2
1995
Investment
RealGDPannualgrowth
(Percent)
MacroeconomicIndicators
96.2
33.7
29
90.7
1,130
100.0
19.9
2.4
2000
4.1
1.5
29.3
31.9
Ͳ0.2
3.8
2001
102.8
32.2
27.1
85.1
1,291
95.5
Ͳ12.7
1.7
121.4
34.8
26.1
89.7
1,252
95.0
8.0
1.0
2002
2.7
8.6
29.1
31.4
6.6
3.1
2002
2001
TableVI.7.RepublicofKorea:SelectedEconomicIndicators,1995Ͳ2004
CompendiumonDebtSustainabilityandDevelopment
155.4
33.9
26.6
92.1
1,192
89.0
19.3
2.0
2003
3.6
5.7
30.0
33.0
4.0
4.6
2003
199.1
33.8
26.1
94.6
1,146
85.3
31.0
4.1
2004
3.6
2.8
30.2
35.0
1.9
2004
CompendiumonDebtSustainabilityandDevelopment
CHAPTERVII
APPROPRIATEINSTITUTIONALSETTINGS
FORPUBLICDEBTMANAGEMENT
JaimeDelgadilloCortez
(WorldBank97)
A. Introduction
Weak institutions dealing with public debt management in transition and emerging economies as
well as external shocks can be major sources of debt distress. While shocks cannot be totally
controlled,theinstitutionalsettingfordebtmanagementcanbestrengthened.Thusvulnerabilityto
debtproblemscanbereducedorbettermanagedwhensolidinstitutionsareinplace.
Publicdebtmanagementcanbedefinedastheprocessofestablishingandexecutingastrategyfor
managingtheGovernment’sdebtportfolioinordertomeetgovernmentfundingrequirements,to
achieveobjectiveswithregardtocostsandrisks,andtomeetotherobjectivesrelatedtodebtsuch
as promoting investment for economic growth and developing the domestic financial market for
governmentsecurities.Effectivedebtmanagementcanalsohelptoensurethatboththeleveland
growthofdebtarefiscallysustainable.98
In emerging and transition economies the main emphasis in debt management is put on the
following:
x
x
x
x
Theproductionofreliabledebtdata;
Marketdevelopment;
Ensuringadequatefinancingfordevelopmentalandsocialneeds;
EnsuringcompliancewithdebtͲserviceobligations;
97
98
TheauthorwasaSeniorEconomicAffairsOfficerofUNCTADwhenhewrotethischapter.
SeeWorldBankandIMF(2002),andDMFAS,EffectiveDebtManagementaRevision,forthcoming.
143
AppropriateInstitutionalSettingsforPublicDebtManagement
x
x
x
x
Controllingcontingentliabilities;
Negotiatingagreementswithcreditors;
Performingcost/riskanalysis;
Designingstrategiestoattainasustainabledebtposition.
An appropriate institutional framework for debt management can contribute to achieving the
objectivesofeffectivedebtmanagement.Institutionalarrangementsshouldfocusonthefollowing:
x
x
x
x
x
x
Governance;
Clarityoftherolesoftheinstitutionsdealingwithdebtmanagement;
Specificationoftheobjectivesforpublicdebtmanagement;
Coordinationofpublicdebtmanagementwithotherpublicpolicies;
TheorganizationalstructureofDMOs;
Transparencyandaccountability.
This paper is divided in two parts. Section II provides an overview of the context of public debt
managementanddescribesthechallengeswhichDMOsandotherareasofpublicdebtmanagement
must meet as part of the broad framework of macroeconomic policies. Section III examines the
different issues related to the institutional framework of debt management including governance,
mandates, accountability and transparency, the separation of executive and operational debt
management,andtheneedforanexecutivedebtmanagementcommittee.Theprincipalfocusisthe
role,organizationandfunctionsofDMOsinlowandmiddleͲincomecountriesfromtheperspective
ofdevelopmentneeds.
B. TheContextofPublicDebtManagement
1. TheChallengesandConstraintsFacingInstitutionsDealingwithDebtManagement
ProͲactive debt management is essential in today’s market conditions. DMOs must face the
challengeofmorecomplexportfoliosofpublicandprivatedebt,theglobalizationofcapitalmarkets,
andthevolatilityofcapitalflows.Furthermore,manyemergingmarketeconomiesobtainsubstantial
financingintheformofequityflows.
External factors such as volatility in the price of export products and exchangeͲ and interestͲ rate
fluctuations, and contagion effects are beyond the DMO’s control. However, in normal
circumstances,DMOscanplayacrucialroleincrisispreventionandresolution.
Inadequate legal arrangements, unclear definition of functions and responsibilities, inappropriate
organizational structures, inadequate staff and insufficient training, and the failure to define
strategicobjectivesandresponsibilitiesarealltoocommonfeaturesofDMOs.Theseweaknessesare
manifest in the absence of strong middle offices which should be equipped to conduct analytical
work required for defining a debt strategy. Unclear debt management objectives and benchmarks
andtheinabilitytoconductDebtSustainabilityAnalysis(DSA)arealsofrequentproblemsofDMOs.
Other deficiencies involve the implementation of strategies and the lack of coordination with
monetaryandfiscalpolicies.
FactorswhicharenotunderthecontrolofDMOsbutwhichnonethelesscanimpedetheirefficient
performance include structural deficiencies in money markets and in the primary and secondary
markets for financial instruments as well as inadequate management of quasiͲfiscal deficits and
144
CompendiumonDebtSustainabilityandDevelopment
internationalreserves.MoreoversomeDMOslack theresourcesormandate(orboth) to transmit
clearmessagestootherlevelsoftheGovernment.
Thisdocumentdoesnotaddressalloftheseissues.Itfocusesonwaysinwhichtheconstraintsand
challenges facing DMOs can be addressed by strengthening the ability of DMOs to manage their
debt portfolios and by strengthening other institutional and legal arrangements for debt
management.
Figure VII.1. illustrates the core activities typically performed by DMOs in lowͲ and middleͲincome
countries. In the former, debt management is concerned principally with captive markets for
domestic debt and concessional financing or grants. Government financing depends heavily on
shortͲterminstrumentswithhighinterestrates.DMOsthusattempttodiversifythedebtportfolio
by developing primary and secondary financial markets and by eventually accessing international
capitalmarkets.
In the case of middleͲincome countries, debt management has a wider role in involving the
managementofthecostsandrisksofamorediversifiedportfolioincludingrecoursetotransactions
inderivatives,providingarangeoffinancialservicestotheGovernment,andeventuallyparticipating
inaintegratedfinancialriskmanagementwithotherpartsoftheGovernment.99
FigureVII.1.
Debt
Management
ExpandedRole
IntegratedCost/RiskManagement
ofLiabilities;
SupplyofFinancial
ServicestoGovernment;
ManagementofFinancialAssets;
TransactionsinDerivativeMarkets;
OperationalRiskManagement.
Debt
Management
BasicRole
AccessInternationalCapital
MarketstoDiversifyDebt
Portfolio;
DevelopDomesticPrimary&
SecondaryMarkets;
IssueShortͲTermGovernment
Securities;TapCaptiveSourcesof
DomesticDebt;
ManageODAandIDAGrants.
MiddleͲIncome
Countries
LowͲIncome
Countries
99
See,forexample,TheChangingRoleofthePublicDebtManager,presentationmadebyPhillipAnderson,WorldBank,in
UNCTAD(2005).
145
AppropriateInstitutionalSettingsforPublicDebtManagement
2. PublicDebtManagementasanIntegralPartoftheFrameworkofMacroeconomic
Policies
Figure VII.2. below indicates the relationship between public debt, external debt and the real
domesticsectorsoftheeconomy.Publicinvestmentprogramsarefinancedwithinternalorexternal
resourcesintheformofloansorgrants.
Debt flows are recorded in the Balance of Payments. DSA relates public debt information to the
balanceofpaymentsdataandmacroeconomicvariablesincludinggrowthobjectivesandeconomic
and social programs. Information regarding debt stocks and flows (including debt service and
disbursements), interest rates, exchange rates, capitalͲaccount movements, etc. is combined to
determinethesizeofthefinancinggap.Thisneedstobefilledwithacombinationofexternaland
domesticfinancingand,ifnecessary,withdebtrestructuringordebtrelief.
Given the importance of domestic debt and the necessity of incorporating it into the overall
managementofgovernmentliabilities,theconceptofTotalPublicDebtisusedinthispaper,which
thusincludesbothexternalanddomesticdebt.
FigureVII.2.
Flow chart for public debt management in developing economies
within macroeconomic framework
EXTERNAL SECTOR
Imports - Exports
B.O.P.
Fin
an
Gr
ow cing
th
pr
oj
Ob
jec ects
tiv
es
s
ipt
n ts
e
m t
m
ay res
e
p
s
te
e
ur
vic in
sb
er l &
Di
t s ipa
b
c
De Prin
t
en
re
ce
PUBLIC DEBT
EXTERNAL
Internal Financing
Requirements
Tax and revenues
Local currency
Local counterpart funds
Public investment programs
REAL SECTOR
G.D.P.
External Financing requirements
Loans and grants
BUDGET
DOMESTIC DEBT
PUBLIC FINANCIAL SECTOR
Sincepublic debt is often the largest liability in the national balance sheets of low and middleͲ
income countries, its management cannot be seen as isolated from the overall macroeconomic
managementofacountry.Publicdebthasmajorlinkageswithmacroeconomicandfinancialstability.
146
CompendiumonDebtSustainabilityandDevelopment
The objectives of fiscal, monetary and public debt policies should be coordinated to achieve and
maintaindebtsustainabilityandfiscaldiscipline.
Moreover, an effective system of overall financial management is essential not only for
macroeconomic management butalsoforprovidingreliablefinancialinformationtopreventfraud
andwaste.
Publicdebtmanagementaspartofsuchasystemcansignificantlycontributetotheattainmentof
its objectives. Relevant information on debt management strategies, DSA, and public debt stocks
andflowsareessentialingredientsofsoundpublicfinances.Thisisparticularlytrueinthecontextof
budgetingandexpendituremanagement.Inthisarea,Governmentswanttoimproveplanningand
budgetformulation;setrealisticandachievablespendingceilings;improvespending prioritization;
monitorcommitmentsanddisbursements;andensureaccurateandtimelyinformationflowsamong
governmentinstitutions.
As is noted in the World Bank’s Public Expenditure Management Handbook, theory and practice
showthatreformofacountry’sinstitutionsoffinancialmanagementͲbothformalandinformal–
can have a decisive influence on budgetary outcomes at three levels. At the first level, the
introduction of institutional reforms in public financial management can improve aggregate fiscal
discipline and planning as well as the traditional control functions of public expenditure through
budgetparameters.Atthesecondlevel,thesereformscanimprovetheplanningfunctionofpublic
expendituremanagementthroughimprovementsinthecapacitytoallocateresourcesinaccordance
withstrategicprioritiesandbaselinedataonpriorexpendituresandrevenuepatterns.Atthethird
level, the reforms contribute to political decisionͲmaking concerning the allocation of scarce
resourcestoselectedpriorities.
Figure VII.3. illustrates the flows of information between a consolidated DMO and its main
stakeholders: the Ministry of Finance, the Central Bank, the National Planning Agency, and its
externalcreditorsincludingInternationalOrganizations.Incomingandoutgoinginformationtoand
from the DMO contribute to a close and coordinated relationship regarding debt management
amongkeyinstitutions.Thisfacilitatestheimplementationofdebtmanagementoperationssuchas
debtservicepayments,andenablestheDMOtoassisttheGovernmentinconductingaDSA.
147
AppropriateInstitutionalSettingsforPublicDebtManagement
FigureVII.3.IntegratedDebtManagementInformationFlows
CONGRESS
Budget
Debt Service Projections
Debt Service
Central Bank
Fiscal Targets
BOP Information
Private Debt
National Planning /
Development Office
Ministry of Finance :
Debt Stocks & Flows
Debt Exchange Rates
Interest Rates
- Budget
Public Investment
Social & Economic Projects
Grants
Disbursement Projections
Marcoeconomic Objectives
DMO
International
Organizations
Creditors
Stocks & Flows
Internal/External Financing
Requirements
Records on Balances
Public/Private Entities
(guarantees)
- Treasury
- Accounting
Information on guarantees
C. TheRoleandOrganizationofaDMO
1. GovernanceIssues:LegalFramework,MandatesandJurisdiction,andAccountability
andTransparency
The legal framework for a DMO should identify the authority that borrows and issues new debt,
invests,andundertakestransactionsontheGovernment’sbehalf.Theorganizationalframeworkfor
debt management should be well specified, and should ensure that mandates and roles are well
articulated. In accordance with IMFͲWorld Bank guidelines, the legal framework should cover the
appropriatedelegationofauthoritytodebtmanagers.
An Act of Parliament usually stipulates that the Minister of Finance may delegate to the DMO
specificpowers,dutiesandfunctionsthatpertaintothemanagementofpublicdebt.Thishelpsthe
DMOtoconcludetransactionsinatimelymanner,totakeadvantageofmarketopportunities,and
toensurethatprojectsarefundedinthefiscalyearinwhichtheywerebudgetedfor.Thesebenefits
also apply to concessional funding inasmuch as loan negotiations may be concluded but projects
maynotbeimplementedontimeduetodecisionͲmakingdelayscausedbyinadequatedelegation.
ThedegreeofdelegationmaydependonthenumberoffinancialtransactionstheDMOundertakes
in a year. Finance Ministers face difficult decisions in determining how much decisionͲmaking
authorityshouldbevestedinothers,whatspecificborrowingandinvestmentpowerstodelegate,
and whom to delegate this power to. Given the importance of these issues, Governments tend to
move cautiously. Debt management objectives should be clearly defined and published. These
objectives should encompass all types of government obligations, including domestic and external
debt and contingent liabilities. The policies regarding risk and cost should also be disclosed. The
responsibilities and roles attached to each institution involved in debt management should be
clearly spelled out, since otherwise lack of coordination can lead to uncertainties and higher
148
CompendiumonDebtSustainabilityandDevelopment
transaction and borrowing costs. Consolidation of the authority for debt management in a single
wellͲstructuredDMOcanenormouslycontributetotransparencyandaccountability.
Legislation on fiscal responsibility can bring the debt management objectives in line with fiscal
targets. This is particularly important when local Governments have a certain degree of
independencetoincurdebts,bothforeignanddomestic,especiallyinfederalcountries.
Delegationimpliesaccountability.Therefore,itishighlydesirablethattheActofParliamentspecifies
thattheMinisterofFinancepreparesanannualorsemiͲannualreportwithParliamentonactivities
related to public debt management, and on plans regarding public debt management in the next
fiscalyear.Thishelpspromotetransparencyandaccountability,andencouragesadomesticdebate
on these issues. Investors and international financial institutions would also be able to acquire a
better knowledge of the Government’s future funding plans and of its developmental priorities.
(MajorfeaturesofthelegalframeworkforDMOsandofitsresponsibilitiesaresummarizedinBoxes
1and2.)
BOXVII.1.DMOs’MainFunctionsandResponsibilities
x
Toimplementthedebtstrategy,debtmanagementpolicies,procedures,benchmarksandguidelines
prescribedintheregulationsdesignedattheappropriatelevelofGovernment;
x
ToissuedebtortocontractdebtonbehalfoftheMinisterofFinance,toparticipateinDSAtogether
withotherareasoftheGovernment,toalerttheauthoritiesconcerningsituationsofunsustainable
debt,andtorecommendtimelyadjustmentswhenneeded;
x
Tomaintainatimelyandreliabledatabaseonpublicdebtandtoconductregulardatavalidation;
x
Tominimizecostsandrisksassociatedwithpublicborrowingandpubliclyissuedguarantees;
x
To order debt service payments to the financial agent of the central Government through the
Treasuryand/ortheBudgetDepartment,fortheloansandbondissuesofthecentralGovernment.
x
Togenerateandprovidereliableandtimelyinformationonpublicdebtpoliciesanddatatoavariety
ofusersandtothepubliconaperiodicbasis;
x
Toprovidegovernmentguaranteesafterriskevaluation.TheCongressonannualbasisthroughthe
budget law normally authorizes guarantee coverage. The DMO should monitor all forms of
contingentliabilities;
x
Tomonitortheloansandbondsissuesofpublicentitiesandenterprises;
x
To monitor debt incurred at the subͲnational level, including the loans and bonds issues of local
Governmentsandentitiescontrolledbythem;
x
Tomonitorgrants,privateexternaldebtandonͲlendingtobothpublicandprivateentities;
x
To ensure that the provisions of international agreements with creditors (Paris Club, London Club,
otherbilateralcreditors,multilateralcreditors,etc.)arecompliedwith.
149
AppropriateInstitutionalSettingsforPublicDebtManagement
Sound practice under the heading of accountability requires regular auditing of the financial
transactions undertaken bythe debtmanagerstoassesstheircompliancewithgenerallyaccepted
accountingpracticesandwiththeGovernment’sportfoliomanagementpolicies.Thisauditingwould
reviewtherisksintheportfolioandcompliancewiththeriskmanagementframework.Itcouldalso
facilitate the establishment of multiͲyear targets for debt. The results of the audits would be
disclosedinthereportstotheMinisterofFinanceandParliament.
BOXVII.2.LegalandRegulatoryProcessesforDebtIssuancebyaDMO
A DMO needs to be able to operate in accordance with rules which ensure that debt issuance is
consistent with specified borrowing limits, and which do nothing to undermine the confidence of
lendersandinvestorsconcerningtheobligationtoserviceandrepaygovernmentdebt.Thedelegation
ofauthorityshouldbeclear,asshouldbeaccountabilityandreportingobligations.ManyGovernments
have in place legislation of this kind. Usually the legislation authorizes the Minister of Finance to
conductallborrowingandrelatedfinancialtransactionsonbehalfoftheGovernmentandestablishesa
maximum amount of new funding and guarantees that can be extended over a specified period
(generally one year). This avoids the need for specific authorizations from Parliament for individual
transactions, which might increase the role of political factors in the decision making and delay the
executionoftransactions.
2. Policies,ProceduresandOperations
Risks of losses from inadequate operational controls should be managed according to sound
business practices, including wellͲarticulated responsibilities for staff, clear monitoring and control
policies,andadequatereportingarrangements.
x
x
x
x
x
x
Debt management activities should be supported by an accurate and comprehensive
managementinformationsystemwithpropersafeguards.
StaffinvolvedindebtmanagementshouldbesubjecttoacodeͲofͲconductandconflictͲofͲ
interestguidelinesregardingthemanagementoftheirpersonalfinancialaffairs.
A framework should be developed to enable debt managers to identify and manage the
tradeͲoffs between expected costs and risks in the Government’s debt portfolio. Portfolio
benchmarksshouldreflectthelevelofriskthatisacceptabletotheDMO.
Aspartofriskassessment,debtmanagersshouldregularlyconductstresstestsofthedebt
portfolioonthebasisofeconomicandfinancialshockstowhichtheGovernment–andthe
countrymoregenerally–arepotentiallyexposed.
In order to help/guide decisions and reduce Government’s risk, debt managers should
considerthefinancialandotherriskcharacteristicsoftheGovernment’scashflows.
Theresponsibilityforidentifyinganddevelopingplanstomanageoperationalrisksalsolies
with the DMO which should have a plan to minimize damages caused by such risks. (For
moredetailonoperationalrisksseeAnnex1.)
150
CompendiumonDebtSustainabilityandDevelopment
BOXVII.3.TypesofRisk
MarketRisk.Theriskassociatedwithchangesinmarketindicators,suchasinterestrates,exchangerates,commodity
prices.ForbothdomesticandforeigncurrencydebtchangesininterestratesaffectdebtͲservicingcostsonnewissues
and on floating rate debt at the rateͲreset dates. The market risk of debt denominated in or indexed to foreign
currencies is due to the vulnerability of debtͲservicing costs as measured in domestic currency to exchange rate
movements.Bondswithembeddedputoptions(i.e.rightsforinvestorstosellthebondstotheissuerataspecified
priceduringacertainperiod)canexacerbatemarketandrolloverrisks.
RolloverRisk.Theriskthatdebtwillhavetoberolledoveratahighcostor,inextremecases,cannotberolledoverat
all.Totheextentthatrolloverriskislimitedtotheriskthatdebtmighthavetoberolledoverathigherinterestratesit
maybeclassifiedasatypeofmarketrisk.However,becauserolloverriskcanleadto,orexacerbate,adebtcrisis,itis
oftentreatedseparately.Managingthisriskisparticularlyimportantforemergingmarketcountries.
LiquidityRisk.Therearetwotypesofliquidityrisk.Oneconcernstherisksofsituationsinwhichaborrowerdoesnot
have access to liquid assets when they are needed. The otherrefers to the risk of penalty rates of interest or other
costswhenaborrowerwantstoexitapositionthroughthesaleofassetsforwhichthemarketisilliquid.Thisriskis
particularlyrelevanttothemanagementofliquidassetsandliabilitiesandtotheuseofderivativescontracts.
CreditRisk.TheriskofnonͲperformancebyaborrowersorbyoneofthecounterpartiestootherfinancialcontracts.
ThisriskarisesinvariouscontextssuchastheacceptanceofbidsinauctionsofsecuritiesissuedbytheGovernment
andinrelationtocontingentliabilitiesandderivativecontracts.
SettlementRisk.Thepotentiallossthatacounterpartycouldsufferasaresultofthepossibilitythatitdoesnotreceive
funds or other assets, for reasons other than default, from another counterparty in accordance with an agreed
timetable.
Operational Risk.Thisincludesarangeofdifferenttypesofrisksduetoinvolvementinvariouskindsofbusiness.It
includesrisksduetotransactionerrorsinthevariousstagesofexecutingandrecordingtransactions,toinadequacies
orfailuresininternalcontrols,systemsandservices,andtotheeffectsofnaturaldisasters.Itmaybedefinedtoinclude
reputationalriskandlegalrisk.
DMOs can implement a “Code of Conduct for Staff and Management” that rests on a tripod of
professionalismandintegrity;honesty,faithfulness,efficiency,staffcourtesyinofficialconduct;and
dignifiedconductinprivatelife.
x
Professionalism and integrity requires staff to openly demonstrate professionalism and
integrityinexecutingthepoliciesandprogramsoftheDMO;
x
Honesty,faithfulness,efficiency,andcourtesyinofficialconductrequiresstafftokeepfaith
to their official responsibilities by not allowing personal considerations or activities to
interfere with official duties, maintaining constancy and sincerity of purpose, being resultͲ
oriented,andrespectingthepeopleitdealswith;
x
Dignifiedconductinprivateliferequiresstafftoexerciserestraintintheirprivatelivesandin
theconductofprivateactivitiesthatcouldhavebearingsontheirofficialengagements.
Annualworkplansshouldbetightlyintegratedwithdebtstrategywork.Thereisastrategyhierarchy
extendingfromoverallstrategicdebtmanagementobjectivestoannualdebtmanagementreviews
andplansconsistentwiththeoverallobjectivesandtooperationalplansforindividualworkareasto
giveeffecttotheannualstrategy.
151
AppropriateInstitutionalSettingsforPublicDebtManagement
3. TheSeparationofExecutiveandOperationalDebtManagement
TheExecutiveDebtManagementfunctions(thatisthepolicy,regulatoryandresourcingfunctions100)
are the responsibility of the Minister of Finance and other high government officials, such as the
Heads of the DMO, National Planning, and Budget and Treasury Offices. These functions may be
subjecttooveralldirectionandcoordinationthroughahighlevelbodywhichcouldbedenominated
astheExecutiveDebtManagementCommittee(EDMC).
Theroleofthe EDMCistoapprovedebt management guidelinesand theprinciplestoimplement
them. It meets at intervals to analyse the DMO’s performance and evaluate compliance with
establishedregulationsandtargets.TheGovernoroftheCentralBankcanbepartofthisCommittee
tohelptoensurecoordinationbetweenmonetarypolicy,debtmanagementandfiscalpolicy.DayͲ
toͲday operations are delegated by the EDMC to the DMO and then reported to and coordinated
withtheMinisterofFinance.
An organizational structure for effective debt management is shown schematically in figure VII.5.,
andtheproposedcompositionoftheEDMCinfigureVII.6.ItsdifferentfunctionsarespecifiedinBox
VII.4.
FigureVII.5.EffectivePublicDebtManagement
ExecutiveDebtManagementCommittee
(HeadedbyMinisterofFinance)
ExecutiveDebtManagement
OperationalDebtManagement
DebtManagementOffice
Front Office
MiddleOffice
BackOffice
The DMO must have a clear mediumͲterm strategy, performance indicators, and strict monitoring
andcontrolfunctions.Thesefunctionsshouldnotberelatedonlytodebtissuanceanddebtservice.
Theyshouldalsoencompasseffectivemanagementoftherisksassociatedwiththedebtstructure
andensuringcompatibilitywiththefiscaltargets,whilereducinggovernmentfinance’svulnerability
toshocks.
Based on previous work on effective debt management developed by DMFAS/UNCTAD and other
international organizations such as the World Bank and the IMF and best practices in debt
management implemented in several countries, the recommended Executive Debt Management
(EDM)functionscanbesummarizedasfollows:
100
SeeDMFAS/UNCTAD(1993).
152
CompendiumonDebtSustainabilityandDevelopment
(a)
EDMfunctionsforexternaldebtincludetheestablishmentofdebtsustainabilitystandards;
determination of borrowing needs and limits, and desired terms and borrowing sources;
formulation of guidelines for debt operations such as debt conversions, buyͲbacks, onͲ
lending,etc;apolicyframeworkforgovernmentguaranteesandcontingentliabilities;and
arrangementsandregulationsforborrowing,disbursements,anddebtservice.
(b)
EDMfunctionsfordomesticdebtconcerntheformulationofdebtmanagementobjectives
and strategy; establishing borrowing ceilings according to budgetary and fiscal goals;
development of a benchmark debt structure; determination of the volume and types of
instrumentstobeusedandtheirmaturity,timing,frequency,andsellingtechniques;and
developmentofcommunicationlinkageswithstakeholders.
(c)
Operational Debt Management is the responsibility of the DMO itself. Basic functions
underthisheadingincluderecording,operating,monitoring,controlling,coordinatingand
negotiating public debt. These functions are best performed within the framework of a
Back,Middle,andFrontOfficetypeoforganization.Separationoffunctionsinthiscontext
helps promote the independence of those designing strategies and monitoring them
(Middle Office) from those registering debt and performing operations (Back Office) and
fromthosecarryingoutnegotiationsanddebttransactions(FrontOffice).
FigureVII.6.PossibleCompositionoftheExecutiveDebtManagementCommittee
Head of
Treasury
Head of
Budget
Head of DMO
Secretary
Minister of
Finance
Chairman
Economic Adviser
to the President
Governor of
Central Bank
Head of Planning
Office
153
AppropriateInstitutionalSettingsforPublicDebtManagement
Consolidationofthedebtmanagementfunctionsinasingleofficecanleadtoefficiencygains.Thisis
crucially important to avoid fragmentation of the debt strategy. When conducting DSA and risk
analysisitisimportanttohaveanintegratedviewofthetotaldebtportfolio.
The terms of reference of the DMO should incorporate all functions related to the contracting of
domesticandexternaldebt.Therefore,theorganizationalstructuremusthaveunitsresponsiblefor
theregistrationandmanagementofbothtypesofliabilities.(FordetailsseeBoxVII.1).
TheorganizationalstructureoftheDMOshouldbebasedonaFunctionsManualthatdeterminesits
role, responsibilities and functions together with a staff table detailing job descriptions and
responsibilities.ThefunctionsofeachelementoftheDMOstructureshouldbeclearlyspecifiedhere.
There should be effective coordination and information sharing within the DMO embodied in an
internalcommunicationsstrategy.
TheDMOshouldhavethepersonnelrequiredforefficientresponsetoitsmandates,andapolicyof
adequateremunerationtoattractandretainqualifiedstaff.
BOXVII.4.FunctionsoftheExecutiveDebtManagementCommittee
x
ApprovethedebtmanagementstrategyoverthemediumͲterm;
x
Decideonsectorsthatwillhaveaccesstoexternalordomesticfinancingandonwhatterms;
x
Definethelevelandcharacteristicsofdomesticdebtissuesforfiscalpurposes;
x
Establishborrowingceilingsbydebtorandcreditorcategories;
x
Establishguidelinesforextendinggovernmentguarantees;
x
Definetherequiredmixofexternalanddomesticindebtednessandthedesiredamortizationprofiles;
x
DecideondebtrestructuringsproposedbytheDMOtoconformwiththedebtstrategy;
x
Providelaws,guidelinesandregulationsforeffectivedebtmanagement;
x
Define institutional framework for the DMO and other institutions involved in debt management
operations,includingpropercoordinationofactivities;
x
Put in place the organizational framework for the DMO, including information flows, functions, and
schedulesofduties;
x
Through the Budget Law, for each fiscal year, establish the debtͲservice targets and ceilings on
indebtednessforforeignanddomesticdebt;
x
Establishbenchmarksforcertaindebtindicators,suchasdebtservicetoexports,stockofpublicdebtto
GDP,debtservicetogovernmentrevenues,etc.;
x
Definepolicies,includingthosecoveringsalaries,careerperspectivesandallowances,forattractingand
retainingDMOstaffwithrelevantqualifications;
x
PutinplacetrainingprogramsforDMOstaff;
x
Supportimprovements,maintenanceandextensionsofthedebtdatabase.
154
CompendiumonDebtSustainabilityandDevelopment
4. TheFront,Middle,andBackOfficesofaDMO
(a)BackOffice
The Back Office centralizes all aspects of the registration, monitoring and control of
disbursements/subscriptions, of the execution and management of public debt service operations,
andoftheproductionofstatisticalinformation.Thefunctionscomprisetheadministrationofthefull
cycleofthelifeofacontract/instrumentfromthesignature/issuetoitsfullpayment.
TheOfficeisresponsibleforthemanagementoftherecordsofdebtholdersandfortheregistration
of government debt instruments. Forecasts of forthcoming debtͲservice payments for domestic as
well as external debt need to be produced and sent to the financial agent of the Government for
compliancewiththedebtͲserviceobligations.
The Office performs the basic functions which permit all other operational functions to be carried
out.Thedistributionoftaskstocomplywiththesefunctionscouldbedividedintothefollowing:(1)
theAreaofRegistrationconcerningtheregistryofdebtinformationinthedatabase;(2)theAreaof
AccountingOperations,whichismainlyconcernedwiththeissueofthePaymentOrders;and(3)the
AreaoftheDatabaseAdministratorinchargeofthesystemandnetworkmaintenanceincludingthe
requiredinformationtechnology.ThesefunctionsshouldberegulatedbyaProceduresManualthat
setsnormsfortheflowofinformationintheoperativecycleandthatlinkstheoperationalactivities
withthestructureandfunctionsoftheDMO.
UnderAreaofRegistrationgrants,onͲlending,guaranteeddebtandcontingentliabilitiesshouldbe
registeredandmonitoredclosely,asshouldprivatenonͲguaranteeddebtanddebtincurredatthe
subͲnationallevel.
TheBackOfficenormallyhastodealwithlargeͲscalerequirementsforinformationonpublicdebt.
International organizations, different areas of the executive and legislative branches, researchers,
and the media require reliable and continuously updated debt information. Transparency and
efficiencyingeneratinginformationisthusakeytaskofthe“AreaofRegistration”.101
AnimportantactivityofthispartoftheBackOfficeistoconductdatavalidationatregularintervals
inordertoensurethereliabilityofthedatabase.Normally,thisalsorequiresregularreconciliations
of data with creditors. The dissemination of information on public debt should be closely
coordinatedwiththeFrontandMiddleOffices.
The preparation of payment orders to service public debt can be performed by the Area of
AccountingOperations,withpaymentschedulesgeneratedbythefunctionalgroupsoftheAreaof
Registration,althoughtheactualaccountingfordebtservicepaymentsisnotnecessarilydoneinside
the DMO. With a reliable debt database and a debt system, the preparation of payment orders
shouldberapidandefficient.Thepaymentorderscanbegeneratedandprinteddirectlyfromthe
databasesystem,basedonthedebtsystem’sinformation.Theprocessofdebtservicepaymentsto
creditorsisperformedafterthereconciliationofthecreditors’requestswiththeamountsscheduled
inthedebtsystem’sdatabase.Thiscentralizationofpublicdebtregistrationandmonitoringandthe
operative process for the issue of the payment orders represent important savings through the
reductionofprocessingtimeandtheeliminationofpenaltiesforlatepayments.
101
FormoredetailsonthistopicseeInformationandTransparencyinDebtManagement,presentationbyUdaibirS.Das,
IMF,inUNCTAD(2005).
155
AppropriateInstitutionalSettingsforPublicDebtManagement
TheAreaofAccountingOperationsoftheBackOfficeshouldalsoensurethatbudgetaryprovisions
exist for (external and domestic) public debt service, including contingent liabilities, and that
sufficientsumsareallocatedtoreserves.
A wellͲorganized Back Office will have a structure for ensuring the efficient flow of information,
adequate business processes, and the quality of information produced. Therefore, the structure
shouldbeorganizedwithadistributionoffunctionsthatclearlydefinesandestablishesthesources
offinancingandthecoͲordinationofthedifferententitiesinvolved.
Under the Area of the Database Administrator the monitoring and control of publicͲdebt
information should be based on a methodical centralization of public credit operations in the
database of the DMO. The debt database should contain upͲtoͲdate information on the domestic
and external debt registered in the system. The registration of operations of domestic or external
financinginthedatabasesystemisinitiatedwiththeopeningofloanfilesclassifiedbythetypeand
use of the financing, creditor, debtor, and executing agency. This registration will facilitate an
adequatecontrolofthemanagementofdisbursementsandpaymentsmadeduringthefiscalyearas
wellastheprojectionoffuturedebtservice.Thecontrolandmonitoringoftheregistrationofnew
loansinthedatabaseshouldbecarriedoutbytheheadofeachfunctionalgroup.Thestatusofthe
database and its evolution should be evaluated in regular meetings among the heads of the
functionalgroupswhohavetheresponsibilityofexecutingtheworkprogram.
The input of loan information to the database should be monitored and controlled periodically by
the head of unit of the Back Office of the DMO. The technicians of the units should have the
responsibilitytorunlistsofloanstoverifytheconsistencyoftheinformationandtocorrectthemif
necessary. The Database Administrator should perform the validation of the consolidated debt
information periodically. Errors and inconsistencies in the information can be detected through
consolidatedreports,andtheheadofeachgroupshouldbenotifiedforcorrections.Thestatusof
the database and its consistency should be analysed in periodic meetings with the heads of units
whohavetheresponsibilityofexecutingtheagreedworkprogram.
Confidence in the debt information processed and reported by the DMO has a direct relationship
withthequalityofdatathatisenteredinthesystem.Inordertoensuretimelinessandhighquality,
theprocessingandreportingofdebtinformationshouldberegulatedbyaresolutionorlegalnorm
thatinstructsalltheentitiesofthepublicsectortorespondtodatarequirementsoftheDMO.
ThecontrolandsupervisionfunctionsofaDMOrequirethatdebtinformationbecollectedwithout
obstacles.ThiswillguaranteethattheauthoritieshaveaccesstoupͲtoͲdatedetailedandaggregated
information. Therefore it is important that the DMO establishes direct contact with the executing
agenciesorusersofresourcesandcreditors.Informationfromthesesourceswillbereconciledwith
thatreceivedfromothersourcesincludingtheCentralBank.
It is also important that the institution in charge of monitoring the public investment program
providesallitsinformationtotheDMO.Thisinformationwillguaranteethattheprojectionsofdebt
servicearecompatiblewith estimates ofdisbursementsforinvestmentprojects.Thiswill enhance
thequalityoftheestimatesprovidedbytheDMOforthepreparationoftheGovernment’sbudget.
(b)TheMiddleOffice
The main function of the Middle Office is to conduct the analytical work required for assisting
executivemanagementlevelsindesigningadebtstrategyandaframeworkforriskmonitoringand
156
CompendiumonDebtSustainabilityandDevelopment
control.Regulardebtportfolioreviewshouldbepartoftheactivitiesofthisofficeincoordination
withothergovernmentoffices.
Otherimportantrolesofthisofficearethegenerationofmanagerialreportsonpublicdebtforusers
insidetheDMOandtheGovernment,andthepublicationandotherdisseminationofstatisticsand
other information related to policies concerning external and domestic debt. The preparation of
debt information and reports should be in accordance with standard requirements and specific
requests. The Procedures Manual should specify the reports that are required by various
governmententitiesandbyexternaluserssuchastheWorldBank,theIMF,regionaldevelopment
banks,ParisClubcreditors,civilsocietyandotherprivateparties.
The work program of the Middle Office should include the preparation of monthly and quarterly
managerial reports for the Ministry of Finance. The reports would comprise the stock of debt
outstandingandtransactionsthattookplaceduringspecificperiods.Thisworkprogramshouldalso
besupportedbythepreparationanddistributionofaStatisticalBulletinofPublicDebt.
The work program should also include estimates mainly used for analytical purposes such as
projectionsofdisbursementsandpublicdebtservicewithvariousassumptionsconcerninginterest
andexchangerates.Theusefulnessofsuchestimateswillbeenhancedbythedevelopmentofthe
capacitytoincorporatedebtdataintotheframeworkofbalanceͲofͲpaymentsandmacroeconomic
data analysis. Such an expanded framework will facilitate the design and implementation of debt
strategies.
ForthisworkitmaybeusefultoestablishanAnalyticalFunctionandaRiskAnalysisFunction.
TheAnalyticalFunctionwillperformportfolioanalysisinamacroeconomicframeworkandanalysis
of longͲterm debt sustainability on a regular basis. DSA needs to include fiscal sustainability and
should also include scenario analysis of ways of meeting mediumͲ and longͲterm social and
economicneeds.ThiscanbeaccomplishedbyusingdifferentDSAanalyticaltools,suchastheWorld
Bank/IMFdebtdynamicstemplates,DebtͲProorDSM+.102
Performance of sensitivity analysis with different assumptions about exchange rates and interest
rates allows the Middle Office to provide information about the impact of different debt service
scenariosonfiscalandmonetaryvariables.Proposeddebtmanagementtargetsregardingcurrency
composition and amortization profiles are also part of the Middle Office’s responsibilities. This
function will provide a basis for the Minister or the EDMC to evaluate the macroeconomic debt
strategy and amend it, if necessary. The Analytical Function would also allow the DMO to adopt
strategieswithinthemandategiventoitandtoproposestrategychangestotheMinisterofFinance.
TheRiskAnalysisFunctionwillberesponsiblefortheevaluationandestablishmentofcostandrisk
limits for the debt portfolio. This can be accomplished by scenario analysis involving different
assumptionsconcerningnotonlyexchangeandinterestratesbutalsoothermajormacroeconomic
variablessuchasglobaleconomicgrowthaswellasgrowthandpricesinthecountry’smajorexport
markets.
102
Information on these analytical tools can be obtained through the websites of Debt Relief International or the
DMFAS/UNCTADProgramme.
157
AppropriateInstitutionalSettingsforPublicDebtManagement
(c)TheFrontOffice
ThefunctionsoftheFrontOfficearemostlyrelatedtothegatheringoffinancialresourcestocover
resourceutilizationandneedsrelatedtothepublicdebtordebt.Theythusincludealltheprocesses
involvingthenegotiationandcontractofnewborrowing.
ThefrontofficeperformsthefunctionsthatcouldbedescribedasthoseofanEDMCSecretariat,i.e.
toensurethatthelaw,therulesandregulations,andtheguidelinesissuedbytheEDMCareapplied
andfollowed.ForthispurposetheFrontOfficerequiresproperlegaladvice.
Two major functions of the Front Office are the Implementation/Monitoring/Negotiating Function
andtheGovernmentSecuritiesMarketFunction.
The Implementation/Monitoring/Negotiating Function is responsible for the following up and
implementationofthedecisionstakenattheexecutivelevelandforensuringthatimplementation
bytheGovernmentisestablishedinaccordance.Inlowerincomecountriesthisfunctionisalsolikely
tocoverattractingOfficialDevelopmentAssistance(ODA)andgrants.
The Government Securities Market Function comprises numerous responsibilities regarding the
development of markets for government debt, and carrying out issuance, redemption, and other
tasksrelatedtomanagementoftheGovernment’sdebt.
Theseincludethefollowing:
x
x
x
x
x
x
The development of securities market regulation to support the issuance and trading of
governmentsecurities;
The development of market infrastructure to help increase market liquidity and reduce
systemicrisk;
Strengtheningthedemandforgovernmentsecuritiesbybuildingthepotentialinvestorbase;
Improving the quality government securities in primary and secondary markets through
extendingmaturitiesandconsolidatingthenumberofissues;
MatchingtheGovernment’sfinancingneedswiththetermstructureofitsdebt;and
Creatingefficientchannelsforthemarketinganddistributionofgovernmentsecurities.
In its Government Securities Market Function the Front Office should aim as far as possible at the
separationofinstrumentsusedfordebtmanagement,ontheonehand,andformonetarypolicy,on
the other. When the market for government securities is limited to shortͲterm instruments, it
conflictsbetweenthepursuitoftheobjectivesofmonetaryanddebtpoliciesaredifficulttoavoid.
InmoresophisticatedDMOstheFrontOfficehasrolesinvolvingderivativestransactions,integrated
riskmanagement,accessingtheinternationalcapitalmarkets,andprovidingvariousotherfinancial
servicestotheGovernment.
5. InstitutionalLocationoftheDMO
AunifiedDMOwithconsolidatedfunctionsregardingoperationaldebtmanagementappearstobe
themostappropriatesettingforeffectivedebtmanagement.103Theexistenceofasingleinstitution
in charge of implementing debt policies permits a greater attention and concentration to debt
management issues and helps to ensure a clear separation between fiscal, monetary and debt
managementpolicies.
103
SeealsoCurrie,Dethier,andTogo(2003).
158
CompendiumonDebtSustainabilityandDevelopment
RegardingthelocationoftheDMOwiththeGovernmentpresentpracticesvary.
Separate DMOs are more frequent in developed economies with sophisticated financial markets.
UnderthisarrangementDMOsimplementthedebtstrategiesdeterminedbytheMinisterofFinance
as an agency of the Government. 104 To ensure Government monitoring and control of debt
management, for example, through an EDMC, clear governance, legal and institutional
arrangements are put in place and strategic objectives and benchmarks for debt management
established.
ThemainadvantagesofseparateDMOscanbesummarizedasfollows:
(a)
(b)
(c)
(d)
Greaterefficiencyinmanagingdebt;
Moreindependencefrompoliticalinfluence;
Thepossibilityofattractingqualifiedstaffatbettersalaries;and
Latitude for the application of privateͲsector management practices and debt
techniques.
AnexampleoftheorganizationalstructureofaseparateDMOisprovidedinBoxVII.5.
DMOsinsidetheMinistryofFinance(MOF)aremorecommoninlessdevelopedeconomies,where
morecoordinationisneededbetweendebtmanagementandotherpoliciesowingtothevitaland
strategicroleoftheformer.
Advantagesofthisarrangementarethefollowing:
(a)
Greater coordination of debt management with the core activities of the MOF,
suchasfiscalandbudgetarypolicies;
(b)
MoreflexibilityinmanagingcontingentliabilitiesandonͲlending;and
(c)
Facilitationofhandlingissuesrelatedtothefiscalsustainabilityofdebt.
Fiscaldiscipline,socialandeconomicgrowth,anddebtsustainabilityareinextricablyintertwinedin
lessdevelopedeconomies.105
IneithercasetheMOFisultimatelyaccountableforincurringdebtonbehalfoftheGovernmentand
delegatessomeofitsauthoritytotheDMOforthispurpose.Whenthedebtmanagementactivities
are consolidated in a single office with an appropriate organizational structure and governance
arrangements, there are no great dissimilarities between a separate DMO and a DMO inside the
MOF.
104
105
RecentExperiencesintheOrganizationofDebtManagementOffices,presentationbyFredJenseninUNCTAD(2005).
SeeBorresenandCosioͲPascal(2002).
159
AppropriateInstitutionalSettingsforPublicDebtManagement
BoxVII.5.TheSeparateDMOofNigeria
ThereareafewdevelopingcountriessuchasNigeriawithseparateDMOs.Thefigurebelowshows
thearrangementsoftheNigerianDMO.
Supervisory Board
(Executive Debt Management)
Public debt committee headed by
Minister of Finance
(Executive Debt Management)
DirectorGeneral (OPERATIONAL
(Operational Debt Management)
Front Office
Middle Office
Back Office
160
Internal Audit
Corporate
Affairs
Department
CompendiumonDebtSustainabilityandDevelopment
References
AsieduE(2003).DebtReliefandInstitutionalReform:afocusonHeavilyIndebtedPoorCountries.
TheQuarterlyReviewofEconomicsandFinance,Kansas,May.
BankforInternationalSettlements(2003).SoundPracticesfortheManagementandSupervisionof
OperationalRisk,Basel,February.
Borresen P and CosioͲPascal E (2002). Role and Organization of a Debt Office. DMFAS Programme
TechnicalPaperNº1(UNCTAD/GDS/DMFAS/2).NewYorkandGeneva,UnitedNations.
Cohen D et al. (2004). Beyond the HIPC Initiative. France, Investment Development Consultancy,
March.
CommonwealthBusinessForum(2003).AchievingSustainableDevelopment:ChallengesforBusiness
andGovernments.Abuja,December.
Contaduria General de la Nacion (2005). The Integrated System of Financial Information. Buenos
Aires,February.
Currie E, Dethier JJ and Togo E (2003). Institutional Arrangements for Public Debt Management.
WorldBankResearchPaper3021,Washington,DC.
DeredzaC(2004).ConceptualisingaSovereignForeignBorrowingPolicyFramework.ForumMEFMI,
Harare,March.
DMFAS/UNCTAD(1989).EffectiveDebtManagement(UNCTAD/RDP/DFP/DMS/2).Geneva.
DMFAS/UNCTAD(1993).EffectiveDebtManagement(UNCTAD/GID/DMS/15)Geneva.
International Monetary Fund (IMF), Bank for International Settlements (BIS), the Commonwealth
Secretariat (Comsec), Eurostat, the Organization for Economic CoͲoperation and
Development (OECD), the Paris Club Secretariat, the United Nations Conference on Trade
andDevelopment(UNCTAD)andtheWorldBank(2003).ExternalDebtStatistics:Guidefor
CompilersandUsers.Washington,DC.
IMF(2003).ManualonFiscalTransparency.Washington,DC.
IMF and the Hong Kong Monetary Authority (2000). Sovereign Assets and Liabilities Management.
Washington,DC,November.
IMF(2004).SovereignDebtStructureforCrisisPrevention.Washington,DC,July.
MagnussonT(2001).TheInstitutionalandLegalBaseforEffectiveDebtManagement.UNCTADThird
InterͲRegionalDebtManagementConference,Geneva.
MehranH,ed.(1985).ExternalDebtManagement.Washington,DC,InternationalMonetaryFund.
MEFMI and World Bank (2001). Public Debt Management, Cash Management and Domestic Debt
MarketDevelopment.TanzaniaWashington,June.
Noel M (2000). Building SubͲnational Debt Markets in Developing and Transition Countries, a
framework for Analysis, Policy Reform and Assistance Strategy. For the World Bank
preparation of the Manual on Domestic Debt Markets Development Ͳ The Policy Issues.
Washington,DC,January.
UNCTAD (2005). Presentations on Debt Management. UNCTAD’s Fifth InterͲRegional Debt
ManagementConference.Geneva,June.
161
AppropriateInstitutionalSettingsforPublicDebtManagement
UNCTAD (2004).EconomicDevelopmentinAfrica–DebtSustainability:OasisorMirage, NewYork
andGeneva,August.
UNCTAD(2003).ProceedingsoftheThirdInterͲRegionalDebtManagementConference,Geneva3–6
December2001.GenevaandNewYork,UnitedNations.
UNDP (1997). A Report on the Joint UNCTADͲWorld Bank Programme of Debt Management,
ManagementDevelopmentandGovernanceDivision.NewYork,DiscussionPaper,4,UNDP.
WheelerG(2004).SoundPracticeinGovernmentDebtManagement.Washington,DC,WorldBank.
WorldBankandIMF(2001).GuidelinesforPublicDebtManagement.Washington,March.
WorldBankandIMF(2001a).DevelopingGovernmentBondMarkets:AHandbook.Washington.
WorldBankandtheIMF(2002).GuidelinesforPublicDebtManagement:AccompanyingDocument.
Washington,November.
World Bank and the IMF (2003). Amendments to the Guidelines for Public Debt Management.
Washington,November.
World Bank and the IMF (2004). Debt Sustainability in LowͲIncome CountriesͲProposal for and
OperationalFrameworkandPolicyImplications.Washington,February.
162
CompendiumonDebtSustainabilityandDevelopment
Abbreviations
ALM BIS BOP CB ComSec DMFAS DMOs DSA DSM+ EDM EDMC GDP GͲ8 GͲ77 IDA IFMS INTOSAI IT/IS IMF MDGs MEFMI MOF ODA OECD SAI UN UNCTAD AssetandLiabilityManagement
BankforInternationalSettlements
BalanceofPayments
Centralbank
CommonwealthSecretariat
DebtManagementandFinancialAnalysisSystem
DebtmanagementOffices
DebtSustainabilityAnalysis
DebtSustainabilityModelPlus
ExecutiveDebtManagement
ExecutiveDebtmanagementCommittee
GrossDomesticProduct
GroupofEight
Groupof77
InternationalDevelopmentAgency
IntegratedFinancialManagementSystems
InternationalOrganizationofSupremeAuditInstitutions
InformationTechnology/InformationSystems
InternationalMonetaryFund
MillenniumDevelopmentGoals
Macroeconomic&FinancialManagementInstituteofEasternandSouthernAfrica
MinistryofFinance
OfficialDevelopmentAssistance
OrganizationforEconomicCooperationandDevelopment
SupremeAuditInstitutions
UnitedNations
UnitedNationsConferenceonTradeandDevelopment
163
AppropriateInstitutionalSettingsforPublicDebtManagement
Annex
ManagingOperationalRisks
Operations risks arise in the areas that provide support services to the management of public debt.
Auditorswouldrecognizethefollowingoperationalriskswhentheyexaminetheorganizationalstructure
ofpublicdebtmanagement.
(a)
Lack of separation of duties or functions. Public debt transactions must be independently
processed,confirmed,valued,andreviewed,andmonitoredbyanindependentadministrative
office.
(b)
Inadequate staff expertise.Supervisors must have the proper expertise to avoidbecoming a
“rubberstamp”tothoseresponsiblefordebttransactions.Supportstaffisusuallythefirstline
ofdefensetouncovererrorsandirregularitiesthatmayoccurinprocessingdebttransactions.
(c)
Productrisk.Newdebtinstrumentscanbetoocomplexorpoorlyunderstood.Thiscanleadto
theinabilityofsupportstafftoprocess,value,andcontrolnewdebtinstruments.
(d)
System and technology risks. These risks exist when staff fails to stay up to date in its
understanding of technological developments associated with new information systems or
adopts computerized information systems without “reͲengineering” their debt management
practices.
(e)
Proceduresrisks.Theserisksexistwhenthedebtmanagementfunctionsdonothavewritten
procedures and the work flow is not structured in a predictable and wellͲ designed manner
withproperaudittrails.Thesewrittenproceduresbecomemoreimportant,themorecomplex
debtinstrumentsare.
(f)
Disaster recovery risks. These risks exist when the debt organization has not planned for
alternativesites,computerresources,communications,resources,tradingfacilities,andother
supportservicesinthecaseofadisaster.Thoseresponsiblefordebttransactionsmusthave
alternativeremotetradingandtechnologysites.
(g)
Documentation risks. These risks exist when debt transactions do not have wellͲdesigned
agreements that are legally authorized, properly executed and supported by appropriate
confirmationinatimelymanner.Legaldepartmentsandsupportstaffmustmaintainmaster
agreementsandsupportingconfirmations.
(h)
Valuationrisks.Theserisksexistwhenthesupportstaffcannotperform,atleastonaregular
basis,anindependentvaluationofalldebtinstrumentsorifthevaluationofthesupportstaff
differs from the valuation of the Supreme Audit Institutions (SAI) or an independent third
party.
Source:INTOSAIGuidanceforPlanningandConductinganAuditofPublicDebt.
164
CompendiumonDebtSustainabilityandDevelopment
CHAPTERVIII
CREDITRATINGAGENCIESANDTHEIRPOTENTIAL
IMPACTONDEVELOPINGCOUNTRIES
MarwanElkhoury
(IndependentConsultant)
A. Introduction
Credit rating agencies (subsequently denoted CRAs) specialize in analysing and evaluating the
creditworthiness of corporate and sovereign issuers of debt securities. In the new financial
architecture CRAs are expected to become more important in the management of both corporate
and sovereign credit risk. Their role has recently received a boost from the revision by the Basel
CommitteeonBankingSupervision(BCBS)ofcapitalstandardsforbanksculminatinginBaselII.
ThelogicunderlyingtheexistenceofCRAsistosolvetheproblemoftheinformationalasymmetry
between lenders and borrowers regarding the creditworthiness of the latter. Issuers with lower
credit ratings pay higher interest rates embodying larger risk premiums than higherͲrated issuers.
Moreover,ratingsdeterminetheeligibilityofdebtandotherfinancialinstrumentsfortheportfolios
ofcertaininstitutionalinvestorsduetonationalregulationsthatrestrictinvestmentinspeculativeͲ
gradebonds.
The rating agencies fall into two categories, recognized and nonͲrecognized. The former are
recognized by supervisors in each country for regulatory purposes. In the United States only five
CRAs(ofwhichthebestknownareMoody’sandStandard&Poor’s)arerecognizedbytheSEC.The
majority of CRAs such as the Economist Intelligence Unit (EIU), Institutional Investor (II), and
Euromoneyare“nonͲrecognized”.
165
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
ThereisawidedisparityamongCRAs.Theymaydifferinsizeandscope(geographicalandsectoral)
of coverage. There are also wide differences in their methodologies and definitions of the default
risk,whichrenderscomparisonbetweenthemdifficult.
RegardingtheirrolevisͲàͲvisdevelopingcountries,thesovereignratingisparticularlyimportant.As
defined by Nagy (1984), “Country risk is the exposure to a loss in crossͲborder lending, caused by
events in a particular country which are Ͳ at least to some extent Ͳ under the control of the
Government but definitely not under the control of a private enterprise or individual”. Under this
definition all forms of crossͲborder lending in a country, ͲͲ whether to the Government, a bank, a
privateenterpriseoranindividualͲͲareincluded.Countryriskisthereforeabroaderconceptthan
sovereignrisk.ThelatterisrestrictedtotheriskoflendingtotheGovernmentofasovereignnation.
However, sovereign and country risk are highly correlated as the Government is the major actor
affectingboth.Moreover,thereonlyrareexceptionstotheprincipleofthesovereignceiling;i.e.the
debtratingofacompanyorbankbasedinacountrycannotexceedthecountry’ssovereignrating.
ThefailureofbigCRAstopredictthe1997Ͳ1998AsiancrisisandtherecentbankruptciesofEnron,
WorldComandParmalathaveraisedquestionsconcerningtheratingprocessandtheaccountability
ofCRAsandhaspromptedlegislatorstoscrutinizeratingagencies.Thisreportgivesanoverviewof
thesovereigncreditratingindustry,analysesitsimpactondevelopingcountriesandassessessome
oftheCRAs’shortcomingsinthecontextofconcernsthathaverecentlybeenraised.
B. CreditRatingAgenciesintheInternationalFinancialSystem
1. AsymmetryofInformationandCRAsas“Opinion”Makers
A credit rating compresses a large variety of information that needs to be known about the
creditworthiness of the issuer of bonds and certain other financial instruments. The CRAs thus
contribute to solving principalͲagent problems by helping lenders “pierce the fog of asymmetric
information that surrounds lending relationships and help borrowers emerge from that same
fog”(White(2001)).
CRAs stress that their ratings constitute opinions. They are not a recommendation to buy, sell or
holdasecurityanddonotaddressthesuitabilityofaninvestmentforaninvestor.Ratingshavean
impactonissuersviavariousregulatoryschemesandby determiningtheconditionsandthecosts
under which they access debt markets. Regulators have outsourced to CRAs much of the
responsibility for assessing debt risk. For investors ratings are a screening tool that influences the
compositionoftheirportfoliosaswellastheirinvestmentdecisions.
2. CreditRatingsandBaselII
Regulatory changes in banks’ capital requirements under Basel II have resulted in a new role for
creditratingagencies.Ratingscanbeusedtoassigntheriskweightsdeterminingminimumcapital
chargesfordifferentcategoriesofborrower.UndertheStandardizedApproachtocreditriskBaselII
establishes credit risk weights for each supervisory category which rely on “external credit
assessments”(seeBoxVIII.1.).Moreover,creditratingsarealsousedforassessingrisksinsomeof
theotherrulesofBaselII.
TheimportanceofratingsͲbasedregulationsisparticularlyvisibleintheUnitedStates,whereitcan
betracedbacktothe1930s.Theseregulationsnotonlyaffectbanksbutalsoinsurers,pensionfunds,
166
CompendiumonDebtSustainabilityandDevelopment
mutual funds and brokerͲdealers by restricting or prohibiting the purchase of bonds with “low”
ratings, i.e. nonͲinvestment grade or speculativeͲgrade ratings.106While ratingsͲbased regulations
arelesscommoninEurope,theyarepartofthenewCapitalRequirementsDirectivethattheEUwill
implementthroughBaselII.
BOXVIII.1.BaselII
ThemajorobjectiveofBaselIIistorevisetherulesofthe1988BaselCapitalAccordinsuchawayasto
alignbanks’regulatorycapitalmorecloselywiththeirrisks,takingaccountofprogressinthemeasurement
andmanagementoftheserisksandtheopportunitieswhichtheseprovideforstrengthenedsupervision.
Under Pillar 1 of Basel 2 regulatory capital requirements for credit risk are calculated according to two
alternative approaches, the Standardized and the Internal RatingsͲBased. Under the Standardized
Approach(SA)themeasurementofcreditriskisbasedonexternalcreditassessmentsprovidedbyexternal
creditassessmentinstitutions(ECAIs)suchascreditratingagenciesorexportcreditagencies.Underthe
internal ratingsͲbased approach (IRBA), subject to supervisory approval as to the satisfaction of certain
conditions,banksusetheirownratingsystemstomeasuresomeorallofthedeterminantsofcreditrisk.
Underthefoundationversion(FIRBA)bankscalculatetheprobabilityofdefault(PD)onthebasisoftheir
ownratingsbutrelyontheirsupervisorsformeasuresoftheotherdeterminantsofcreditrisk.Underthe
advancedversion(AIRBA)banksalsoestimatetheirownmeasuresofallthedeterminantsofcreditrisk,
includinglossgivendefault(LGD)andexposureatdefault(EAD).
Under the regulatory capital requirements for operational risk there are three options of progressively
greater sophistication. Under the Basic Indicator Approach (BIA) the capital charge is a percentage of
banks’grossincome.UndertheStandardizedApproach(SAOR)thecapitalchargeisthesumofspecified
percentagesof banks’gross income from eightbusiness lines (or alternatively for two of thesebusiness
lines,retailandcommercialbanking,ofdifferentpercentagesofloansandadvances).UndertheAdvanced
Measurement Approach (AMA), subject to the satisfaction of more stringent supervisory criteria, banks
estimatetherequiredcapitalwiththeirowninternalsystemsformeasuringoperationalrisk.
Pillars 2 and 3 of Basel 2 are concerned with the supervisory review of capital adequacy and the
achievementofmarketdisciplinethroughdisclosure.
Various writers such as Reisen (2002) have expressed the view that the Basel II Accord may
destabilizeprivatecapitalflowstodevelopingcountries.Thiswouldbetrueifthecloserlinksunder
Basel II between the levels of banks’ regulatory capital and their assessment of credit risks
accentuated proͲcyclical fluctuations in their lending. Moreover the same link may also result in
higher interest rates than under the 1988 Accord for less creditworthy developingͲcountry
borrowers. The ratings of CRAs may contribute to unfavorable effects under both headings. As
discussed below, changes in these ratings sometimes follow closely cyclical changes in economic
conditions.Moreoverowingtotheirlowcreditratingscertaindevelopingcountriesmaybeassigned
higher weights for credit risk than under 1988 Capital Accord and thus be charged higher rates of
interestontheirborrowing.
106
The major CRAs have their own ratings schemes which differ for different categories of debt – longͲand shortͲterm,
bank and nonͲbank Ͳ and in the case of Fitch’s ratings for banks include the likelihood of external support, should this
becomenecessarytoenablethemtocontinuemeetingtheirfinancialobligationsonatimelybasis.Thebestknownratings
arethoseofStandardandPoor’sandMoody’sforlongͲtermdebt,whichvarybetweenAAAandBBBͲforinvestmentgrade
forStandardandPoor’s(AaaͲBaa3forMoody’s)andbetweenBB+andCCforspeculativegradeforStandardandPoor’s
(Ba1ͲCforMoody’s).Formoredetailsseetable1ofAnnex2.
167
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
C. CRAs’ProceduresandMethods
1. QuantitativeandQualitativeMethods
The processes and methods used to establish credit ratings vary widely among CRAs. Traditionally
CRAs have relied on a process based on a quantitative and qualitative assessment reviewed and
finalized by a rating committee. More recently there has been increased reliance on quantitative
statistical models based on publicly available data with the result that the assessment process is
more mechanical and involves less reliance on confidential information. No single model
outperformsalltheothers.Performanceisheavilyinfluencedbycircumstances.
A sovereign rating is aimed at “measuring the risk that a Government may default on its own
obligationsineitherlocalorforeigncurrency.Ittakesintoaccountboththeabilityandwillingnessof
a Government to repay its debt in a timely manner” (Moody’s, Special Comment (2006)). The key
measureincreditriskmodelsisthemeasureoftheprobabilityofdefault,PD,butexposureisalso
determined by the expected timing of default and by the recovery rate, RE, after default has
occurred.
(a)
S&P ratings seek to capture only the forwardͲlooking probability of the occurrence of
default.Theyprovidenoassessmentoftheexpectedtimeofdefaultorofmodeofdefault
resolutionandrecoveryvalues.
(b)
BycontrastMoody’sratingsfocusontheExpectedLoss,EL,whichisafunctionofbothPD
andtheexpectedrecoveryrate,RE.ThusEL=PD.(1ͲRE).
(c)
Fitch’s ratings also focus on both PD and RE (Bhatia, 2002). They have a more explicitly
hybridcharacterinthatanalystsarealsoremindedtobeforwardͲlookingandtobealertto
possiblediscontinuitiesbetweenpasttrackrecordsandfuturetrends.
The credit ratings of S&P and Moody’s are assigned by rating committees and not by individual
analysts. There is a large dose of judgment in the committees’ final ratings CRAs provide little
guidanceastohowtheyassignrelativeweightstoeachfactor,thoughtheydoprovideinformation
on what variables they consider in determining sovereign ratings. Identifying the relationship
betweentheCRAs’criteriaandactualratingsisdifficult,inpartbecausesomeofthecriteriaused
areneitherquantitativenorquantifiablebutqualitative.Theanalyticalvariablesareinterrelatedand
the weights are not fixed either across sovereigns or over time. Even for quantifiable factors,
determiningrelativeweightsisdifficultbecausetheagenciesrelyonalargenumberofcriteriaand
thereisnoformulaforcombiningthescorestodetermineratings.
InassessingsovereignriskCRAshighlightseveralriskparametersofvaryingimportance:economic,
political, fiscal and monetary flexibility and the debt burden (see Box VIII.2.). Economic risk
addresses the ability to repay its obligations on time and is a function of both quantitative and
qualitativefactors.Politicalriskaddressesthesovereign’swillingnesstorepaydebt.Willingnessto
pay is a qualitative issue that distinguishes sovereigns from most other types of issuers. Partly
becausecreditorshaveonlylimitedlegalredress,aGovernmentcan(andsometimesdoes)default
selectively on its obligations, even when it possesses the financial capacity for debt service. In
practice,politicalriskandeconomicriskarerelated.AGovernmentthatisunwillingtorepaydebtis
usually pursuing economic policies that weaken its ability to do so. Willingness to pay, therefore,
encompasses the range of economic and political factors influencing government policy (see Box
VIII.2.).
168
CompendiumonDebtSustainabilityandDevelopment
Broadly speaking, the economic variables aim at measuring three types of performance: (1)
measuresofdomesticeconomicperformance,(2)measuresofacountry’sexternalpositionandits
ability to service its external obligations and (3) the influence of external developments. Bhatia
(2002) notes that CRAs’ analysis prior to the Asian financial crisis focused on traditional
macroeconomic indicators with limited emphasis on contingent liability and international liquidity
considerations.MoreoverprivateͲsectorweaknesseswerenotincludedintheanalysisofsovereign
rating.
Inpractice,asmallnumberofvariablesͲͲGDPpercapita,realGDPgrowthpercapita,theconsumer
priceindex(CPI),theratioofgovernmentfiscalbalancetoGDP,andgovernmentdebttoGDPͲͲhave
a large impact on credit ratings.(The relationship between these indicators and S&P’s ratings are
illustrated in figures 1Ͳ5 of Annex 1.). By and large, higher GDP per capita lead to higher ratings;
higher CPI to lower ratings, the lower the rating, the lower the government balance as a ratio to
GDP;higherfiscaldeficitsandgovernmentdebtinrelationtoGDPalsolowerratings.
169
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
BoxVIII.2.S&PSovereignRatingsMethodologyProfile
Politicalrisk
x
x
x
x
x
x
Stabilityandlegitimacyofpoliticalinstitutions
Popularparticipationinpoliticalprocesses
Orderlinessofleadershipsuccessions
Transparencyineconomicpolicydecisionsandobjectives
Publicsecurity
Geopoliticalrisk
Incomeandeconomicstructure
x
x
x
x
x
x
x
Prosperity,diversityanddegreetowhicheconomyismarketͲoriented
Incomedisparities
Effectivenessoffinancialsectorinintermediatingfunsavailabilityofcredit
CompetitivenessandprofitabilityofnonͲfinancialprivatesector
Efficiencyofpublicsector
ProtectionismandothernonͲmarketinfluences
Laborflexibility
Economicgrowthprospects
x
Sizeandcompositionofsavingsandinvestment
x Rateandpatternofeconomicgrowth
Fiscalflexibility
x
x
x
x
x
Generalgovernmentrevenue,expenditure,andsurplus/deficittrends
RevenueͲraisingflexibilityandefficiency
Expenditureeffectivenessandpressures
Timeliness,coverageandtransparencyinreporting
Pensionobligations
Generalgovernmentburden
x
x
x
x
Generalgovernmentgrossandnet(ofassets)debtasapercentofGDP
Shareofrevenuedevotedtointerest
Currencycompositionandmaturityprofile
Depthandbreadthoflocalcapitalmarkets
Offshoreandcontingentliabilities
x
x
SizeandhealthofNFPEs
Robustnessoffinancialsector
Monetaryflexibility
x
x
x
x
x
Pricebehaviorineconomiccycles
Moneyandcreditexpansion
Compatibilityofexchangerateregimeandmonetarygoals
Institutionalfactorssuchascentralbankindependence
Rangeandefficiencyofmonetarygoals
Externalliquidity
x
x
x
x
Impactoffiscalandmonetarypoliciesonexternalaccounts
Structureofthecurrentaccount
Compositionofcapitalflows
Reserveadequacy
Externaldebtburden
x
x
x
x
Grossandnetexternaldebt,includingdepositsandstructureddebt
Maturityprofile,currencycomposition,andsensitivitytointerestratechanges
Accesstoconcessionallending
Debtserviceburden
Source:S&P,“SovereignCreditRatings:APrimer”,October2006.
Notes:NFPEs:NonͲfinancialpublicsectorenterprises.
170
CompendiumonDebtSustainabilityandDevelopment
2. EmpiricalAssessmentsofCreditRatingDeterminants
Anumberofeconomistshaveestimatedeconometricallythedeterminantsofcreditratingsforboth
matureandemergingmarkets(CantorandPacker(1995,1996),Haqueetal.,(1996,1997),Reisen
andvonMaltzan(1999),JuttnerandMcCarthy(2000),andBhatia,(2002)).Inthesestudiesasmall
numberofvariablesexplain90percentofthevariationintheratings:
x
x
x
x
x
x
GDPpercapita;
GDPgrowth;
Inflation;
TheratioofnonͲgoldforeignexchangereservestoimports;
TheratioofthecurrentaccountbalancetoGDP;
Defaulthistoryandthelevelofeconomicdevelopment.
Indeed, a single variable, GDP per capita, explains about 80 percent of the variation in ratings
(Borenszstein and Panizza (2006)). It is worth noting that the fiscal position, measured by the
average annual central government budget deficit/surplus ratio to GDP, in the three years before
the rating year and the external position measured by the average annual current account
deficit/surplus in relation to GDP, in the three years before the rating year, were found to be
statisticallyinsignificant.
Whileincludingpoliticaleventscanimprovetheexplanatorypoweroftheregressions,theexclusion
ofpoliticalvariablesdoesnotbiastheparameterestimates(Haqueetal.,1996;CantorandPacker,
1996). In addition, for developingͲcountry ratings, two other variables adversely affected ratings
independentlyofdomesticeconomicfundamentals(Haqueetal.,1996,1997):
x
x
Increasesininternationalinterestrates;
Thestructureofitsexportsanditsconcentration.
JüttnerandMcCarthy(2000)foundastructuralbreakinratingsassessmentin1997inthewakeof
theSouthͲEastAsiancrisis.“[…]Econometricestimatesmayconveywrongormeaninglesssignalsto
investorsduringaratingcrisis…thereisnosetmodelorframeworkforjudgmentwhicharecapable
of explaining the variations in assignment of sovereign ratings over time” (Jüttner and McCarthy
(2000)).Theauthorsaddinafootnotethatthismeansthatinaglobalfinancialcrisisratingsmodels
might become completely obsolete since a stable relationship between rating and their
determinantsmightbeimpossibletoidentify.
In their analysis of the determinant of ratings during the Asian crisis, Jüttner and McCarthy found
thatthefollowingvariables:
x TheCPI;
x Theratioofexternaldebttoexports;
x Adummydefaulthistory,and;
x Theinterestratedifferential;
x Therealexchangerate.
Neither the interest rate differential nor the real exchange rate were found to be significant
determinantspriortotheAsiancrisisthusindicatingthatthesevariablesmayhavebeenoverlooked
by the agencies before the crisis. Variables denoting financial strength were not found to be
significant determinants of sovereign ratings even one year after the Asian crisis. However, these
variables were subsequently included in ratings assessments by the major CRAS following their
unsatisfactoryperformanceduringAsiancrisis.
171
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
3. RatingDifferences,Notching,SolicitedandUnsolicitedRatings
Although CRAs have different concepts and measurements of the probability of default, various
studieswhichhavecomparedMoody’sandS&P’ratings,havefoundagreatsimilarityforinvestment
graderatings(CantorandPacker,1996;AmmerandPacker,2000).InthecaseofspeculativeͲgrade
issues,Moody’sandS&Passigndivergentratingsmuchmorefrequentlytosovereignbondsthanto
corporate bonds. The literature also finds clear evidence of differences in rating scales once we
movebeyondthetwolargestagencies.Forexample,ratingsforthesameissuertendtobelowerfor
thetwolargestagenciesthanforotheragenciessuchasFitchorDuffandPhelps.
Some of these differences can be explained by sample selection bias. The analysis of Cantor and
Packer(1996)pointstoonlylimitedevidenceofsignificantselectionbiasandsignificantevidencefor
differences in rating scales between larger and small CRAs. Regardless of ratings differences, the
market appears to reward issuers with a lower interest costs when a third rating is assigned,
especiallywhentheratingishigher(BCBS(2000)).
Fitch and the EganͲJones Ratings Company have accused the two big CRAs of practicing the
“notching”, a practice whereby S&P and Moody’s would initiate an automatic downward of
structuredsecurities,ifthetwoagencieswerenothiredtoratethem(EganͲJonesRatingsCompany,
2002). Moody’s response to Fitch’s accusations is that unsolicited ratings usually result in a lower
rating for debt securities because of either a lack of information or the use of different
methodologiestodeterminetheprobabilityofdefault.
Unsolicitedratingsraisepotentialconflictsofinterest.BothMoody’sandS&Pstatethattheyreserve
therighttorateandmakepublicratingsforUnitedStatesSECͲregisteredcorporatebonds,whether
or not requested by an issuer. If the issuer does not request the rating, the rating will simply be
based on publicly available information. If the issuer requests the rating, then it provides
informationtotheratingagencyandpaysthefees.Manynewentrantsinthecreditratingindustry
issueunsolicitedratingstogaincredibilityinthemarket.SomeissuershaveaccusedCRAsofusing
unsolicitedratingsandthethreatoflowerratingsinduceissuerstocooperateintheratingprocess
andpaythefeesofsolicitedratings.107
Since2001,Moody’sclaimsthatithasnotdoneanyunsolicitedratinginEurope.S&Palsoclaimsnot
to do any unsolicited rating outside the United States. As unsolicited ratings are based on public
information and thus lack issuer input, the issue of unsolicited ratings could be addressed by
requiring CRAs to disclose whether it has been solicited or not. Both Moody’s and S&P already
specify in their ratings whether the rating has been solicited and give issuers the opportunity to
participateatanystageoftheprocessiftheywish.
D. ImpactofRatings
1. CostsandBenefitsofObtainingaRating
As mentioned earlier, the primary purpose of obtaining a rating is to enhance access to private
capital markets and lower debtͲissuance and interest costs. Theoretical work (Ramakrishnan and
Thakor, 1984; Millon and Thakor, 1985) suggests that credit rating agencies, in their role as
information gatherers and processors, can reduce a firm’s capital costs by certifying its value in a
107
SEC,‘Concept’Release.RatingAgenciesandtheUseofCreditRatingsundertheFederalSecuritiesLaws,Securitiesand
ExchangeCommission.ReleaseNos.33Ͳ8236;34Ͳ47972;ICͲ26066.
172
CompendiumonDebtSustainabilityandDevelopment
market,thussolvingorreducingtheinformationalasymmetriesbetweenpurchasersandissuers.For
sovereign borrowers, there is evidence of a clear correlation between bond spreads and rating
grades,asshowninFigureVIII.1.,(BIS(2006)):thelowertherating,thehigherthespread.
FigureVIII.1.BondSpreadsbyRatings
Source:BISQuarterlyReview,March2006fromJPMorgan
ChaseEMBIGlobalDiversified(EMBIGD).
There are other indirect benefits from ratings for low income countries, namely to foster FDI, to
promote more vibrant local capital markets, and increase publicͲsector financial transparency
(StandardsandPoor’s(2004)).Asaresult,evensomesovereignsthatdonotintendtoissuecrossͲ
borderdebtintheimmediatefutureseekcreditratingsfromCRAs.
Foremergingmarkets,thereisanimportantexternalityofobtainingarating,thatofthe“sovereign
ceiling”effect.Borenzsteinetal.(2006)findthat,althoughithasbeenrelaxedsince1997,theeffect
ofthesovereignceilingremainsstatisticallyhighlysignificant,especiallyforbankcorporations,being
moreimportantforbanksthatresideincountrieswithahighlevelsofsovereigndebtandsmaller
forbankswithstrongforeignparents.
2. BoomsandBusts:FinancialCrisesinEmergingMarketsandtheProͲcyclicalityof
Ratings
The 1997Ͳ1998 Asian crisis highlighted CRAs’ potential for reinforcing boomsͲandͲbusts of capital
flows. As ratings were lagging instead of leading market events and overͲreacted during both the
preͲandpostͲcrisisperiods,theymayhavehelpedtoamplifythesecycles.
Severalempiricalstudiesshowthatsovereignratingsaresticky,laggingmarketsentimentandoverͲ
reactingwithalagtoeconomicconditionsandthebusinesscycle.Larrain,ReisenandvonMaltzan
(1997)havefoundthatratingsarecorrelatedwithsovereignbondyieldspreads.Intheaftermathof
the1994Ͳ1995Mexicancrisis,theauthorsfindatwoͲwaycausalitybetweensovereignratingsand
market spreads. Not only do international capital markets react to changes in the ratings, but the
ratingssystematicallyreact,withalag,tomarketconditionsasreflectedinthesovereignbondyield
spreads.Thisstudyalsoindicatesahighlysignificantannouncementeffectwhenemergingmarkets
sovereign bonds are put on review with negative outlook. Moreover, the study finds a significant
173
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
negative effect of rating announcements: following a rating downgrade investors readjust their
portfolios.Positiveratingannouncements,bycontrast,donotseemtohaveasignificanteffecton
bondspreads.
Moody’s more recent (2003) report on proͲcyclicality claims that the relative stability of credit
ratings compared to marketͲbased indicators suggests that ratings were more likely to dampen
ratherthantoamplifythecreditcycle,andthatmostratingchangesreflectedlongͲlastingchanges
in fundamental credit risk rather than temporary cyclical developments. The relationship between
creditratingsandcyclicality–andthustheimpactofchangesintheCRAs’practicesinresponseto
shortcomingsrevealedbythecrisesofthe1990sͲthusremainsanopenempiricalquestion.
3. AccuracyandPerformanceofRatings
CRAs’failuretopredicttheMexicanandAsianfinancialcriseswasdue,amongotherthings,tothe
fact that contingent liability and international liquidity considerations had not been taken into
accountbyCRAs..
ConcerningtheAsiancrisis,Moody’sacknowledgedthatithadbeenconfrontedwithanewsetof
circumstancesrequiringaparadigmshiftinthefollowingareas:
x
x
x
x
x
Greater analytic emphasis on the risks of shortͲterm debt for otherwise creditworthy
countries;
Greateremphasisontheidentityandcreditworthinessofacountry’sshortͲtermborrowers;
Greaterappreciationoftherisksposedbyaweakbankingsystem;and
GreaterattentiontotheidentityandlikelybehaviorofforeignshortͲtermcreditors;
Increased sensitivity to the risk that a financial crisis in one country can lead to contagion
effectsforothercountries.
AbalancehastobefoundinthetradeͲoffbetweenaccuracyandstability.Ratingagenciesareaverse
toreversingratingswithinashortperiodoftime.BothMoody’sandS&Pintendtheirratingstobe
stable measures of relative credit risk. Moody’s claims that this corresponds to issuers’ as well as
institutionalinvestors’wishesandthatits“desireforstableratingsreflectstheviewthatmorestable
ratingsare“better”ratings.
Bhatia (2002) has measured “failures” based on ratings stability. With exceptions for some of the
lowestratingshedefinesa“failedrating”asonethatisloweredorraisedby“threeormorenotches
within 12 months. The choice of three notches is related to the small probability of a three notch
ratingchangeamongCRAs.ApplyingtheBhatiadefinitionofratingfailuretothelongͲtermforeign
currency sovereign ratings of S&P and Moody’s in 1997Ͳ2002, shows that S&P and Moody’s both
experienced failures during the Asian crisis; S&P also failed during the Russian and Argentinean
crisis; and Moody’s failed during the Russian but not the Argentinean crisis (see table VIII.1.).
Bhatia’sfailuredefinitionsuggeststhatratingfailureswerelessprevalentin1999Ͳ2002thanin1997Ͳ
1998.
174
CompendiumonDebtSustainabilityandDevelopment
TableVIII.1.SovereignRatingsFailureStatistics,1997Ͳ20021/
Source:Bhatia,2002,Box5.
In response to criticism concerning such failures, Moody’s has introduced “Watchlist” and S&P’s
launched “Outlook” reports in order to alleviate the tension between accuracy and stability by
providingtimelywarningsoflikelyratingchanges.
Ratings performance can also be compared with market indicators. IMF (1999) conducted an
analysisofyieldspreadsinrelationtotheAsiancrisisandfoundthatoneyearaheadofthecrisisin
Thailand,Indonesiaand the Republic ofKorea,sovereignspreadswere quitelowͲof theorderof
100Ͳ150basispoints.InRussiaandBraziltheywerehigherͲabout300basispoints.Thus,inrelative
terms the markets were in broad agreement with the CRAs with respect to these countries,
indicating a higher risk of default for Russia and Brazil than for the Asian countries. Moreover,
spreads did not widen much initially in response to the onset of the Asian crisis, a pattern
conformingtothatoftheratings.Thustheperformanceoffinancialmarketsbroadlyparalleledthat
ofthemajorCRAs.
4. ImpactofRatingsonPoliciesPursuedbyBorrowingCountries
Forborrowingcountriesarating downgradehasnegativeeffectsontheiraccesstocredit andthe
costoftheirborrowing(CantorandPacker,(1996)).Althoughpreciseinformationisnotavailableon
the way in which macroeconomic policies are taken into consideration by CRAs in establishing
sovereign ratings, it is reasonable to assume that orthodox policies focusing on the reduction of
inflation and government budget deficits are favored. There is a risk, therefore, that in order to
avoidratingdowngradesborrowingcountriesadoptpoliciesthataddresstheshortͲtermconcernsof
175
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
portfolioinvestorsevenwhentheyareinconflictwithlongͲtermdevelopmentneeds.However,this
isanissuewhichhasnotbeenthesubjectofsystematicresearch.
E. PublicPolicyConcerns
1. RecentRegulatoryInitiatives
InviewofthecriticalroleplayedbyCRAsinthemodernfinancialarchitecture,policyͲmakershave
recentlyfocusedonsomeshortͲcomingsarisingfromthefollowingconcerns:
x
x
x
x
Barrierstoentryandlackofcompetition;
Conflictsofinterest;
Transparency;
Accountability.
These concerns have been raised by the International Organization of Securities Commission,
(IOSCO), the United States Securities and Exchange Commission, (SEC), the European Commission
Committee of European Securities Regulations, (CESR), and by the United States Congress and
Senate.
OnthebasisofSection702oftheSarbanesͲOxleyActof2002theUnitedStatesCongressmandated
theSECtoissueaReportontheRoleandFunctionofCreditRatingAgenciesintheoperationofthe
SecuritiesMarkets.Thiswastoaddressseveralissuespertainingtothecurrentroleandfunctioning
of CRAs including the information flow in the creditͲrating process, barriers to entry artificially
created by the Nationally Recognized Statistically Rating Organizations (NRSRO) designation in the
UnitedStates,andconflictsofinterestorabusivepractices.
AreviewoftheconceptofNRSROwasalreadyunderwayattheSEC.InJune2003,theSECissueda
Concept Release seeking comments with respect to whether CRAs’ ratings should continue to be
usedforregulatorypurposes,andifso,whethertheNRSROcertificationprocedurewasappropriate
as well as more generally what should be the adequate level of regulatory oversight for CRAs. In
April2005,theSECreleasedaProposedRuleaimingatinsuringahigherleveloftransparencywith
respecttotheNRSROconcept.
The technical committee of the IOSCO issued three reports in September 2003: (i) Report on the
Activities of Credit Rating Agencies; (ii) Statement of Principles Regarding the Activities of Credit
Rating Agencies; (iii) and Report on Analyst Conflicts of Interest. These reports highlighted the
important role CRAs play in financial markets, and aimed at ensuring greater reliability for their
ratings.InDecember2004,theIOSCOpublisheditsCodeofConductFundamentalsforCreditRating
Agencies (the IOSCO Code) which aimed at developing “governance rules” for CRAs to ensure (i)
quality and integrity of the rating process, (ii) independence of the process and avoidance of
conflicts of interest and (iii) greater transparency in the methodology of ratings and adequate
treatment of confidential information. However, the IOSCO Code did not address the issue of
enforcementoftheCode,recommendingthatCRAsadopttheserulesvoluntarily.
InresponsetoIOSCO’sCodeofProfessionalConduct,Moody’sandS&PpublishedtheirownCodeof
Professional conduct in the second half of 2005, thus aligning their policies and procedures with
IOSCO’s Code. In the spring of 2006, Moody’s and S&P published their first report on the
implementationoftheCodeofconduct.Hereitwasstatedthat,evenbeforetheSECandIOSCOhad
recommended new rules of conduct in 2003, the two agencies had already established internal
176
CompendiumonDebtSustainabilityandDevelopment
codes of conduct and procedures to prevent and manage potential conflict of interest and to
safeguardtheindependenceandobjectivityoftheirratingprocesses.
ConsiderationoftheissuesrelatedtoCRAsbytheUnitedStatesCongresseventuallyculminatedin
the Credit Rating Agency Reform Act which was signed into law in early September 2006. This
amended the Securities Exchange Act of 1934 to redefine an NRSRO as any CRA that has been in
business for at least three consecutive years and is registered under the Act. It also prescribed
procedural requirements for mandatory NRSRO registration and certification. It granted the SEC
exclusiveenforcementauthorityoveranyNRSROandauthorizedtheSEC(i)totakeactionagainstan
NRSRO that issued credit ratings in contravention of procedures, criteria and methodologies
included in its registration application, and (ii) to censure, or limit, suspend or revoke the
registrationofanNRSROforviolationsoftheAct.
IntheEU,theEnronandParmalatcollapsesprompteddiscussionsonCRAreliability.Inresponsetoa
call by Commission for advice the CESR released in March 2005 “CESR’s Technical Advice to the
EuropeanCommissiononpossibleMeasuresConcerningCreditRatingAgencies”.
2. IssuesofConcern
(a)Barrierstoentryandlackofcompetition
IntheUnitedStatesthereareonly5CRAsdesignatedbytheSECasNRSROs:A.M.Best.,Dominion
BondRatingService(DBRS),Fitch,Moody’sInvestorsService(Moody’s)andtheStandard&Poor’s
(S&P)divisionofMcGrawHill.DBRSisCanadianͲbasedwitharegionalscopeandtheonlynonͲU.S.
NRSRO designated agency. A.M.Best is a global agency which rates the debt only of insurance
companies. Thus there are three global NRSROs providing a comprehensive service in the United
States,ofwhichtwoagencies,Moody’sandS&P,controlover80percentofthemarket.Themean
numberofCRAsrecognizedamongtheBCBS’membercountriesisaroundsixandtherearebetween
130Ͳ150creditratingagenciesintheworld.However,onlyasmallnumberofCRAsarerecognized
internationallyandthenumberhasnotchangedmuchsincethe1970s(BCBS,2000).
AccordingtotheUnitedStatesDepartmentofJustice,theNRSROdesignationhasactedasabarrier
to entry in a catchͲ22 manner108. A new rating agency cannot obtain national recognition without
NRSROstatusanditcannotobtainNRSROstatuswithoutnationalrecognition.Inthewordsofthe
RapidRatingstestimonybeforetheCommitteeonFinancialServices(H.R.2990(2005b,p.8)),“the
effect of this catchͲ22 has been to preserve a duopoly that has thwarted competition and
innovation”.
In an effort to increase competition and improve the quality of credit ratings Representative
FitzpatrickintroducedH.R.2990,TheCreditRatingAgencyDuopolyReliefActof2005.Hebelieved
thattheSECNRSROdesignationconstitutedan“insurmountableandartificialbarriertoentry…[…]
Lack of competition in the industry has led to inflated prices, stifled innovation, lower quality of
ratings,anduncheckedconflictsofinterestandanticompetitivepractices.”(H.R.2990(2005a),p.4Ͳ
5)).ThisbillwasthebasisoftheCreditRatingAgencyReformActof2006(H.R.2990(2005b)).
Inits2005reporttotheEUCommissionmentionedabovetheCESRalsostatedthatnewCRAsfacea
numberofbarrierstoentryandexistingCRAsfaceanumberofnaturalbarrierstoexpansion.Issuers
usually only desire ratings from those CRAs that are respected by investors and which tend to be
only those with a long performance record (CESR (2005), paras. 247Ͳ248). The CESR report
108
http://www.sec.gov/rules/concepts/s71203/rapid110603.htm#P69_8177#P69_8177.
177
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
concludedthat“theimpactofregulatoryrequirementsoncompetitionisnotclearandthereforeit
cannot conclude that any regulatory requirements would either increase or decrease the entry
barrierstotheratingindustry.ThusCESRdoesnotrecommendtheuseofregulatoryrequirements
asameasuretoreduceorremoveentrybarrierstothemarketforcreditratings”(CESR(2005),para.
252).TheCESRrecommendeda“waitͲandͲsee”attitudeandimplementationofIOSCO’s“Code”.
In a response to such initiatives Moody’s stated that it “has supported eliminating regulatory
barrierstoentry”.But,withregardtocompetitionissues,Moody’sarguesthatthe“costlynatureof
executive time” would not allow issuers to have many different ratings. Because of network
externalities, only a small number of CRAs would be favored by investors, who would desire
“consistencyandcomparabilityincreditopinions”.NewlyestablishedCRAswouldneedtimetogain
credibilityinthemarket.
S&Palsorecommendeditssupportto“amoreopenandtransparentprocesstodesignateNRSROs,
reduce barriers to entry and ensure that the markets remain the ultimate judge of the rating
process”(StandardsandPoor’s(2003)).However,S&PdidnotbelievethatthewholeNRSROprocess
shouldbewithdrawn.
(b)Potentialconflictsofinterest
InitsSeptember2003“ReportofAnalystConflictsofInterest”,IOSCOhighlightedpotentialconflicts
ofinterestfacingtheindustrythatcaninterferewiththeindependenceandobjectivityofitsanalysis.
Conflictsofinterestmayarisewhenaratingagencyoffersconsultingorotheradvisoryservicesto
issuersitratessinceissuerscouldbeundulypressuredtopurchaseadvisoryservicesinreturnforan
improved rating. The report also drew attention to the issue of “notching” by CRAs, i.e. lowering
ratings for issues which they had not rated, and that of “solicited” versus “unsolicited” ratings,
whereaggressivetacticsmightbeusedtoinducepaymentsforaratinganissuerdidnotrequest.
TheIOSCOCodeaddressesthefirstoftheseissueswiththefollowingrecommendation:“Thecredit
ratingaCRAassignstoanissuerorsecurityshouldnotbeaffectedbytheexistenceoforpotential
forabusinessrelationshipbetweentheCRA(oritsaffiliates)andtheissuer(oritsaffiliates)orany
otherparty,orthenonͲexistenceofsucharelationship”(IOSCOCode(2004),Section2,para.2.2).
This principle has been integrated into Moody’s and S&P own Codes of Professional Conduct
(StandardsandPoor’s(2003)).”
(c)Transparency
ManymarketparticipantshaveexpressedconcernoverthelackoftransparencyoverCRAs’ratings
methodologies, procedures, practices and processes. In this context the IOSCO Code stresses the
following : “In order to promote transparency and improve the ability of market participants and
regulators to judge whether a CRA has satisfactorily implemented the Code Fundamentals, CRAs
shoulddisclosehoweachprovisionoftheCodeFundamentalsisaddressedintheCRA’sowncodeof
conduct. CRAs should explain if and how their own codes of conduct deviate from the Code
Fundamentals and how such deviations nonetheless achieve the objectives laid out in the Code
FundamentalsandtheIOSCOCRAPrinciples.Thiswillpermitmarketparticipantsandregulatorsto
draw their own conclusions about whether the CRA has implemented the Code Fundamentals to
theirsatisfaction,andtoreactaccordingly”(IOSCOCode(2004),p.2).
IOSCOrequirestheCRAs’methodologiestobecomepublictoenhancetransparencyinanindustry
whichisveryopaqueinnature.CESRgoesfurtherandproposes,asanalternativetoselfͲregulation,
“theneedtointroducesomespecificrulesonfairrepresentationwhichwouldestablishaminimum
178
CompendiumonDebtSustainabilityandDevelopment
levelofdisclosureonthoseelementsandassumptionswhichmakeclearformarketoperatorsand
investors to understand how a specific rating was determined by a credit rating agency” (CESR
(2005),para.117).
Thenatureandextentofinformationmadeavailabletothepublicstillvariesfromagencytoagency.
SincethepublicationoftheIOSCOCodeanditsintegrationintotheCRAs’ownCodeofConduct,the
CRAshaveincreasedthenumberoflengthyresearchreportsandpublicationsontheirwebsitesand
published some of the criteria used to assess credit risk in their bid to improve transparency.
However, the view is still widespread that CRAs’ methodologies, the variables and weights which
theyemploy,andthecriteriausedinthedeliberationsofratingscommitteesremainopaquetoboth
investorsandborrowers.TheCESRsummedupthecontinuingproblemwhenitstatedthat:“Credit
ratingagenciesshouldaimfortransparencyasthebestwayforwardtoenableinvestorsandissuers
tounderstandthequalityandobjectivityofthecreditrating.Creditratingagenciesshouldtherefore
implementmeasure2.7oftheIOSCOCode”.
(d)Accountability
Thereisnomechanismtoprotectinvestorsand/orborrowersfrommistakesmadebyCRAsorany
abuse of power on their part. This is true even if reputational interests and competition provide
incentives for generating quality financial information. In order to promote transparency and
improvetheabilityofmarketparticipantsandregulatorstojudgewhetheraCRAhassatisfactorily
implementedwhatitpledgesitisdoing,theIOSCOCoderecommendsonlythatCRAsgivefulleffect
totheCodebypublishingtheirown,adheringtoitandjustifyingpubliclyanydeviationbetweenthis
codeandtheiractivities.
There remains the need for more formal regulation to address market failures in the form of
imperfectcompetitionandprincipalͲagentproblemsinthecreditratingindustry.TheCESRtechnical
reportclearlyputsitsfingerontheissueinvolved:“Thereasonforhavingaregulatorymechanism
shouldratherbethatthereexistssomemarketfailurethathastobedealtwith.Inessenceallthe
issuesdiscussedinthepreviouschapterarisebecausetheexistenceofconflictsofinterestsbetween
the CRAs and the issuers and/or the users of ratings (the investors). These types of conflicts of
interests between professional players on the financial markets are natural and exist in numerous
areasofthemarkets.Theybecomeespeciallyapparentintheratingmarketbecauseofthelackof
balanceofpowerbetweenthedifferentplayers.IssuersarerelativelyweakcomparedtotheCRAs
becauseoftheirdependenceontheratingstheyget.Investorshavenothistoricallyinvestedlarge
resources in improving rating agencies behavior, perhaps because there was insufficient
transparencyontheway CRAsoperated tofacilitatethis.ThismeantthatCRAshistoricallyhave a
verystrongposition.WhattheIOSCOCodeistryingtodoistorebalancetheinterestsbetweenthe
differentplayers”(CESR(2005),para.260).
Rousseau(2005)–notinreferencesͲsumsupconcernsovertheresulting“accountabilitygap”as
follows:“ThisaccountabilitygapisworrisomeforCRAsaswellasmarketparticipants.Fortheformer,
theaccountabilitygapmayaffecttheircredibilityinthemarketplace.Forthelatter,itisofparticular
concern given the role that CRAs play in capital markets...There is a need for a […] mechanism to
takeoverifreputationfails.”
ForthefirsttimeinthehistoryofratingsintheUnitedStatestheCreditRatingAgencyReformActof
2006 has clearly designated the SEC to monitor CRAs’ compliance with new securities laws and
regulations. The SEC will be able to act as deemed necessary and to study and report to
congressional committees any problems faced in the future with anything relating to the credit
ratingindustry.
179
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
F. Conclusions
CRAsplayakeyroleinfinancialmarketsbyhelpingtoreducetheinformationalasymmetrybetween
lenders and investors, on one side, and issuers on the other side, about the creditworthiness of
companies (corporate risk) or countries (sovereign risk). CRAs’ role has expanded with financial
globalization and has received an additional boost from Basel II which incorporates the ratings of
CRAsintotherulesforsettingweightsforcreditrisk.
Inmakingtheirratings,CRAsanalysepublicandnonͲpublicfinancialandaccountingdataaswellas
information about economic and political factors that may affect the ability and willingness of a
Governmentorfirmstomeettheirobligationsinatimelymanner.However,CRAslacktransparency
anddonotprovideclearinformationabouttheirmethodologies.
Ratingstendtobesticky,laggingmarkets,andthentooverͲreactwhentheydochange.ThisoverͲ
reactionmayhaveaggravatedfinancialcrisesintherecentpast,contributingtofinancialinstability
andcrossͲcountrycontagion.Moreovertheactionsofcountrieswhichstrivetomaintaintheirrating
gradesthroughtightmacroeconomicpoliciesmaybecounterͲproductiveforlongͲterminvestment
andgrowth.
The recent bankruptcies of Enron, WorldCom, and Parmalat have prompted legislative scrutiny of
theagencies.Criticismhasbeenespeciallydirectedtowardsthehighdegreeofconcentrationofthe
industry,whichintheUnitedStateshasreflectedaregistrationandcertificationprocessintheform
ofNRSROdesignation biasedagainstnewentrants. The effectofsuch concentration hasbeen the
absenceofthedisciplineenforcedbycompetitionandalowlevelofinnovation.
IntheUnitedStatespolicyactionhasincludedthe2006CreditRatingAgencyReformActwhichhas
overhauled the regulatory framework by prescribing procedural requirements for NRSRO
registrationandcertificationandbystrengtheningthepowersoftheSEC.
At the international level the main initiative has been the publication by IOSCO of its Code of
Conduct.ThisCodeaimsatdevelopinggovernancerulesforCRAstoensurethequalityandintegrity
oftheratingprocess,theindependenceoftheprocessandtheavoidanceofconflictsofinterest,and
greater transparency. In its 2005 Technical Advice to the European Commission on possible
Measures Concerning Credit Rating Agencies the CESR recommended the implementation of the
IOSCOCodeandadoptionofa“waitͲandͲsee”attitude.
Definitiveassessmentoftheseinitiativeswouldstillbepremature.Theindustrywillreceiveafillip
from implementation of Basel II. The major CRAs will undoubtedly seek a substantial share of the
new business which will result. The promotion of competition may require policy action at the
nationalleveltoencouragetheestablishmentofnewagenciesandtochannelbusinessgeneratedby
newregulatoryrequirementsintheirdirection.Regulatoryactionatthenationallevelmayalsobe
necessary to ensure that the agencies operate in accord with levels of accountability and
transparencymatchingtherecommendationsoftheIOSCOCode.
180
CompendiumonDebtSustainabilityandDevelopment
References
Ammer J and Packer F (2000). How Consistent are Credit Ratings? A Geographic and Sectoral
Analysis of Default risk. Board of Governors of the Federal Reserve System, International
FinanceDiscussionPaper,668,June.
BaselCommitteeonBankingSupervision(BCBS)(2000).CreditRatingsandComplementarySources
ofCreditQualityInformation.BaselCommitteeonBankingSupervisionWorkingPapers, 3,
Basel,August.
BankforInternationalSettlements(BIS)(2006).QuarterlyReview,June.
BeersDandCavanaughM(2005).SovereignCreditRatings:APrimer.NewYork,Standard&Poor’s,
RatingsDirect,September.
BeersDandChambersJ(2004).SovereignDefaultsSettoFallAgainin2005.StandardandPoor’s,
RatingsDirect,September.
Beers D (2004). Credit FAQ: The Future of Sovereign Credit Ratings. London, Standard and Poor’s,
March.
BergAandPattilloC(1999).AreCurrencyCrisesPredictable?IMFStaffPapers,Vol.46,June.
BhatiaAV(2002).SovereignCreditRatingsMethodology:anEvaluation.IMFWorkingPaper02/170,
InternationalMonetaryFund.
Blum J and Hellwig M (1995). The Macroeconomic Implications of Capital Adequacy Requirements
forBanks.EuropeanEconomicReview,39(3).
BorioCEV,FurfineCandLoweP(2001).ProcyclicalityoftheFinancialSystemandFinancialStability:
IssuesandPolicyoptions.BISPapers,1,March.
Borensztein E, Cowan K and Valenzuela P (2006). The “Sovereign Ceiling Lite” and Bank Credit
Ratings in Emerging Markets Economies. Mimeo, Washington, DC, InterͲAmerican
DevelopmentBank,SantiagodeChile,BancoCentraldeChile.
BorenszteinEandPanizzaU(2006).TheCostofSovereignDefault.Workingpaper.WashingtonDC,
InterͲAmericanDevelopmentBank.
BrandLandBaharR(1999).RatingsPerformance1998.Standard&Poor’sCorporation.
CantorRandPackerF(1996).Determinantsandimpactofsovereigncreditratings.EconomicPolicy
Review,FederalReserveBankofNewYork,vol.2,October.
CantorRandPackerF(1996).MultipleRatingsandCreditStandards:DifferencesfoOpinioninthe
CreditRatingIndustry.StaffReports12,FederalReserveBankofNewYork.
Cantor R and Packer F (1995). Sovereign Credit Ratings. Current Issues in Economic and Finance,
FederalReserveBankofNewYork,Vol1(3),June.
CantorRandPackerF(1994).TheCreditRatingsIndustry.QuarterlyReview,FederalReserveBankof
NewYork,Summer/Fall.
CESR(2005).CESR’sTechnicalAdvicetotheEuropeanCommissiononPossibleMeasuresConcerning
CreditRatingAgencies.TheCommitteeofEuropeanSecuritiesRegulations(CESR/05Ͳ139b),
March.
Claessens S and Embrechts G (2002). Basel II, sovereign Ratings and Transfer Risk: External versus
InternalRatings.UniversityofAmsterdamandRabobankInternational,theNetherlandsfor
181
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
presentation at the conference Basel II: an Economic Assessment. Basel, Bank for
InternationalSettlements,May17Ͳ18.
Cornford A (2005). Basel II: The Revised Framework of June 2004. UNCTAD Discussion Paper, 178,
April.
CornfordA(2001).TheBaselCommittee’sProposalsforRevisedCapitalStandards:Mark2andThe
StateofPlay.UNCTADDiscussionPaper,156,September.
Daly K and Fischer D (2006). Sovereign Ratings History Since 1975. Standard and Poor’s,
RatingsDirect,April.
EganͲJonesRatingsCompany(2002).StatementofEganͲJonesonCreditRatingAgenciestotheSEC.
November10.
Estrella A (2000). Credit Ratings and Complementary Sources of Credit Quality Information. Basel,
BaselCommitteeonBankingSupervision,WorkingPapers,3,August.
Ferri G (2003). More Analysis: Agencies Invest Enough in Less Developed Countries ? University of
Bari.
Ferri G, Liu LG and Majnoni G (2000). The Role of Rating Agency Assessments in Less Developed
Countries: Impact of the Proposed Basel Guidelines. The World Bank, University of Bari.
At:.www.siteresources.worldbank.org/INTABCDEWASHINGTON2000/Resources/ferrietal.pdf
FerriG,LiuLGandStiglitzJ(1999).TheProcyclicalRoleofRatingAgencies:EvidencefromtheEast
AsianCrisis.EconomicNotesbyBancaMontedeiPaschidisienataSpA,vol.28(3).
FitchRatings(2006).FitchSovereignRatingTransitionandDefaultStudy.SpecialReport,23February.
GoldsteinM,KaminskyGL,ReinhartCM(GKR)(2000).AssessingFinancialStability:anEarlyWarning
SystemforEmergingMarkets.WashingtonDC,InstituteforInternationalEconomics.
Haque NU, Mathieson D and Mark N (1998). The relative importance of Political and Economic
VariablesinCreditworthinessRatings.IMFWorkingPaper(WP/98/46).
HaqueNU,MathiesonDandMarkN(1997).RatingtheRatersofCountryCreditworthiness.Finance
&Development,InternationalMonetaryFund,March.
HaqueNUetal.(1996).TheEconomicContentofIndicatorsofDevelopingCountryCreditworthiness.
IMFStaffPapers,vol.43(4),December.
H.R.2990(2006).AnAct,ToImproveRatingsQualitybyFosteringCompetition,Transparency,and
Accountability In The Credit Rating Agency Industry. U.S. House of Representatives, 109th
Congress,2ndSession.
H.R.2990(2005a).ReformingCreditRatingAgencies:TheSEC’SNeedforStatuoryAuthority.Hearing
Before the Subcommittee on Capital Markets, insurance and Government Sponsored
Enterprises of the Committee on Financial Services, U.S. House of Representatives, 109th
Congress,1stSession,SerialNo.109–66,12April.
H.R.2990(2005b).TheCreditRatingAgencyDuopolyReliefAct.HearingBeforetheSubcommittee
onCapitalMarkets,insuranceandGovernmentSponsoredEnterprisesoftheCommitteeon
Financial Services, U.S. House of Representatives, 109th Congress, 1st Session, Serial No.
109Ͳ66,29November.
IMF (1999). International Capital Markets: Developments, Prospects and key Policy Issues,
September1999.Availableat:www.imf.org/external/pubs/ft/icm/1999/index.htm.
182
CompendiumonDebtSustainabilityandDevelopment
IOSCO (2003). IOSCO Statement Of Principles Regarding The Activities Of Credit Rating Agencies,
Statement of the Technical Committee of IOSCO, Report on the Activities of Credit Rating
Agencies,ReportoftheTechnicalCommitteeofIOSCO(IOSCOPD153),September.
IOSCO (2003). Report on Analyst Conflict of Interest, Report of the Technical Committee of IOSCO
(IOSCOPD152).
IOSCO (2003). IOSCO Statement of Principles Regarding The Activities of Credit Rating Agencies,
StatementoftheTechnicalCommitteeofIOSCO(IOSCOPD151).
Japan Center for International Finance (1999). Characteristics and Appraisal of Major Rating
CompaniesͲͲFocusingonRatingsinJapanandAsia.April.
JuttnerandMcCarthy(2000).ModellingaRatingCrisis.Sydney,MacquarieUniversity,unpublished,
2Ͳ22.
Kaminsky G and Schmukler SL (2002). Emerging Markets Instability: Do Sovereign Ratings Affect
CountryRiskandStockReturns?WorldBankEconomicReview,OxfordUniversityPress,vol.
16(2),August.
KaminskyGL,ReinhartCandVeghC(2004).WhenItRains,ItPours:ProͲcyclicalCapitalFlowsand
MacroeconomicPolicies.NBERWorkingPaper,10780.
Kaminsky GL and Reinhart C (1999). The Twin Crises: The causes of Banking and BalanceͲofͲ
paymentsProblems.AmericanEconomicReview,89(3),June.
Kaminsky GS, Lizondo and Reinhart CM (1998). Leading Indicators ofCurrency Crises. Staff Papers,
InternationalMonetaryFund,Vol.45,March.
Larrain GH, Reisen and von Maltzan J (1997). Emerging Market Risk and sovereign Credit Ratings.
Paris,OECDDevelopmentCentreTechnicalPapers,124,April.
Lehmann A (2004). Sovereign Credit Ratings and Private Capital Flows to LowͲIncome Countries.
AfricanDevelopmentReview,16(2).
Liu LG and Ferri G (2001). How do Global Credit Rating Agencies Rate Firms from Developing
Countries?ADBInstituteResearchPaper,26.
Millon MH and Thakor AV (1985). Moral Hazard and Information Sharing: A Model of Financial
InformationGatheringAgencies.JournalofFinance,40.
MartinMandRoseͲInnesC(2004).PrivateCapitalFlowstoLowͲIncomeCountries:Perceptionand
Reality.CanadianDevelopmentReport,Chapter2.
Moody’s(2001).RevisedCountryCeilingPolicy:RatingMethodology.June.
Moody’s,SpecialComment(1998).WhitePaper:Moody’sRatingRecordintheEastAsianFinancial
Crisis.Moody’sInvestorsService,May.
Moody’s Special Comment (2003). Are Corporate Bond Ratings ProͲcyclical? Moody’s Investor’s
Service,May.
Moody’sSpecialComment(2006).AGuidetoMoody’sSovereignRatings.August,p.1.
MoraN(2005).SovereignCreditRatings:GuiltybeyondReasonableDoubt?AmericanUniversityof
Beirut.
Nagy PJ (1984).Country Risk: How to Assess, Quantify, and Monitor it? Euromoney Publications,
London.
183
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
PackerFandReynoldsE(1997).TheSamuraiBondMarket.CurrentIssuesinEconomicandFinance,
June.
PowellA(2004).BaselIIandDevelopingCountries:SailingThroughtheSeaofStandards.WorldBank
PolicyResearchWorkingPaper,3387,September.
Ramakrishnan RTS and Thakor AV (1984). Information Reliability and a Theory of Financial
Intermediation.ReviewofEconomicStudies,51.
ReinhartCM(2002).Default,CurrencyCrisesandSovereignCreditRatings(NBERWP8738).
ReisenH(2002).RatingssincetheAsianCrisis.DiscussionPaper,2002(2),UnitedNationsUniversity.
Reisen H and von Maltzan J (1999). Boom and Bust in Sovereign Ratings. OECD Technical Papers,
148.Availableat:www.oecd.org/pdf/M00006000/M00006204.pdf.
Rousseau S (2005). Enhancing the Accountability of Credit Rating Agencies: The Case for A
DisclosureͲBasedApproach.CMI,UniversityofMontreal,p.37.
SachsJ,TornellAandVelascoA(1996).FinancialCrisisinEmergingMarkets:theLessonsfrom1995.
BrookingsPapersonEconomicActivityno.1,Washington,BrookingsInstitution.
Setty G and Dodd R (2003). Credit Rating Agencies: Their Impact on Capital Flows to Developing
Countries.FinancialPolicyForum,SpecialPolicyReport,6.
Smith Rand WalterI(2001).RatingAgencies:IsthereanAgencyIssue?SternSchoolofBusiness,
NewYorkUniversity.
Standards and Poor’s (2003). Standards and Poor’s Supports a New, More Transparent NRSRO
DesignationProcess.PressRelease,July.
Standards and Poor’s (2004). Credit FAQ: The Future of Sovereign Credit Ratings. David Beers,
London.
StandardandPoor’sandUNDP(2005).SovereignRatingsinAfrica.September.
SyA(2004).Ratingtheratingagencies:anticipatingcurrencycrisesordebtcrises?In:CantorRed.,
RecentResearchonCreditRatings(specialissue),JournalofBankingandFinance,28(11).
Varma P et al. (2003). Sovereign Bond Defaults, Rating Transitions, and Recoveries (1985–2002).
Moody’sSpecialComment,February.
WhiteLJ(2001).TheCreditRatingIndustry:AnIndustrialOrganizationAnalysis.Paperpresentedat
the Conference on “The Role of Credit Reporting Systems in the International Economy”,
WashingtonD.C.,TheWorldBank,1–2March.
World Bank (2006). Global Development Finance: The Development Potential of Surging Capital
Flows.TheWorldBank.
184
CompendiumonDebtSustainabilityandDevelopment
AnnexI.SovereignRatingsMethodologyProfile
Figure1.GDPperCapita
Source:S&P,Oct.2006,“SovereignCreditRatings:APrimer”.
Figure2.RealGDPGrowthperCapita
Source:S&P,Oct.2006,“SovereignCreditRatings:APrimer”.
185
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
Figure3.ConsumerPriceIndex(CPI)
Source:S&P,Oct.2006,“SovereignCreditRatings:APrimer”.
Figure4.GeneralGovernmentBalanceasPercentageofGDP
Source:S&P,Oct.2006,“SovereignCreditRatings:APrimer”.
186
CompendiumonDebtSustainabilityandDevelopment
Figure5.NetGeneralGovernmentDebtasPercentageofGDP
Source:S&P,Sept.2005,“SovereignCreditRatings:APrimer”.
187
CreditRatingAgenciesandtheirPotentialImpactonDevelopingCountries
AnnexII
Table1.RatingSymbols
RATING SYMBOLS FOR LONG-TERM AND SHORT-TERM DEBT
Interpretation
Moody's
S&P
Fitch
Long-Term Short-Term Long-Term Short-Term Long-Term Short-Term
INVESTMENT-GRADE RATINGS
Highest Credit Quality
Aaa
AAA
AAA
High Credit Quality
Strong Payment Capacity
Adequate Payment Capacity
Last Rating in Investment-Grade
Aa1
Aa2
Aa3
A1
A2
A3
Baa1
Baa2
Baa3
Prime-1
Prime-2
Prime-3
AA+
AA
AA-
A1+
AA+
AA
AA-
A+
A
A-
A1
A+
A
A-
BBB+
BBB
BBB-
A2
A3
BBB+
BBB
BBB-
F1
F2
F3
SPECULATIVE-GRADE RATINGS
Speculative
credit risk developing
due to economic changes
Ba1
Ba2
Ba3
Higly Speculative,
credit risk present
with limited margin of safety
B1
B2
B3
High Default Risk,
capacity depending on sustained,
favorable conditions
Caa1
Caa2
Caa3
BB+
BB
BBNot Prime
B+
B
BCCC+
CCC
CCCCC
C, D
Ca, C
Default,
although prospect of partial recovery
B
BB+
BB
BB-
B
B+
B
BC
D
CCC+
CCC
CCCCC
C, D
C
D
Source:BasedonS&P,Moody’sandFitch.
188
CompendiumonDebtSustainabilityandDevelopment
Table2.RatingAgenciesRecognizedinVariousCountries
Source:BCBS(2000),Table2,p.46.
Note:Table2showstheratingagenciesrecognizedbythebankingsupervisorsinBCBScountriesand
selected nonͲmembers. The total number of agencies recognized in each country is shown in
therightͲhandcolumn.Itisevidentthereisconsiderabledisparityinthenumberofrecognitions
grantedbysupervisors.ThebigthreeCRAs,S&P,Moody’sandFitch,arerecognizedbyallBCBS
membersandalmostallnonͲBCBScountriesshown.
189
CompendiumonDebtSustainabilityandDevelopment
CHAPTERIX
PURSUINGSUSTAINABLEDEVELOPMENT
STRATEGIES:THECASEOFTHEBALANCEOF
PAYMENTRULESINWTO
RobertHowse,AleneSmithandAllanF.Smith
(UniversityofMichigan)
A. Introduction
1. Equity
InSectionIIIoftheMillenniumDeclarationentitled“DevelopmentandPovertyReduction,”United
NationsMemberStatescommittedthemselvesto“tocreateanenvironmentͲatthenationaland
global levels alike Ͳ which is conducive to development and to the elimination of poverty.” This
depends on “good governance within each country”, “good governance at the international level,
andontransparencyinthefinancial,monetaryandtradingsystems.”Hence,they“arecommittedto
anopen,equitable,ruleͲbased,predictableandnonͲdiscriminatorymultilateraltradingandfinancial
system.”
The concept of equity in international trade and financial rules and institutions has not been
explicitly defined and is the subject of debate and speculation among philosophers and political
theorists. Economists are often skeptical of whether the trade and financial systems should be
understood at all in terms of justice rather than as instruments of economic policy coordination.
Nevertheless,itwillbeobservedthattheactualrulesoftendodepend,explicitlyorimplicitly,ona
conceptoffairness.Forinstance,oneoftherulesthatwillbediscussedinthispaper,containedin
ArticleIVoftheIMFArticlesofAgreementrequiredthatIMFMembersnot“manipulateexchange
rates or the international monetary system in order to prevent effective balance of payments
adjustmentortogainanunfaircompetitiveadvantageoverothermembers.”
191
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
The concept of equity is thus inescapable in the interpretation and application of the law of the
internationaltradeandfinancialsystems.Thequestioniswhethertherearelegalandpolicysources
thatallowustogiveadefinitemeaningtothisconceptasweapplyittoparticularrulesanddisputes
inthetradeandfinancialsystems.
One ingredient of equity that is widely reflected in international instruments concerning trade,
financeanddevelopmentisthenotionthatrulesshouldbeadjustedtotheindividualsituationsof
countrieswithrespecttotheirdevelopmentneeds.Thus,thereiswidespreadagreementthatformal
legalequality,treatingeveryonethesameregardlessoftheirparticularsituation,isnotequitable.At
thesametime,thereisdisagreementamongstatesonhowmuchdifferentialtreatmentisjustified
inagivensituation.
Thereisaninterestingparallelismbetweentheconceptionofequityastreating“unlikes”differently
and the recognition in recent economic literature that—contrary to what was implied in the
WashingtonConsensusformula—thereisnotasingleformulaorpathwaytodevelopmentthatwill
workforallcountries.109
Anotherdimensionofequityreflectedininternationalhumanrightsinstrumentsisthatofvoiceand
participation.Theseinstrumentssuggestthatpeopleshouldnothaveavisionofdevelopmentforced
onthemordecidedbyothers.TheDeclarationontheRighttoDevelopment,forexample,stipulates
thattheRighttoDevelopmentincludes“freeandmeaningfulparticipationindevelopment.”
Closelyrelatedtothenotionofequityistheconceptofsocialandeconomicjusticeexpressedinthe
United Nations Covenant on Social and Economic and Cultural Rights. Article 11 of the Covenant
provides: “1. The States Parties to the present Covenant recognize the right of everyone to an
adequatestandardoflivingforhimselfandhisfamily,includingadequatefood,clothingandhousing,
and to the continuous improvement of living conditions. The States Parties will take appropriate
steps to ensure the realization of this right, recognizing to this effect the essential importance of
international coͲoperation based on free consent.” While not all states Ͳ most notably the United
StatesͲhaveembracedtherightsintheCovenantastreatyorcustomaryinternationallaw,eventhe
UnitedStateshasparticipatedintheDeclarationontheRighttoDevelopment,whichincorporates
toalargeextentandaffirmstheserights.Aconcreteimplicationofthisnotionofequityisthatthe
rules of the international trade and financial system should, at a minimum, not undermine, and
ideally should facilitate, the ability of states to discharge their obligations under the Covenant to
implementsocialandeconomicrights.
Finally, equity has been considered by United Nation Member States to imply a fair global
distribution of burdens and benefits from the operations of the international trade and financial
system.Thisgoesbeyondanotionsimplythatthesystem(s)shouldenablestatestoachievesocial
and economic justice within their borders to a conception of global solidarity. According to the
MillenniumDeclaration,solidarityrequiresthat“globalchallengesmustbemanagedinawaythat
distributes the costs and burdens fairly in accordance with basic principles of equity and social
justice.”110
In its examination of WTO rules and jurisprudence as they relate to the balanceͲofͲpayments and
other international financial issues this paper will draw on the dimensions of equity articulated
above.
109
Rodrik D (2001). The Global Governance of Trade as if Development Really Mattered. United Nations Development
Programme.
110
See also Beviglia Zampetti A (2005). Progressing Towards a Just Future Through the MDGs: What is the Role of an
“Equitable”MultilateralTradingSystem?Draft:12.October.
192
CompendiumonDebtSustainabilityandDevelopment
TheWTOAgreementdefinesthegoalofthemultilateraltradingsystemintermsoftheprinciplethat
“relationsinthefieldoftradeandeconomicendeavorshouldbeconductedwithaviewtoraising
standardsofliving,ensuringfullemploymentandalargeandsteadilygrowingvolumeofrealincome
and effective demand, and expanding the production of and trade in goods and services, while
allowingfortheoptimaluseoftheworld’sresourcesinaccordancewiththeobjectiveofsustainable
development...”Clearly,thegoalsofraisingstandardsoflivingandensuringfullemploymentare
closely linked to the conception of social and economic justice in the UN Covenant on Social
EconomicandCulturalRights.
2. Coherence
Arguablycoherenceisalogicalimplicationoftherecognitionofequityasafundamentalelementof
theinternationaltradeandfinancialsystems.Coherencerefers,firstly,to therulesand policiesof
theinstitutionswhereequityisarticulatedanddefinednormativelyand,secondly,totherulesand
policies of the international trading and financial systems themselves. Inequity may result from
uncoordinatedrulesbetweenthetradingandthefinancialsystems.
Forexample,theIMFmayrequireacountrytoimproveitsbalanceofpayments.However,therules
of the trading system may not permit the use of certain instruments for doing so. There may be
goodreasonswhytheseinstrumentsareconstrainedbytherulesoftheinternationaltradingsystem.
However,intheabsenceofabroadandpalatablerangeofpolicyoptionsfortrade,thecountrymay
pursue the goal specified by the IMF through recourse to policy instruments that threaten social
equity,andresultinpovertyandunemployment.
AnearlyexplicitattempttoaddresscoherenceattheWTOistheUruguayRoundDeclarationonthe
ContributionoftheWorldTradeOrganizationtoGreaterCoherenceinGlobalEconomicPolicymaking.
Paragraph2oftheDeclarationreads:
“Trade liberalization forms an increasingly important component in the success of the
adjustment programs that many countries are undertaking, often involving significant
transitional social costs. In this connection, Ministersnote the role of the World Bank and the
IMF in supporting adjustment to trade liberalization, including support to net foodͲimporting
developingcountriesfacingshortͲtermcostsarisingfromagriculturaltradereforms.”
ThemostimportantpartoftheDeclarationisarguablytobefoundinParagraph5:
“The interlinkages between the different aspects of economic policy require that the
international institutions with responsibilities in each of these areas follow consistent and
mutually supportive policies. The World Trade Organization should therefore pursue and
developcooperationwiththeinternationalorganizationsresponsibleformonetaryandfinancial
matters, while respecting the mandate, the confidentiality requirements and the necessary
autonomy in decisionͲmaking procedures of each institution, and avoiding the imposition on
Governments of crossͲconditionality or additional conditions. Ministers further invite the
DirectorͲGeneral of the WTO to review with the Managing Director of the International
Monetary Fund and the President of the World Bank, the implications of the WTO’s
responsibilitiesforitscooperationwiththeBrettonWoodsinstitutions,aswellastheformssuch
cooperation might take, with a view to achieving greater coherence in global economic
policymaking.”
Paragraph 2 draws attention to the significant social costs of trade liberalization and economic
reform. But the Declaration does not extend the idea of coherence to cooperation with those
193
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
internationalinstitutionswhosemandateisdevelopmentpolicysuchasUNCTADandUNDPnorto
thoseconcernedwithsocialequityandjusticesuchastheUnitedNationshumanrightsinstitutions.
Instead, the need for coherence is limited to those international institutions “responsible for
monetaryandfinancialmatters”.However,accordingtoParagraph5theseinstitutionsaretoavoid
“crossͲconditionality,” under which a Government is forced to pursue harmful or inconsistent
policies to meet uncoordinated conditions or restrictions imposed by different international
economicinstitutions.
In 1996, the WTO entered into a cooperation agreement with the IMF, which provides for
consultationsandtheexchangeofinformationbetweenthetwoorganizations.Paragraph10ofthe
agreementprovides:
“The Fund’s staff shall consult with the WTO Secretariat on issues of possible inconsistency
between measures under discussion with a common member and that member’s obligations
undertheWTOAgreement.TheWTOSecretariatshallconsultwiththeFund’sstaffonissuesof
possible inconsistency between measures under discussion with a common member and that
member’sobligationsundertheFund’sArticlesofAgreement.”
ThisprovisionisimportantforcoherencesinceitappearstostrengthenandimplementParagraph5
of the Uruguay Round Declaration on Coherence by requiring consultations about possible
inconsistency while measures are still “under discussion”. As discussed below, in at least one
prominent WTO dispute it is clear that this consultation requirement had not been followed, a
circumstance to which unfortunately the Appellate Body of the WTO did not attach much
importance.
TheWTODohaDeclarationprovidedfortheestablishmentofanewWorkingGrouponTrade,Debt
andFinance:
“36.Weagreetoanexamination,inaWorkingGroupundertheauspicesoftheGeneralCouncil,
oftherelationshipbetweentrade,debtandfinance,andofanypossiblerecommendationson
steps that might be taken within the mandate and competence of the WTO to enhance the
capacityofthemultilateraltradingsystemtocontributetoadurablesolutiontotheproblemof
external indebtedness of developing and leastͲdeveloped countries, and to strengthen the
coherence of international trade and financial policies, with a view to safeguarding the
multilateraltradingsystemfromtheeffectsoffinancialandmonetaryinstability.”
TheWorkingGrouphasexaminedanumberoftheissuesdiscussedbelowbuthasbeenunableto
makeanyconcreterecommendations.A2005report111summarizestheviewsofindividualMembers
of the Working Group but concludes only with the recommendation that the Group continue its
activitiesintothefuture.
B. ExchangeControlsandConvertibility
The GATT rules on exchange controls and convertibility reflect the international financial and
monetary system designed at Bretton Woods at the end the World War II. The postͲwar Bretton
Woods arrangements contemplated a system of fixed exchange rates tied to gold. When a
temporary imbalance of payments occurred (i.e. where a country could not meet payments for
imports with its receipts of foreign currency from export sales without selling gold for foreign
111
World Trade Organization (2005). Report of the Working Group on Trade, Debt and Finance to the General Council.
(WT/WGTDF/4),October.
194
CompendiumonDebtSustainabilityandDevelopment
currency),thiswouldbefinancedbyborrowingfromtheInternationalMonetaryFund.Inthecaseof
astructuralorpersistentimbalance,acountrywoulddevalueitscurrencyunderthesupervisionof
theIMF,whichmightrecommenddomesticpolicyadjustmentstoensurethatfurtherdevaluations
werenotrequiredsubsequentlyinordertomaintainthebalanceofpayments.
TheBrettonWoodssystembrokedownin1971whentheUnitedStatesunilaterallyoptedoutofthe
systemwhenitannouncedthesuspensionofconvertibilityofthedollarintogold.TheresultiswellͲ
summarizedina2004UNCTADdocument:
“Unfortunately, after the breakdown of the Bretton Woods system at the beginning of the
1970s, the world monetary system slipped back into the kind of “monetary chaos” that had
characterizedthepreͲwarperiodanditsdismaleconomicandpoliticaloutcomes.Nevertheless,
the liberalization of the trading system, even after the end of the BrettonWoods system, was
pushed forward by policymakers as if a consistent approach on the monetary side, i.e. a
coherentmonetaryorder,wouldhaveexisted.Onlyrecently,withtheAsiancrisisaswellaswith
the Latin American currency turmoil, have the shortcomings of the “monetary chaos” and its
repercussionsonthetradingsystembeenacknowledged,evenbymainstreameconomictheory
andtheWTO.Butinstabilityisonlypartofthestory.....ifchangesintheinternationalvalueof
moneyareinnowayrelatedtothefundamentalsofcountrieswithopenmarketsforgoodsand
capital,traditionaltradetheoriesquicklylosetheirgrasponrealityandtradeliberalizationloses
muchofitsallegedjustification.”112
In the case of developing countries progress towards convertibility and the removal of exchange
controlswasamajorfeatureoftheeconomicorthodoxyinthe1980sand1990s.Suchreformswere
thought to have the effect of encouraging foreign investment and creating domestic financial
systems as well as access to the global financial networks that would underwrite growth and
development.
TheAsianandLatinAmericanfinancialcrisesofthe1990sledtorethinkingofthisorthodoxy.Well
knowneconomistssuchasJagdishBhagwatiandJosephStiglitzmaintainedthattoorapidfinancial
liberalization contributed to the crises, which led to widespread human misery in a number of
countries,expressedtheirsupportforcapitalcontrolsasaninstrumentforstemmingapanicflight
ofshortͲtermcapital.113114
TheGATTrulesconcerningexchangemeasuresandconvertibilityarecontainedinArticleXVofthe
GeneralAgreement:
x
Article XV:4 states that “Contracting parties shall not, by exchange action, frustrate the
intentoftheprovisionsofthisAgreement,norbytradeaction,theintentoftheprovisions
oftheArticlesofAgreementoftheInternationalMonetaryFund.”
x
According to the Interpretative Note Ad Article XV: “The word “frustrate” is intended to
indicate, for example, that infringements of the letter of any Article of this Agreement by
exchangeactionshallnotberegardedasaviolationofthatArticleif,inpractice,thereisno
appreciabledeparturefromtheintentoftheArticle.Thus,acontractingpartywhich,aspart
of its exchange control operated in accordance with the Articles of Agreement of the
International Monetary Fund, required payment to be received for its exports in its own
currencyorinthecurrencyofoneormoremembersoftheInternationalMonetaryFundwill
112
World Trade Organization (2004). Economic Policy Challenges in an Open Economy: Coherence between Trade and
Finance. Communication of UNCTAD to the WTO Working Group on Trade, Debt and Finance (WT/WGTDF/W/27),
November.
113
BhagwhatiJ(2004).InDefenseofGlobalization.Oxford,OxfordUniversityPress,199Ͳ200.
114
StiglitzJ(2002).GlobalizationandItsDiscontents.NewYork,W.W.Norton.
195
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
nottherebybedeemedtocontraveneArticleXIorArticleXIII[oftheGATTonquantitative
restrictions]. Another example would be that of a contracting party which specifies on an
importlicensethecountryfromwhichthegoodsmaybeimported,forthepurposenotof
introducing any additional element of discrimination in its import licensing system but of
enforcingpermissibleexchangecontrols.”
x
ArticleXV:9oftheGATTprovides:“NothinginthisAgreementshallpreclude:(a)theusebya
contracting party of exchange controls or exchange restrictions in accordance with the
Articles of Agreement of the International Monetary Fund or with that contracting party’s
specialexchangeagreementwiththeCONTRACTINGPARTIES,or(b)theusebyacontracting
partyofrestrictionsorcontrolsonimportsorexportsthesoleeffectofwhich,additionalto
theeffectspermittedunderArticlesXI,XII,XIIIandXIV,istomakeeffectivesuchexchange
controlsorexchangerestrictions.”
x
AccordingtoArticleXVI:2oftheGATT,thereshallbedeferenceto“thedeterminationofthe
Fundastowhetheractionbyacontractingpartyinexchangemattersinaccordancewiththe
ArticlesofAgreementoftheInternationalMonetaryFund,...”
Taken together, these provisions suggest that, where measures have been taken with respect to
exchange controls or restrictions, even if such measures would otherwise be considered trade
restrictionsbecauseoftheireffectonimportandexporttransactions,theintentoftheGATTisnot
toimposedisciplinesbeyondthoserequiredbytheIMF.
Itisinaccuratetoviewtheseprovisions,assomecommentatorshave,essentiallycedingjurisdiction
to the IMF. According to this view, when an exchange measure is not consistent with the IMF
Articles, the “safe haven” of Article XV disappears and the measure may well then fall afoul of a
provisionoftheGATTsuchasArticleXI.Thus,whenacountrydisagreeswiththeFundonthebest
course for solving a financial crisis, including one that does not worsen the plight of the least
advantaged, the GATTpermitscountry tobe“punished”throughbeingfoundinviolation ofGATT
rules.InsuchcasestheGATT/WTOwouldbecomearesidualenforcerfortheIMF.
ArguablythiswasnottheintentoftheGATTframers.Firstofall,priortotheWTOtheGATTdispute
settlementsystemcontainedmanydiplomaticsafetyvalves.Secondly,theoriginalIMFArticleswere
premisedonaworldoflargelyfixedexchangeratesadjustedthroughIMFsupervision.However,in
today’s world of speculationͲdriven currency markets and the widespread liberalization of capital
controls(generallyendorsedbytheIMF)thereisnoagreedinternationalstandardagainstwhicha
currencycanbeviewedasoverͲorunderͲvalued,thustriggeringareasonableobligationtoadjust
economicfundamentalsthroughmeansthatdonotimposeunreasonablecostsonothercountries.
Inthisworldrecoursetoexchangerestrictionsmaybeajustifiableoptionforacountryseekingto
avoidacurrencycrisisortoprotectitselffromthecontagioneffectsofacrisiselsewhere.Thiscan
beillustratedwiththecaseofMalaysia.
In September 1998 Malaysia decided to defy the IMF’s advice and to impose selective capital
controlsinordertohelptoresolveitsfinancialcrisisaswellastoenablethemaintenanceofafixed
exchangerate.KaplanandRodrikconcludethat,incomparisonwithothercountriesthatfollowIMF
prescriptions, and taking into account differences in those countries’ situations, “the Malaysian
policywasmoresuccessfulinaccomplishinganimmediatereductionininterestrates,stabilizingthe
currency, and stemming financial panic. This eased, for the short term at least, worries that the
banking system would go under and that there would be a devaluation spiral. The turnaround in
marketconfidencewascorrespondinglymorerapid.Inaddition,fiscalpolicywasonbalancemore
196
CompendiumonDebtSustainabilityandDevelopment
expansionary.Alltheseinturnspurredconsumptionandeconomicactivity.”115Fromtheperspective
of equity, the observation of Kaplan and Rodrik that the Malaysian approach resulted in
employment and incomes falling less in other Asian countries which followed IMF prescriptions is
particularlyrelevant.
Malaysia’s capital controls would not, on their face, have violated the original IMF Articles of
Agreement, which the drafters of the GATT had in mind when they linked the safe haven for
exchangemeasurestoIMFdisciplines.OnlycurrentͲaccountrestrictionsareclearviolationsofthese
Articles,andIMFMemberStatesretaintherighttomaintaincontrolsoncapitalaccounttransactions.
AsUpakbiexplains:“Whenreadtogether,...[ArticleVII:3(b)andArticleXIV:2oftheIMFArticlesof
Agreement] suggest . . . [w]hereas, in respect of what is regarded as current international
transactions (current account transactions), restrictions are ab initio disallowed (subject to limited
exceptions);inrespectof the capitalaccount,thereverseisthecaseͲrestrictionsaremaintainable
(again,subjecttolimitedexceptions).Giventhisdifferenceinthetreatmentofthetwoaccounts,a
distinction between what falls within their respective ambits appears crucial. . . In part, [current
transactions,accordingtoArticleXXXoftheIMFArticlesofAgreement]envisage“allpaymentsdue
inconnectionwithforeigntrade.”116
AdifficultywiththeseprovisionsistheirassumptionofaclearͲcutdistinctionbetweencapitaland
current account controls. They do not strictly speaking address current account, tradeͲrelated
measures aimed at preventing circumvention of capital controls, such as artificial or overͲinvoiced
tradetransactionswithinmultinationalenterprises.
With respect to exchange rates, the IMF Articles of Agreement provide that an IMF Member shall
not“manipulateexchangeratesortheinternationalmonetarysysteminordertopreventeffective
balanceofpaymentsadjustmentortogainanunfaircompetitiveadvantageoverothermembers.”
(IMF Agreement Article IV, Section 1(iii)). Currency manipulation as such is defined in the
surveillanceprovisionsoftheIMFArticlesas“protractedlargeͲscaleinterventioninonedirectionin
theexchangemarket.”
However, the concept of an “unfair” competitive advantage is not defined. Recently the United
StateshasputconsiderablepressureonChinatoeitherrevalueor“float”itscurrency,onthetheory
that the exchange rate was “unfair” Ͳ rigged through official intervention to keep Chinese exports
artificially cheap. The implication is that a “fair” exchange rate would that determined by the
currency markets. However, in a world of floating exchange rates, where there is no objective
standard,theconceptofa“correct”exchangerateisanillusiveone.
However,inthiscasetheinternationalcommunitymayrequireabroaderbenchmarkwhichincludes
a relevant conception of “equity.” Should China’s purchase of dollars with Renminbi be evaluated
solely in the context of the objective of maintaining an exchange rate that constitutes an “unfair”
competitive advantage? Arguably, under an approach to fairness influenced by conceptions of
sustainabledevelopmentandtherighttodevelopment,oneessentialaspectofthequestionwould
bewhetherChina’sexchangeͲratepolicyrepresentsalegitimateexerciseofitsrighttodevelopment,
andthusthewayinwhichthepolicyfiguresinitsdevelopmentneedsandstrategies.Forexample,
as Erik Denters argues, pegging the Renminbi to the dollar may well have encouraged foreign
investment,acrucialpartofChina’sdevelopmentstrategy.117Atthesametime,onewouldhaveto
115
KaplanEandRodrikD(2001).DidtheMalaysianCapitalControlsWork?NBERWorkingPaper,8142:27.
Ukpabi UC (2004). Juridical Substance or Myth over BalanceͲofͲPayments: Developing Countries and the Role of the
InternationalMonetaryFundintheWorldTradeOrganization.MichiganJournalofInternationalLaw,26(2):710.
117
DentersE(2003).ManipulationofExchangeRatesinInternationalLaw:TheChineseYuan.ASILInsights,November.
116
197
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
ask how far China’s exchangeͲrate measures undermine the development policies of other WTO
Members.
In GATT/WTO practice and jurisprudence, the justifiability of measures under Article XV has been
consideredonanumberofoccasionsinrelationtomanagementofthebalanceofpayments.
x
AccordingtotheGATTAnalyticalIndex,“DuringtheReviewSessionin1954Ͳ55,Italybrought
a complaint concerning action by Turkey providing export bonuses for certain agricultural
productsandlevyingspecialimporttaxesoncertaingoodsdeemedlessessentialinorderto
providethenecessaryfundsforthebonuses.Italystatedthattheexportsubsidieshadnot
beennotifiedasrequiredbyArticleXVI:1andthattheimporttaxeswereinconsistentwith
ArticleII:1(b).Turkeystatedthataspartofareformofitsforeignexchangesystem,ithad
establishedanEqualizationFundwhichwasfinancedbythesaleofimportpermits,andthat
thissystemhadbeenapprovedbytheInternationalMonetaryFund.Arepresentativeofthe
Fund confirmed that the practices under question were multiple currency practices under
the Fund Articles of Agreement and that in a Decision concerning Turkey the Fund had
stated that it did not object to the temporary continuance of these practices and would
remaininconsultationwithTurkeyonthesepractices.”118
x
In 1998 in the ArgentinaͲTextiles and Apparel case, Argentina argued that a 3 perͲcent ad
valoremtaxthatitcollectedwasforpurposesoffundingthecollectionofaccuratestatistical
dataonimportandexporttransactionsaspartofitsoverallunderstandingwiththeIMFon
stabilization and adjustment.119In its ruling the panel held that there was no exception
undertheGATTthatwould,forthesereasons,limitArgentina’sobligationsunderArticleVIII
withregardtocustomsfees.Thepaneldidnotconsiderwhether,giventhatArgentinawas
maintaining the tax in the context of its arrangements with the IMF, the tax could be
deemedtobeanexchangemeasurewithinthemeaningofXV:9oftheGATT.Thefactthat
thetaxappliedtoallimportsindicatesthatitwasnotintendedasaprotectionistmeasureto
shelterArgentineindustriesfromcompetitionwhilelendingplausibilitytoitsconnectionto
Argentina’s exchange arrangements. The Appellate Body upheld the panel’s approach.
Argentina had argued that the Declaration on Coherence and the subsequent Agreement
betweenthe1996IMFandtheWTO,referredtoabove,were“legislativedevelopments”in
the WTO which had the effect of creating a metaͲnorm of avoidance of “crossͲ
conditionalities,” such that its relations with the IMF would require a state to engage in
conductthatwouldviolateWTOlaw.TheAppellateBodyfirstofallobservedthatArgentina
hadnotshowntothepanel’ssatisfactionthatthetaxhadbeenrequestedofitbytheIMFor
therewasaconflictoflegalobligation,i.e.thatArgentinahadalegallybindingagreement
withtheIMFthatwouldbeviolatedifitdidnotimposethetax.
ThefindingsoftheAppellateBodyintheArgentinaͲTextilesandApparelcasesuggestanarrowand
formalistic view of the problem of coherence and conflicting conditionalities. In many cases the
IMF’s requirements are of a general nature, and linked to the achievement of certain results. The
IMFleavestheinstrumentalitiestothecountry’sGovernment.ThattheIMFhasnotrequested“x”
policydoes notmeanthat“x”policydoesnotresultfromrequirementsimposedby theIMF—the
policyinquestionmaybeoneoftheonlyfeasiblewaysofsatisfyingtheIMFdemandsatreasonable
social cost. Moreover the notion of legal conflict suggested by the Appellate Body is equally
problematic. It reduces the challenge of coherence to a notion of avoiding conflicting treaty
requirements.However,internationallawisnottheonlyoreventheprimaryleverthattheIMFuses
118
WorldTradeOrganization(1995)GuidetoGATTLawandPractice.Geneva,1:439.
World Trade Organization (1998). ArgentinaͲMeasures Affecting Imports of Footwear, Textiles, Apparel and Other
Items,ReportoftheAppellateBody.(WT/DS56/AB/R),(adopted22April1998),paras.69Ͳ74.
119
198
CompendiumonDebtSustainabilityandDevelopment
to“enforce”conditionality;rather,theIMFwillsimplynotdisbursefurtherfundstoacountrythat
does not meet its conditions, regardless of whether those conditions are formalized as legal
requirementsorexpressedasistypicallythecasein“memoranda”or“lettersofintent”.120121
In its decision the Appellate Body placed considerable emphasis on the notion that neither the
Declaration on Coherence nor the subsequent Cooperation Agreement between the WTO and the
IMF added to or diminished the rights and obligations contained in the WTO Agreements. The
Appellate Body noted that the effect of crossͲconditionalities or possible conflicts between
measuresthatmightresultfromIMFprogramsandWTOobligationswasspecifiedas“consultation”
betweentheFundandtheWTO.Yetwhenitconsideredwhetherthepanel’sfailuretoconsultwith
theFundconstitutedaviolationofitsobligationtomakeanobjectiveassessmentofthematter,the
Appellate Body ignored the consultation requirement as set out in the Paragraph 10 of the
Agreement between the IMF and the WTO. The thrust of the Declaration on Coherence and the
subsequent Agreement between the IMF and the WTO is that issues that arise from possible
inconsistencies between measures taken in relation to Fund programs on the one hand and WTO
obligationsontheotherought,atleastinthefirstinstance,tobeaddressedthroughconsultations
andcooperationbetweentheWTOSecretariatandtheFund.
Insummary, asinterpretedin thepracticeofWTO disputesettlementinthecases discussedhere
and in others, the WTO rules on exchange actions are likely to be permissive regarding any
macroeconomicpolicyinterventionthathastheexplicitblessingof,orisspecificallyrequiredbythe
IMF. However, where a WTO Member takes an action that the Fund is not prepared to endorse
explicitly, or that it has not required, and such an action falls generally with the kind of exchange
measurescoveredbyArticleXV,thereissomethingclosetoapresumptionthattheWTOruleshave
beenviolated.Yet,acompletereadingoftheagreementestablishingtheWTOandofIMFrulesand
proceduressuggeststhattheydonotnecessarilyjustifythispresumption.
C. TradeRestrictionsforBalanceͲofͲPaymentsPurposes122
ArticlesXIItoXIVoftheGATTelaborateacomplexcodedesignedtogovernanddisciplinetheuseof
import restrictions for balanceͲofͲ payments purposes. Article XII:1 states the basic right of any
Contracting Party to impose quantitative restrictions in derogation from Article XI “in order to
safeguarditsexternalfinancial positionanditsbalanceofpayments”.ArticleXII:2establishesthat
suchrestrictionsshallbelimitedtowhatis“necessary:(i)toforestalltheimminentthreatof,orto
stop,aseriousdeclineinmonetaryreserves,or(ii)inthecaseofaContractingPartywithverylow
monetaryreservestoachieveareasonablerateofincreaseinitsreserves”.Suchrestrictionsmustbe
progressivelyrelaxedasthebalanceofpaymentsimproves.
Furthermore, Contracting Parties “undertake, in carrying out their domestic policies, to pay due
regardtotheneedformaintainingorrestoringequilibriumintheirbalanceofpaymentsonasound
andlastingbasis”(ArticleXII:3).Atthesametime,noContractingPartyisobligatedtotakedomestic
balanceͲofͲpaymentsmeasuresthatwouldthreatentheobjectiveoffullemployment).Aprocessof
consultations is envisaged with the GATT Council concerning any new restrictions or increase in
120
See Eldar O (2005). Reform of IMF Conditionality; a Proposal for SelfͲImposed Conditionality. IILJ Working paper, 10.
NewYorkUniversityLawSchool,CentreGlobalAdministrativeLawSeries.
121
Siegel DE (2002). Legal Aspects of the IMF/WTO Relationship: The Fund’s Articles of Agreement and the WTO
Agreements.AmericanJournalofInternationalLaw,96:561Ͳ581.
122
The following draws from Michael J, Trebilcock and Howse R (2005). The Regulation of International Trade. Third
Edition,Routledge,LondonandNewYork,ch.5“Trade,ExchangeRatesandtheBalanceofPayments.”
199
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
restrictions,withperiodicreviewofthenecessityofthetrademeasuresandtheirconsistencywith
Articles XII–XIV. In addition, Article XII contains provisions on dispute settlement, including the
authorizationofretaliationwhereaPartypersistsintraderestrictionsthathavebeenfoundbythe
ContractingPartiestoviolatetheGATT.
ArticlesXIIIandXIVcontain,respectively,therequirementthatmeasurestakenpursuanttoArticle
XII:1 be implemented on a nonͲdiscriminatory basis and certain narrow exceptions to this nonͲ
discrimination requirement, e.g. where discriminatory exchange controls have been authorized by
theIMF(seethediscussionofsubstitutabilitybelow).
In the case of developing countries, there is a much broader exemption from GATT disciplines for
traderestrictionsundertakenforbalanceͲofͲpaymentsreasons.ArticleXVII:2(b)statestheprinciple
that developing countries should have additional flexibility “to apply quantitative restrictions for
balanceofpaymentspurposesinamannerwhichtakesfullaccountofthecontinuedhighlevelof
demandforimportslikelytobegeneratedbytheirprogramsofeconomicdevelopment”.
This suggests that even though a developing country could address its balance of payments
difficultiesthroughexchangeͲrateadjustmentsortightermacroeconomicpolicies,itshouldnotbe
expectedtodosoinviewoftheharmtodevelopmentthatmaycomefromtheresultantdeclinein
needed imports. It is recognized that quantitative restrictions will allow a developing country to
conserveitslimitedforeigncurrencyresourcesforpurchasesofimportsnecessaryfordevelopment
–whereasadevaluationofitscurrencywouldresultinallimportsbecomingmoreexpensive.Inthis
connectionitbearsemphasisthatbalanceͲofͲpaymentsrestrictionsingeneralmaybediscriminatory
with respect to products although not with respect to countries. Indeed, it is explicitly stated that
“the contracting party may determine (the) incidence (of restrictions) on imports of different
productsorclassesofproductsinsuchawayastogiveprioritytotheimportationofthoseproducts
whicharemoreessentialinthelightofitspolicyofeconomicdevelopment”(ArticleXVIIIB(10)).
In 1979 the Contracting Parties, without formally amending the General Agreement, made the
“Declaration on Trade Measures taken for BalanceͲofͲPayments Purposes”, which expanded the
ambitofArticlesXII–XIVandXVIIIbeyondquantitativerestrictionstoinclude“allimportmeasures
takenforbalanceofpaymentspurposes”.
TheUnderstandingontheBalanceofPaymentsProvisionsoftheGeneralAgreementonTariffsand
Trade 1994 (BOP Understanding), incorporated in the Uruguay Round Final Act, is aimed at
improvingGATT/WTOdisciplineregardingtrademeasurestakenforbalanceͲofͲpaymentspurposes.
Memberscommitthemselvestopublish,assoonaspossible,timeͲschedulesfortheremovalofsuch
trade measures. Furthermore in perhaps the most important modification of the existing GATT
regimeMemberscommitthemselvestogivepreferencetotrademeasuresofapriceͲbasednature,
such as tariff surcharges, and to only resort to new quantitative restrictions where “because of a
criticalbalanceͲofͲpaymentssituation,priceͲbasedmeasurescannotarrestasharpdeteriorationin
theexternalpaymentsposition”(Articles2,3).
PursuanttotheUnderstanding,on31January1995theWTOGeneralCouncilestablishedtheWTO
Committee on BalanceͲofͲPayments Restrictions. From its inception through 2003, the Committee
has conducted consultations with numerous Members concerning the existence and possible
reductionandphaseͲoutoftheirbalanceͲofͲpaymentsrestrictions.Insomeinstances,withrespect
for example to India and Tunisia, there was controversy within the Committee itself as to how
rapidly the balanceͲofͲpayments situation of the country could reasonably permit the removal of
measures.
200
CompendiumonDebtSustainabilityandDevelopment
Dissatisfied with the lack of consensus on India’s use of balanceͲofͲpayments based trade
restrictionstheUnitedStateschallengedIndia’scontinueduse oftraderestrictionsforbalanceͲofͲ
paymentsreasonsindisputesettlement,claimingviolationsoftheGATTandtheBOPUnderstanding.
A key issue here was the relationship between the mandate of the BOP Committee and the
jurisdictionoftheWTOdisputesettlementorgans.Indiaarguedthat,giventheexplicitroleofthe
Committee in the surveillance of the challenged measures, the dispute panel should defer to that
process.ThepanelbelowfoundthatthecompetenceoftheBOPCommitteeandthatofthepanel
werenotmutuallyexclusiveinthesematters.Indiaappealedthisfinding.
The Appellate Body (AB) first observed that, according to Article 1.1 of the Dispute Settlement
Understanding (DSU), the dispute settlement procedures in the DSU apply generally to disputes
broughtunderthedisputesettlementprovisionsofthecoveredagreements(inthiscaseArticleXXIII
oftheGATT 1994). Moreoverone couldnotinferanylimitation ontherightsofaccesstodispute
settlementundertheDSU,oronthecompetenceofpanelstointerpretandapplythebalanceͲofͲ
payments provisions of the GATT, from the grant of competence to review Article XVIII:B
justificationsforsuchrestrictionstotheCONTRACTINGPARTIES.
India, however, had argued that GATT practice with respect to Article XXIII precluded access to
dispute settlement regarding trade restrictions maintained for balanceͲofͲpayments purposes. The
BOP Understanding limited the competence of the dispute settlement organs in balanceͲofͲ
paymentsdisputesinfavorofthatoftheMembership,sittingastheBOPCommittee.Thedistinction
thatIndiadrewwasbetweendisputesaboutthe“application”ofbalanceofpaymentsmeasuresand
thosethatconcernedthesubstantivejustificationofthemeasures.
TherewerealsodifferencesbetweenIndiaandtheABoverthescopeofdevelopmentpolicieswhich
couldjustifyTraderestrictionsforbalanceͲofͲpaymentsreasons.
India argued that under Article XVIII balanceͲofͲpayments restrictions are to be removed as the
conditions to which they were addressed improve only so long as the removal was not likely to
provoke the return of those conditions. Moreover under a further proviso of Article XVIII a
developing country should not be required to remove balanceͲofͲpayments import restrictions, if
doing so could require a change in that country’s development policies.123India’s reliance on this
provisionrequiredtheABtodeterminewhatisadevelopmentpolicyandwhetherremovalbyIndia
ofitsbalanceͲofͲpaymentsrestrictionswouldrequireachangeinthesepolicies.
In its ruling the AB relied on a judgment of the IMF that India did not need to change its
developmentpoliciesbecauseitcouldaddresstheconsequencesofremovingitsimportrestrictions
through“macroeconomic”policies.However,thisrulingisquestionableonvariousgrounds.
Had the AB considered development policy informed by a conception of equity that includes the
notion that development policy is a matter in the first instance for participation of those who are
affected,itwouldhaveanalyzedthelegalissuequitedifferently.
x
Firstly,theABwouldnothaveacceptedthatoneinstitution,particularlythetechnocratsin
thatinstitution,have“ownership”ofthemeaningofa“development”policy.
x
Secondly,theABwouldnothaveembracedthestarkcontrastbetweendevelopmentpolicy
andmacroeconomicpolicy.Thiscontrastimpliesthatdevelopmentpolicyisrestrictedtoa
series of techniques that “experts” view as formulae for “development,” rather than
123
The following draws on Howse R (2004). Mainstreaming the Right to Development into International Trade Law and
PolicyattheWorldTradeOrganization.(E/CN.4/Sub.2/2004/17),3Ͳ20.UnitedNations,Geneva.
201
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
including all those policies that people—in this case, those of India—view as affecting the
fulfillmentoftheirapproachtodevelopment.Fromtheperspectiveofequity,asinformedby
the social and economic rights recognized in the UN Covenant on Social, Economic and
CulturalRights,macroeconomicpoliciesareclearly“developmentpolicies.”
x
Thirdly, on the question of whether India should be required to change its development
policyinordertobeabletoremovethebalanceͲofͲpaymentsrestrictionswithoutareturn
to crisis conditions, for the purposes of both equity and coherence the AB ought to have
solicited the views of a broader range of institutions and social actors—at a minimum the
internationalorganizationswithexpressmandatesregardingdevelopmentsuchasUNCTAD
andtheUNDP.
x
Finally,theABmighthavetakenaccountoftheselfͲdeclaratorycharacterofArticleXVII.B,
i.e.thatitempowersIndiatochartitsowncourseindevelopmentpolicy.Thisimpliesthat
the provision is not intended to invite the dispute settlement organs to examine de novo
India’sjudgmentthatremovaloftherestrictionswouldrequireachangeinitsdevelopment
policy.
D. TradeFinancingandEquity
Increasing exports is recommended as part of policy packages for addressing indebtedness and
balanceͲofͲpayments difficulties since, unlike macroeconomic deflation, it actually increases
employment and reduces poverty. Trade financing is crucial to many export transactions. Yet the
very economic conditions that export receipts are needed to address may make access to such
financingdifficult,particularlyfordevelopingcountriesthathavesufferedfinancialcrises.Thisissue
isbroachedinthe2005ReporttotheWTOGeneralCounciloftheWorkingGrouponTrade,Debt
andFinance.124Ina1999WTOstudyFingerandShulnechtexplaintheimportanceofgovernmentͲ
backed export credit agencies in trade financing as follows: “the commercial and political risk of
international trade transactions is often much larger than for domestic transactions. . . . wellͲ
functioningECAs[ExportCreditAgencies]areprobablyevenmoreimportantfordevelopingcountry
exporters [than for industrial country exporters in developed countries]. [DevelopingͲ country
exporters](andtheirbanks)areoftenrelativelysmalland,therefore,lessabletogeneratetheirown
information on commercial and political risk abroad. They are also often likely to obtain less
favorablefinancingtermsbecauseofmistrustbyimportersfromothercountries.”125
WTOrules,however,arenotconcernedwithfacilitatingdevelopingcountryexportsthroughexport
financing, but rather with disciplining or curbing such financing to the extent it is viewed as an
export subsidy. The relevant provisions are paragraphs (j) and (k) of Annex I (“Illustrative List of
Export Subsidies”) to the WTO Agreement on Subsidies and Countervailing Measures (SCM
Agreement). Paragraph (j) states that the following would be examples of prohibited export
subsidies: “The provision by Governments (or special institutions controlled by Governments) of
export credit guarantee or insurance programs, of insurance or guarantee programs against
increasesinthecostofexportedproductsorofexchangeriskprograms,atpremiumrateswhichare
inadequatetocoverthelongͲtermoperatingcostsandlossesoftheprograms.”
124
World Trade Organization (2005). Report of the Working Group on Trade, Debt and Finance to the General Council.
(WT/WGTDF/4).10October.
125
Finger KM and Schulknecht L (1999). Trade, Finance and Financial Crises. World Trade Organization Special Studies.
WorldTradeOrganization,Geneva.9Ͳ10.
202
CompendiumonDebtSustainabilityandDevelopment
Furtherexamplesaretobefoundinparagraph(k)ofAnnexI:“ThegrantbyGovernments(orspecial
institutions controlled by and/or acting under the authority of Governments) of export credits at
ratesbelowthosewhichtheyactuallyhavetopayforthefundssoemployed(orwouldhavetopay
iftheyborrowedoninternationalcapitalmarketsinordertoobtainfundsofthesamematurityand
othercredittermsanddenominatedinthesamecurrencyastheexportcredit),orthepaymentby
themofallorpartofthecostsincurredbyexportersorfinancialinstitutionsinobtainingcredits,in
sofarastheyareusedtosecureamaterialadvantageinthefieldofexportcreditterms”.
This characterization is subject to following important exception in paragraph (k): “Provided,
however, that if a Member is a party to an international undertaking on official export credits to
which at least twelve original Members to this Agreement are parties as of 1 January 1979 (or a
successor undertaking which has been adopted by those original Members), or if in practice a
Memberappliestheinterestratesprovisionsoftherelevantundertaking,anexportcreditpractice
whichisinconformitywiththoseprovisionsshallnotbeconsideredanexportsubsidyprohibitedby
this Agreement.” The international undertaking referred to here is the OECD Export Credit
Arrangement. By incorporating this Arrangement in paragraph (k) the WTO SCM Agreement
essentiallydrawsalinebetweenprohibitedandpermissibleformsofexportfinancingbasedonan
AgreementnegotiatedbyandfordevelopedcountriesinadevelopedͲcountryforum,theOECD.
The benchmarks in paragraphs (j) and (k) for deciding whether or not a trade financing measure
shouldbeclassifiedasanexportsubsidypresupposethematurecapitalmarketsandsophisticated
riskͲspreading and allocationͲvehicles typical of fully developed economies. Whether they are also
appropriatefordevelopingcountries,especiallyonesthathavehadaccesstoprivatecapitalseverely
limited due to debt and/or other financial crises is questionable. The Center for International
EnvironmentalLawnotesconcerningtheOECDArrangement:“TheArrangementcanbeunderstood
asacartelͲlike,priceͲfixingmechanism,wherethelargestlendersofexportcreditsestablishlimits
oncompetition…Itisanagreementbytherichestcountriesintheworld,andthereforeitsprovisions
aretailoredfortheirneeds.”126
ImplicationsinpracticeoftheSCMAgreementcanbeillustratedbytheBrazilͲAircraftcase,where
theissuewasthesaleforexportofcommuterjetssupportedbyexportcreditsbybothBraziland
Brazil’scompetitorCanada.
IntheBrazilͲAircraftcaseBrazilarguedthat“duetothehighlevelofriskperceivedbyinternational
markets with respect to Brazilian borrowers, the cost to EMBRAER and to Brazilian financial
institutionsofraisingfundstofinanceexportsofBrazilianregionalaircraftishigherthanthecostto
Bombardier and Canadian financial institutions of raising funds to finance exports of Canadian
regionalaircraft.BecausePROEXpaymentsmerelyoffsetinpartthathighercostoffunds,allowing
export credit financing for Brazilian regional aircraft on terms that are closer to, but still less
favorablethan,thoseavailableforcompetingCanadianregionalaircraft,thosepaymentsarenotin
Brazil’sviewusedtosecureamaterialadvantageinthefieldofexportcreditterms.”Inotherwords,
Brazilwasarguingthattheparticularfinancingbarriersfacedindevelopingcountriesshouldbeused
to determine the benchmark against which an export credit is assessed to decide whether it is an
unfairexportsubsidy.(Para7.21)
The panel curtly and almost scornfully rejected Brazil’s approach. Most disturbingly, it suggested
that Brazil’s argument that the baseline of the “marketplace” in paragraph (k) be adjusted to the
circumstancesandneedsofdevelopingcountrieshadtoberejectedbecausetheparagraphwasnot
126
CenterofInternationalEnvironmentalLaw(2003).ExportCreditAgenciesandtheWorldTradeOrganization.DraftIssue
Brief.November:4Ͳ5.
203
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
a provision concerning special and differential treatment (for developing countries) (Para. 7.32).
Article 27 of the SCM Agreement does provide limited relaxation of WTO disciplines applying to
export subsidies for developing countries. Nevertheless, the interpretative approach of the panel
suggeststhat,evenwherespecialanddifferentialtreatmentexistsinaWTOAgreement,theother
provisionsshouldbeinterpretedinamannerthatisblindastotheequitiesasbetweendeveloped
anddevelopingcountrymembers.
Article 27.1 of the SCM Agreement states a general principle much broader than the specific
exceptions and limitations of Article 27.2Ͳ15: “Members recognize that subsidies may play an
importantroleineconomicdevelopmentprogramsofdevelopingcountryMembers.”However,the
panel tendentiously characterized Brazil’s approach to the meaning of “used to secure a material
advantage” in para. (k) as a “general lowering” of SCM disciplines, which might be harmful to
developingcountriesasawhole.ButonamorereasonableinterpretationBrazil’sargumentwasnot
intendedtoleadtoanacrossͲtheͲboardloweringofdisciplines,butrathertotakeintoaccountthe
differenceinfinancialmarketconditionsofaparticulardevelopingcountryinrelationtothoseofits
developedͲcountrycompetitors.Itishardtoseehowsuchanapproachcouldbeharmfultoother
developing countries, manyofwhichface muchmoreseriousstructuraldisadvantagesin termsof
accesstofinancingthanBrazil.
The Appellate Body compounded the indifference to developingͲcountry concerns and challenges
shownbythepanel.Althoughparagraph(k)refersonlytotheOECDArrangementasa“safehaven”
in terms of the disciplines of that paragraph, the Appellate Body used the benchmarks of the
ArrangementastheappropriatemethodologyfordetermininginBrazil’scasewhethertheratesof
interestonitsexportcreditsweresuchastoleadtotheconclusionthatthey“areusedtosecurea
materialadvantage”.
Intheaftermathofthisdecision,somedevelopingcountrieshavejustifiablyputparagraph(k)ofthe
SCMAgreementontheagendaofthepresentDohaRoundofnegotiations.127
E. TheGeneralAgreementonTradeandServices(GATS),Balanceof
Payments,andDebtSustainability
The regulation of banks and other financial institutions is critical to management of debt and
financial crises, especially from an equity perspective. The collapse of financial intermediaries can
destroythesavingsandjobsofordinarycitizens.Thus,theWTOhasaspecialsetofrulesthatapply
toliberalizationoffinancialserviceswithinthegeneralcontextofGATS.
Before considering these special rules, it is important to understand the provisions of the general
WTOframeworkforservicesliberalization,theGATS,whichmayapplytothemanagementofdebt
and financial crises. The GATS applies to trade in services through four modes: (1) crossͲborder
delivery; (2) presenceoftheconsumerinthe territoryofthevendor(e.g. tourism,education,and
health care); (3) commercial presence of the vendor in the consumer state; and (4) crossͲborder
movement of workers engaged in providing services. Certain obligations in the GATS apply to all
services trade in these four modes. There are also general exceptions, including in relation to
balanceͲofͲpayments measures (which are examined below). Many of the most important
obligationsinGATS,suchastherulesapplyingtothegrantingofMarketAccesstoforeignsuppliers
127
CIEL,supra.n.20.
204
CompendiumonDebtSustainabilityandDevelopment
andaccordingthemtheNational(i.e.nonͲdiscriminatory)Treatmentobligationapplyonlywherea
specifiedservicesectorislistedinaWTOMember’sscheduleofspecificcommitments.
Article XI:1 of GATS creates a general rule that “a Member shall not apply restrictions on
international transfers and payments” applicable to sectors where a Member has made specific
commitmentsArticleXI:2states:“NothinginthisAgreementshallaffecttherightsandobligationsof
the Members of the International Monetary Fund under the Articles of Agreement of the Fund,
including the use of exchange actions which are in conformity with the Articles of Agreement,
providedthataMembershallnotimposerestrictionsonanycapitaltransactionsinconsistentlywith
its specific commitments regarding such transactions, except under Article XII [of GATS] or at the
requestoftheFund.”
ThelanguageofXI:2indicatesanextremelyimportantdifferencebetweenGATTandGATS.However
narrowlyorrestrictivelyinterpreted,therelevantprovisionsoftheGATT,aswehaveseen,contain
only disciplines on currentͲaccount measures. However, under the GATS a Member’s specific
commitmentsmaypreventitfrominstitutingcapitalͲaccountcontrols.Tounderstandtheflexibility
undertheGATSwithregardtocapitalcontrolsitisthereforenecessarytolookcarefullyatArticleXII,
the balanceͲofͲpayments exception. This exception can only be utilized after satisfying a very
complexandlongseriesofconditions.ThiscanbeillustratedfromthetextofArticleXIIisasfollows:
ArticleXII:RestrictionstoSafeguardtheBalanceofPayments
1.
In the event of serious balanceͲofͲpayments and external financial difficulties or threat
thereof, a Member may adopt or maintain restrictions on trade in services on which it has
undertakenspecificcommitments,includingonpaymentsortransfersfortransactionsrelatedto
such commitments. It is recognized that particular pressures on the balance of payments of a
Memberintheprocessofeconomicdevelopmentoreconomictransitionmaynecessitatetheuse
ofrestrictionstoensure,interalia,themaintenanceofaleveloffinancialreservesadequatefor
theimplementationofitsprogramofeconomicdevelopmentoreconomictransition.
2.
Therestrictionsreferredtoinparagraph1:
(a) (b)
(c)
(d)
(e)
shallnotdiscriminateamongMembers;
shall be consistent with the Articles of Agreement of the International
MonetaryFund;
shallavoidunnecessarydamagetothecommercial,economicandfinancial
interestsofanyotherMember;
shallnotexceedthosenecessarytodealwiththecircumstancesdescribedin
paragraph1;
shallbetemporaryandbephasedoutprogressivelyasthesituationspecified
inimproves.
3.
In determining the incidence of such restrictions, Members may give priority to the
supply of services which are more essential to their economic or development programs.
However, such restrictions shall not be adopted or maintained for the purpose of protecting a
particularservicesector.
4.
Anyrestrictionsadoptedormaintainedunderparagraph1,oranychangestherein,shall
bepromptlynotifiedtotheGeneralCouncil.
205
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
5.
(a) Members applying the provisions of this Article shall consult promptly with the
CommitteeonBalanceͲofͲPaymentsRestrictionsonrestrictionsadoptedunderthisArticle.
(b) TheMinisterialConferenceshallestablishprocedures128forperiodic consultations
withtheobjectiveofenablingsuchrecommendationstobemadetotheMemberconcernedasit
maydeemappropriate.
(c) Such consultations shall assess the balanceͲofͲpayment situation of the Member
concerned and the restrictions adopted or maintained under this Article, taking into account,
interalia,suchfactorsas:
i)
the nature and extent of the balanceͲofͲpayments and the external
financial;
ii)
difficulties;
iii) the external economic and trading environment of the consulting
Member;
iv) alternativecorrectivemeasureswhichmaybeavailable.
(d) Theconsultationsshalladdressthecomplianceofanyrestrictionswithparagraph2,
inparticulartheprogressivephaseoutofrestrictionsinaccordancewithparagraph2(e).
(e) In such consultations, all findings of statistical and other facts presented by the
International Monetary Fund relating to foreign exchange, monetary reserves and balance of
payments,shallbeacceptedandconclusionsshallbebasedontheassessmentbytheFundofthe
balanceͲofpaymentsandtheexternalfinancialsituationoftheconsultingMember.
6.
IfaMemberwhichisnotamemberoftheInternationalMonetaryFundwishestoapply
the provisionsofthisArticle,the MinisterialConferenceshallestablishareviewprocedureand
anyotherproceduresnecessary.
AnumberoffeaturesofArticleXIIareworthyofspecialattention.
x
First of all, XII:1 gives developing or transitional economies a clear right to take measures
thatprovidealeveloffinancialreserves“adequate”fortheMember’sprogramofeconomic
transition or development. Thus, Article XII:1 affirms that development goals are the
legitimatebasisforaWTOMemberdeterminingthekindsofbalanceofpaymentsmeasures
itneeds.
x
Whereasthemeasuresmust“notexceedthosenecessary”todealwith“seriousbalanceͲofͲ
payments and external financial difficulties or threat thereof,” Article XII:3 affirms that a
Member “may give priority to the supply of services which are more essential to their
economicordevelopmentprograms.”
x
Moregenerally,theconceptof“necessity”oughttobeinterpretedinthecontextofArticle
XII as a whole, which gives considerable emphasis to an individual Member’s approach to
development. Article XII can be read as not requiring a Member to use alternative policy
measures, even if these are less restrictive of services trade, where such measures would
underminetheconceptofequityimplicitorexplicitintheMember’sdevelopmentprogram.
128
Itisunderstoodthattheproceduresunderparagraph5shallbethesameastheGATT1994procedure.
206
CompendiumonDebtSustainabilityandDevelopment
UnliketheprovisionsoftheGATT,ArticleXIIoftheGATSspecifiesthatdeferencetotheIMFextends
only to statistics and facts and conclusions drawn from such statistics and facts. Therefore, a
judgmentabouttheconsistencyofmeasureswiththeIMFArticlesmaybemadeindependentlyat
theWTO.
ArticleXIIenvisagesconsultationsonbalanceͲofͲpaymentsmeasuresintheCommitteeonBalance
of Payments Restrictions. However, according to the logic of the IndiaͲBalance of Payments case
discussedabove,sincetheGATSprovidesnoexceptionfromdisputesettlementforArticlesXIand
XIIoftheGATS,theexistenceoftheCommitteeonBalanceͲofͲPaymentswouldnotleadtoremoval
orrestrictionofpanelandABjurisdiction.
ItisimportanttoappreciatetheextenttowhichGATSspecificcommitmentsmayimplylimitstothe
ability to impose capital account controls. Footnote 8 of Article XVI:1 reads: “If a Member
undertakesamarketͲaccesscommitmentinrelationtothesupplyofaservicethrough[mode1]and
if the crossͲborder movement of capital is an essential part of the service itself, that Member is
thereby committed to allow such movement of capital. If a Member undertakes a marketͲaccess
commitmentinrelationtothesupplyofaservicethrough[mode3],itistherebycommittedtoallow
relatedtransfersofcapitalintoitsterritory.”Situationswhere“movementofcapitalisanessential
part of the service itself” would apply most obviously to certain kinds of financial services (for
example,mutualfunds),buttheotherkindofsituationmentionedinFootnote8ismuchbroader,
applyingtoallcaseswheretheserviceisbeingsuppliedthroughacommercialpresenceintheWTO
Member. Nevertheless, in such circumstances, the requirement of liberalization seems limited to
inwardmovementofcapital.
Some kinds of controls over (outbound) capital might be viewed as conditions on who can supply
services(numberofservicesuppliers)inviolationofXVI:2(a),oras“limitationsonthetotalvalueof
servicetransactionsorassets”inviolationofXVI:2(b)or“totalnumberofserviceoperationsorthe
total quantity of service output” in violation of XVI:2(c). This possibility would follow from an
extremelybroadinterpretationofXVI:2(a)and(c)bytheABintheUSͲGamblingcase.Essentially,the
ABsuggestedthattoviolateeitherprovision,measuresneednottaketheexplicitformsdescribedin
thoseprovisions,providedthattheyhavecomparableeffectsonrestrictingmarketaccessandare
quantitative in nature (Report of the Appellate Body, paras. 232, 247). Since capital controls are
clearly measures that are quantitative in nature, they may well have effects on the number of
servicesuppliersorthetotalvalueofservicestransactionsorassetsunderArticleXVI.
Commitments with respect to financial services are governed by the Annex on Financial Services.
TheAnnexcontainsthefollowingprovision:
DomesticRegulation
(a) Notwithstanding any other provisions of the Agreement, a Member shall not be
preventedfromtakingmeasuresforprudentialreasons,includingfortheprotectionof
investors,depositors,policyholdersorpersonstowhomafiduciarydutyisowedbya
financialservicesupplier,ortoensuretheintegrityandstabilityofthefinancialsystem.
WheresuchmeasuresdonotconformwiththeprovisionsoftheAgreement,theyshall
not be used as a means of avoiding the Member’s commitments or obligations under
theAgreement.
The first sentence of this provision appears to allow any measure “to ensure the integrity and
stabilityofthefinancialsystem”withouttheneedtoshowthatthemeasureisnecessaryortheleast
restrictive of trade in services. The second sentence, however, seems drafted in a manner to
207
PursuingSustainableDevelopmentStrategies:TheCaseoftheBalanceofPaymentRulesinWTO
underminetheregulatoryflexibilitygrantedinthefirstsentence,inthatitqualifiestheuseofthe
provision as an exception to GATS commitments and obligations. Thus, where a measure is not in
conformitywithGATS,it“shallnotbeusedasameansofavoidingtheMember’scommitmentsor
obligationsundertheAgreement.”
Itisdifficulttodiscerntheexactimplicationofthisqualifyingorconditionallanguage.Onepossible
reading would be that it imports an intent requirement into 2(a), namely the notion that the
measuresmustbegenuinelyintendedto“ensuretheintegrityandstabilityofthefinancialsystem”
ratherthantoprotectdomesticfinancialindustries.Suchanintentrequirementmightbedifficultto
applyinthecaseofafinancialcrisis,whereensuringthesurvivalofdomesticfinancialinstitutions
maywellbepartandparcelofensuringthe“integrityandstabilityofthefinancialsystem”itself.
Finally,anycommitmentorobligationunderGATSissubjecttothegeneralexceptionsinArticleXIV
ofGATS.Thus,whetherornotaMember’smeasuremeetsthecriteriasetforthinArticleXIIofthe
GATS or the Annex on Financial services, the measure may still be justified if “necessary” for the
protection of human life or health or of public moral or public order. According to footnote 5 of
ArticleXII,Thepublicorderexceptionmaybeinvoked“onlywhereagenuineandsufficientlyserious
threatisposedtooneofthefundamentalinterestsofsociety.”Inthiscontextitisnoteworthythat
intheUnitedStatesͲGamblingcasetheABupheldthepanelapproachthatsuggestedtheremustbe
somedeferencetoaWTOMember’sowndeterminationofthemeaningofpublicmoralsandpublic
order(AppellateBodyReport,paras.296Ͳ297).
AsageneralmatterotherpoliciessuchasexchangeͲratestabilization,depreciationorappreciation
undertakeninresponsetoafinancialcrisismaybeunsustainableintheabsenceofcapitalcontrols.
Experience with applicable parts of the GATS will thus eventually play a role in determining the
rangeofmacroeconomicpolicyresponses.
208
CompendiumonDebtSustainabilityandDevelopment
F. Conclusions
WTO rules on exchange actions and the balanceͲofͲpayments justifications for trade restrictions
clearlyreflectaconceptionofequitythattakesintoaccounttheparticularneedsandsituationsof
developing countries. In certain, carefully defined matters the WTO rules entail deference to
judgmentsoftheIMF.
However,inactualdisputesettlementelementsintherulesthatreflectequitytowardsdeveloping
countrieshavebeenminimizedorignored.Moreover,thedisputesettlementorganshavegonewell
beyondtheexplicitlimitsofdeferencetotheIMF,deferringsuchtotheIMFeveninsuchamatteras
the meaning of a country’s “development policy” (the IndiaͲBalance of Payments case). Since
developing countries have limited representation and voice in the IMF, from the perspective of
equityasparticipationindecisionͲmakingconcerningdevelopmentthesetendenciesofthedispute
settlementorgansseemdifficulttojustify.
More generally, the concept of coherence reflected in relevant WTO instruments and activities
directed towards balanceͲofͲpayments and exchange matters is too narrowly focused on relations
between the IMF and the WTO, and does not include cooperation with other international
institutionsconcernedwithequityindevelopment.Theconceptofcoherenceshouldberevisedto
accommodate the relationship with equity implied in the Millennium Declaration and related
instruments.
Moreover even within the narrow conception of coherence embraced in the WTO, the agreed
mechanism for avoiding crossͲconditionalities, namely obligatory consultations between the WTO
SecretariatandtheIMFpriortoeithertakingdecisionsthatcouldresultincrossͲconditionalities,has
not been closely followed. A review should be undertaken of the justification for not using this
process and of the extent to which avoidance of crossͲconditionalities has been achieved in
experiencesofar.
InthecaseoftheGeneralAgreementonTradeinServices(GATS),thereisarealpossibilitythata
WTO Member’s specific commitments combined with the general obligations of the GATS could
meanthataMember’sadoptionofcapitalcontrolsconstitutesaGATSviolation,eventhoughsuch
controls may be necessary to address a financial crisis in a manner consistent with social and
economic justice. In view of the exceptions and limitations in the GATS that could none the less
justifysuchmeasuresthereisacaseforthedrawingupofguidelinesinthisareawhichtakeaccount
ofequityinthetradeandfinancialsystemintheinterpretationofsuchlimitsandexceptions.This
task should be undertaken by international institutions with a mandate related to equity in
development.
As exemplified by the BrazilͲAircraft case, the rules on export subsidies in the SCM Agreement
appear to limit the capacity of developing countries to provide support for export transactions
through export credits. This reflects the use of market benchmarks devised for and by developed
countries in the OECD Arrangement for the assessment of export credit arrangements.
Considerationshouldbegiventoanalternativeapproachwhichwouldtakeintoaccountstructural
differences between the financial markets of developed and developing countries as well as the
specialchallengesregardingaccesstocapitalmarketsforexportfinancingfacingcountriesthathave
facedfinancialordebtcrises.
209
CompendiumonDebtSustainabilityandDevelopment
CHAPTERX
RISKASSOCIATEDWITHTRENDSINTHE
TREATMENTOFSOVEREIGNDEBTINBILATERAL
TRADEANDINVESTMENTTREATIES
AldoCaliari
(CenterofConcern)
A. Introduction
ThereisagrowingtrendinFreeTradeAgreementsfortheinclusionofprovisionsthatsubjectpolicy
towardsthefinancialsectortolegaldisciplinesenshrinedintradeandinvestmentagreementsand
totheassociateddisputeͲsettlementmechanisms.Thistrendplaceslimitsontheusebydeveloping
countries of several tools designed to build and preserve stable and healthy financial sectors
responsive to national development priorities and supportive of trade. The limits are capable of
increasingdevelopingcountries’vulnerabilitytofinancialanddebtcrises.
B. SovereignDebtinBilateralTradeandInvestmentTreaties
In bilateral Free Trade Agreements recently negotiated by the United States Government a
controversialissuehasbeentheinsistenceoftheUnitedStatesonpursuinginclusionofclausesthat
wouldapplytosovereigndebtissuedbythepartiesprinciplessuchasNationalTreatmentandMostͲ
FavoredͲNation(MFN)TreatmentwhicharepartofbilateralinvestmenttreatiesandofGATT/WTO
rulesfortradeingoodsandservices.
A review of some recent treaties reveals at least two different approaches to the treatment of
sovereigndebt.
211
RiskAssociatedwithTrendsintheTreatmentofSovereignDebtinBilateralTradeandInvestmentTreaties
1. SovereignDebtExplicitlyExcludedfromApplicationofthePrinciples
UnderNAFTA,investmentcoversasweepingarrayoftypesofownershipinterests,includingloans
and securities. However, in conformity with Article 1416 in the section on Financial Services,
“investment means “investment” as defined in Article 1139 (Investment Definitions), except that,
withrespectto“loans”and“debtsecurities”referredtointhatArticle:(a)aloantoordebtsecurity
issuedbyafinancialinstitutionisaninvestmentonlywhereitistreatedasregulatorycapitalbythe
Partyinwhoseterritorythefinancialinstitutionislocated;and(b)aloangrantedbyordebtsecurity
owned by a financial institution, other than a loan to or debt security of a financial institution
referredtoinsubparagraph(a),isnotaninvestment;”
Tothisisaddedthefollowing:“forgreatercertainty:(c)aloanto,ordebtsecurityissuedby,aParty
orastateenterprisethereofisnotaninvestment”(author’sitalics).
Therefore,underNAFTA,sovereigndebtsareexplicitlyexcludedfromthedefinitionofinvestment.
2.
SovereignDebtExplicitlyIncludedwithintheScopeofApplicationofInvestment
Principles
In the 2003 United StatesͲChile Free Trade Agreement (FTA) specific principles on investment are
explicitly applicable to sovereign debt. The United StatesͲChile FTA contains a broad definition of
investmentbasedonthefollowingstandardadoptedbytheUnitedStatesinitsmostrecentBilateral
InvestmentTreaty(BIT)Model.129
“Investmentmeanseveryassetthataninvestorownsorcontrols,directlyorindirectly,thathasthe
characteristicsofaninvestment,includingsuchcharacteristicsasthecommitmentofcapitalorother
resources,theexpectationofgainorprofit,ortheassumptionofrisk.Formsthataninvestmentmay
takeinclude:
x
x
x
x
x
x
x
x
Anenterprise;
Shares,stock,andotherformsofequityparticipationinanenterprise;
Bonds,debentures,loans,andotherdebtinstruments;
Futures,options,andotherderivatives;
Rights under contract, including turnkey, construction, management, production,
concession,orrevenueͲsharingcontracts;
Intellectualpropertyrights;
Rights conferred pursuant to domestic law, such as concessions, licenses, authorizations,
andpermits;and
Other tangible or intangible, movable or immovable property, and related property rights,
such as leases, mortgages, liens, and pledges; but investment does not mean an order or
judgmententeredinajudicialoradministrativeaction…”
Thisdefinitiongenerallyincludes“bonds,debentures,loansandotherdebtinstruments”.130Inwhat
represents a significant departure from NAFTA, the treaty explicitly makes the agreement’s
129
ThisdefinitionhasbecomestandardblueprintfortheUSnegotiatingpositionintreaties.SeeUnitedStates2004Model
BIT,Art.1
130
Usuallywithafootnotethatclarifies“Someformsofdebt,suchasbonds,debentures,andlongͲtermnotes,aremore
likely to have the characteristics of an investment, while other forms of debt, such as claims to payment that are
immediatelydueandresultfromthesaleofgoodsorservices,areleslikelytohavesuchcharacteristics.”
212
CompendiumonDebtSustainabilityandDevelopment
provisions applicable to sovereign debts issued by the Chilean Government.131The same rules are
contained in the Central America Free Trade Agreement (CAFTA).132Thus, the United StatesͲChile
FTAandCAFTAmakeNationalTreatmentandMFNTreatmentapplicabletosovereigndebtsissued
bytheGovernmentsofthecountriesinvolved.
3. The“Elliptic”InclusionofDebtintheUnitedStatesͲUruguayFTA
TheUnitedStatesͲUruguayFTA(signedin2004)raisesaninterestingquestionbecauseitsprovisions
could lead to reͲinterpretation of previous treaties. The FTA contains the standard definition of
investment as including “Bond, debentures, other debt instruments and loans” It also contains, in
AnnexF,aclausewithlanguagesimilartothefirstpartoftheNAFTAarticleabove.133
Uptothispointofthetext,althoughthereisnoexplicitexclusionasintheNAFTAsupplementary
clause, the agreement seems to imply that sovereign debt is excluded from the definition of
investment.
However,thisdoesnotappear tobe the case.Annex Gofthe UnitedStatesͲUruguayFTAheaded
“SovereignDebt”,readsasfollows:
“1.NoclaimthatarestructuringofadebtinstrumentissuedbyUruguaybreaches
an obligation under Articles 5 through 10 may be submitted to, or if already
submitted continue in, arbitration under Section B, if the restructuring is a
negotiated restructuring at the time of submission, or becomes a negotiated
restructuringaftersuchsubmission.”
Thiswouldappeartomeanthatsovereigndebtis,indeed,includedinthescopeofthedefinitionof
investmentforthepurposesoftheTreaty.Italsowouldopenthewayfortheinterpretationthat,
absent an explicit exclusion, sovereign debt is considered to fit into the scope of the definition of
investment.Thiscouldhavetheconsequenceofleadingtoanexpansionofthescopeoftheterm,
“investment”, in treaties worded similarly to the United StatesͲUruguay FTA, such as the United
StatesͲSingaporeFTA.
C. ImplicationsforSovereignDebtProblemsofIncludingNationalTreatment
andMFNTreatmentinFTAs
WhatarethepossibleimplicationsofapplyingNationalTreatmentandMFNTreatmenttosovereign
debt?
131
SeeAnnex10ͲB(Annextothechapterofthetreatythatdealswithinvestment):“ThereschedulingofthedebtsofChile,
orofitsappropriateinstitutionsownedorcontrolledthroughownershipinterestsbyChile,owedtotheUnitedStatesand
thereschedulingofitsdebtsowedtocreditorsingeneralarenotsubjecttoanyprovisionofSectionAotherthanArticles
10.2and10.3”Articles10.2and10.3intheTreatyrefertoNationalTreatmentandMostͲFavoredͲNationTreatment.
132
SeealsoUgarteche(2004,14Ͳ18and34Ͳ35).
133
Art.4reads:
“(b)Investmentmeans“investment”asdefinedinArticle1,exceptthat,withrespectto“loans”and“debtinstruments”
referredtointhatArticle:(i)aloantoordebtinstrumentissuedbyafinancialinstitutionisaninvestmentinafinancial
institutiononlywhereitistreatedasregulatorycapitalbythePartyinwhoseterritorythefinancialinstitutionislocated;
and(ii)aloangrantedbyordebtinstrumentownedbyafinancialinstitution,otherthanaloantoordebtinstrumentofa
financialinstitutionreferredtoinsubparagraph(b)(i),isnotaninvestmentinafinancialinstitution”.
213
RiskAssociatedwithTrendsintheTreatmentofSovereignDebtinBilateralTradeandInvestmentTreaties
These principles were originally developed in different historical contexts. National Treatment
featured from the first half of the twentieth century onwards in treaties of friendship, commerce
andnavigation(FCNtreaties),i.e.bilateraltreatiescoveringmiscellaneoussubjectssuchasaccessto
ports,tariffs,thepowersandresponsibilitiesofconsuls,andprotectionagainstappropriation.The
lastoftheseheadingstypicallyincludedprovisionsconcerningnationaltreatment,i.e.guaranteesof
nonͲdiscriminatory treatment of foreign firms. MFN clauses were included in reciprocal trade
agreementsnegotiatedbetweentheUnitedStatesandvariouscountriesunderaprogramlegislated
in 1934. Under the MFN clauses included in these agreements each of the parties bound itself to
extendtotheothertariffconcessionsatleastasgreatasthoseextendedtothemostfavorednation
with which it traded. Both National Treatment and MFN Treatment were included in the GATT as
principlesapplyingtotradeingoods.
The extension of National Treatment and MFN Treatment to other subjects is neither
straightforward nor uncontroversial.134Indeed, their extension to sovereign debt raises issues that
could be more harmful to developing countries than those considered under their traditional
applicationtoforeigninvestment.Adiscussionofanumberoftheseissuesfollows:
1.
DismantlingToolsNeededfortheRecoveryoftheLocalEconomyinPostͲCrisis
Situations
The application of National Treatment to sovereign debt would restrict the ability of the debtor
Government to take certain policy measures aimed at the recovery of the local economy in the
aftermath of financial crises. National Treatment in this context means that foreign creditors are
offeredtreatmentindebtrestructuringsnolessfavorablethanthatofferedtodomesticcreditors.135
However,thereareseveralreasonswhyacountryrestructuringitssovereigndebtafterafinancial
crisismightneedtoresorttoofferingpreferentialconditionstodomesticcreditors.
x
In a financial crisis, domestic creditors often suffer a double adjustment. First, they are
typically forced to accept a “haircut” on their claims, which means that the value of their
loansarereducedbyacertainpercentage.Secondly,theyoftensuffercostsrelatedtothe
internaladjustment,suchashighinterestrates.Infact,theimpactofdebtrestructuringon
domesticcapitalmarketsand,inturn,ontheresumptionofgrowthandrepaymentcapacity
needs to be taken into account in assessing the consequences of debt crises (Machinea,
2004:188)..”
x
Dealing with domestic before foreign debt might also allow the Government to return
rapidlytodomesticcapitalmarketsduringwhatislikelytobeasustainedinterruptioninits
accesstointernationalcapitalmarkets(IMF,2002:13).
x
The debtor may also need to accord priority to domestic debt in order to protect the
financialsystem.TheIMFhassaidthat“therestructuringofcertaintypesofdomesticdebt
may have major implications for economic performance, as a result of its impact on the
financialsystemandtheoperationofdomesticcapitalmarkets”(IMF,2002:13).Sovereign
debt restructuring typically has a double impact on the financial system. On the one hand
134
SeeKhor(2002),whostates:“ItiscertainlynotclearthattheprinciplesoftheWTO(includingNationalTreatmentand
MostͲFavoredͲNationtreatment)thatapplytotradeingoodsshouldapplytoinvestment,northat,ifapplied,theywould
benefitdevelopingcountries.”SeealsoChangandGreen(2003),ActionAid(2003),OxfamInternational(2003).
135
ThisisimportantinthecontextofthedevelopingcountrysignatoriesofCAFTA,since,withtheexceptionofHonduras,
animportantshareofpublicdebtinallthesecountriesisowedtodomesticcreditors.Insomeofthem,likeCostaRica,
domesticdebtisactuallyhigherthanexternaldebt.
214
CompendiumonDebtSustainabilityandDevelopment
financialinstitutionsareweakenedbytheimpactontheircapitallevelsofthereductionin
the value of bonds. On the other hand, debt restructuring is associated with a general
increase in uncertainty, which can inflict widespread damage on the creditworthiness of
firms (Machinea, 2004: 188Ͳ189). Thus, in such cases special treatment to domestic debt
mayenablethedebtortoprotect“acoreofthebankingsystembyensuringtheavailability
ofassetsrequiredforbankstomanagecapital,liquidityandexposuretomarketrisks”(IMF,
2002:13).
x
Asovereigndebtormayalsoneedtoaccordspecialtreatmenttodomesticdebtorsforthe
samereasonsthatcanleadittoaccordspecialtreatmenttonationalsectorsandindustries
aspartofanationaldevelopmentstrategyandtheachievementofdevelopmentgoals.
x
In the IMF’s view sheltering domestic investors from the full impact of debt restructuring
maybenecessaryinorderto“garnersupportforanambitiousadjustmentprogram”(IMF,
2002:13).
2. PreventingtheStatefromPayingSalariesandPensionsinDebtCrises
TheapplicationofNationalTreatmenttosovereigndebtmeansthattheGovernmentwillbeunable
toprioritizedomesticdebtassociatedwithmeetingwages,salariesandpensionobligations.Inother
words, the Government is bound to treat these debts in the same way as foreign debts held by
transnationalbanksandinstitutionalinvestors.Ifitsresourcesareenoughtocoveronlyaportionof
its debts, the state will not be able to choose to direct those funds to meeting these priorities, at
leastnotaslongasitdoesnotdevoteequalamountforpaymentstoforeigncreditors.
Unlikeanindebtedprivatecompany,anindebtedsovereignhashumanͲrightsobligationsandsocial
responsibilitiestowardsitspeople.Thismeansthat,indealingwithsovereigndebt,thereareissues
that cannot be addressed by strict analogies with bankruptcy principles applicable to the private
sector.ThusproposalsofcivilsocietyforarulesͲbasedframeworkhavetypicallycalledforrecourse
toanalogieswithframeworkswhichaccommodatetheoverallmissionthatthestateisexpectedto
fulfill. Such frameworks include Chapter 9 of United States Bankruptcy Law applicable to
municipalities. Even the IMF’s muchͲcriticized Sovereign Debt Restructuring Mechanism proposal
excluded“Wages,salariesandpensions”fromitsapplication(IMF,2003:24).
3. ReducingtheLeverageofDebtorsinaDebtRestructuring
ByfirstgatheringthesupportofdomesticcreditorsaGovernmentcanacquiresubstantialcloutfor
thenegotiationsoverdebtrestructuringwithothercreditors.Theofferofpreferentialconditionsto
thesedomesticcreditorscanbecriticalinthiscontext.ThusiftheprincipleofNationalTreatmentis
applied to sovereign debt, this avenue for the indebted country to strengthen its negotiating
positioniseffectivelyforeclosed.
TheofferofpreferentialconditionstodomesticcreditorswascrucialtoenhancingtheGovernment’s
leverageinArgentina’snegotiationswithitscreditorsafteritsDecember2001default.InSeptember
2003theGovernmentreleaseditsinitialproposedconditionsfordebtrestructuring,whichincluded
a 75ͲperͲcent haircut for bond holders. The Government contended that this was the size of the
reduction that would enable it to recover sustainable economic growth, while ensuring that its
promisesofpaymentwerekept.SomegroupsofbondͲholdersquicklyrejectedthisoffer,claiming
thatitwaswoefullyinsufficientand,inthelightofthecountry’smostrecentgrowthfigures,below
thecapacityofthecountrytorepay.ThecreditorsalsostronglylobbiedtheG7which,directlyand
215
RiskAssociatedwithTrendsintheTreatmentofSovereignDebtinBilateralTradeandInvestmentTreaties
through the IMF, put pressure on Argentina to improve its offer.136With pressure mounting from
thesequarters,ArgentinaturnedtodomesticpensionfundswithanofferofinflationͲlinkedbonds
thatrepresentedanimprovementovertheoffermadetotheotherbondholders.Bythusgranting
these institutions preferential conditions, Argentina was able to reach agreement with creditors
holdingmorethan17percentofitstotaldebt.Thiswasacriticalfirststepingarneringthesupport
of a majority of creditors that eventually totaled 76 per cent. However, the offer of preferential
treatmenttodomesticpensionfundswouldnothavebeencompatiblewiththeprincipleofNational
Treatment.
4.
CreationofaPrivilegefortheDebtOwned(orAcquired)byCreditorsfromthe
Party
ApplicationofNationalandMFNTreatmentonlytocreditorsofcountriesthatarepartiestobilateral
investmenttreaties(whichhaveinrecentyearslargelyreplacedtheFCNtreatiesmentionedearlier)
would have the discriminatory result of granting seniority to creditors from such countries over
thosefromother countries.Thiswouldaffect therightsofbondͲholdersfromnonͲparty countries
without their consent since they are, by definition, excluded from intervening in the negotiations
under the bilateral agreement. For these bondͲholders such treatment might be equated to an
involuntarydebtswapunderwhichtheyfindthemselvesholdingadowngradedinstrument.
D. InvestorͲStateLawsuitsandSovereignDebt
OneeffectofapplyingtheprinciplesofinvestmenttreatiestosovereigndebtisthatGovernments
that violate investor protections can face expensive lawsuits. As under NAFTA and numerous
bilateral investment treaties, CAFTA grants private foreign investors the right to bypass domestic
courtsandsueGovernmentsininternationaltribunals(Peterson,2004:3).
Such“investorͲstatelawsuits”arehighlycontroversialforanumberofreasons(Peterson,2004and
2004a). Many arbitration tribunals operate with a lack of transparency, having no obligation to
disclose relevant documents or allow any form of public participation. The system for choosing
arbitrators has also drawn criticism as the arbitrators can be drawn from the ranks of practicing
investment lawyers and there is no obligation to appoint arbitors who will be independent in the
senseofnothavingastakeinhowthetreatyisinterpreted.
Moreover,arbitraltribunalsdonothavetopayregardtolegalprecedents(Peterson,2004:6).This
feature, which creates a lot of uncertainty in the investment arena, could become particularly
troublesomewhenappliedtosovereigndebtcrises.Indeed,themainrationaleformoresystematic
arrangements for handling sovereign debt defaults has been the need to provide greater
predictability for both debtors and creditors in the messy process of exiting sovereign debt crises.
Clearly,theexistingsystemofarbitrationtribunalswoulddoapoorjobataddressingthoseconcerns
andwouldinjectadditionaluncertaintyintoexistingarrangementsforthefollowingreasons:
136
InitsIMFagreementtheArgentinegovernmenthadpromisedto“negotiateingoodfaith”andwassingledoutinsome
G7statementsasnotcomplyingwithsuchapledge.Privatecreditorsmaintainedthatnegotiationsingoodfaithrequired
theagreementof80percentofcreditors,whilethegovernmentofArgentinaclaimedthatafigureabove65Ͳ70percent
would suffice. It was incongruous that the IMF and G7 countries, which were themselves amongst the creditors, should
haveunilaterallyattemptedtodefinetheconditionsofanacceptabledebtrestructuring.
216
CompendiumonDebtSustainabilityandDevelopment
x
The application of the principles of National Treatment and MFN to sovereign debt might
givethesearbitraltribunalstheauthoritytodefinedifficultquestionsthatarguablybelong
tothedomesticjurisdictionofstates.
x
The application of these principles might also open the way for the application of other
more general principles that are becoming common in investment treaties, such as
“minimumstandardoftreatment”or“fairtreatment.Asillustratedaboveinthediscussion
ofArgentina’sdebtrenegotiation,thereisnorulesͲbasedframeworktodeterminewhatis
an“acceptable”levelofrepaymentor“negotiationingoodfaith”,etc.indebtnegotiations
andrestructurings.Noristhereanycertaintythatprinciplesorrulesoriginallyformulatedin
thecontextofbankruptcylawwillbeappliedbyanarbitrationtribunal.
x
These general principles are contentious even in the context of investment treaties. That
minimum or fair standards of treatment apply only to investors, while considerations
involving workers and other human rights as well as the environment, which might
counterbalancethem,arenotgivenequalweightisasourceofcontroversy.
x
CloselyrelatedtopointsraisedinsectionC.2isthepointthatapplicationofthesegeneral
principlestosovereigndebtwouldnottakeaccountoftheresponsibilityoftheGovernment
ofthedebtorcountrytoitspopulation.
E. ConcludingRemarks
The existing regime for dealing with sovereign debt crises lacks a rulesͲbased, multilateral
framework.Thisleavesdebtorsvulnerabletopowerasymmetriesascomparedwithcreditors.These
asymmetries would be reinforced by extension of the definition of the investment instruments
coveredinbilateralinvestmenttreatiestoincludeallormostdebtinstruments,particularlythosefor
sovereigndebt.Therehavealreadybeenmovestowardsamoreinclusivedefinitionofinvestmentin
somerecenttreaties.Thishastheconsequencethatdebtinstrumentsaresubjecttoprinciplessuch
as National Treatment and MFN Treatment which were originally developed to handle problems
arisingunderbilateralinvestmenttreatiesandgoodstradeundertheGATT,andnotdebtcrises.A
notableexceptiontotherecenttendencyforextendingsuchprinciplestodebtistheNAFTA,which
explicitly excludes sovereign debt from the definition of investment. In view of the dangers to
developing countries from the extension of principles designed for foreign investment and goods
tradetodebtinstruments,theNAFTAapproachfurnishesasuperiormodelforthefuture.
217
RiskAssociatedwithTrendsintheTreatmentofSovereignDebtinBilateralTradeandInvestmentTreaties
References
ActionAid(2003).UnlimitedCompanies.TheDevelopmentalImpactsofanInvestmentAgreement
attheWTO.UnitedKingdom,June.
ChangHJandGreenD(2003).TheNorthernWTOAgendaonInvestment:DoasWeSay,NotasWe
Did.Geneva,SouthCentreandCAFOD,June.
Correa C (1999). Key Issues for Developing Countries in a Possible Multilateral Agreement on
Investment.In:InternationalMonetaryandFinancialIssuesforthe1990s,5.NewYorkand
Geneva,UnitedNations.
IMF (2003). Proposed Features of a Sovereign Debt Restructuring Mechanism. Washington DC, 12
February.
IMF(2003a).IMFDiscussesPossibleFeaturesofaSovereignDebtRestructuringMechanism.Public
InformationNotice,03(06).WashingtonDC,7January.
IMF (2002). Sovereign Debt Restructuring MechanismͲFurther Considerations. Prepared by the
InternationalCapitalMarkets.WashingtonDC,14August.
IMF (2002a). The Design of the Sovereign Debt Restructuring Mechanism. Washington DC, 27
November.
IMF (2002b). IMF Board Discusses Possible Features of a New Sovereign Debt Restructuring
Mechanism.PublicInformationNotice,02(106).WashingtonDC,24September.
Khor(2002).TheWTO,thePostͲDohaAgendaandtheFutureoftheTradeSystem:ADevelopment
Perspective.ThirdWorldNetwork.Malaysia,May.
Machinea JL (2004). Reestructuración de la deuda: nuevas propuestas para viejos problemas. In:
Gobernabilidadeintegraciónfinanciera:ámbitoglobalyregional.SantiagodeChile,ECLAC.
OxfamInternational(2003).TheEmperor’sNewClothes.BriefingPaper,46,April.
Peterson LE (2004). Bilateral Investment Treaties and Development PolicyͲMaking. International
InstituteforSustainableDevelopment,Winnipeg.
Peterson LE (2004a). UK Bilateral Investment Treaty Programme and Sustainable Development.
ImplicationsofBilateralNegotiationsonInvestmentRegulationataTimewhenMultilateral
TalksareFaltering.London,TheRoyalInstituteofForeignAffairs,February.
UgartecheO(2004).VeinteAñosDespués...LaDeudadelosPaísesAndinos.Lima,ProgramaLaboral
deDesarrollo,September.
218