29
1 The 2017 Version of the IMF and World Bank’s LIC Debt Sustainability Framework: “Significant Overhaul” or Obsolete? Brian Pinto 1 June 1 2018 1 The author thanks Hans Weisfeld for promptly clarifying questions on IMF (2017a). An early version of this paper was shared with the IMF and the World Bank. While this paper draws upon the work of the ADF Policy Innovation Lab, the views herein are those of the author alone acting in his independent capacity and should not be attributed to the African Development Bank, its Executive Directors or its Management.

The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

1

The2017VersionoftheIMFandWorldBank’sLICDebtSustainabilityFramework:

“SignificantOverhaul”orObsolete?BrianPinto1

June12018

1TheauthorthanksHansWeisfeldforpromptlyclarifyingquestionsonIMF(2017a).AnearlyversionofthispaperwassharedwiththeIMFandtheWorldBank.WhilethispaperdrawsupontheworkoftheADFPolicyInnovationLab,theviewshereinarethoseoftheauthoraloneactinginhisindependentcapacityandshouldnotbeattributedtotheAfricanDevelopmentBank,itsExecutiveDirectorsoritsManagement.

Page 2: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

2

Acronyms/AbbreviationsADF:AfricanDevelopmentFundBoP:BalanceofPaymentsCAD:CurrentaccountdeficitCI:CompositeIndicatorCPIA:CountryPolicyandInstitutionalAssessmentDRM:DomesticRevenueMobilizationDS:DebtSustainabilityDSA:DebtSustainabilityAnalysisDLP:DebtLimitsPolicy(oftheIMF)[check]NCBP:Non-concessionalBorrowingPolicy(oftheWorldBank)[check]DSA:DebtSustainabilityAnalysisDSF:DebtSustainabilityFrameworkEM:EmergingMarketFDI:ForeignDirectInvestmentForexdebt:DebtdenominatedinthehardcurrenciesFX:ForeignexchangeFXRA:ForeignexchangereserveadequacyHIPC-MDRI:HeavilyIndebtedPoorCountries-MultilateralDebtReliefInitiativeMCLs:ModeratelyConcessionalLoansMDBs:MultilateralDevelopmentBanksLIC:Low-IncomeCountryLIDC:Low-IncomeDevelopingCountryNR:NaturalResourceNRX:NaturalResourceExportingODA:OfficialDevelopmentAssistancePFM:PublicFinanceManagementpp:percentagepointPPG:PublicandPubliclyGuaranteedPV:PresentvalueSDGs:SustainableDevelopmentGoalsSOE:State-ownedenterprise(publicenterprise)SSA:Sub-SaharanAfricanToT:TermsofTradeTPD:TotalPublicDebtTPDD:TotalPublicDebtDistressWB:WorldBankWP:WorkingPaperXDD:Externaldebtdistress

Page 3: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

3

Summary

Thispapercritiquesthe2017versionoftheIMFandWorldBank’sLICDSFcontainedinIMF(2017a).Itusesaccumulatedemergingmarketanddevelopingcountryexperience,especiallyfromAfrica,andeconomiclogictoshowthatthe2017DSFisobsoleteandshouldnotthereforebeimplemented.Itrecommendsanalternative,simplerapproachthatrespondstotheexigenciesoftheday:thatanewpublicdebtcrisisinADFcountries(AfricanLIDCsinIMFparlance)isunfoldingevenaspublicinvestmentneedsformeetingtheSDGsandcreatingjobsforAfrica’sburgeoningyouthremainvast.

TheframeworkcomesfromtheADFPolicyInnovationLab,whichbeganworkinlate2015onthedebtsustainabilityimplicationsoflessconcessionalODA(intermsofvolumes,termsorboth)goingforward.Itbecameclearfromthiswork(ADFLab2016,FelinoandPinto2017),whichlookedat33ADFcountries,thattheLICDSFwouldnotbesuitablebecauseofitsoverwhelmingfocusonPPGexternaldebtrisks:theattentionneededtoshifttopublicdebtanditsdynamicsasthefundamentaldriverofexternaldebtdistress.Furthermore,thetusslebetweendebtsustainabilityanddevelopmentneededattentionatatimewhenODAwasshowingsignsofbecominglessplentiful.TheworkoftheLabfedintothereportoftheHighLevelPanelonreinvigoratingAfricandevelopmentfinance(HLP2017).

Evidencehasgrown,includingfromtheIMF—seeIMF(2017b)andespeciallytheexcellentreportinIMF(2018a)—thatpublicdebtsustainabilityproblemshaveintensifiedinADFcountries,investmentneedsarevast,andODAisdwindling.Themarkethasdisplacedtheofficialsectorindefiningthemarginalborrowingcostofagrowingnumberofgovernments.Onereasonisthatofficialfundingsourcesarenotenoughforthehugepublicinvestmentsneededinhumancapitalandinfrastructure.AnotherreasonuncoveredbycountryvisitsbytheADFPolicyLabisthat,asoneofficialputit,“themarketdoesnotaskquestions”—althoughitdoespriceforriskandislessforgivingthanofficiallenders.

Indeed,amajorissueisthemisuseofpublicresourcesamongAfricanLIDCs,includingthosefromnaturalresourcewealthandmarketborrowings.Themainfactorsdrivingthedeteriorationinpublicdebtdynamicsincludehighinterestratesonmarketborrowings,growthslowdowns,exchangeratecollapsesandgrowingprimaryfiscaldeficits.Thelastfactorincorporatesatusslebetweenpooruseofpublicfunds,insufficientDRMandweakPFMsystemsontheonehandandspendingpressuresfromthevastupfrontinvestmentneedsinconnectionwithinfrastructureandSDGsontheother.

Forexample,atoneextreme,GhanaandMozambiquefaceobviouspublicdebtsustainabilityproblemsmagnifiedbyweakPFMsystems.Attheotherextreme,Ethiopia,KenyaandRwandacapturetheconstanttensionbetweenmacroeconomicsustainabilityandvastinvestmentneedsininfrastructurepresentinallADFcountries.Thistensionisreflectedinhighfiscaldeficitsthatspilloverintoburgeoningcurrentaccountdeficitsandexternaldebtdynamics.

NowassumethattheLICDSFdidnotexistandhadtobedesignedfromscratch,whichiswhatthe“significantoverhaul”claimedinIMF(2017a)implies.Itshouldataminimumdothefollowing:

1. Concentrateonpublicdebtlevelsanddynamicsbecausethisiswherethefundamentalproblemlies.Inaddition,examinevulnerabilityfromcurrentaccountdeficitsandexternaldebtaccumulationrecognizingthesearesymptomsofrisingfiscaldeficitsandthegrowingunsustainabilityinpublicdebtdynamics

Page 4: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

4

2. Recognizethatthemarketandnonconcessionalfundingsourcesincreasinglydeterminethemarginalcostofgovernmentborrowing,evenincountrieswhichhavenotissuedEurobonds,thoughnotalwaysintransparentways(recallMozambique’s“hiddendebt”uncoveredinApril2016,ataboutthesametimethatitissueditsfirstEurobondinconnectionwithrestructuringtheguaranteeddebtofthestate-ownedtunafishingcompany.ChadborrowedUSD1.45billionin2014fromGlencoreandfourbankstoberepaidwithcrudeoilshipments)

3. Acceptthat,inspiteoftheHIPC-MDRIdebtwriteoff,concessionalityhasnotpreventedthere-emergenceofpublicdebtsustainabilityproblemsorledtoasolidfoundationformeetingtheSDGs.Thereasonisthatconcessionalitycannotcompensatefortheseriousgapsinleadership,economicgovernance,fiscalandfinancialinstitutionsandtheprivateinvestmentclimate.ThiscallsforatougherpolicydialoguecoordinatedacrosstheIMF,WBandAfDBinsteadofthepresentfragmentedapproach

4. Buildinadiscussionofhowtoreconciledebtsustainabilitywithdevelopment.Themarketisunlikelytobeasuitablefinancingsourceforthevastpublicinvestmentneeds,callingformorefrontloadingofODA—buthowtoachievethisgiventhefiscalconstraintsdonorsface?TheanswermustinvolvesomeincreaseinthepricingofODAandraisingthebarforaccess,bothofwhichcanbefacilitatedbythefactthatcostlymarketborrowingsnowplayasignificantroleamongADFcountriescombinedwithheterogeneityintheirabilitytousepublicresourceswell.

Notwithstandingtheaboverequirementsdictatedbytheaforementionedexigenciesoftheday,the2017versionoftheDSF(IMF2017a)toberolledoutonJuly12018continueswiththeinheritedfocusonassessingdistressinrelationtothepresentvalueofPPGexternaldebt.ThiswasentirelyappropriatewhentheLICDSFwasoriginallyintroducedtomakeacaseforHIPC-MDRIdebtrelief:mostpublicdebtwasthenconcessionalPPGexternaldebtfromofficialsources,therebyalsojustifyingtheuseofthepresentvalueofdebt.Butthisisnolongertrue.Whilethe2017DSFincludestotalpublicdebt,itdoessoignoringmarketrealitiesandcontinuingtheantiquatedfocusonthepresentvalueofdebt.Itshouldthereforenotbeimplementedandanalternativedevelopedthataccordsprimacytopublicdebtanditsdynamics,andrecognizestheever-presenttensionbetweendebtsustainabilityanddevelopmenteveninthebest-governedADFcountries.ThefeaturesofsuchanalternativearesketchedoutandillustratedwiththerecentLICDSFexperienceofEthiopia.

Page 5: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

5

I.INTRODUCTION

The2017IMF-WBReviewoftheLICDSFwaslaunchedinearly2016tooverhaultheDSF.AconsultationseminarwasheldonApril12,2016,duringthe2016SpringMeetingsatIMFHQinWashingtonDCtodiscussthescopeandkeytopicstobeincludedinthereview.2Theseincluded:

• Allowingforachangedfinancinglandscapewithmoremarketdebtfromdomesticandexternalsourcesandashiftawayfromofficialfinancingsources

• Strengtheningtheanalysisofpublicdebtbyincorporatingdomesticdebt(whichistypicallycontractedatcommercialterms)

• Reflectingmarketrisksbetter(debtrollovers,yieldspreadsrelativetobenchmarks,foreignexchangereserveadequacy)

• Needforbetteranticipationofrisksandmorerobustriskratings.

Insum:theDSFneededtogetinclosertouchwiththenewfinancingrealitiesinLICs,callingforgreateremphasisonpublicdebt,thegrowingroleofmarket-basedfinancingandforeignexchangereserveadequacy(FXRA).Thereview,whichtookmorethanoneyearandinvolvedajointIMF-WorldBankteam,culminatedinanIMFStaffReportdatedAugust222017(IMF2017a).ItwasapprovedbytheIMF’sBoardinSeptember2017andisduetoberolledoutonJuly12018bytheWorldBankandIMF.

ThispapercritiquestheIMF-WB2017reviewoftheLICDSF(“2017DSF”)fromtheperspectiveoftheclientcountriesoftheADF,namely,theSub-SaharanAfricancountriesincludedinthelistof“Low-IncomeDevelopingCountries”(LIDCs)containedinAnnexTable1ofIMF(2018a).Itsmainconclusionisthatthe2017DSFfallsshortofthe“significantoverhaul”itclaimstobeandisobsolete.Hence,implementingitwouldbecounterproductive.Themainreasonsarethefollowing:

1. ContinuedobsoletefocusonPPGexternaldebt.ThisfocuswasentirelyappropriatewhentheLICDSFwasintroducedin2004tomakeacaseforHIPC-MDRIdebtrelief.Atthattime,officialexternalborrowingspredominatedinpublicdebt,makingPPGexternaldebtanobviousvariabletofocuson.Butoncedebtreliefwasprovided,AfricanLIDCgovernmentsstartedborrowingfromthemarkets,domesticandinternational.Thiswouldnecessitateashifttorisksfromtotalpublicdebt,includingitsdomesticcomponent,andawayfromPPGexternaldebt.

2. Unsatisfactoryincorporationoftotalpublicdebt.The2017DSFallowsforrisksassociatedwithtotalpublicdebtbysimplygraftingdomesticdebtontothePVofexternalpublicdebt,whichcontinuestodefinetheframeworkforriskassessment.Thiscausestwoproblems.First,forAfricanLIDCsaswithmostEMsbeforethem,thecausalflowisfromunsustainablepublicfinancestoexternaldebt:largefiscaldeficitsandadversepublicdebtdynamicsspilloverintocurrentaccountdeficitsandexternaldebtdynamics.Tocontinuetofocusonexternaldebtisthereforetoconfusethesymptomwiththemalady.Second,theinsistenceonretainingthePVofexternaldebt(computedusinga5%discountrate)asthevariableofinterestisoutdatedintworespects:(i)itobscurestheinsightsthatcanbeaffordedbysimplylookingatthenominaldebt-to-GDPratio,inwhichcaseitbecomeseasytointerpretthedynamicsbasedonprimaryfiscalbalancesandr-g,theinterestrate-growthrate

2Theauthorwaspresentduringthisconsultationseminar.

Page 6: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

6

differential;and(ii)itisinconsistentinthatmarketdebtisincludedatfacevalue,notatitspresentvaluediscountedat5%becausethiswouldgiveanumberlargerthanthenominaldebtvalue(sincetheinterestrateonsuchdebtistypicallyabove5%).Butifonebelievedthat5%wasanappropriatediscountrateforassessingthe“true”burdenofdebt,thenitshouldbeappliedacrosstheboard.ThesearepowerfulreasonsforabandoningthePVofdebtconceptandshiftingtonominaldebt.Besides,eventhepoorestandmostfragileAfricanLIDCsnowborrownonconcesionallyatthemargin,bringingustothenextpoint.

3. Failuretoincorporatemarketsignals.ForasignificantsetofADFcountries,themarketnowdeterminesthemarginalcostofgovernmentborrowing.3ThoseissuingEurobondshavesovereignratingsfromthemajorcreditratingagencies.Likewise,domesticdebtistypicallyissuedoncommercialterms—EthiopiaisanotableexceptionandKenyaintroducedinterestrateceilingsinlate2016—andmostADFcountrieshaveahighdegreeofopencapitalaccountopenness.The2017DSFignoresthisinformation,stickingtothePVofdebtandassessingriskbasedonthresholdslinkedtotheCPIAandnowanewCompositeIndicatorbasedonpast,narrowlydefinedinstancesofdebtdistress(linkedtosuchvariablesasarrearstotheIMF).Instead,moreattentionneedstobepaidtomarketsignalsondevaluationanddefaultrisks.

4. Unsuitabilityforreconcilingdebtsustainabilityanddevelopment.IMF(2017a)notes:“CSOshaveshownincreasedinterestintheLICDSFandhaveadvocateditsuseasatooltohelpdeterminehowtofinancethelargedevelopmentneedsofLICsinthecontextoftheeffortstoachievetheSustainableDevelopmentGoals(SDGs).”Thisclaimnotwithstanding,theestimatedfinancingneedsfortheSDGsaresovastthattheyarelikelytoblowpublicdebtandCADsustainabilityoutofthewaterevenforthebest-managedADFcountries,suchasRwanda,KenyaandEthiopia.Considerablemodificationisneeded,notjustintheDSF,butthewholearchitectureofODA,forabetterreconciliationofdevelopmentanddebtsustainability.

Further,theLICDSFeveninits2017versiontendstotreatdebtsustainabilityasanendinitself,notasaplatformfordevelopment,aswillbecomeclearfromthecountryexamplesinthenextsection.Moreover,theinsistenceonconcessionalityisoutofsyncwiththefactthatthemarketnowdeterminesthemarginalcostofborrowingforagrowingnumberofcountries.Indeed,ADFcountries,withtheexceptionofthemostfragile,shouldeasilybeabletocopewithlessconcessionalODAiftherootcausesofunsustainabledebtareaddressed:inabilitytodeliveronprimaryfiscaldeficittargets,poorexpenditurecompositionandmisuseofpublicresourcesandnaturalresourcewealth,weakgrowthpolicyandexchangeratecollapsesrelatedtopoliticalandmacropolicyrisksthatmagnifytheburdenofforexdebt.ThesefindingsinFelinoandPinto(2017)areechoedintheIMF’sownexcellentMarch2018LIDCdebtpaper(IMF2018a),whichalsonotesthatgrantsandremittancesaredeclining.ThesefactorscallforaseachangeinthepricingandallocationofODAandthenatureofpolicydialogueinordertoincreasethechancesforbetterdevelopmentoutcomesinSub-SaharanAfrica.

ThispaperpresentsanalternativeDSAapproachforAfricanLIDCsbasedontheanalysisoftheADFPolicyLabdescribedintwoworkingpapers:ADFLab(2016)andFelinoandPinto(2017).ThesepapersuseexactlythesamedataasintheIMF-WBDSAsbutemployasimplerapproachwhichislikelytobemorerobustgiventhevariousdataissuesnotedinIMF(2018a).Ultimately,DSAframeworkscanonlyprovidehypothesesthatthenrequiretestingviacountryvisitsanddiscussionswithvariousstakeholders.Suchanapproachhasbecomeurgentgiventhe

3Weusetheterms“AfricanLIDCs”and“ADFcountries”interchangeably.

Page 7: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

7

deteriorationinpublicdebtsustainabilitynotedinFelinoandPinto(2017—henceforth,FP2017)andconfirmedinIMF(2018a).AnewdebtcrisisisunfoldinginAfricaandadifferentDSAapproachisneededthathelpsreconciledebtsustainabilityanddevelopmentwhileallowingforcountyheterogeneity.

ThenextsectionprovidesthreecountryexamplestoillustratetheshortcomingsoftheLICDSF,whichareunlikelytobeaddressedbyits2017version.SectionIIIdescribesthemainreformsproposedinthe2017DSFfollowedbyacritiqueinsectionIV.SectionVproposesanalternativeapproachillustratedbytheEthiopianexperienceinsectionVI,whilesectionVIIconcludes.

II.THREECOUNTRYEXAMPLES

ThissectiondiscussesrecentDSAsforGhana,RwandaandEthiopiatoillustratechallengesassociatedwiththeLICDSFthatareunlikelytoberemediedbythe2017DSF.TheADFPolicyLabvisitedtwoofthesecountriesduringitswork:GhanaandRwanda.

Twopointscanbedistilledfromthecountryexamples:

First,theLICDSFdoesnotgivesufficientprominencetopublicdebtanditsdynamics.ToanticipatethecritiqueinsectionIV,the2017DSFretainstheobsoletefocusonPPGexternaldebt.AddingondomesticdebtinthemannerproposedinIMF(2017a)toanalyzerisksfromtotalpublicdebtdoesnotalterthisconclusion.TheDSFneedstobecomemoremarketorientedandincorporatehard-wonlessonsfromEMsinitsprescriptions.

Second,theLICDSFtendstotreatdebtsustainabilityasanendinitself.Thisisobviousfromallthreecountryexamples,withimportantquestionsbeingleftoffthetable.Amoresuitableapproachwouldtreatdebtsustainabilityasaplatformfordevelopment,notasanendinitself.ElementsofanalternativeapproachwillbespeltoutinsectionV.

Athirdpointisworthmaking.AgrowingbodyofevidenceindicatesthatpublicdebtsustainabilityproblemsarenowentrenchedinseveralAfricanLIDCsandthatanewdebtcrisisissprouting.Forexample,IMF(2018a,pp42-3)lists8ADFcountriesasbeingindebtdistressasofend2017basedontheLICDSF:Chad,Eritrea,Somalia,SouthSudan,Sudan,andZimbabwe.Inaddition,MozambiqueandRepublicofCongowereindefaultandthereforeclassifiedasindebtdistress.Afurther7ADFcountrieswereathighriskofdebtdistressasofend2017:Cameroon,CentralAfricanRepublic,Ethiopia,Ghana,Mauritania,TheGambiaandZambia.Moreover,“Onlyintwowerelargerfiscaldeficitslinkedtohigherpublicinvestment,eitherinfull(Cameroon)orinpart(Zambia).”Ethiopiatoofallsintothiscategory,asisapparentfromIMF(2018b).ThisprecarioussituationimpartsameasureofurgencytodesigningaDSFthatcanbelinkedtodevelopmentfinancingneedsandcanbecontinuouslyupdatedinsteadofwaitingforthenormalArticleIVconsultationcycle,whichcouldresultinneedlessdelaysindiagnosisandpolicyprescription.

Ghana

GhanawasheldupastheposterchildfortheHIPC-MDRIdebtwrite-offin2007,whenitissuedthefirstofitsfive[check]Eurobondswiththegoalofusingitsnewlydiscoveredfiscalspacetopromotegrowthanddevelopment.StevenRadeletgaveanupbeataccountof17“emergingcountries”inhis2010bookonEmergingAfrica,openingwithaglowingsummaryofGhana’sachievementsovertheprevious15years.

Page 8: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

8

InJanuary2017,withGhanaalreadydeeplymiredinmacroeconomicdifficulty,a“fiscalhole”of4%ofGDPwasdiscoveredfollowingtheinstallationofanewgovernment.Publicdebthadreached74%ofGDPatend-2016comparedtoaprojected70.4%intheIMF-WorldBankOctober2016debtsustainabilityanalysis.

Thenegativedebtoutcomeshouldnothavecomeasabigsurprisegivenobviouslyadversepublicdebtdynamicsinearlieryearsandaweaktrackrecordonusingpublicresourceswell—infact,electionissuesincludedcorruption,electricityshortagesandinsufficientattentiontopublicvalueformoney,resultingintheincumbentgovernmentbeingvotedout.Inotherwords,largefiscaldeficitswerenotassociatedwithinvestmentsthatwouldspurfuturegrowthandtaxes.

MoresurprisingisthefailuretocomeupwithadecisiveplantoaddressGhana’spersistentfiscalimbalanceinspiteofitstrackrecord.Forexample,theDSAinIMFcountryreport16/16(January2016)recommendedastrategytolowergovernmentinterestpaymentsbysubstitutingEurobondsforlocalcurrencydebtaspartof“debtmanagement”,anapproachthathassubsequentlycontinued,asindicatedintheDSAinIMFcountryreport16/321(October2016).Thisignoredthestrongsignalsfromthemarketondefaultrisk—forexample,thesecondarymarketyieldonGhana’s2020Eurobondwas15%inearly2016.ItalsoignoredacostlylessonfromEMsonavoidingthebuildupofcurrencymismatchesonthegovernment’sbalancesheet.Preferencehasbeengiventoshort-termfinancialengineeringploysthatincreasefiscalriskoverrecognizingandaddressingthecountry’sfundamentalfiscalproblem.

Indeed,theADFLab’svisittoGhanainFebruary2017indicatedrelyingheavilyonrevenuesfromnewoilfieldsaswellasissuingEurobonds(toavoidsoaringinterestcostsinthedomesticmarket)asastrategyforloweringthepublicdebt-to-GDPratio,reflectingtheanalysisintheOctober2016DSA.Newoilproductionwouldraisegrowthrates,primarysurplusesandforeignexchangereserves,whileputtingdownwardpressureonthecedi/USDrate.ShiftingtoEurobondswouldlowerinterestratepaymentsreportedinthefiscaldeficit(domesticborrowingcostswereoftheorderof25%)butatthecostofincreasingcurrencymismatchesandenhancingcurrencyrisksonthegovernment’sbalancesheet.NowhereintheOctober2016DSAisthewasteofpublicresourcesorthemisuseofnaturalresourcewealthmentioned.Debtsustainabilityistreatedmyopicallyasanendinitself.

OnlyinFebruary2017wereweaknessesinfiscalgovernancehighlightedbyattributingthe“fiscalhole”to“significantpublicspendingcommitmentsthatbypassedPFMsystems”.4Improvingmatterswouldrequireimplementing“thenewgovernment’sintentionstoreducetaxexemptions,improvetaxcomplianceandreviewthewidespreadearmarkingofrevenues”;inshort,ensuringtheintegrityofPFMsystemsandselectingandexecutingpublicinvestmentscarefully.

ItisatleastmildlyshockingthatastarofthepostHIPC-MDRIerahasnotbeenabletoimplementbasicPFMsystemsinspiteof10yearshavinggoneby.Italsounderlinestheimportanceofshiftingtheonusonre-attainingpublicdebtsustainabilitytotheGhanaianleadersthemselvesandawayfromexternalbodiesliketheIMF,whichcandolittlewithoutcommitmentandleadershipfromtheverytopinGhanaitself.Indeed,the4%ofGDPfiscalholeoccurredundertheverynoseoftheIMF,withGhanahavingbeenanExtendedCreditFacility(ECF)IMFprogramcountrysinceApril2015.

4QuotesfromIMF’sFebruary102017PressRelease17/43.

Page 9: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

9

Rwanda5

RwandaisthehighestratedADFcountrybasedontheCPIAandranked56thoutof190countriesinthe2017WorldBank’sDoingBusinesssurvey,secondinAfricaafterMauritius.IthasbeenselectedfortheG20CompactwithAfrica,thegoalofwhichistofosterforeigndirectinvestmentfromadvancedcountriesbyaddressingimpedimentsinboththehostandhomecountries.

AspartofitsVision2020strategyaimedatattainingmiddle-incomestatusby2020,significantpublicinvestmenthasbeendirectedto“programstoimprovesocialoutcomes,increaseagriculturalproductivityandtransformtheeconomytohighervalueaddedactivities,improvegenderequality,andfosterfinancialinclusion,amongotherthings”.

TheIMF2017ArticleIVreportnotes:"Growth-enhancinginfrastructureinvestmenthascontinued,withprojectscarefullychosentomaintainRwanda’slowdebtriskrating,whiledomesticrevenuesandexternalfinancinghaveincreasedanddomesticfinancingwaslimited.Supervisoryframeworkshavebeenstrengthened,andexchangerateflexibilityhasbeenthecenterpieceofadjustmentpoliciestoaddressexternalimbalances."

Nevertheless,currentaccountdeficits(CAD)areunsustainablyhigh,at14.4%ofGDPin2016withprojectionsof10.1%for2017and11.3%for2018.TheIMFreportstwoCADnumbers:thestandardoneandonewithoutbigprojects.Eventhelatterisexceptionallylarge,closetodoubledigitsasapercentageofGDP.Infact,S&PdowngradedRwandafromB+fromBinSeptember2016onaccountofitsswellingCAD,althoughtheIMFDSFcontinuestoratethecountryatlowriskofdebtdistress.

Rwanda’sexperiencecapturesthetensionbetweenvastupfrontpublicinvestmentneedsfordevelopmentandpublicdebtsustainabilityeveninthebest-governedADFcountries.Inspiteofitsbeingratedatlowriskofdebtdistress,thedatapresentedintheIMF2017ArticleIVreportposeapuzzleinthattheactualincreaseinexternaldebttoGDPbasedontheDSA(Table4,baselinescenario)overthethreeyears2014-16isalmost15percentagepointsofGDPlessthantheamountonewouldexpectbasedonthefundamentaldriversofexternaldebt,namely,thenon-interestCAD(NICAD),netFDIchangeandendogenousdebtdynamics(whichcapturetheeffectsofinterestrates,exchangeratesandgrowthrates).

Inotherwords,hadwestartedwiththeexternaldebt-to-GDPratioof26%attheendof2013andaddedthe“Identifiednetdebt-creatingflows”overthenextthreeyearsshowninTable4oftheDSA,wewouldhaveendedupwithanexternaldebt-to-GDPratioof53.4%attheendof2016insteadofthemoresoothing39%showninthetable.Thislargediscrepancyrequiresanexplanation:eithertheNICADandothervariablesdeterminingthepathofexternaldebtarebeingoverstatedordebtisbeingaccumulatedoff-balancesheet.6Yetthissignificantdiscrepancy,whichshowsupasaseriesoflargenegativenumbersinthe“Residual”lineofTable4doesnotmeritanymentionintheDSAapartfromagenericfootnotetoTable4.Thisexampleillustratestheriskthatpopulatingtemplatesoftenbecomesamechanicalexercise.Onemustask:wouldEthiopia’sriskratinghavebeenloweredhaditsexternaldebtratiobeen53%in2016insteadof39%?

Ethiopia

5QuotesfromJuly2017ArtIVReport(IMFCountryReport17/217).6Notethatiftheactualincreaseinexternaldebtisgreaterthantheamountidentifiedbytheprecedingfactorsthatcouldinprinciplebeexplainedbyfront-loadedborrowing;buttheconverseispuzzling.

Page 10: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

10

InJanuary2018,EthiopiawasdowngradedtohighriskofdebtdistressbasedontheLICDSF.Theverynextmonth,Moody’saffirmeditsB1(B+equivalent)ratingwithastableoutlook.ThisissimilartotheratingsforKenyaandRwanda,bothofwhichareratedatlowriskofdebtdistressbytheLICDSF.7

TheproximatereasonforEthiopia’sdowngradebytheIMFandWBisthattheratioofPPGexternaldebtservicetoexports(DSE)wentaboveitsthresholdinthemostrecentbaselineprojectionscenariocontainedinIMFCountryReport18/18releasedinJanuary2018(IMF2018b).ThereasonfortheDSEbreachisonaccountofmaturingprincipalonnon-concessionalloans,withthematurityreductionofoneparticularbilateraldepositatthecentralbankfrom8to6yearsbeingpivotal.8Inaddition,foreignexchangereservesfellandtherewere"widespreadforeignexchangeshortages",inotherwords,internationalliquidityisanissue.

NowconsiderwhathappenswhendatafromTable2(selectedindicators)andTable5b(balanceofpayments)inIMF(2018b)arecombined.Forexample,in2016/17,thetradedeficitwas16.1%ofGDPofwhich13.1pporover80%wasonaccountofthepublicsectordeficit(differencebetweeninvestmentandsaving)whileonly3.1ppwereonaccountoftheprivatesector.Moreover,netprivatetransferswere6.9%ofGDPwhileofficialtransferswerejust1.8%,meaningthatthe"public"componentofthecurrentaccountbalancewas-11.3%ofGDPpartiallyoffsetbytheprivatecomponentof+3.8%ofGDP,resultinginanoverallbalanceof-8.1%ofGDP(thatis,aCADof8.1%ofGDP,includingadeficitof0.6%ofGDPonnetincomefromabroad,seeTable5b).Thus,thepublicsectordeficit,itselfdrivenbylargeinfrastructureinvestments,drivestheCADandexternaldebt.Therefore,ifEthiopiaisathighriskofexternaldebtdistress,thefundamentalcausemustlieinthepublicfinances.Thismeansthatpublicdebtanditsdynamicsneedtobegivenacentralpositioninanyassessmentofdebtdistress.

ButtheDSAinIMF(2018b)hasnothingtosayaboutpublicdebtexceptthatitstaysbelowitsthresholdanddoesnot"flagadditionalrisks".Indeed,twocrucialquestionsforEthiopia,butunlikelytobeaddressedevenbythe2017DSF,arethefollowing:

1. Whatwouldpublicdebtdynamicslooklikeintheabsenceoffinancialrepression(issuingdomesticdebtatnegativerealinterestratescombinedwithcapitalcontrols)andrealexchangerateovervaluation?Boththesecouldmaskunsustainablepublicdebtdynamicsbykeepinginterestratesartificiallylowandunderstatingtheburdenofforeigncurrencydebt.

2. Willthebigpublicinvestmentspayoffbywillboostingfuturegrowthandtaxesandtherebyensuringsolvencyviaadequatefutureprimaryfiscalsurpluses?

BothquestionsarevitalbutunlikelytobepickedupwithoutanewframeworkforDSAasdiscussedinsectionsVandVI,wheretheEthiopiancaseisexploredfurther.

III.REFORMSEMBEDDEDINTHE2017DSF

TheLICDSFwasintroducedin2005andhasbeenreviewedthricepriortothe2017Review,mostrecentlyin2012,whencountry-specificinformationanddomesticdebtvulnerabilities

7Theratingagenciesarenotwithouttheirowncredibilityproblems.Butthediscrepancyisnoteworthy.8Seefootnote7intheDSAinIMF(2018).

Page 11: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

11

werehighlighted.The2017Review(thefourth)ismotivatedbythefollowingbroadconsiderations:9

1. Needtopaymoreattentiontomarket-basedfinancingrisksfromdomesticandforeign(Eurobond)sources(mentionedthroughoutIMF2017a)

2. Recognitionthatdebtprojectionshavetendedtounderestimatedebtburdensforbothexternalandtotalpublicdebtwithunexpectedchangesindebtratiosdrivenprimarilyby“fiscaldeviationsandbalanceofpayments(BoP)shocks”(IMF2017ap.9)

3. Inlikevein,andcontinuingtoemphasizetheprimacyoffiscalandpublicfinanceshocks,IMF(2017ap.11)notesthatshocksfromtheprimaryfiscaldeficittendtobeunderestimated.

4. TheDSFhasnotperformedwellinidentifyingXDDepisodesoutofsample.IMF(2017ap.18)notesthatperformancedeterioratessharplywhen7extrayearsofdataareincluded.

5. Staffjudgmenthasbeenappliedinonlyabout25%oftheDSAssincetheinceptionoftheDSFaccordingtoIMF(2017ap15).Moreover,thenewfeatureinthe2012reviewtoallowforadeeperassessmentoftotalpublicdebtanditsimpactontheoverallriskratinghasbeen“onlysporadicallyused”.Theseneedtochangetoallowfordeeper“attemptstodisentanglerisks”.

Mainreforms

Themain2017reformscanbegaugedfromacomparisonofFigure6onpage21withthefigureinAppendix1ofIMF(2017a).Theseinclude:

1. Classifyingcountriesasweak,mediumorstrongbasedonanewcompositeindicatorCIinsteadofexclusivelyontheCPIA

2. Eliminatingthe“probabilityapproach”tosettingdebtthresholds(theprobabilityapproachreferstousingcountryspecificinformationwhencountriesareontheedgeofbreachingadebtburdenthresholdtobetterinformtheriskrating)

3. Eliminatingremittance-augmentedthresholds104. Introducing“realismtools”tocrosschecktherobustnessofmacroeconomicprojections

andbaselinestresstests.

Anotherchangeistoshortentheprojectionhorizonfrom20yearsto10yearsforthemechanicaldeterminationofriskratings,asmostriskthresholdbreachestypicallyhappeninthefirstfiveyears(seep.31ofIMF2017a).

ImplementationofrevisedDSF

DrawingmainlyuponAnnexIIItoIMF(2017a),themainstepsinimplementingthe2017DSFareasfollows:

Step1:IdentifyaseriesofXDDepisodesovertheperiod1970-2015.98episodeswereidentifiedbasedonthenewmethodologyshownininFigureAIII.1(versus76undertheoldmethodology).AstrikingfeatureofthemethodologyisitsIMF-centricnature.

9Theinterpretationandassessmentofthe2017ReviewcontainedinthispaperareinformedbytheADFLab’sownexplorationofDSoverthe18-monthperiodNovember2015toJune2017.10ThischangeiswelcomeonconceptualgroundsbecauseremittancesmayincreaseFXreservesbuttheydonotaugmentthegovernment’scapacitytoserviceitsdebtunlessremittanceflowscanbetaxed.Wereturntothispointlater.

Page 12: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

12

Step2:Runaprobitmodelforeachof4debtburdenindicators.TheseindicatorsallpertaintoPPGXDandincludePVofXDtoGDPorexportsanddebtservice(interestplusmaturingprincipalonPPGXD)torevenueorexports.Theprobitmodel(whichestimatestheprobabilityofXDDforaparticularcountryinaparticularyear)usesthefollowingcontrolorexplanatoryvariables:

1. Debtburdenindicator2. CPIA3. Countrygrowthrate4. Reserves(scaledbyimports)5. Reservessquared6. Remittances(scaledbynominalGDP)7. Worldgrowth(proxyforexternalshocks)

Adisappointingfeatureistheabsenceofdiscussionofalternativesthatmighthavebeenpursued,forexample,variablespertainingtounsustainablepublicdebtdynamicsthatfeedintoXDD.Also,therationaleforusingremittancesisnotclear.IMF(2017a)claimsthisenhancesrepaymentcapacity.Remittancesenhanceinternationalliquidity,butunlesstheycanbetaxedtheydonotincreasethegovernment’srepaymentcapacity.Likewise,thereisnodiscussionofendogeneity.IstheCPIAhighbecauseXDDhasbeenavoidedorhasXDDbeenavoidedbecausetheCPIAishigh?Similarly,theinclusionofworldgrowthisnotclearlyjustified.Whynottermsoftradeshocks?

Step3:Classifycountriesasweak,strongormediumbasednotsimplyontheCPIA(asearlier)butonacompositeindicatorCIbasedonvariables2to7inthelistinStep2andusingthecoefficientforeachvariableaveragedoverthefourprobitregressions(oneforeachofthefourdebtburdenindicators).TheclassificationisbasedonevenlyspacedpercentilesfortheCI.Notethatitisnotclearwhyacountrywithhigherremittancesshouldbeclassifiedas“strong”ifremittancesresultfromadiasporathathasleftthecountrybecauseoflackofopportunitiesthere.ThevirtueoftheCPIAisthatitspolicyandinstitutionalimplicationsareeasytograsp.Thisisapuzzling“reform”.

Step4:Obtainathresholdforeachdebtburdenindicatorbyfirstchoosingacutoffprobabilityofdebtdistress.ThiscutoffprobabilityischosentostrikeadesiredbalancebetweenType1(missedcrises)andType2(falsealarms)errors.TheninverttheestimatedprobitregressionequationtoobtainthedebtburdenthresholdcorrespondingtothecutoffprobabilityasafunctionoftheCI(asshownonp26ofAnnexIII)andvaluethesethresholdsatthesamepercentilesfortheCIwhichdefinethecountryclassificationsoastoobtainanumericaldebtthresholdforeachcountryclassification.

Step5:DefineapredictionruleforpredictingXDDbasedonabreachofthedebtburdenthresholds(the2017DSFwouldsignalhighriskofXDDifanyofthethresholdsisbreached).

Step6:ThissteppertainstosettingbenchmarksforthePVoftotalpublicdebt(TPD=PVofPPGexternaldebtplusdomesticpublicdebt.Inpractice,“publicdebt”forADFcountriesusuallyincludesjustthedebtofthecentralgovernment;thedebtofSOEstendstobeexcluded.)Aruleisdefinedfordomesticdebtdefaultepisodes,explicit(rare)orimplicit(forexample,basedonhighinflationorfinancialrepression,morecommon).TheresultsareshownTableAIII.7,butitisnotclearwhichofthe18definitionsofdebtdistressiseventuallyused.SufficeittosaythatthresholdsforTPDaregraftedontothoseforthefourXDthresholdstoensureconsistencyand

Page 13: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

13

obtainbenchmarksforTPDforeachofthethreecountryclassifications;theapproachisdescribedinsectionDofAnnexIII.

Tosumup:The2017Reviewcouldhavetakenabolderapproachinsteadofcontinuingtobeconstrainedbyexternaldebtdistress(XDD),inparticular,byengaginginadeeperdiscussionoftheunderlyingcausesofsuchdistress,whicharelikelytoemanatefromfiscalimbalancesandTPD.Inthiscase,emphasizingpublicdebtasbothafundamentalcauseofmacroeconomicdistressandthefocusofcorrectivepolicybecomeslogical.

IV.CRITIQUEOFTHE2017DSF

Partofthecritiquehasbeenintegratedintothedescriptionofthe2017DSFinsectionIII.ItisIMF-centricinitsdeterminationofexternaldebtdistressepisodesatatimewhenmarketdebthasbecomeimportantandremainsanchoredintheoriginalframeworkofXDD.Tweakingthe2012DSFtoobtainanupdatedversionthatfitsadditionalhistoricaldatabetterdoesnottellushowthe2017DSFwillperformoutofsample.Forthis,wewillneedtowaitanother10years.

Acripplingdrawbackofthe2017DSFisthatitdoesnotaccordsufficientimportancetopublicdebtasthefundamentaldriverofCADsandexternaldebtdynamics.Thisalonewouldmaketheimplementationofthe2017DSFcounterproductive.Further,theadditionalsophisticationintroducedcoulddilutetheincentivesfortheuseofstaffjudgmentanddiscretion,whicharevitallyimportantgiventhedeterioratingdebtsituationamongAfricanLIDCs.

WhilethestatementinIMF(2018a)thattheLICDSF“hasbeenthecornerstoneofassessmentsofriskstodebtsustainabilityinLICs”mayhavebeentrueofthepast,theclaimthatthe2017DSFisa“significantoverhaul…bringingsignificantadvantagestoallstakeholders”fallsshort.

Thefirstproblemisthatthe2017DSFworkswithintheconfinesoftheinheritedDSF,whichbyitsnatureprecludesasignificantoverhaul.The2017ReviewneededtorevisitthehistorybehindtheintroductionoftheLICDSFin2004-5andaskwhethertheoriginalreasonsstillapplied,alongthelinesofazero-basedbudgetingapproach.ItneededtoaskwhatthefundamentalcausesofXDDepisodesare.TheseareunlikelytobelowremittancesorFXreservesorworldgrowth,variablesincludedintheprobitregressionsdiscussedabove.Thefollowinghypothesiswouldhavebeenworthinvestigating:thevastmajorityofXDDepisodesamongAfricanLIDCsaredrivenbyunsustainablepublicfinances.

ThereisnomoreeloquenttestimonysupportingthishypothesisthantheexecutivesummaryoftheIMF’sownOctober2017RegionalEconomicOutlookreportonSub-SaharanAfrica(IMF2017b):“Thenumberoflow-incomecountriesindebtdistressorfacinghighriskofdebtdistressincreasedfrom7in2013to12in2016,andalloftheregion’sfrontiermarketsorothercountrieswithcreditratings,exceptNamibia,havebeendowngradedbelowinvestmentgrade.Thedebtincreasehasbeendrivenbyawideninginfiscaldeficits,slowgrowth,theslumpincommodityprices,andexchangeratedepreciationsinsomecountries.Whilecurrentaccountshaveimprovedandexchangemarketpressureseasedsomewhat,internationalreservesarebelowadequacylevelsinmanycountries.”

TheideathatXDDistypicallydrivenbyfiscalimbalancesandunsustainablepublicdebtissupportedbyEMexperience,wheremostofthesovereigndebtcriseshavebeencausedeitherbyunsustainablepublicdebttrajectoriesleadingtoeventualdefaultsonexternaldebt(andpossiblyevenondomesticdebt)orbybailoutsoftheprivatesector(seeforexample,Pinto2014

Page 14: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

14

chapter8).ItisalsosupportedbythecountrycasestudiesinAnnexIIofIMF(2017a).InthethreeexamplesgivenofmissedcrisesinGhana,MalawiandMozambique,primaryfiscaldeficits,publicdebt,terms-of-tradeshocks(tobaccopricesforMalawi)andoptimisticgrowthforecastsallplaymajorroles.

Similarly,correctivepolicyisalmostalwayslikelytohingeonthepublicfinancesandbettergrowthpolicy.ThisislikelytobethecaseevenwhentheimmediatecauseoftheXDDepisodeispurelyexogenous,suchasaToTshock.Ontheotherhand,apartfromgrowthandtheCPIA,thepolicyimplicationsoftheothercontrolvariablesintheprobitmodelinsectionIIIarehardtofathom.

Second,IMF(2017a)couldhavebenefitedfromamoresystematiccost-benefitanalysisofimplementingtheDSF.Onegetstheimpressionthatperformancewasnotgreat;butthiscouldhavebeenbetterdocumented.Besides,onewouldhavelikedtoknowwhattheaveragecostofproducingaDSAisintermsofstaffweeksandcosts.Hasitbeenworththeeffort?ThesequestionsareimportantbecausetheLICDSFhasbeenonacoursetowardsobsolescenceeversincethecompletionoftheHIPC-MDRIprocesssome10yearsagoforthreereasons:(i)thecontinuedemphasisonPPGXDtotheneglectoftotalpublicdebt;(ii)theslowrecognitionthat,afterHIPC-MDRIdebtrelief,governmentshavebeenrelyingonheavilyonmarketfinance(notjustEurobondsbutalsodomesticdebt,thelatterpossiblyapplyingtothefragilecountriesaswell);and(iii)thecontinuedrelianceonCPIA-linkeddebtburdenthresholdswithoutpayingattentiontomarketsignalsondefaultanddevaluationrisk.

Third,the2017DSFfailstoaddresstheshortcomingsidentifiedinsectionIandconveystheimpressionthattheDSFisanendinitself.Thiscanbedemonstratedintwoways.

• ThedisconnectbetweentheriskratingsforXDDandmarketassessmentsofrisk.Forexample,Rwanda,atopperformerbasedontheCPIAandwhichcarriesalowriskofXDD,hasonlyaBsovereigncreditratingfromS&P,5levelsbelowinvestmentgradeonaccountofitsgrowingCADnecessitatedbyimportsformeetinginfrastructureandotherdevelopmentneeds.

• Analmostexclusiveemphasisoncrisisavoidance,asillustratedbythediscussiononpage17ofIMF(2017a)andespeciallyfootnote10onthedistresssignalsused,whichareevidentlygearedtotheinterestsofexternalofficialcreditors.

Fourth,theorientationoftheDSFshouldberadicallyalteredtomakeitmoredevelopment-relevant.

• Thefirstchangeshouldbetomakepublicdebtanditsdynamicsthecentralfocus.ThereasonisthatmostoftheXDDriskprobablyemanatesfromthepublicsector(governmentplusSOEs)forADFcountries.Therecouldbeamuchshorterhorizon,say5years,fordiscussingpublicdebtdynamicssupplementedwithadiscussionofCADsandFXRA.Atthesametime,risksfromtheprivatesector’sbalancesheetshouldbemanagedandmonitoredbyrestrictingexternalprivateborrowingaswellascurrencymismatchesonbanks’balancesheets.ThisneedstobedoneproactivelybasedonEMexperience:preventionofdebtcrisesisfarlesscostlythanthecure.11

• ThesecondchangeshouldenhancetheDSFasaninputintoapolicydialoguefordevelopment,asopposedtomerecrisisavoidanceoraninputintotheDLPoftheIMFandNCBPoftheWorldBank.Adistinctionneedstobemadebetweennaturalresource

11AizenmanandPinto(2013).

Page 15: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

15

exporting(NRX)countriesandtherestowingtotheexistenceoftheNRcurseinSSA,asarguedinFP2017.ThisisamajortopicgiventheprevalenceofhydrocarbonsaswellasotherexhaustibleNRsuchascopperandironoreinSSA,buthastendedtobesoft-pedaledinrecentyears.NRXcountriesshouldfaceamuchhigherbarforaccesstodevelopmentfinancetocreateanincentiveforthemtousetheirownwealthwell.Indeed,asshowninFP2017,countriesrichinNRtendtoheadlinegroupsthataremarkedeitherbyunsustainablepublicdebtdynamics(suchasGhana,MozambiqueandZambia)orpovertyandfragility(ofthe9ADFcountriesintheLabstudywithapercapitalessthan$500,FP2017notesthat5areNR-richandall9havepublicdebtsustainabilityproblems).

Onthepositiveside,theemphasisonjudgmentaspartofthe2017DSFreformsiswelcome.Asapracticalmatter,exercisingjudgmentrestsontwopremises:thoroughfamiliaritywiththenuts-and-boltsoftheunderlyingalgebraicequationsofthemodelandtheabilitytomodifyflexiblytohighlightcountry-specificaspects;andexperience,includingathoroughunderstandingofhowaparticularcountry’seconomyworks.Thediscussionatthebottomofpage15ofIMF(2017a)showsthatdiscretionisrarelyused.Unfortunately,templatesliketheDSFtendtobecomesubstitutesforthinkingandthisriskincreasesthemoresophisticatedthetemplateis.

V.ANALTERNATIVEFRAMEWORKFORAFRICANLIDCDEBTSUSTAINABILITY

ThealternativetoDSF2017proposedhereshouldbeseennotasafixedtemplatebutanapproachallowingforcountryspecificitythatrespondsnimblytochangingcircumstances.Inthisspirit,thealternativeapproachrestsonafewprincipleswhicharemotivatedbymacroeconomicbasicsandlessonsfromEMs,andtheparticularcircumstancesdefiningthedebtsituationinAfricatoday:

Principle1:Basedondevelopmenttheory/empiricsandmacroeconomicaccounting101,onecansetoutthefollowing"causalhierarchy":

Weakleadershipandgovernance(lowtransparency,highcorruption)plusweakinstitutions(especiallyfiscalandfinancial)=>unsustainablepublicfinances=>unsustainablespilloversintocurrentaccountdeficitsandexternaldebtdynamics.

Conclusion:TheattentioninDSAsmustshiftfromdebtdistressassociatedwithPPGexternaldebttodistresslinkedtopublicdebtanditsdynamics,correspondingtoashiftfromsymptomtomalady.

Principle2.EvidencefromFP2017andIMF(2018a)indicatesdeep-seatedpublicdebtsustainabilityproblemsamongAfricanLIDCs.Theunderlyingfactorsincludeslowinggrowth,aweakprivateinvestmentclimate,largeprimarydeficitswithbadexpenditurecompositionandexchangeratecollapseslinkedtoweakmacromanagementandpoliticalinstability.ThesefactorsarenotgoingtobealleviatedbyaninsistenceonconcessionalityinODA,whichhasfailedtopreventthere-emergenceofunsustainablepublicdebtinthepostHIPC-MDRIeraortocontributetoastrongfoundationforlong-rungrowth.Moreover,mostAfricancountriesarenatural-resourcerichbuthaveaweaktrackrecordinusingthiswealthtosupportdevelopment.

Conclusion:ConcessionalityisoverratedinthesensethatitwillnotcompensateforthefundamentalproblemsAfricanLIDCsfaceandcoulddilutetheincentivesfordifficultreform.

Page 16: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

16

TheAfricanLIDC-donorrelationshipneedstoberevisitedwithamorefocusedandtougherpolicydialoguewiththeonusshiftingtoAfricanleaders.

Principle3.Debtsustainabilityisnotanendinitselfbutaplatformforthesustainabilityoflong-rungrowthanddevelopment.Eventhebest-governedAfricanLIDCisgoingtofaceconsiderabletensionbetweenpublicdebtandCADsustainabilityontheonehandandthevastinvestmentsneededininfrastructureandhumancapitaltomeettheSDGsontheother.Atthesametime,ODAisdwindlingwiththemarketnowdefiningthemarginalcostofborrowingforasignificantsetofAfricanLIDCs.

Conclusion:ThebiggestchallengeforAfricanLIDCstodayisreconcilingdebtsustainabilityanddevelopment.Thiswouldrequireacombinationof:(i)re-pricingODAtoreflectthenewmarketrealities,(ii)findingmechanismstofrontloadODAwithoutincreasingthefiscalburdenondonors,and(iii)launchinganEnhancedPolicyDialogueconductedcollaborativelyacrosstheAfDB,IMFandWBtomaximizetheimpactofanalyticalandpolicywork.12

SimplifyingthePublicDSA13

Inlinewithprinciples1and2,theDSAshouldconcentrateonthevariablesdrivingpublicdebtdynamicsinamannerthatistransparent.Thismeanshighlightingthethreevariablesdrivingdynamics:theprimaryfiscaldeficit(totalrevenueminusnon-interestspending),theinterestrateonpublicdebtandtheGDPgrowthrate.ThesevariablesdeterminethetrajectoryofthepublicdebttoGDPratio(the“debtratio”).Forexample,ifthegovernmentisrunningaprimarydeficitandinterestratesexceedgrowthrates,thedebtratiowillgrowwithoutlimitintheabsenceofcorrectivefiscalpolicy,eventuallyresultinginacrisis.Similarly,ariseintheratiooftheprimarydeficittoGDP,oranincreaseininterestratesoracollapseingrowthratesorsomecombinationoftheseoutcomeswillworsendebtdynamicsandcouldprecipitateacrisis.

Inaddition,theexchangerateisakeyvariableimpactingpublicdebtdynamicsinADFcountries:theaverageshareofpublicdebtdenominatedinforeigncurrency(“forexdebt”)iscloseto70%.Largecurrencydepreciationssuchaswehavewitnessedoverthepastfewyearsquicklyincreasetheburdenofforexdebtandthedebtratio.Therefore,theframeworkshouldincorporatecalculationsthatenabletheimpactofcurrencyvolatilityonforexdebttobecapturedbyseparatinginterestratesonlocalcurrencyandforexdebt.14

Similarly,debtratioscanbeadverselyaffectedbycontingentliabilitiesfromstate-ownedenterprisesandcommercialbanks.Ifthestate-ownedenergycompany,orafailingbank,needsabailout,thiswilladdtothepublicdebtburden.MoresystematicinformationisneededonSOEdebtsandrelatedsolvencyissues,asemphasizedinIMF(2018a).

Makingdebtdynamicstransparentmeansthefollowing:

ShiftattentiontoNominalPublicDebt:The2017DSFlikeitspredecessorssetsriskthresholdsfortheratioofthePVofpublicdebttoGDP(seeforexampleTableAIII.8inAnnexIIIofIMF2017a).ThePVofpublicdebt(definedbytheIMFasthePVofPPGexternaldebtplusnominaldomesticdebt)isappealingwhenmostdebtisfromofficialsourcesandisconcessional,thatis,contractedatbelow-marketinterestrates.UsingadiscountratehigherthantheofficialinterestratethenenablesagrantelementtobecalculatedasthedifferencebetweenthenominalvalueofthedebtanditsPV.However,forEurobondsorcommerciallocalcurrencydebt,sucha12TherationalefortheserecommendationsisspeltoutindetailinHLP(2017)anddiscussedlaterinthissection.13TheapproachhereisbasedonADFLab(2016)andFP2017.14ThenumericalexamplefromEthiopiabelowwillshowexactlywhatthismeans.

Page 17: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

17

calculationwouldyieldanegativegrantelement(becausetheinterestrateonsuchdebtwouldtypicallybehigherthanthediscountrate).Forthisreason,theIMFincludessuchdebtatitsnominallevel.15Butthiscreatesaninconsistencybecauseitunderstatesthetrueburdenofmarketdebtassumingthat5%isindeedtheappropriatediscountrateandnotarbitrary.Usingnominaldebtavoidsthisproblemandenablesustoeasilyinterpret(r-g),theinterestrate-growthratedifferential.

Levelsversusdynamics:Thelevelofpublicdebtanditsdynamicsbothneedexamination.ThiswouldenableearlywarningbygettingawayfromthebinaryclassificationinherentintheLICDSF(somedegreeofdebtdistressversusnodistressbasedonsomethresholdbeingbreached)basedondebtlevels:onceadebtdistressepisodetakesplace,itisalreadytoolate.Ontheotherhand,lookingatpublicdebtdynamicsoverahistoricalperiodandhowthesearelikelytoevolveoverashortprojectionperiod,say5years,providesampleopportunityfordiscussingcorrectivepolicy.ThisnodoubthappensinthecaseofIMF-supportedprograms.Butitdoesnotflowfromthecore2017DSF.

Concentrateonhistorywithashortprojectionperiod:Thereasonforconcentratingonhistoryisahighdegreeofpathdependence:thenextfewyearsarelikelytobesimilartothepastfew.BasedonEMexperience,weknowthatgrowingoutofadebtproblemisextremelyrare.Instead,countriesusuallyhavetomakeharddecisionstoreduceprimaryfiscaldeficitsandlowercountry(macroeconomicandpolitical)risk.Longlagsmaybeinvolvedbeforereformsareseenascredibleandgrowthresumes.Moreover,seriousdebtdatagapsargueinfavorofsimplerapproachesandshorterprojectionperiods.

ShifttheonusforreformtoAfricancountrygovernments:InsteadofbuildinginoptimisticfiscalconsolidationandgrowthscenariosintoDSAprojectionsthatseldommaterialize—seetheexcellentaccountinMooneyanddeStoyres(2017)—theapproachshouldshifttotakingahardlookathistoryand,whereapplicable,askingcountrygovernmentshowtheyintendtorestoresustainabledebttrajectories.Thisisextremelyimportantincasesofpoortrackrecordsonusingpublicresourceswellandwheretheprobabilityofdifficultreformsbeingimplementedislow.Insuchcases,asimplifiedDSAfollowedbycountryvisitstoascertainthegovernment’sownstrategyislikelytobemoreeffectivethancontinuingtoinsistonconcessionalborrowingthatcoulddiluteincentivesfordifficultreform.AtougherdialogueandhigherbarforaccesstoODAareneeded.

DevelopmentandDebtSustainability

Inspiteofafallinthepercentageofthepopulationlivinginpoverty,50millionmorepeoplewereaddedtothepoorinAfricaduring2000-15owingtorapidpopulationgrowth.In2015,420millionAfricanswerelivinginabsolutepovertybasedontheWorldBank’s2011yardstick,morethanhalftheworld’spoor.Peopleundertheageof25makeup60%ofthecontinent’spopulation.With500millionmorepeopleexpectedby2030andanother250milliongoingintotheworkforce,therestorationoffastgrowthisessentialtoavoidasocio-economicdisaster.Evenifgrowthweretoaverage6%duringthe2020s,thiswouldstillleave400millionpoorpeoplein2030;ifgrowthweretostagnateinthe3-4%range,thenumbersofthepoorwouldswellto500million,puttingthefirstSDGto“eradicateextremepovertyforallpeopleeverywhereby2030”outofreach(HLP2017).

Twoobservationsareparamountinthisconnection.

15SeeforexampleAnnexIIinIMF(2015).

Page 18: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

18

First,ADFcountriesneedlargeamountsoffrontloadeddevelopmentfinancetoplugtheinfrastructuredeficitiftheyaretogrowfastenoughtomeettheSDGs,continuetoalleviatepovertyandcreatejobsforAfrica’sburgeoningyouthpopulaton.UNCTAD(2016)estimatesthattheSDGscouldrequireinvestmentsofbetween$600billionand$1.2trillionperyear.RecentworkbytheAfricanDevelopmentBankindicatesthatinfrastructurealonerequire$130-170billionperyear,leavingafinancinggapofasmuchas$108billion(AfricanEconomicOutlook2018).Theseneedsexistatbothatthenationalandregionallevelstointegrateseveralsmalleconomiestogethergiventhearbitrarycarvingupofthecontinent.Themarketisclearlynotaviablesourceforsuchfinancingonvolume,costandmaturitygrounds,makingthefrontloadingofODAtheonlyremainingalternative.

Box1:Africa’sDevelopmentandDebtSustainabilityChallenges

LessonsfromEMsandtheTransitionCountriesofCentralandEasternEurope:Fast-growingcountriestendtoself-financetheircapitalaccumulationwithhighsavingsratesandrelativelylowcurrentaccountdeficitsandevensurpluses.Inadditiontostrongfiscalandfinancialsectorinstitutions,countrieshaveespousedhardbudgetconstraintsforthepublicandprivatesector,competitionincludingfromimports,andcompetitiverealexchangerates.FollowingthespateofEMcrisesinthelate1990sandearly2000s,thesecountrieshaveself-insuredbystrengtheningpublicsectorbalancesheets,increasingFXreservesandbetterregulationofcapitalflowsandexternalborrowingbytheprivatesector.

ADFcountriesarenotinapositiontoself-financetheirgrowthbecauseoftheirrelativelylowincomesandmassivedevelopmentfinancingneeds.Theirspecialdevelopmentchallengesaresetoutbelowinthreemodules:growth;debtsustainability;anddevelopmentfinance.

Growth:Thisrequiresacombinationofcapitaldeepening(increaseincapital-laborratios)andafastrateoftotalfactorproductivity(TFP)growthtosupporteconomicgrowthandoverallwelfare.Inprinciple,Africashouldbenefitfromahighrateofreturntocapitalgivenitsscarcitythereanditsdistancefromtheglobaltechnologicalfrontier.Herearethespecialchallengesitfaces:

1. Loweringcountryriskthroughbettergovernanceandinstitutions,improvedvolatilitymanagement(especiallyfromdomesticcorruptionandpoliticalshocks)andensuringsustainablepublicfinancestoraiserisk-adjustedreturns

2. Engineeringabigpushoninfrastructureandskills3. Promotingregionalintegrationofinfrastructureandtradegivensmallaveragecountrysizeand

fragmentationoftheinternalmarket.

Debtsustainability:AconstanttensionexistsonthecontinentbetweenpublicdebtandCADsustainabilityandfinancinginfrastructure(nationallyandregionally)andtheSDGs,evenforthebetterperformingcountries.Themarketisunlikelytobeanappropriatefinancingsourceonvolume,costandmaturitygrounds.AstrongcaseexistsforfrontloadingODA,evenonmoderatelyconcessionalterms,whichwillremainfarsuperiortothemarketalternativesnowbeingpursued.Butcountriesmustdemonstratetheirwillingnessandabilitytostrengthenfiscalinstitutionsandpublicfinancemanagementandshowthatresourcescanbeusedwell,includingthosefromNRwealth.

Developmentfinance:ADFcountriesneedlargeamountsoffrontloadeddevelopmentfinancetoplugtheinfrastructuredeficitandmeettheSDGs.Thisisbothatthenationalandregionallevel.DonorscanhelpwithoutincreasingtheirfiscalburdenbyadoptingthetriplepackageofsecuritizingpartoftheannualflowofgrantsviaaBigBond,moderatelyconcessionalloansandanenhancedpolicydialogue(seetextfordetails).SomecountrieswillstillneedaccesstohighlyconcessionalODA,callingforatwo-tierapproachasoutlinedinHLP(2017)andFP2017.ThebiggestchallengeistoachievegreatercollaborationacrossMDBsandbilateraldonors.

Source:HLP(2017),Pinto(2014).

Page 19: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

19

Second,eventhebest-governedcountryhasvastpublicinvestmentneedswithoutwhichitwillbehardifnotimpossibletocreateafoundationforlong-rungrowthanddevelopment.Butborrowingtofinancetheseinvestments,especiallyfromthemarketswiththeirrelativelyhighinterestratesandshortmaturitiesrelativetothelongpaybackperiodsforinfrastructureinvestments,willfurtherexacerbatethealreadyworseningpublicdebtsustainabilityinAfrica.Box1abovesummarizesthenexusofdebtsustainabilityanddevelopmentchallengesinAfrica.

ReconcilingdebtsustainabilityanddevelopmentwithODAdwindlingcallsforharddecisions.First,ODAneedstobepricedandallocatedbetterandthereisclearscopeforthisgiventhefactthatthemarketdefinesthemarginalcostofborrowingforagrowingnumberofADFcountriesandheterogeneityamongcountriesinusingpublicresourceswell.Second,amechanismhastobedesignedtofrontloadODAwithoutincreasingtheburdenondonors.HLP(2017)recommendsthisbedonebyconvertingaportionofthegrantsnowgoingtoAfricaintoa“BigBond”.Forexample,ifapproximately$5billionofthe$45intheannualgrantstoAfricaissetaside,thiscouldsupporttheissuanceofa30-year$100billionbondbydonors:thepresentvalueofanannualstreamof$5billionforthenext30yearsis$100billionwhendiscountedat3%,whichisapproximatelyequaltotheprevailingyieldontheUS30-yearbond.TheproceedsofthisBigBondcouldbedisbursedthroughmoderatelyconcessional40-yearloansat3-4%(MCLs),whichwouldbesuperiortomarketborrowingoncostandmaturityconsiderations.Third,suchfrontloadingwouldneedtobeaccompaniedbyanenhancedpolicydialoguetoensurethatpublicresourcesareusedwell,withaccesstofrontloadedfinancesprovidingthecarrot.Fourth,thefocusoftheexistingsystemonhighlyconcessionalfinanceshouldshifttothemorevulnerableandfragileADFcountries.

Apragmaticwayofproceedingwouldbetoclassifycountriesaccordingtotwocriteria:whetherthemarginalcostofborrowingdeterminedbynon-concessionalborrowing(includingthemarket)orofficialcreditors;andwhetherpublicdebtsustainabilityproblemsmorelikelytobelinkedtopoorresourceuseandweakPFMsystemsormassivepublicinvestmentsininfrastructure,asshowninthematrixbelow.

UnsustainableDebtDynamicsdrivenprimarilyby:

Marginalcostofborrowingdeterminedby:

PoorresourceuseandweakPFM

Largepublicinvestmentsininfrastructure

Market

A

B

Officialcreditors

D

C

GhanaandMozambiqueareexamplesofcountrieswhichwouldbeinquadrantA.ThesecountrieswouldneedtomeetstringentcriteriainordertoqualifyforaccesstoanykindofODA,giventhegrowingscarcityofthelatterandtheinabilityofthesecountriestousepublicresources,includingfromtheirNRwealth,well.CountriesclassifiedasfallingintoquadrantBin

Page 20: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

20

thematrixwouldbeidealcandidatesforfrontloadedODAthroughMCLs:theywouldbenefitfromMCLsrelativetothecostsofmarketfinanceandprovidesomeassurancepublicresourceswillbeusedwell,therebypotentiallyignitingaracetothetopthroughsignalingeffects.AnillustrativelistofcountriesforQuadrantBisEthiopia,Kenya,Rwanda,Senegal,Cameroon,Côted’IvoireandTanzania.Similarly,countriesinquadrantCcouldbecandidatesforsuchfinance:theymaynothaveissuedEurobonds,buttheirmarginalcostofborrowingcouldbedeterminedbydomesticborrowingsorloansfromcommercialbanks.

QuadrantDwouldprobablyincludemostofthefragilecountries.Whileitmayseemlogicaltorestricthighlyconcessionalfinancetothesecountries,theincentiveeffectsmayposeadilemma.TheresultsinFP2017indicatethatthesecountriestypicallyhavethelowestCPIAscoresandareoftenNRrich,suggestingthefragilitycouldbeself-inflicted.NorisitobviousthatthemarginalborrowingcostofcountriesinquadrantDisdeterminedbyofficialcreditors:Chad,whichisclassifiedasADF-onlyandasa“fragileandconflict-affectedstate”,borrowedUSD1.45billionin2014fromGlencoreandfourbankstoberepaidwithcrudeoilshipments.Thisloanhadtoberestructuredin2015whenoilpricesfellasrepaymentswereabsorbingagrowingfractionoffiscalrevenues,asreportedinthisReutersarticle.

Thediscussioninthecontextofthematrixstressestheneedtomoveawayfromaone-size-fits-allDSAtemplateandtothinkofhardeningaccesscriteriafordwindlingODA.TheADFLab’svisitsto5countriesinthecontextofitsdebtsustainabilityworkindicatethatsuch“hardeningofbudgetconstraints”wouldnotcomeasashocktoAfricanLIDCs.ThesentimentwasoftenexpressedthatODAisnotlastforeverandthatwhatisparamountishowpublicresourcesused,notwhetherloansarehighlyormoderatelyconcessional.

ForeachADFcountry,thefollowinginformationwouldneedtobegathered:

1.UnderstandtheSDGanddevelopmentfinanceneedsofthecountrywell:whataretheprioritypublicinvestmentstoreduce(ifnotclose)theinfrastructuregapinpowerandtransport?Istherearosterofproperlyvettedavailableflagshippublicinvestmentprojects?Whataretheneedsforinvestinginhumandevelopmentandskills?Anunsettlingdisconnectistheshortageofbankableinfrastructureprojectsinacontinentwheretheinfrastructuredeficitismassive.

2.Attempttodefinethemarginalcostofgovernmentborrowings,notjustfromEurobondsandthroughdomesticT-bills,butalsousinginformationonbridgeloansfromcommercialbanksandloansfromChina,thetermsofwhichareoftennottransparent.

3.UnderstandcontingentliabilitiesbetteraswellastheroleofSOEsintheprovisionofinfrastructureservices.

4.PreparearosterofkeyregionalinfrastructureprojectsthatcanbefundedbypartoftheproceedsfromtheBigBond,asnotedinHLP(2017).SuchprojectscouldchangethenarrativeaboutAfricabysignalingitisopenforbusinessandattractingtheinterestofinfrastructurecompaniesworldwide.Togetanideaoftheimpact:arecentstoryinIndia’sEconomicTimesnoted“nearly15globalplayers,includingSamsungConstructionandChinaConstruction”arecompetingwithdomesticcompaniestosecureacontractforanINR26,000croreinternational

Page 21: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

21

conventionandexpocenter.16INR26,000croretranslatesto$4billion.Thinkofwhatevena$20billionBigBondcouldachieveforAfrica.

VI.ETHIOPIA:ACONCRETEILLUSTRATIONOFTHEALTERNATIVEAPPROACH

TheEthiopianvignetteinsectionIIdemonstratesconclusivelythathighpublicsectordeficits,linkedtolargepublicinvestmentsininfrastructure,havespilledoverintocurrentaccountdeficitsandexternaldebtdynamics.ThiscallsforadeeperexaminationofEthiopia’spublicdebtdynamics,whichwereexaminedinFP2017usingdatafromtheOctober2016IMFCountryReport16/322,theoneimmediatelyprecedingtheJanuary2018report(IMF2018b)thatdowngradedEthiopiatohighriskofdebtdistress.TheassessmentinFP2017madethreepoints:

1. Publicdebtdynamicsweremaskedbyfinancialrepression(issuingbirrdebtathighlynegativerealinterestrates)andrealovervaluation.Bothimprovetheappearanceofpublicdebtdynamicsbutthesustainabilityofsuchpoliciesisquestionableandinvolvesseriousdistortions(includingmisallocationofinvestment)whilepotentiallyhurtingcompetitivenessandgrowth.

2. Ethiopiaepitomizestheconstanttensionbetweentheneedforlarge,upfrontpublicinfrastructureinvestmentsinADFcountriesandpublicdebtsustainability.Theimmediateeffectoftheinvestmentsistoraisepublicindebtedness(albeitartificiallyloweredinEthiopia’scasethroughfinancialrepressionandrealovervaluationofthebirr)andcurrentaccountdeficits:Ethiopia'sCADin2016/17was8.2%ofGDPinspiteofsignificantremittances.However,thepublicinvestmentswillpayoffonlyoverthelongrun.

3. Thecrucialquestioniswhetherthepublicinvestmentswillindeedpayoff:havetheybeencarefullyselectedanddotheyhavesufficientlyhigheconomicratesofreturntoboostfuturegrowthandtaxestherebyensuringsolvencyviaadequateprimaryfiscalsurplusesinthefuture?Thisvitalquestionneedsananswer.

ThebigquestionthereforeiswhethertheEthiopianpublicsectorisfacingaliquidityorasolvencyproblem.Whileitdoesnotexplicitlyaddressthisquestion,IMF(2018b)appearstothinktheproblemisliquidity,notsolvency.Thiscanbeinferredfromthereasonforthedowngrade:thematurityreductionofoneparticularbilateraldepositatthecentralbankfrom8to6yearstippedtheratioofdebtserviceonPPGexternaldebttoexports(DSE)overitsthreshold,triggeringthedowngrade(seefootnote7intheDSAinIMF2018b).AlsomentionedarefallingFXreservesand"widespreadforeignexchangeshortages".

TheIMFreportisalsosanguineaboutprospectivereturnsonthelargepublicinvestmentsmade.Box1intheDSAnotes:"ThecommencementoftheAddis-Djiboutirailwaylinewillimprovetradelogisticsandreducethetransportationcostofmovinggoodsinandoutofthecountry.Itwilltakeonly10hoursforthenewrailwaytotakegoodsbetweenEthiopiaandDjibouti,asignificantimprovementoverthe3-4daysbytruckcurrently.Further,theHawassaIndustrialParkandsecondphaseoftheBole-LemiIndustrialParkhavestartedoperationsandaresetto

16TheEconomicTimesNovember29,2017.“GlobalfirmslineupforIndia’slargestconventioncentre;Rs26,000-crprojectmaycreate5Ljobs”.

Page 22: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

22

increasemanufacturingexportsandcontributetothediversificationofexports.Investmentsinhydro-power,industrialparks,exportprocessingzones,andpublicpoliciestoencourageFDIandprivateinvestmentinlightmanufacturingindustriesareexpectedtosupportexportgrowthanddiversification."Whileencouraging,thisassessmentcomesacrosslikeabest-casescenario.

Twostepsareinvolvedinthesolvencyversusliquiditydetermination.Thefirstisadeeperdiveintopublicdebtdynamics.Thesecondistoestablishthatthebigpublicinvestmentsthathavebeenmadearelikelytopayoff.

Ethiopia’spublicdebtdynamicsrecast

TobeconfidentthatEthiopiadoesnotfaceasolvencyproblem,onewouldneedtoexaminepublicdebtdynamics.AllthatIMF(2018b)saysaboutthisinthecontextofthedowngradeisthatpublicdebtstaysbelowitsthresholdanddoesnot"flagadditionalrisks".However,therealquestioniswhatpublicdebtdynamicswouldhavelookedlikeintheabsenceoffinancialrepressionandrealovervaluation.ThisquestioncanbeansweredbetterbyfollowingthesimplifiedalternativeDSAapproachreferredtoearlierinthissection.Table[xx]usesexactlythesameinformationasinIMF(2018b)torecastthepublicDSAinawaythatmakesthedynamicsmoreevident.17

Table[xx]:Ethiopia’sPublicDebtDynamics2015-17BasedonIMF(2018b)

Ethiopia

2014/2015 2015/2016 2016/2017

PublicDebt/GDP% 61.20 59.60 53.80 Forexshare% 51.31 54.53 54.65

Primarydeficit/GDP% 9.50 5.90 5.20 NominalGDPgrowth% 22.32 22.32 37.70 CompositeNominalInterestrate% 5.60 5.13 9.51 ofwhich:localcurrency% 4.58 3.49 3.53 ofwhich:Forex% - 6.70 14.50

RealGDPgrowth% 10.30 10.40 8.00 CompositeRealinterestrate% -4.78 -5.11 -14.11

Externaldebt/GDP% 35.10 35.60 31.60 ofwhich:publicshare% 89.46 91.29 93.04

Non-interestcurrentaccountdeficit/GDP% 11.40 9.60 7.70 Source:ComputationsbyauthorbasedondatafromTable3oftheDSAinIMF(2018b). TheformulasfortheimpliedcompositeinterestratesshowninTable[xx],whichincludetheimpactofexchangeratemovementsonthestockofforexdebt,arederivedinAppendix1.Take2016/17asanexample.Thenominalinterestrateonlocalcurrencydebtwas3.5%whilethatonforexdebtwas14.5%,givingacompositenominalinterestrateof9.5%.ThiswasfarlowerthannominalGDPgrowthof37.7%!Thesameappliestothetwoearlieryears,resultinginrealgrowthratesfaraboverealinterestratesforallthreeyears.Asaresultofthismassive

17Appendix2setsoutafewquestionsneedingclarificationwithregardtotheDSAinIMF(2018b).

Page 23: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

23

differential,thepublicdebt-to-GDPratiofellfrom61%in2014/15to54%in2016/17inspitelargeprimaryfiscaldeficits.Debtdynamicsarethereforebeingdrivenbyrapidgrowthconnectedtomassivepublicinfrastructureinvestmentscombinedwithexceptionallylowinterestratesonaccountofcurrencyovervaluationandfinancialrepression.

Asitturnsout,thebirrwasdevaluedby15%inOctober2017(theexchangeratewentfrom23.4177to26.9215birrperUSD,ariseofabout15%,onOct102017accordingtoReuters).Furthermore,accordingtoIMF2018b,thegovernmenthasbeentakingstepssinceOctober2016tocurbborrowingbythegovernmentandSOEs.ThisindicatesconcernsaboutcompetitivenessandtheCADaswellasrestrictionsonfiscalspacegivenalltheborrowingandinvestmentthathavealreadybeenundertaken.Ontheimpactofthedevaluation,Table3oftheDSAinIMF(2018b)showsa11.3%realdepreciationin2017/18comparedtoacumulativerealappreciationofsome17%overthepreviousthreeyears.Thiswilladd3.1percentagepointstothepublicdebt-to-GDPratioin2017/18(incontrasttoareductionof3.1ppin2016/17,aswingofover6ppofGDP).

Onfinancialrepression,thecentralbank,NBE,requiresbankstopurchaseNBEbillsequalto27%oftheirloanissuance(IMF2018bparagraph20)carrying3%interest(inflationisprojectedat12%in2017/18)inordertofundtheDevelopmentbankofEthiopia.TheIMFisarguingforamorecompetitivebankingsystemandamovetowardsmarket-determinedinterestrates,whichareclearstepsintherightdirection(accordingtoReuters,theinterestrateondepositswasraisedfrom5%to7%inOctober2017alongwiththedevaluation).Similarly,amovetowardsaflexible,marketdeterminedexchangeratewouldbeastepintherightdirection,asnotedinIMF(2018b)(accordingtointernetreports,theblackmarketratewas31.80-33birrversustheofficialrateof27.22perUSDinearly2018).

Thismovetowardsmarket-determinedinterestandexchangerates,whichwouldprobablybeneededtoimprovetheinvestmentclimateespeciallyforFDI,couldhaveasignificantimpactonthedebttrajectory.Forexample,supposethecompositerealinterestratewere3%foreachoftheyears2015/16and2016/17--areasonablelevelforaLIC--insteadofthe-5.1%and-14.1%respectivelyshowninTable[xx].ThengiventheactualrealgrowthratesandprimarydeficitsfortheseyearsshowninTable[xx]above,thedebt-to-GDPratiowouldhavebeen65.3%ofGDPattheendof2016/17insteadofthe53.8%showninthetable.Thisisalargedifference.

Microfoundationsforgrowth

Thesecondstepinthesolvencyversusliquiditydeterminationcallsformoreinsightintotheprospectivereturnsonthepublicinvestmentsininfrastructureandthemicrofoundationsforgrowth.Doingsowouldrequireattemptinganswerstothefollowingtypesofquestions:

1. WhatareEthiopia’sremainingprioritypublicinvestmentsininfrastructureandhumancapital?

2. WhatspecificallyaretheFDIandprivatesectorinvestmentsthatareanticipated?Theprivateinvestmentrateisshownasover20%ofGDPinIMF(2018b),whichisrelativelyhigh.Inwhatsectorsisitinvesting?Alsonotethat(Ethiopiaisrankedat161among190countriesontheDoingBusinessIndex).WhatlessonscanEthiopialearnfromothercountriesonthesuccessfactorsforIndustrialParksandEPZs?

3. Whatistheplanformovingtomarketinterestratesandexchangeratesandwhatwilltheimpactbeonpublicdebtanditsdynamics?

Page 24: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

24

4. Isthebirracrediblestoreofvalue(notethatthereisablackmarketfortheUSDandadeeperunderstandingisrequiredofwhethercurrencysubstitutionexistsandwhetherpeopleprefertoholdtheirsavingsoutsidethebankingsystemonaccountofnegativerealinterestrates)?

5. Whatarethekeystructuralreformspending(includinginincreasingcompetition,privatization,pricingofpowerandotherinfrastructureservices)?

6. Isthereastrongframeworkformanagingvolatilityfromexogenoussources(includingdrought)anddomesticsources(includingsocialinstabilityandcorruption)?Inthiscontext,encouragingmoveshavestartedinpursuitofnationalreconciliationaftertheinstallationofanewPrimeMinisterinApril2018.

ThesequestionscouldbepursuedjointlybytheIMF,WBandAfDBinthecontextofanenhancedpolicydialoguewiththeEthiopiangovernment.IfsuchadialogueestablishesthatEthiopiaindeedfacesonlyaliquidityproblem,thenitwouldbeasolidcandidateforthe40-yearmoderatelyconcessionalloansinthecontextoffrontloadingODAthroughtheBigBonddiscussedinsectionV.

VII.CONCLUSIONS

Eventhoughthe2017LICDSFhasbeenapprovedbytheIMF’sBoardwithimplementationsetforthesecondhalfof2018,theargumentsadvancedhereindicateitisobsoleteforthefollowingreasons:

1. AcontinuedfocusonPPGexternaldebtinspiteofgrowingmarketborrowings,domesticallyandexternally,withthemarketnowdefiningthemarginalborrowingcostforagrowingnumberofADFcountrygovernments

2. AtreatmentoftotalpublicdebtthatdoesnotrecognizethecausalflowfromunsustainablepublicfinancestoexternaldebtdistressandsimplytagsondomesticpublicdebttoPPGexternaldebt.Besides,retainingtheconceptofthepresentvalueofdebtisunhelpfulbecauseofthearbitrarydiscountrateof5%,theinconsistentapplicationwithmarketdebtbeingincludedatfacevalueandtheobscuringofdebtdynamics,inparticular,theimpactof(r-g),theinterestrate-growthratedifferential

3. InsufficientprominencegiventomarketsignalsondefaultanddevaluationriskandignoringlessonsfromEMsintheDSF’sprescriptions,forexample,focusingonfinancialengineeringinsteadoffiscalfundamentalsasinthecaseofGhanaandgenerallybeingexcessivelyoptimisticaboutprospectsforfiscalconsolidationandgrowthgoingforward

4. ThetendencytotreatdebtsustainabilityasanendinitselfandtheunsuitabilityoftheLICDSFforreconcilingdebtsustainabilityanddevelopmentatatimewhenpublicdebtproblemshavere-emergedandinvestmentneedsfortheSDGsremainvast.

ThepaperprovidesevidencefromrecentDSAsonGhana,EthiopiaandRwandainsupportoftheprecedingclaims.Itthenrecommendsanalternative,simplerframeworkthatusesthesamedataasintheLICDSFbutstartswithadetailedexaminationofpublicdebtdynamicsthatisbasedonnominaldebtandpermitsaneasyinterpretationofthekeyvariablesdrivingdebtdynamics.Thiscouldandshouldbecomplementedbyananalysisofcurrentdeficits,externaldebtandforeignexchangereserveadequacy.Thefocuswouldbepredominantlyonhistorywithashortprojectionperiodandthegoalofidentifyingkeyreformstobediscussedwiththegovernment.Thiscouldbedoneinthecontextofanenhancedpolicydialogue,recognizingthat

Page 25: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

25

theonusforpositivechangerestswiththegovernmentandcannotbedictatedbythedonors,withahigherbarbeingsetforaccesstodwindlingODA.

ThisalternativeapproachisillustratedwithEthiopia’srecentdowngradetohighriskofexternaldebtdistressinJanuary2018andthelogicalquestionsthatthisraiseswithpublicdebtconsideredasnotposinganyriskaccordingtoIMF(2018b).Thealternativeapproachwouldtakeahardlookatpublicdebtanditsdynamicsasthedriverofexternaldebtdistress,therebyshiftingattentionfromsymptomtomalady.Thisapproachwillalsobehelpfulinattemptingtoreconciledebtsustainabilityanddevelopment.

InconsideringthealternativeapproachtoDSA,itbecomesabundantlyclearthatthewholeframeworkforODAtoAfricanLIDCsmayhavetoberevisited.Inparticular,reconcilingdebtsustainabilityanddevelopmentinanenvironmentofgrowingdebtvulnerabilityanddeceleratinggrowthwithODAdwindlingcallsforharddecisions.First,ODAneedstobepricedandallocatedbetterandthereisclearscopeforthiswiththemarketdefiningthemarginalcostofborrowingforagrowingnumberofADFcountriesandheterogeneityamongcountriesinusingpublicresourceswell.Second,amechanismisneededforfrontloadingODAwithoutincreasingtheburdenondonors.HLP(2017)recommendsthisbedonebyusingaportionoftheannualgrantsnowgoingtoAfricatosupporttheissuanceofa“BigBond”,asdiscussedinsectionV.Third,frontloadingODAwouldneedtobeaccompaniedbyanenhancedpolicydialoguetoensurethatpublicresourcesareusedwell,withaccesstofrontloadedfinancesprovidingthecarrot.Fourth,highlyconcessionalfinanceshouldbestrictlyrestrictedtothemostvulnerableandfragileADFcountries.

Tosumup,fortheLICDSFtoserveasagenuine“cornerstone”fordebtsustainabilityassessments,anurgentshiftinemphasistowardspublicdebtanditsdynamicsisneeded.Thisisrequiredbothforeconomicreasons(unsustainablepublicfinancesarelikelytobethemainfactorbehindexternaldebtdistressepisodes)anddevelopmentreasons(economicgovernance,whichiscrucialforlong-termdevelopment,isliabletofinditsmostimmediateexpressioninthemanagementofthepublicfinances,includingequitablerevenuemobilizationandtransparentlyvettedpublicinvestments).Besides,massivepublicinvestmentsareneededtomeettheSDGs.Bothreasonsmakeacompellingcasefortheshifttowardspublicdebt.

Inparallel,HLP(2017)hasrecommendationsonattemptingtobetterreconciledebtsustainabilityanddevelopmentwithfront-loadedODAbasedonsecuritizingaportionofannualgrants,moderatelyconcessionalloansandanenhancedpolicydialogue.Therecouldalsobeafocusonfundingpotentialsuccessstoriestoignitearacetothetop,asdiscussedinthecontextofthematrixinsectionV.Thesearesimpleandlogicalideas.ThebigchallengeishowtogetacceptancefortheseideasgiventhecurrentODAarchitecture.AsKeynesnotedintheprefacetohisGeneralTheory,“Thedifficultylies,notinthenewideas,butinescapingfromtheoldones.”

Page 26: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

26

ReferencesADFLab.2016.“DebtsustainabilityimplicationsofhardenedMDBlendingtermsforAfricancountries.”WorkingPaperSeriesNo.1.ADFLabWP1,March.

AfricanDevelopmentBank,OECDandUNDP.2017.AfricanEconomicOutlook2017.

AfricanDevelopmentBank.2018.AfricanEconomicOutlook2018.

Aizenman,JoshuaandBrianPinto.2013.“ManagingFinancialIntegrationandCapitalMobility—PolicyLessonsfromthePastTwoDecades.”ReviewofInternationalEconomics.21(4),636-653.

IMF.2015.StaffGuidanceNoteontheImplementationofPublicDebtLimitsinFund-SupportedPrograms.

IMF.2016.Ghana:ThirdReviewUndertheExtendedCreditFacilityReview.IMFCountryReportNo.16/321.October.

IMF.2017a.ReviewoftheDebtSustainabilityFrameworkforLowIncomeCountries:ProposedReforms.PublishedOctober2,2017.

IMF.2017b.RegionalEconomicOutlook.Sub-SaharanAfrica:FiscalAdjustmentandEconomicDiversification.October.

IMF.2018a.“MacroeconomicDevelopmentsandProspectsinLow-IncomeDevelopingCountries—2018.”IMFPolicyPaper.March.

IMF.2018b.“TheFederalDemocraticRepublicofEthiopia:StaffReportforthe2017ArticleIVConsultation--PressRelease;StaffReport;andStatementbytheExecutiveDirectorforTheFederalDemocraticRepublicofEthiopia.”IMFCountryReportNo.18/18.

Felino,LuisaTeixeiraandBrianPinto.2017.DebtSustainabilityandDevelopmentImplicationsofModeratelyConcessionalLendingTermsforADFCountries.ADFLabWP3,June.

HighLevelPanel.2017.ReinvigoratingAfricanConcessionalFinance.ReportoftheHighLevelPanelonTransformingTrustintheAfDBGroupIntoInfluence.

Mooney,HenryandConstancedeStoyres.2017.“DebtSustainabilityAnalysesforLow-IncomeCountries:AnAssessmentofProjectionPerformance.”IMFWorkingPaperWP/17/220.

Okonjo-Iweala,Ngozi.2012.ReformingtheUnreformable:LessonsfromNigeria.CambridgeMA,andLondon:TheMITPress.

Pinto,Brian.2014.HowDoesMyCountryGrow?EconomicAdviceThroughStory-Telling.OxfordUniversityPress.

UNCTAD.2016.EconomicDevelopmentinAfricaReport2016.DebtDynamicsandDevelopmentFinanceinAfrica.

WorldBank.2017.CPIAAfrica.AssessingAfrica’sPoliciesandInstitutions.July.

Page 27: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

27

Appendix1:ComputationofInterestRatesonPublicDebtThisannexcontainsthederivationsforthethreenominalinterestratesincludedinthedebtsustainabilitycomponentofthecountrytablesinFelinoandPinto(2017).Theseare(i)thecompositenominalinterestrate,(ii)theinterestrateondomestic(localcurrency)debtand(iii)theinterestrateonforex(foreigncurrency)debt.Asthenamesuggests,thecompositenominalinterestrateonpublicdebtisaweightedaverageoftheothertworates,afterconvertingtheinterestrateonforexdebtintoalocalcurrency-equivalentusingexpostinterestparity,thatis,bytakingintoaccounttheimpactofnominalexchangeratemovements.Adepreciationmeansacapitallossonforexdebt,increasingitsburdeninlocalcurrencyterms,whileanappreciationmeansacapitalgain,loweringtheburdenofforexdebtinlocalcurrencyterms.Theweightsusedarethesharesofdomesticandforexdebtattheendofthepreviousyear.18

Theprecedingdescriptionsuggeststhattheinterestratesondomesticandthelocalcurrency-equivalentinterestrateonforexdebtneedtobecomputedbeforethecompositerate.Todoso,onewouldneedtheprecisecurrencycompositionofforexdebt:howmuchinUSD,howmuchinEUR,JPYetcetera.Thiscurrencybreakdownistypicallynotavailable.Theproceduredevelopedforthisreportsidestepsthisproblembyreversingtheorder:itfirstcalculatesthecompositeinterestrateonpublicdebtandthenobtainsthelocalcurrency-equivalentrateonforexdebtusingtheinformationprovidedontheinterestrateondomesticdebtandtheweightsofforexandlocalcurrencydebtintotalpublicdebt.AllthedatacomefromthejointIMF-WorldBankDSAforeachcountry.

CompositeNominalInterestRateonPublicDebt

TheIMF-WorldBankPublicDebtSustainabilityAnalysistablehas,foreachyear,the"identifieddebt-creatingflows"duringthatyear.Thishastwocomponents:theprimarydeficit;andtheso-called"AutomaticDebtDynamics"orADD.Together,theprimarydeficitplusADDcapturetheimpactofthefiscaldeficit(primarydeficitplusinterestpaymentsondebt),GDPgrowthandtheimpactofexchangeratemovementsontheforexcomponentofthepublicdebt.

Byitself,ADDcapturestheimpactofinterestrates,exchangeratesandgrowthratesondebtdynamics.IntheIMF’spublicdebttable,ADDisgivenbytheequation:

ADD=Contributionfrominterestrate/growthratedifferential+Contributionfromrealexchangeratedepreciation.

Inturn,thefirsttermontherighthandsideoftheaboveequationisgivenby:

Contributionfrominterestrate/growthratedifferential=contributionfromaveragerealinterestrate+contributionfromrealGDPgrowth.19

Indiscretetime,thedecomposition,intovariouscomponents,oftheincreasesinthedebt-to-GDPratiofromyeartoyearisgivenbytheequation:

(1) 𝑑! − 𝑑!!! = 𝑝𝑑! +(!!!!!)(!!!!)

𝑑!!!,where:

ddenotesthepublicdebt-to-GDPratio,pdistheratiooftheprimaryfiscaldeficittoGDP,risthecompositerealinterestondomesticandforexdebt,includingtheimpactofrealexchangeratechanges,gistherealgrowthrateandtdenotestheyear.20Intermsofthediscussionabove,thesecondtermontherighthandsideofequation(1)equalsADD.

18Forthetechnicallyinclined:thederivationsfollowfromthediscrete-timedifferenceequationforpublicdebtexpressedasaratioofGDP.Thechangeinthedebt-to-GDPratiorelativetothepreviousyearisdeterminedbytheprimarydeficit,theGDPgrowthrate,interestratesandexchangeratemovementsduringtheyearinquestion.Theinterestratecalculationimplicitlyassumesone-yeardebtasasimplification.19NotethatcontributionfromrealGDPgrowthisgivenby−[𝑔/(1 + 𝑔)]𝑑!!!,wheregisrealGDPgrowthinyeartanddt-1isthepublicdebt-to-GDPratioattheendofyear(t-1).20Foracompletederivation,seeAnnex2inPinto(2014).

Page 28: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

28

Nowtherealgrowthrate,g,aswellasinflationmeasuredbytheGDPdeflator,𝜋,aregiveninthe“Keymacroeconomicandfiscalassumptions”atthebottomoftheIMF-WB’sDSAtableforpublicdebt.Thiscanbeusedtocomputethenominalgrowthrate,G,givenbytheequation: 1 + 𝐺 = 1 + 𝑔 1 + 𝜋 .

Thenextstepistousethefactthatthesecondtermontherighthandside(whichequalsADD)canalsobewrittenas:

(2) (!!!!!)(!!!!)

𝑑!!! = 𝐴𝐷𝐷.

Inequation(2),𝑖!isthecompositenominalinterestrateondomesticandforexdebtandalso,byconstruction,capturestheimpactofexchangeratemovementsontheforexcomponentofpublicdebt.Theformulausedforthevariouscountrytablesinthisreporttocomputethecompositenominalinterestrateisobtainedbyrearrangingequation(2):

(3) 𝑖! = 𝐺! +!""!!!!

(1 + 𝐺!).

NominalInterestRatesonLocalCurrencyandForexDebt

The“Keymacroeconomicandfiscalassumptions”atthebottomoftheIMF-WB’sDSAtableforpublicdebtalsoincludetherealinterestrateondomesticdebt.Letusdenotethisas𝑟!.Thenitsnominalequivalentisgivenbytheequation:

(4) 𝑖! = 1 + 𝑟! 1 + 𝜋 − 1,where𝜋isinflationonceagainmeasuredbytheGDPdeflator.21

Thefinalstepistocalculatethelocalcurrency-equivalentoftheinterestrateonforexdebt,whichwedenote𝑖!.Thisisgivenimplicitlybytheformula:

(5) 𝑖 = 𝑤𝑖! + 1 − 𝑤 𝑖!.

In(5),𝑖isthecompositenominalinterestrateonpublicdebtcomputedinaccordancewith(3)andwistheweightoflocalcurrency(domestic)debtintotalpublicdebtattheendofthepreviousyear.Theonlyunknownis𝑖!,anditcanbesolvedforusing(5).22

CompositeRealInterestRateonPublicDebt

Thecompositerealinterestrate,r,isgivenimplicitlybytheequation:

(6) 1 + 𝑟 = (1 + 𝑖)/(1 + 𝜋),where:

𝑖isthecompositenominalinterestrategivenbyequation(3)and𝜋isinflationmeasuredbytheGDPdeflator.

21ThereasonforusingtheGDPdeflatortoobtainthenominalinterestrateisthatwearelookingattheratioofdebt-to-GDPandthesamepricethatisusedtoconvertrealgrowthintonominalgrowthshouldbeusedtoobtainthenominalinterestrate.22Notethat𝑖!isthelocalcurrency-equivalentinterestrateonforexdebt.Togiveanexample:supposeallforexdebtisinUSDandcarriesa(dollar)interestrateof2%.Inotherwords,𝑖$ = 0.02.Nowsupposethenominalexchangerate,expressedasthelocalcurrencypriceoftheUSD,goesupby10%duringtheyear.Then𝑖!wouldbe12%takingintoaccountthecurrencydepreciation(itwouldactuallybeslightlyhighergiventhediscretetimesettinginwhichtheformulasinthisannexhavebeenpresented).

Page 29: The 2017 Version of the IMF and World Bank’s LIC …...2018/06/01  · MDRI debt relief: most public debt was then concessional PPG external debt from official sources, thereby also

29

Appendix2:AfewquestionsontheEthiopianDSAinIMF(2018b)BeforeproceedingtothequestionsontheDSAinIMF(2018b)below,theseriousdiscrepanciesindebtdatarelativetothepreviouscountryreportforEthiopia,IMFCountryReport16/322,areworthnoting.Forexample,thepublicdebt-to-GDPratiofor2014/15isshownas61.2%inIMF(2018b)whereasitwas50.7%ininthepreviouscountryreport.ThisillustratesthedatachallengesonpublicdebtinADFcountries.

(i)MoreclarityisneededonthenumbersinTable3oftheDSAforpublicdebtinIMF(2018b):

(a)Table3showspublicdebt-to-GDPat53.8%in2016/17butparagraph1oftheDSAputsthenumberat57%.

(b)Table3showstheprimarydeficitat5.2%ofGDPfor2016/17.Texttable1showsthefiscaldeficitat3%ofGDPfor2016/17whileTable3bonpage22showsinterestpaymentsat0.5%ofGDP.Thismeansaprimarydeficitof2.5%forthegeneralgovernment.Whatdoesthebalanceof2.7%ofGDPrepresent?AretheselossesofSOEs?HowaretheSOEsperforming(apartfromEthiopianAirlines)?HoweffectivehavethestepstocurbSOEborrowingbeen?

(c)Howshouldoneinterprettheresidualof7.5%ofGDPfor2014/15inTable3?

(d)Doestherealdepreciationof11.3%for2017/18inTable3(under"Keymacroeconomicandfiscalassumptions")incorporatetheeffectoftheOctober201715%devaluation?

(ii)WhatexactlyisthebilateraldepositatNBEreferredtoinparagraph12andfootnote7oftheDSA?Whatisthesizeandwhataretheterms?

(iii)ThenumbersinTextTable1oftheDSAdon'tseemtomesh.Forexample,iftheratioofexportstoGDPgoesupfrom8.3%in2017/18to9.1%in2018/19andGDPgrowsat8.5%in2018/19,thiswouldrequireexportstogrowatover18%;butthenumbershownforexportgrowthis13.4%.Perhapsdifferentdeflatorsarebeingused?