The-European-Sovereign-Debt-Crisis-and

PIGSorLambs?TheEuropeanSovereignDebtCrisisandtheRoleofRatingAgenciesManfredGärtner&BjörnGriesbach&FlorianJungPublishedonline:11June2011#InternationalAtlanticEconomicSociety2011AbstractThispaperaskswhetherratingagenciesplayedapassiveroleorwereanactivedrivingforceduringEurope’ssovereigndebtcrisis.Weaddressthisbyestimatingrelationshipsbetweensovereigndebtratingsandmacroeconomicandstructuralvariables.Wethenusetheseequationstodecomposeactualratingsintosystematicandarbitrarycomponentsthatarenotexplainedbypreviouslyobservedproceduresofratingagencies.Finally,wecheckwhethersystematic,aswellasarbitrary,partsofcreditratingsaffectcreditspreads.Wefindthatbothdoaffectcreditspreads,whichopensthepossibilitythatarbitraryratingdowngradestriggerprocessesofself-fulfillingpropheciesthatmaydriveevenrelativelyhealthycountriestowardsdefault.KeywordsRatingagencies.Sovereigndebt.Creditrisk.Eurozone.Paneldata.DebtcrisisJELG24.H63.F34IntroductionImportantlessonslearnedfromtherecentglobalfinancialcrisisarethatthejudgmentofprivateratingagenciescanhaveahugeimpactonmacroeconomicoutcomes—andthatitcanbeutterlymistaken.1GiventhesepastfailingsconcerningIntAdvEconRes(2011)17:288–299DOI10.1007/s11294-011-9302-71SeeforexampleCrotty(2009)orGoodhart(2008).WearegratefultoAndreasKleiner,SimonKnausandGiovanniMellaceformanyhelpfulcommentsandsuggestions.M.Gärtner(*):B.Griesbach:F.JungInstituteofEconomics,SchoolofEconomicsandPoliticalScience,UniversityofSt.Gallen,Bodanstr.1,CH-9000St.Gallen,Switzerlande-mail:[e-mail]-mail:[e-mail]-mail:[e-mail]mortgageloans,itwouldbesurprisingifmarketparticipantsagainrelyonthesameratingagencieswhenassessingthedefaultrisksofgovernmentsinthecurrentEuropeansovereigndebtcrisis.Itcouldevenbecataclysmicifthesesovereigndebtratingsweredrivinggovernmentbondyieldsirrespectiveofthedevelopmentoftheunderlyingeconomicfundamentals.Thiswouldputthefateofentirenationsintothehandsofprivateagenciesbecausebadratings,whicharenotinlinewitheconomicfundamentals,couldbejustifiedexpostviaself-fulfillingprophecies.Then,eveninnocentlambscouldbeturnedinto,andtreatedlike,pigs.GiventheimportanceoftheseissuesforthestabilityoftheEuropeanMonetaryUnion(EMU)weelaborateontheroleofratingagenciesinthecurrentsovereigndebtcrisis,focusingontheso-calledPIGScountries,i.e.,Portugal,Ireland,GreeceandSpain.ThequestionwhetherratingagenciesplayedapassiveroleorwereanactivedrivingforceduringEurope’sdebtcrisisisaddressedinthreesteps.First,wetrytodevelopanunderstandingofhowsovereigndebtratingsareformed.Wedothisbyestimatingrelationshipsbetweenratingsandmacroeconomicandstructuralvariables.Second,weusetheseequationstodecomposeactualratingsintoasystematicandanarbitrarypart,thelatterbeingdefinedaswhatisleftunexplainedbypreviouslyobservedproceduresofratingagencies.Finally,wequantifytheeffectofthesystematicandthearbitrarypartofacountry’ssovereigndebtratingonitsgovernmentbondyieldstoassesstheendogenousandthediscretionaryimpactofratingagenciesonmarketoutcomesduringtheEuropeandebtcrisiswithspecialconcernforthePIGScountries.Asourresultssuggest,thePIGScountrieswerenotonlyratedworseduringthecrisisthanallothercountriesinoursampleof26OECDcountries,butthismarkdownalsoresultedinsignificantlyhigherinterestratesongovernmentbonds,whichthemselvesaggravatedtheEuropeandebtcrisis.Ourpapercontributestoseveralstrandsintheliterature.First,itaddstothebroadliteraturefollowingCantorandPacker(1996)thattriestoexplainsovereigndebtratings,whichissummarizedbyMelliosandPaget-Blanc(2006).MostcloselyrelatedtoourstudyistheworkofFerrietal.(1999),MulderandPerrelli(2001)andEl-Shagi(2010)whodiscusstheroleofratingagenciesduringtheAsianCrisisattheendofthe1990s.Byemployingorderedregressionstoexplainsovereigndebtratings,ouranalysisisalsorelatedtoHuetal.(2002),BlockandVaaler(2004)andAfonsoetal.(2007,2009).ItseemsthatnostudyexistswhichanalyzestheroleofratingagenciesduringthecurrentEuropeansovereigndebtcrisisusingsuchmethods.Ourpaperalsoaddstotheliteraturethatfindsasignificantimpactofratingsongovernmentbondspreads.FollowingtheworkofKaminandvonKleist(1999)andEichengreenandMody(2000),whoidentifiedsucharelationshipfordevelopingcountriesinthe1990s,Gomez-Puig(2006)findsanegativerelationshipbetweenbondspreadsandratingsforEMUmembersbetween1996and2001.ThisisconsistentwiththeresultsthatManganelliandWolswijk(2009)obtainedforEMUcountriesbetween1999and2008.Morerecentpaperssuggestthisrelationalsoholdsinthefaceofthecurrenteconomiccrisis,seeforexampleAttinasietal.(2009)orSgherriandZoli(2009).Wedonotonlycontributetothisliteraturebyexplaininggovernmentbondspreadswithmorerecentsovereigndebtratings,butalsobyconductingGrangercausalitytestsonthecausalrelationbetweenbothvariables,whichhasrarelybeendonebefore.RoleofRatingAgencies289Thepaperisstructuredasfollows.Thenextsectionprovidesinformationontheemployeddata.Thenwetrytodeterminehowsovereigndebtratingsarebuilt,followedbyalookathowratingsandtheircomponentsaffectsovereigncreditspreads.Inthefinalsectionweoffe