complement38mmm.com

Thispaperpresentspreliminaryfindingsandisbeingdistributedtoeconomistsandotherinterestedreaderssolelytostimulatediscussionandelicitcomments.
TheviewsexpressedinthispaperarethoseoftheauthorsandarenotnecessarilyreflectiveofviewsattheFederalReserveBankofNewYorkortheFederalReserveSystem.
Anyerrorsoromissionsaretheresponsibilityoftheauthors.
FederalReserveBankofNewYorkStaffReportsStaffReportNo.
580October2012TobiasAdrianAdamB.
AshcraftShadowBanking:AReviewoftheLiteratureREPORTSFRBNYStaffAdrian,Ashcraft:FederalReserveBankofNewYork(e-mail:tobias.
adrian@ny.
frb.
org,adam.
ashcraft@ny.
frb.
org).
ThispaperwaspreparedfortheNewPalgraveDictionaryofEconomics.
TheauthorsthankNicolaCetorelliandAndreiShleiferforhelpfulcomments.
TheviewsexpressedinthispaperarethoseoftheauthorsanddonotnecessarilyreflectthepositionoftheFederalReserveBankofNewYorkortheFederalReserveSystem.
AbstractWeprovideanoverviewoftherapidlyevolvingliteratureonshadowcreditintermedia-tion.
Theshadowbankingsystemconsistsofawebofspecializedfinancialinstitutionsthatconductcredit,maturity,andliquiditytransformationwithoutdirect,explicitaccesstopublicbackstops.
Thelackofsuchaccesstosourcesofgovernmentliquidityandcreditbackstopsmakesshadowbanksinherentlyfragile.
Muchofshadowbankingactivitiesisintertwinedwiththeoperationsofcoreregulatedinstitutionssuchasbankholdingcom-paniesandinsurancecompanies,thuscreatingasourceofsystemicriskforthefinancialsystematlarge.
Wereviewfundamentalreasonsfortheexistenceofshadowbanking,explainthefunctioningofshadowbankinginstitutionsandactivities,discusswhyshad-owbanksneedtoberegulated,andreviewtheimpactofrecentreformeffortsonshadowbankingcreditintermediation.
Keywords:shadowbanking,financialintermediationShadowBanking:AReviewoftheLiteratureTobiasAdrianandAdamB.
AshcraftFederalReserveBankofNewYorkStaffReports,no.
580October2012JELclassification:E44,G00,G01,G281Contents1)WhatIsShadowCreditIntermediation2A)Definition2B)Measurement.
3C)Examples52)WhyDoesShadowCreditIntermediationExist10A)InnovationintheCompositionofAggregateMoneySupply.
11B)Capital,Tax,andAccountingArbitrage.
12C)OtherAgencyProblemsinFinancialMarkets.
133)HowDoesShadowCreditIntermediationWork15A)TheSevenStepsofShadowCreditIntermediation15B)CommercialBanksandShadowBanking.
16C)ExternalandIndependentShadowBanking.
18D)Government-SponsoredShadowBanking.
194)WhyDoesShadowCreditIntermediationNeedtoBeRegulated20A)AssetQuality20B)FundingFragility.
22C)LiquidityBackstops.
245)HowShouldShadowCreditIntermediationBeRegulated25A)MoneyMarketReforms:ABCP,Repo,andMoneyMarketMutualFunds.
26B)BankingRegulationReforms:CapitalandLiquidity.
29C)CreditMarketReforms:SecuritizationandCreditRatings.
306)Conclusion.
31Literature3221)WhatIsShadowCreditIntermediationTheshadowbankingsystemisawebofspecializedfinancialinstitutionsthatchannelfundingfromsaverstoinvestorsthrougharangeofsecuritizationandsecuredfundingtechniques.
Althoughshadowbanks—theinstitutionsthatconstitutetheshadowbankingsystem—conductcreditandmaturitytransformationsimilartothatoftraditionalbanks,theydosowithoutthedirectandexplicitpublicsourcesofliquidityandtailriskinsuranceavailablethroughtheFederalReserve'sdiscountwindowandtheFederalDepositInsuranceCorporation.
Shadowbanksarethereforeinherentlyfragile,notunlikethecommercialbankingsystempriortothecreationofthepublicsafetynet.
ThisdefinitioncloselyfollowsthatofPozsar,Adrian,Ashcraft,andBoesky(2010).
A)DefinitionInthetraditionalbankingsystem,intermediationbetweensaversandborrowersoccursinasingleinstitution.
Throughtheprocessoffundingloanswithdeposits,banksengageincredit,maturity,andliquiditytransformation.
Credittransformationreferstotheenhancementofthecreditqualityofdebtissuedbytheintermediarythroughtheuseofpriorityofclaims.
Forexample,thecreditqualityofseniordepositsisbetterthanthecreditqualityoftheunderlyingloanportfolioduetothepresenceofjuniorequity.
Maturitytransformationreferstotheuseofshort-termdepositstofundlong-termloans,whichcreatesliquidityforthesaverbutexposestheintermediarytorolloveranddurationrisks.
Liquiditytransformationreferstotheuseofliquidinstrumentstofundilliquidassets.
Forexample,apoolofilliquidwholeloansmighttradeatalowerpricethanaliquid-ratedsecuritysecuredbythesameloanpool,ascertificationbyacredibleratingagencywouldreduceinformationasymmetriesbetweenborrowersandsavers.
Saversentrusttheirfundstobanksintheformofdeposits,whichbanksusetofundloanstoborrowers.
Saversfurthermoreowntheequityandlong-termdebtissuanceofthebanks.
DepositsareguaranteedbytheFDIC,andaliquiditybackstopisprovidedbytheFederalReserve'sdiscountwindow.
Relativetodirectlending(thatis,saverslendingdirectlytoborrowers),creditintermediationprovidessaverswithinformationandriskeconomiesofscalebyreducingthecostsinvolvedinscreeningandmonitoringborrowersandbyfacilitatinginvestmentsinamorediverseloanportfolio.
Shadowbankingactivityisremovedfromofficialpublic-sectorenhancements,buttypicallyreceivesindirectorimplicitenhancements.
Officialenhancementstocreditintermediationcanbeclassifiedintofourlevelsofstrength:1.
Aliabilitywithdirectofficialenhancementmustresideonafinancialinstitution'sbalancesheet,whereasoff-balance-sheetliabilitiesoffinancialinstitutionsareindirectlyenhancedbythepublicsector.
2.
Activitieswithdirectandimplicitofficialenhancementincludedebtissuedorguaranteedbythegovernment-sponsoredenterprises(GSEs),whichbenefitfromanimplicitcreditputtothetaxpayer.
3Theimplicitnatureofsupportimpliesthattheintermediaryreceivesthebenefitofcreditandputoptionstothepublicsector,buttypicallywouldnotpaytheirfullmarginalsocialcost.
Itisnotsurprisingthat,withsuchasubsidy,theseintermediarieswouldgrowverylarge.
3.
Activitieswithindirectofficialenhancementgenerallyincludetheoff-balance-sheetactivitiesofdepositoryinstitutions,suchasunfundedcreditcardloancommitmentsandlinesofcredittoconduits.
Thefocushereisontheaccountingandcapitalarbitrageactivitiesbyfinancialinstitutions.
Capitalrequirementshavetypicallybeentiedtoaccountingrules,sotransactionstoremoveassetsfromthebalancesheethavehistoricallyreducedregulatorycapital.
Whilerecentaccountingreformhasreducedthescopeforthisformofarbitragegoingforward,itwasanimportantpartofthenarrativeoftherecentcreditcycle.
4.
Finally,activitieswithindirectandimplicitofficialenhancementsincludeassetmanagementactivitiessuchasbank-affiliatedhedgefundsandmoneymarketmutualfunds(MMMFs),aswellasthesecuritieslendingactivitiesofcustodianbanks.
Creditintermediationactivitiesthattakeplacewithoutofficialcreditenhancementsaresaidtobeunenhanced.
Forexample,thesecuritieslendingactivitiesofinsurancecompanies,pensionfunds,andcertainassetmanagersdonotbenefitfromaccesstoofficialliquidity.
Wedefineshadowcreditintermediationtoincludeallcreditintermediationactivitiesthatareimplicitlyenhanced,indirectlyenhanced,orunenhancedbyofficialguaranteesestablishedonanexantebasis.
B)MeasurementToillustratehowshadowbankinghasevolvedoverthepastfewdecades,Figure1presentstheliabilitiesofeachtypeoffinancialbusiness,derivedfromU.
S.
FlowofFundsdata.
Inparticular,itshowsthesharesoftraditionalintermediaries,whichincludedepositoryinstitutions,pensionfunds,andlifeinsurancecompanies,intheoverallfundingofcreditheldbyfinancialbusinesses.
Whilethesetraditionalformsoffinancialintermediationaccountedforalmost100percentoffundingforcreditintermediationinthemid-1940s,theirsharefelltoaslowas40percentin2007andthen,withthecollapseoftheshadowbankingsystem,reboundedto47percent.
Thefiguredocumentsthatasignificantpartofthelong-rundeclineintheroleofliabilitiesoftraditionalintermediariesisdrivenbytheriseoftheshadowbankingsystem,asdefinedabove.
Inparticular,theshareofliabilitiescorrespondingtoMMMFs,repo,openmarketpaper,GSEs,andcorporate/securitizationpeakedat34percentthroughthelastdecade,butdeclinedtoonly26percentattheendof2011.
Illustratingtheimportanceofmaturitytransformationinthefinancialsystemgenerally,andofshadowcreditintermediationinparticular,Figure2breaksouttheliabilitiesofthefinancialbusinessintooneoffourmajorcategories:1)traditionalmaturitytransformation,includingbankdepositsandinterbankliabilities;2)traditionalcredittransformationincludingtermdebtissuedbybanksandbankholdingcompaniesaswellasreservesofpensionsandlifeinsurancecompanies,inadditiontodepositoryloansnotelsewhereclassified;3)shadowmaturitytransformation,includingMMMFs,repo,4openmarketpaper,andsecuritybroker-dealercreditandpayables;and4)shadowcredittransformation,includingGSEs,termdebtissuedbynonbanks,mutualfundshares,REITmortgagedebt,andloanscategorizedas"other.
"Thefiguresuggestsseveralstrikingpatterns.
First,theamountofmaturitytransformationinthefinancialsystemhasbeendecliningoverthepastsixtyyears.
Whilealmost75percentofintermediatedcreditwasfundedbyshort-termbankliabilitiesinthemid-1940s,thatnumberhasfallenaslowas15percentinrecentyearsbeforereboundingto21percentin2011.
Second,thereducedimportanceofbankmaturitytransformationisbeingoffsetpartlybyshadowmaturitytransformation.
Theconsequenceisthatthefractionoftheaggregatemoneysupplyissuedbyshadowintermediarieshasincreasedsignificantly,peakingat45percentintheearly2000sbeforedecliningto28percentin2011,alevelnotseensince1993.
Whilethefigureillustratesthattheamountofcreditfundedthroughshadowintermediationevenatthepeakismodest(approximately10percent),thegrowingimportanceofshadowmoneyintheaggregatesupplyofmoneywasanimportantfactorinamplifyingtheshockstotheeconomymorebroadly.
Third,whilematuritytransformationbyshadowintermediarieshasincreasedovertheperiod,thelargerstoryisclearlytheincreasedroleoftermdebtmarketsinfundingcredit.
Inparticular,theamountofshadowcredittransformationincreasedfromzeroin1945toasmuchas36percentoftotalfinancialsectorliabilitiesin2007beforedecliningto31percentin2011.
TheincreaseinmarketfundingforcreditisdrivennotonlybytheGSEsandsecuritization,butalsobytheincreasedimportanceofmutualfundsandREITs.
Shadowcredittransformationincreasedfromonly5percentoftotalcredittransformationin1945toapeakamountof60percentin2008beforedecliningto55percentin2011.
Figure1:LiabilitiesofFinancialBusinessSource:FederalReserveFlowofFunds,TablesL107andL212.
TraditionalIntermediationreferstonetinterbankliabilities(line28)pluscheckable(line29)andsavings(line30)depositsofdepositoryinstitutionsplusreservesoflifeinsurancecompanies(line43)andpensions(line44)pluscorporatedebt.
Thelatteriscalculatedbysubtractingfromtotalcorporatedebt(line36)theamountissuedbyholdingcompanies(line10)andbanks(line5)fromL212.
ShadowbankingcomponentsarefromL107:MMMFs(line31),repo(line32),commercialpaper(Line34),GSEs(line35),andsecuritybroker-dealercredit(line41)andpayables(line42).
Shadowcredittransformationincludescorporatedebt(lessamountsintraditionalcredittransformation).
0.
0%10.
0%20.
0%30.
0%40.
0%50.
0%60.
0%70.
0%80.
0%90.
0%100.
0%MMMFsRepoCPSecurityBroker-DealerCreditandPayablesGSEsCorporates/SecuritizationREITsandMutualFundsTraditionalIntermediation5Figure2:MaturityandCreditTransformationSource:FederalReserveFlowofFunds,TablesL107andL212.
Traditionalmaturitytransformationincludesnetinterbankliabilities(line28)pluscheckable(line29)andsavings(line30)depositsofdepositoryinstitutions.
Traditionalcredittransformationincludesreservesoflifeinsurancecompanies(line43)andpensions(line44)pluscorporatedebtissuedbybanksandholdingcompaniesplusloansfromdepositoryinstitutionsNEC(line37).
Thelatteriscalculatedbysubtractingfromtotalcorporatedebt(line36)theamountissuedbyholdingcompanies(line10)andbanks(line5)fromL212.
ShadowmaturitytransformationincludesfromL107MMMFs(line31),repo(line32),commercialpaper(line34),andsecuritybroker-dealercredit(line41)andpayables(line42).
ShadowcredittransformationincludesGSEs(line35),REITs(line39),mutualfundshares(line40),andotherloans(line38).
C)ExamplesThetypesofactivities,institutions,andvehiclesthatarepartoftheshadowbankingsystemareconstantlyevolving.
Theexamplesprovidedinthissubsectionarebynomeansexhaustive,buttheydorepresentpartsoftheshadowbankingsystemthathavebeenparticularlyimportantatsomepointintimeandsomethatstillare.
ABCPConduitsAsset-backedcommercialpaper(ABCP)iscommercialpapercollateralizedbyaspecificpooloffinancialassets.
ABCPisusuallyissuedbybankruptcy-remotespecial-purposevehicles(SPVs),suchasABCPconduitsorspecialinvestmentvehicles(SIVs).
BothABCPconduitsandSIVsobtaincreditratingsontheissuedpaper.
ABCPratingsarelargelybasedonthecreditprofileofbanksprovidingcreditandliquiditysupportbycommercialbanks,whileSIVratingsarebasedonthecreditqualityoftheassetsaswellastheoverallfundingstrategyoftheSIV.
Single-sellerABCPconduitsarebackstopstotheworkingcapitalneedsoflargenonbankfinancecompaniesandreceivesuchsupportfromasinglecommercialbank,whilemulti-sellerconduitsfundtheworkingcapitalneedsofsmallernonbanksandreceivethesupportofmultipleinstitutions.
Similarly,SIVscaneitherbeaffiliatedwithasinglebankinginstitution,orobtainsupportfrommultipleinstitutions.
0.
0%10.
0%20.
0%30.
0%40.
0%50.
0%60.
0%70.
0%80.
0%90.
0%100.
0%TraditionalMaturityTransformationShadowMaturityTransformationTraditionalCreditTransformationShadowCreditTransformationOtherLiabilities6ThebankruptcyremotenessofalloftheseentitiesimpliesthatthecollateralbackingtheABCPisexemptfromthepotentialbankruptcyoftheinstitutionthatprovidesthebackuplinesofcreditandliquidity.
ThematurityofABCPisbetweenoneand180days,exposingtheABCPtorolloverrisk,asourceoffragilityforABCPissuersthatwillbediscussedlater.
TherewereafewexamplesofABCPissuersthatdidnotreceiveunconditionalenhancementsfromcommercialbanks.
OneistheCanadianABCPmarket,whereinvestorswereforcedtoholddefaultedpaper.
Inaddition,extendibleABCPeffectivelytransferstherolloverrisktoinvestors,thusrequiringahigherrateofreturn.
Structuredinvestmentvehicles(SIVs)arespecializedfinancialinstitutionsthatconductshadowmaturitytransformation.
OntheassetsideofSIVsaresecuritizedassetssuchasABS,MBS,CDOs,CLOs,CMOs,orfinancialsectordebt.
TheseassetsarefundedthroughissuanceofABCP,medium-termnotes(MTN),orlong-termnotes(LTN).
Inordertoachieveacreditratingontheirliabilities,SIVsobtainbackuplinesofcreditfromcommercialbanks.
SIVswerefirstcreatedin1988,effectivelymovingthefinancingofABSfromthebalancesheetofCitigrouptoanoff-balance-sheetSIV.
WhilesomeSIVsarecloselyassociatedwithparticularfinancialinstitutions,othersoperateindependentlyofanyparticularinstitution.
Sincethefinancialcrisisof2007-09,SIVshavestoppedoperating.
SIVsresemblecommercialbanksinmanyways,butbothassetsandliabilitiesaretradable,andliquidityandcreditbackstopsareprovidedbyprivateinstitutions.
ABCPhasprovidedfundingflexibilitytoborrowersandinvestmentflexibilitytoinvestorssincethe1980s,whenABCPwasusedasawayforcommercialbankstofundcustomertradereceivablesinacapital-efficientmannerandatcompetitiverates.
ABCPbecameacommonsourceofwarehousingforABScollateralinthelate1990s.
Thepermissibleoff-balance-sheetstructurefacilitatedbalance-sheetsizemanagement,withtheassociatedbenefitsofreducedregulatorycapitalrequirementsandleverage.
ABCPfundinghasalsobeenasourceoffee-basedrevenue.
Forcorporateusers,ABCPbenefitsincludesomefundinganonymity,increasedcommercialpaper(CP)fundingsources,andreducedcostsrelativetostrictbankfunding.
Overtime,ABCPconduitsexpandedfromthefinancingofshort-termreceivablesusedascollateraltoabroadrangeofloans,includingautoloans,creditcards,studentloans,andcommercialmortgageloans.
Atthesametime,asthemarketdeveloped,itcametoembedmuchmorematuritymismatchthroughfundinglonger-termassets,warehousedmortgagecollateral,etc.
Securitiesarbitragevehiclesareoneparticularexampleofashadowbankinginstitutionthatperformedsubstantialamountsofmaturitytransformation.
ThesevehiclesusedABCPtofundvarioustypesofsecurities,includingcollateralizeddebtobligations(CDOs),asset-backedsecurities(ABS),andcorporatedebt.
ABCPexperiencedarunthatbeganinthesummerof2007,whenthesponsorofasingle-sellermortgageconduit,AmericanHome,declaredbankruptcy,andthreemortgageprogramsextendedthematurityoftheirpaper.
OnAugust7,BNPParibashaltedredemptionsattwoaffiliatedmoneymarketmutualfundswhenitwasunabletovalueABCPholdings.
Covitz,Liang,andSuarez(2012)usedatafromtheDepositoryTrustClearingCorporation(DTCC)todocumentaninvestorrunonmorethan100programs,one-thirdoftheoverallmarket.
Whilerunsweremorelikelyonprogramswithgreaterperceivedsubprimemortgageexposure,weakerliquiditysupport,andlowercreditratings,thereisalsoevidenceofinvestorrunsthatwereunrelatedtospecificprogramcharacteristics.
7ABSissuersAsset-backedsecurities(ABS)arecollateralizedclaimsonpoolsofloans,mortgages,orreceivables.
ThecashflowandincomefromABSarestructuredintotranches,whichreceivecreditratings.
Forexample,thesuperseniorAAAtranchemightrepresent80percentofthetotalvalueoftheABS,themezzanineBBBtranchemightrepresent15percentofthetotalvalue,andtheremaindermaybeallocatedtoanequitytranche.
SuchpoolingandtranchingoftheABSarereferredtoassecuritizationastheABS'valueissecuritizedbyitscollateral.
Securitizationactivityisattheheartofshadowbanking,asitallowscreditoriginatorstosellpoolsofcredittootherinstitutions,therebytransferringthecreditrisk.
SecuritizedproductssuchasABSaresoldtobanks,shadowbanks,andrealmoneyinvestors.
TheunderlyingassetsofABSconsistofreceivablesfromcreditcards,autoloans,mortgages,andaircraftleases,amongothers.
Evenroyaltypaymentsandmovierevenueshavebeensecuritized.
SecuritizationtechniquessuchasABSrepresentamajorformoffinancialinnovationinrecentdecadesandaretightlylinkedwithboththecreditcycleandthedevelopmentoftheshadowbankingsystem.
Legally,theABSisstructuredasabankruptcy-remoteSPV.
ABStypicallyperformnomaturitytransformation,butdoachievecreditandliquiditytransformation.
Credittransformationisachievedthroughdiversification.
Forexample,theABScollateralmightconsistofsubprimemortgageloans,whilemuchoftheABS'liabilitiesconsistofAAAassets.
LiquiditytransformationoccursbecauseanyindividualmortgageorloanoftheABScollateralmightbeilliquidduetoadverseselectionproblems,yetapoolofsuchassetsmightbeliquid.
However,theliquidityoftheABSdependscruciallyonthebusinesscycle,asABSbecomemoreilliquidduringdownturns,particularlyduringfinancialcrises.
OnespecialformofABSisthecollateralizeddebtobligation(CDO),whichissecuredbyasmallernumberofloansorbybonds.
ForotherformsofABS,collateralconsistsofalargenumberofindividualloans,mortgages,orreceivables.
ForCDOS,however,thecollateralcanbecorporatebonds,structuredcreditproductssuchasABS,orpoolsofagencymortgage-backedsecurities(MBS).
WhenthecollateralofaCDOisABS,itissometimescalledanABSCDO.
WhencollateralisMBS,theCDOiscalledacollateralizedmortgageobligation(CMO).
Therearealsocollateralizedloanobligations(CLOs),whichareCDOswithsyndicatedloansascollateral.
TheunderlyingloansofCLOsareoftenleveragedloans,usedtorestructurethefundingofcorporationstoallowformoreleverage.
Historically,thefirstCMOwasissuedbySalomonBrothersandFirstBostonin1983forFreddieMac,andthefirstCDOwasissuedbyDrexelBurnhamLambertforImperialSavingsAssociationin1987.
ThecreditqualityofABSCDOsisoftenenhancedthroughCDOsontheunderlyingmezzaninetranchesoftheABSthatarere-securitized.
ThisenhancementreducesthecreditriskoftheCDO'scollateralandallowstheissuanceofAAAtranchesfromanunderlyingpoolofmezzaninetranches,whichcaninturnbefundedinshorter-termmarkets.
CDOissuancepeakedin2007andthentotallycollapsedintheaftermathofthe2007-09financialcrisis.
8Tri-partyRepoArepurchaseagreement(repo)isthesaleofsecuritiestogetherwithanagreementthatthesellerwillbuybackthesecuritiesatalaterdate.
Mostrepocontractsareshortterm—betweenoneand90days—althoughtherearereposwithmuchlongermaturities.
Reposareover-collateralized,andthedifferencebetweenthevalueofthecollateralandthesalepriceiscalledtherepohaircut.
Inaddition,therepurchasepriceisgreaterthanthesaleprice,thedifferenceconstitutingthereporate,whichis,ineconomicterms,aninterestrateonacollateralizedloan.
Inarepotransaction,thepartybuyingthecollateralactsasalender.
Thedistinguishingfeatureofatri-partyrepoisthataclearingbankactsasanintermediarybetweenthetwopartiestotherepo.
Theclearingbankisresponsiblefortheadministrationofthetransaction,includingcollateralallocation,markingtomarket,andsubstitutionofcollateral.
Thetri-partystructureensuresthatboththeborrowerandthelenderareprotectedagainstthedefaultoftheother,asthecollateralresideswithathirdparty.
TheU.
S.
tri-partyrepomarketrepresentsamajorsourceoffundingforsecuritybroker-dealers.
Themarketpeakedatslightlyabove$2.
8trillionin2008andiscurrentlyslightlybelow$1.
7trillion.
Investorsintri-partyrepoareprimarilymoneymarketmutualfundsandothercash-richinvestorssuchascorporatetreasuryfunctions,whiletheborrowersarelargesecuritiesdealerswithinventoriesofsecuritiestofinance.
Clearingbanksunwindthesetradeseachafternoonandreturnthecashtotheinvestors.
Butbecausethedealersretainaportfolioofsecuritiesthatneedfinancingona24-hourbasis,theymustextendcredittotheotherdealersagainstthesesecuritiesforseveralhoursbetweenthatafternoonunwindandthesettlementofnewreposintheearlyevening.
Thatway,thosedealerscanrepaytheirinvestorsandavoiddefaultingontheobligations.
SincetheenactmentoftheBankruptcyAmendmentsandFederalJudgeshipActof1984,reposonTreasury,federalagencysecurities,bankcertificatesofdeposits,andbankers'acceptanceshavebeenexemptedfromtheautomaticstayinbankruptcy.
Thebankruptcyexceptionensuredtheliquidityoftherepomarketbyassuringlendersthattheywouldgetspeedyaccesstotheircollateralintheeventofadealerdefault.
In2005,thesafeharborprovisionwasexpandedtoreposwrittenonbroadercollateralclasses,includingcertainmortgage-backedsecurities.
Thisbroadeningofacceptablecollateralfortheexemptionfromtheautomaticstayforreposallowedtherepomarkettofundcreditcollateral—andthusdirectlyfundtheshadowbankingsystem.
Itshouldbenotedthatthetri-partyrepomarketisonlyasubsetofotherrepoandshort-term,collateralizedborrowingmarkets.
Whilebroker-dealersconducttheirfundingprimarilyinthetri-partyrepomarket,theirlendingoccursmainlyinDVP(deliveryversuspayment)repoorGCF(generalcollateralfinance)repo.
Incontrasttoatri-partyrepo,DVPreposarebilateraltransactionsthatarenotsettledonthebooksoftheclearingbanks.
Instead,settlementtypicallyoccurswhentheborrowerdeliversthesecuritiestothelender.
Adrian,Begalle,Copeland,andMartin(2013)discussvariousformsofrepoandsecuritieslending,andFlemingandGarbade(2003)describeGCFrepo,whichisconductedamongdealers.
9Copeland,Martin,andWalker(2011)documentthecollateralcompositioninthetri-partymarket,aswellastherepomarketconventions,usingdatafromJuly2008toearly2010.
Theyshowthat,duringthisperiod,severalhundredbilliondollarsofcollateralinthetri-partyrepomarketconsistedofcollateralsuchasequities,private-labelABS,andcorporatecreditsecuritieswithoutanyeligibilityforpublicsourcesofliquidityorcreditbackstops.
Krishnamurthy,Nagel,andOrlov(2011)complementthisfindingbylookingdirectlyatthecollateralofMMMFs.
Whiletheyfindthatthemajorityofthe$3.
5trillionMMMFs'collateralisofhighquality,theydodocumentseveralhundredbilliondollarsofprivate-labelABSsecuritiesfundedbyMMMFs.
However,theoverallamountofprivate-labelABSfundedintherepomarketbyMMMFsislessthan3percentoftotaloutstanding.
AdrianandShin(2009,2010a)studytheroleofrepoforsecuritybroker-dealersanddocumentthegrowthofthesectorsincethe1980s.
Adistinguishingfeatureofthebalancesheetmanagementofsecuritybroker-dealersistheprocyclicalityoftheirleverage.
Balancesheetexpansionstendtocoincidewithexpansionsinbroker-dealerleverage,whilebalancesheetcontractionsareachievedviadeleveraging.
AdrianandShinshowthatreposplaythecrucialroleinthisleveragecycleofthebroker-dealers:Themajorityoftheadjustmentinbalancesheetsizetendstobeachievedthroughadjustmentsinthesizeoftherepobook.
WhileAdrianandFleming(2005)pointoutthatthenetfundingofdealersintherepomarkettendstobesmall,AdrianandShin(2010a)arguethattheoverallbalancesheetsizeoffinancialintermediariescanbeviewedasanindicatorofmarketliquidity.
Whengrossbalancesheetsarereducedthroughdeleveraging,financialmarketliquiditytendstodryup.
MoneyMarketFundsMoneymarketmutualfundsareopen-endedmutualfundsthatinvestinshort-termsecuritiessuchasTreasurybills,commercialpaper(includingABCP),andrepo.
MMMFswerefirstcreatedin1971inresponsetoRegulationQ,whichrestrictedtheinterestthatcommercialbankscanpayondeposits.
Sincethen,moneymarketfundshaverepresentedanalternativetobankdepositsfrominvestors'pointofview,withyieldsthataretypicallymoreattractivethanbankdeposits.
Themoneymarketsectorpeakedataround$3.
5trillionin2008.
MMMFsareregulatedbytheSECundertheInvestmentCompanyActof1940.
Moneymarketfundsseekastablenetassetvalue(NAV),whichisgenerally$1.
00,meaningthattheyaimnevertolosemoney.
Ifafund'sNAVdropsbelow$1.
00,itissaidto"breakthebuck.
"InSeptember2008,thedayfollowingtheLehmanBrothersbankruptcy,theReservePrimaryFundbrokethebuckandtriggeredarunonMMMFs.
OtherfundmanagersreactedbysellingassetsandinvestingatonlytheshortestofmaturitiesorbyreallocatingtoTreasurybills,therebyexacerbatingthefundingdifficultiesforotherinstrumentssuchascommercialpaperandrepo.
Wermers(2011)investigatesinmoredetailtheroleofinvestmentflowsintoandoutofmoneymarketmutualfunds,focusingparticularlyontheperiodofthefinancialcrisis.
Wermersshowsthatinstitutionalinvestorsweremorelikelytorunthanretailinvestors,andinstitutionalinvestorstendedtospreadsuchrunbehavioracrossvariousMMMFfamilies.
InstitutionalMMMFinvestorscanthusbeviewedasatransmissionchannelforcontagiousruns.
KacperczykandSchnabl(2011)analyzetheimpactoftheorganizationalstructureofMMMFsontheirrisk-takingbehavior.
Inparticular,theyask10howtherisk-takingdiffersbetweenstand-alonefundsandthefundsthatareownedbylargerholdingcompanies,suchabankholdingcompanies.
KacperczykandSchnablfindsignificantdifferencesintherisk-takingofstand-aloneMMMFsrelativetothefundsthathaveimplicitguaranteesfromfinancialconglomerates.
Duringthefinancialcrisisof2008,whensystemicriskincreasedandconglomeratesbecamerelativelymoreexposedtosystemicrisk,stand-alonemutualfundsincreasedtheirrisk-takingbehaviorrelativelymore.
Conversely,intherun-uptothecrisis,whenmeasuredsystemicriskwaslow,MMMFsthatwerepartofconglomeratestookonrelativelymorerisk.
2)WhyDoesShadowCreditIntermediationExistTheterm"shadowbanking"wascoinedbyMcCulley(2007)andwaspickedupbypolicymakers(see,forexample,Tucker(2010)).
Thefirstarticlesonshadowbankingappearedin2008(Pozsar(2008)andAdrianandShin(2009)).
AcomprehensiveoverviewoftheshadowbankingsystemcanbefoundinPozsar,Adrian,Ashcraft,andBoesky(2010).
AnupdateonregulatoryreformsrelatingtoshadowbankingcanbefoundinAdrianandAshcraft(2012).
AcademicstudiesofshadowbankingincludeGortonandMetrick(2011,2012),Gennaioli,Shleifer,Vishny(2012b),Stein(2010),andAcharya,Schnabl,andSuarez(2010).
Inadditiontotheacademicliteraturebyfinancialeconomists,legalscholarshavecontributedtotheshadowbankingliterature(e.
g.
,Schwarcz(2012)andRicks(2010)).
TheFinancialStabilityBoard(FSB)hasinitiatedinternationalworkinggroupsonshadowbanking(seeFSB(2011,2012)).
TheobjectiveoftheFSBistoenhancetheregulationandoversightoftheshadowbankingsystem.
TheFSBisleadingthisworkbecauseoftheglobalreachofshadowbanking,whichactsasaninternationalsystemicrisktransmitterintimesofcrisis.
Inresponsetothetightenedfinancialregulation,internationalshadowbankregulatoryarbitragemightverywellbegrowinginthefuture,makinganadequateregulatoryframeworkandmonitoringsystemimperative.
FSB(2012)presentsaclassificationofshadowbankingworkinggroups,withtheaimtodevelopaframeworkforpolicyrecommendationsandmonitoring.
Theclassificationsare1)banks'interactionswithshadowbanks,2)moneymarketmutualfunds,3)othershadowbankingentities,4)securitizationactivity,and5)securitieslendingandrepos.
Finally,industrygroupsarealsostudyingshadowbanking.
TheInstituteofInternationalFinancehasputforwardaframeworkforpolicymakinginrelationtoshadowbanking.
Inaddition,theSecuritiesIndustryandFinancialMarketsAssociation(SIFMA)hasmultipleworkstreamsonthetopicofshadowbanking.
Therearethreebroadexplanationsfortheexistenceofshadowbanks:A)innovationinthecompositionofaggregatemoneysupply;B)capital,tax,andaccountingarbitrage;andC)otheragencyproblemsinfinancialmarkets.
Wediscusseachoftheseexplanations.
Empirically,theyareintertwined,anditisdifficulttoattributerelativemagnitudestoeachofthem.
11A)InnovationintheCompositionofAggregateMoneySupplyDrawingmotivationfromthenarrativeofGortonandMetrick(2011),itispossibletoviewshadowcreditintermediationasfinancialinnovationinthecompositionofaggregatemoneysupply.
Moneyplaysacrucialroleintheeconomy,actingnotonlyasastoreofvalue,butalsoasaunitofaccountandmeansofexchange.
Therapidlossofconfidenceinthevalueofmoneyhasbeenarootcauseoffinancialpanicsacrosscountriesandovertime.
Theearliestformsofmoney,commoditymoney,weremadeofpreciousmetals,havinginherentvaluebybeingcomprisedofgoldorsilver.
However,commoditymoneywaseventuallyreplacedwithfiatmoney,whichhaslittleintrinsicvalue,butisinsteadbackedonlybytheissuer'spromisetoconvertthenotesintoacommodity.
Inparticular,thebankingsystemoftheearly1800swascharacterizedbybanksthatissuednotesbackedbythepromiseofconvertibilityintogoldorsilvercoin.
Bankingchartersweretightlyrestrictedbystatelegislatures.
IntheFreeBankingEra(1837to1862),therewasfreeentryintothesectorforanybankerwithadequateinitialcapital,butbankswererequiredtodepositstateorfederalgovernmentbondswithfacevalueequaltothevalueofnotesissuedwithastaterepresentative.
Whileonemighthavethoughtthatthepresenceofcollateralwouldhavemadefreebankingstable,theperiodwascharacterizedbyaseriesofpanics,andalmostone-thirdofallfreebanksultimatelyfailed.
Therootcauseofthesepanicsisasubjectofdebateintheacademicliterature;reductionsinthevalueofstatedebtlikelyplayedaprominentrole.
Jaremski(2010)documentsthatfailureratesoffreebanksiscorrelatedwithstatebondprices,butdoesnotfindthesameforcharterbanks.
Rogoff(1985)suggeststhattheexistenceofmarketdiscountsonstatebondsnotonlyreducedconfidencebynoteholders,limitingtheirliquidityandvalue,butalsocreatedscopefor"wildcatbanking,"whereimplicitleveragebetweenthefacevalueofbanknotesandthemarketvalueofstategovernmentbondspermittedbankstohaveextraordinaryleverage.
ThescopeforpanicscreatedbyconcernaboutthevalueofbanknoteswaseliminatedbythepassageoftheNationalBankingActsin1863and1864.
ThislegislationreplacedbanknoteswithanationalcurrencybackedbythedepositofU.
S.
Treasurybonds,enactedaceilingontheaggregatecirculationofnotes,andsetrequiredreservesonbothnotesanddepositsinordertoencouragebankstoholdsaferportfolios.
WhiletheNationalBankingActscreatedconfidenceincurrency,financialinnovationinthecompositionofmoneyintheformofbankdepositshadalreadyoccurred.
Whilebanknotesweresecured,depositsweresecuredonlybythegeneralassetsofthebank.
Whendepositorslostconfidenceinthesolvencyofabank,theywouldinsistthatthebankhonoritsobligationtoconvertdepositsintospecie.
Asbankshadalimitedsupplyofspecieinreserve,theycouldnotaccommodatelargerunsbydepositors,whichincreasedtheincentivesofdepositorstorunatthefirstsignoftrouble.
Stategovernmentsmadenumerousattemptstostabilizedepositsthroughinsuranceschemes,butmostofthemfailed.
Asaresult,theindustrycreatedcollectivesknownasclearinghouses,whichcarefullymonitoredthefinancialconditionoftheirmembersandstoodbehindtheirliabilitiesintheeventofarunbydepositors.
12ThefirstclearinghouseswereestablishedbyNewYorkbanksin1853.
Gorton(1985)documentsthatwhenonememberfacedarun,theclearinghousessuspendedtheproductionofbank-specificfinancialinformationandinsteadpublishedfinancialinformationonallmemberstogether.
Inordertopreventthecostlyliquidationofilliquidassetslikeloans,theclearinghousesissuedloancertificatestomembers,securedbymembers'assets.
Thesecertificatescouldbeusedintheclearingprocessinplaceofcurrency,whichfreedupcurrencytoaccommodatewithdrawalsbydepositors.
Duringthepanicsof1893and1907,theclearinghousesissuedloancertificatesdirectlytothepublic,permittingdepositorstoreplacetheirclaimsonabankwithaclaimontheclearinghouse.
WhilethecreationoftheFederalReservein1913wasintendedtobringstabilitytothebankingsystembyreplacingthesystemclearinghouses,thecentralbankdidnotbegintoactasalenderoflastresortuntilwellaftertheGreatDepression.
Consequently,itwastheenactmentoffederaldepositinsurancein1933thatfirstbroughtstabilitytodemanddeposits.
Overthepastthirtyyears,significantinnovationsinthecompositionoftheaggregatemoneysupplyhavemadethefinancialsystemmorevulnerabletoalossofconfidencebytheholdersofmoney.
Inparticular,moneymarketmutualfundsweredevelopedinthe1970sinresponsetolimitsontheabilityofdepositoryinstitutionstopayinterestoncheckingaccounts,aswellasinresponsetoaneedforlimitsondepositinsurance,whichleftlargedepositorsexposedtobankrisk.
Oneofthemaininvestmentsofmoneymarketmutualfundsisovernightrepurchaseagreements,theequivalentofbanknotessecuredbycollateral,mostoftenU.
S.
Treasuryobligations.
Seekingstability,financialinnovationtransformeduninsureddepositsintoaninstrumentthatlookslikeaninsureddepositintheformofanovernightrepurchaseobligation.
Sunderam(2012)explorestheextenttowhichshadowbankingliabilitiesconstitutesubstitutesforhigh-poweredmoney.
Heshowsinasimplemodelthatshadowbankingliabilitiesshouldconstitutesubstitutesformoneyintheprivatesector'sassetallocation.
Empirically,Sunderamshowsthatshadowbankingliabilitiesrespondtomoneydemand,extrapolatingthatheightenedmoneydemandcanexplainabouthalfofthegrowthofABCPinthemid-2000s.
HealsoconfirmsthatregulatorychangestoABCPplayedasignificantroleinthegrowthoftheshadowbankingsystem,athemethatweturntointhenextsection.
B)Capital,Tax,andAccountingArbitrageWedefineshadowbankingactivitiesasbankingintermediationwithoutpublicliquidityandcreditguarantees.
ThevalueofpublicguaranteeswasrigorouslymodeledbyMerton(1977)usinganoptionspricingapproach.
MertonandBodie(1993)proposethefunctionalapproachtofinancialintermediation,whichisananalysisoffinancialintermediariesinrelationtotheamountofrisk-sharingthattheyachieveviaguarantees.
Pozsar,Adrian,Ashcraft,Boesky(2010)provideacomprehensiveoverviewofshadowbankinginstitutionsandactivitiesthatcanbeviewedasafunctionalanalysisofmarket-basedcreditintermediation.
Manyoftheirinsightsarecomprisedinmapsoftheshadowbankingsystemthatprovideablueprintofthefundingflows.
LevitinandWachter(2011)provideaquantitativeassessment13oftheroleofimplicitguaranteesforthesupplyofmortgages.
ThereisalsoalargeliteraturethatstudiestheimplicitguaranteesoftheGSEs(seePassmore,Sherlund,andBurgess(2005),FrameandWhite(2005),andAcharya,Richardson,Nieuwerburgh,andWhite(2011)).
Acharya,Schnabl,andSuarez(2011)documentthattherapidexpansionofABCPsince2004resultedfromchangesinregulatorycapitalrules.
Inparticular,FASBissuedadirectiveinJanuary2003(FIN46)andupdatedthedirectiveinDecember2003(FIN46A)suggestingthatsponsoringbanksshouldconsolidateassetsinABCPconduitsontotheirbalancedsheets.
However,U.
S.
bankingregulatorsclarifiedthatassetsconsolidatedontobalancesheetsfromconduitswouldnotneedtobeincludedinthemeasurementofrisk-basedcapitalandinsteaduseda10percentcreditconversionfactorfortheamountcoveredbyaliquidityguarantee.
Acharya,Schnabl,andSuarezdocumentthatthemajorityofguaranteeswerestructuredasliquidity-enhancingguaranteesaimedatminimizingregulatorycapital,insteadofcreditguarantees,andthatthemajorityofconduitsweresupportedbycommercialbankssubjecttothemoststringentcapitalrequirements.
Moreover,theauthorsdocumentthatconduitsweresponsoredbybankswithloweconomiccapitalasmeasuredbytheratioofthebookvalueofequitytoassets.
Finally,theauthorsfindthatinvestorsinconduitswithliquidityguaranteeswererepaidinfull,whileinvestorsinconduitswithweakerguaranteessufferedsmalllosses,suggestingtherewasnorisktransferdespitethecapitalrelief.
Themotivationforcapitalarbitrageisconsistentwiththemispricingofexplicitcreditandliquidityputoptionsassociatedwithdepositinsuranceandaccesstoofficialliquidity,aswellasthepresenceofaperceptionthatlargebanksare"toobigtofail,"whichpermitsthemtoengageinexcessiveleveragematuritytransformation.
Thepresenceofminimumcapitalandliquiditystandardsmitigatestheseincentives,andtheabilityofbankstoevadebindingstandardspermitsthemtomaximizethevalueoftheseputoptions.
C)OtherAgencyProblemsinFinancialMarketsAshcraftandSchuermann(2008)describesevenimportantinformationalfrictionsthatexistedinthesecuritizationofsubprimemortgagecreditpriortothefinancialcrisis,althoughthesefrictionscanbegeneralizedtoallsecuritizationtransactions.
Theyincludeasymmetricinformationproblemsbetweenthelenderandoriginator(predatorylendingandborrowing),betweenthelenderandinvestors,betweentheservicerandinvestors,betweentheservicerandborrower,betweenthebeneficiaryofinvestedfundsandassetmanagers,andbetweenthebeneficiaryofinvestedfundsandcreditratingagencies.
Inaddition,asymmetricinformationbetweeninvestorsandissuersresultsinrisk-insensitivecostoffunding.
Forexample,Keysetal.
(2010)documentthatmortgageborrowerswithFICOscoresjustaboveathresholdof620performsignificantlyworsethanborrowerswithFICOscoresjustbelow620.
Asitismoredifficulttosecuritizeloansbelowthatthreshold,theauthorsarguethatthisresultisconsistentwithissuersexploitingasymmetricinformation,disruptingtheotherwisemonotonerelationshipbetweenborrowercreditscoresandperformance.
14Althoughsecuritizationhasarelativelyshorthistory,itisatroubledone.
ThefirstknownsecuritizationtransactionsintheUnitedStatesoccurredinthe1920s,whencommercialrealestate(CRE)bondhousessoldloanstofinanceCREtoretailinvestorsthroughavehicleknownasCREbonds.
WiggersandAshcraft(2012)documenttheperformanceofthesebonds,whichdefaultedinlargenumbersfollowingtheonsetoftheGreatDepression.
Althoughthesharpdeteriorationineconomicconditionsplayedanimportantpartinexplainingtheirpoorperformance,sodidaggressiveunderwritingandsalesofthebondsinsmalldenominationstounsophisticatedretailinvestors.
Duringthe1990snofewerthanfivedifferentsectorsofABSranintotrouble,includingbutnotlimitedtohomeequity,homeimprovementlending,manufacturedhousing,equipmentleasing,andfranchiseloans.
Ineachofthesecases,therewasgenerallymeaningfulriskretentionbyasponsorusingsecuritizationasasourceoffunding.
However,onecommonthemeappearstohavebeentheaggressivepursuitofgain-on-sale-relatedearningsfromsecuritizationinadvanceofaninitialpublicoffering,andthiswasoftenachievedthroughcompetitiononunderwritingstandards.
Incontrast,thechallengesofsecuritizationinthe2000swereconcentratedinmultisectorCDOsin2002aswellasRMBSandCMBSin2005-07.
Thesecreditcyclesweremorelikelytoinvolvefirmsusingsecuritizationforarbitrageandwereusedasasourceoffeeincomewithminimalintendedriskretention,althoughmanyofthemwereleftholdingwarehousesofloansasthefinancialcrisisunfolded.
Over-relianceoncreditratingscancreateproblemswhentheratingagenciesfacetheirownagencyproblems.
Forexample,Mathis,McAndrews,Rochet(2009)analyzeadynamicmodelofratingswherereputationisendogenousandthemarketenvironmentmayvaryovertime.
Theauthors'modelpredictsthataratingagencyislikelytoissuelessaccurateratingsinboomtimesthanitwouldduringrecessionaryperiods.
Moreover,theauthorsdemonstratethatcompetitionamongratingagenciesyieldssimilarqualitativeresults.
XiaandStrobl(2012)documentthattheconflictofinterestcausedbytheissuer-paysratingmodelleadstoinflatedcorporatecreditratings.
Inparticular,theauthorscomparetheratingsissuedbyanissuer-paidratingagencywithaninvestor-paidagencyanddemonstratethatthedifferencebetweenthetwoismorepronouncedwhenissuer-paidagency'sconflictofinterestisparticularlysevere.
Forexample,theissuer-paidagencyhasmorefavorableratingsforfirmswithmoreshort-termdebt,anewlyappointedCEOorCFO,andalowerpercentageofpastbondissuesratedbytheagency.
However,theauthorsfindnoevidencethatthesevariablesarerelatedtocorporatebondyieldspreads,whichsuggeststhatinvestorsmaybeunawareofincentiveproblemsattheissuer-paidagency.
Cohen(2010)documentssignificantrelationshipsbetweenvariablesthatshouldnotaffectaCRA'sviewofthecreditriskofconduit/fusionCMBStransactionsissuedduring2001-07,butthatwouldaffectissuers'andCRAs'incentivesinanenvironmentwhereratingshoppingwaspresent.
153)HowDoesShadowCreditIntermediationWorkPozsar,Adrian,Ashcraft,andBoesky(2010)makeadistinctionbetweenthe"internal,""external,""independent,"and"governmentsponsored"shadowbankingsystem.
Theinternalsystemconsistsofshadowbankingactivitiesconductedundertheauspicesofbankholdingcompanies.
Theexternalsystemcomprisesshadowbankingactivitiesthatareconductedbymajornonbankfinancialinstitutionssuchasnonbank-affiliatedbroker-dealersorinsurancecompanies.
Independentshadowbankinginstitutionsarespecializedshadowbankingvehiclesthatareindependentofanymajorfinancialinstitutions.
Finally,thegovernment-sponsoredshadowbankingsystemconsistsofinstitutionsthatprovidecreditintermediationserviceswithimplicitgovernmentguarantees.
Beforediscussingthevariousshadowbankinginstitutionsindetail,wereviewthe"sevensteps"ofshadowcreditintermediation.
A)TheSevenStepsofShadowCreditIntermediationTheshadowbankingsystemisorganizedaroundsecuritizationandwholesalefunding.
Loans,leases,andmortgagesaresecuritizedandthusbecometradableinstruments.
Fundingisalsointheformoftradableinstruments,suchascommercialpaperandrepo.
Saversholdmoneymarketbalances,insteadofdepositswithbanks.
Theshadowbankingsystemdecomposesthecreditintermediationintoachainofwholesale-funded,securitization-basedlending.
Throughtheshadowintermediationprocess,theshadowbankingsystemtransformsrisky,long-termloans(subprimemortgages,forexample)intoseeminglycredit-risk-free,short-term,money-likeinstruments.
Shadowcreditintermediationisperformedthroughchainsofnonbankfinancialintermediariesinamultistepprocessthatcanbeinterpretedasa"verticalslicing"ofthetraditionalbank'screditintermediationprocessintosevensteps.
Pozsar,Adrian,Ashcraft,andBoesky(2010)explainthesevenstepsofshadowbankcreditintermediation:1.
Loanorigination(autoloansandleases,nonconformingmortgages,etc.
)isperformedbyfinancecompanies.
2.
Loanwarehousingisconductedbysingle-andmulti-sellerconduitsandisfundedthroughasset-backedcommercialpaper(ABCP).
3.
Thepoolingandstructuringofloansintotermasset-backedsecurities(ABS)isconductedbybroker-dealers'ABSsyndicatedesks.
4.
ABSwarehousingisfacilitatedthroughtradingbooksandisfundedthroughrepos,totalreturnswaps,orhybridandrepoconduits.
5.
ThepoolingandstructuringofABSintoCDOsisalsoconductedbybroker-dealers'ABS.
166.
ABSintermediationisperformedbylimited-purposefinancecompanies(LPFCs),structuredinvestmentvehicles(SIVs),securitiesarbitrageconduits,andcredithedgefunds,whicharefundedinavarietyofwaysincluding,forexample,repo,ABCP,MTNs,bonds,andcapitalnotes.
7.
Thefundingofalltheaboveactivitiesandentitiesisconductedinwholesalefundingmarketsbyfundingproviderssuchasregulatedandunregulatedmoneymarketintermediaries(forexample,2(a)-7MMMFsandenhancedcashfunds,respectively)anddirectmoneymarketinvestors(suchassecuritieslenders).
Inadditiontothesecashinvestors,whichfundshadowbanksthroughshort-termrepo,CP,andABCPinstruments,fixed-incomemutualfunds,pensionfunds,andinsurancecompaniesalsofundshadowbanksbyinvestingintheirlonger-termMTNsandbonds.
Notallintermediationchainsinvolveallsevensteps,andsomemightinvolveevenmoresteps.
Forexample,anintermediationchainmightstopatthesecondstepifapoolofprimeautoloansissoldbyacaptivefinancecompanytoabank-sponsoredmulti-sellerconduitfortermwarehousingpurposes.
Inanotherexample,ABSCDOscouldbefurtherrepackagedintoaCDO^2,whichwouldelongatetheintermediationchaintoincludeeightsteps.
Typically,thepooreranunderlyingloanpool'squalityatthebeginningofthechain(forexample,apoolofsubprimemortgages),thelongerthecreditintermediationchainthatwouldberequiredto"polish"thequalityoftheunderlyingloanstothestandardsofmoneymarketmutualfundsandsimilarfunds.
Asaruleofthumb,theintermediationoflow-qualitylong-termloans(nonconformingmortgages)involvedallsevenormoresteps,whereastheintermediationofhigh-qualityshort-tomedium-termloans(creditcardandautoloans)involvedusuallythreesteps(andrarelymore).
Theintermediationchainalwaysstartswithoriginationandendswithwholesalefunding,andeachshadowbankappearsonlyonceintheprocess.
B)CommercialBanksandShadowBankingPerdefinition,creditintermediationactivityonthebalancesheetsofcommercialbanksdoesnotconstituteshadowbanking,asithasaccesstoofficialliquidityandcreditguaranteesbytheFederalReserveandtheFDIC.
However,commercialbankscanbeinvolvedinshadowbankingactivitiesinseveralways.
Forexample,theycanprovidecreditandliquiditylinestoshadowbankentitiessuchasFigure3:TheShadowCreditIntermediationProcessSource:Pozsar,Adrian,Ashcraft,andBoesky(2010)ABCP,repo"FundingFlows"CPABCPRepoABCP,repoCP,repoABCP$1NAVLoansLoansLoansABSABSABSCDOMaturityandLiquidityTransformationLoanOriginationLoanWarehousingABSIssuanceABSWarehousingABSCDOIssuanceABSIntermediationWholesaleFundingCredit,MaturityandLiquidityTransformationCredit,MaturityandLiquidityTransformationCreditTransformation(Blending)Credit,MaturityandLiquidityTransformationCreditTransformation(Blending)Credit,MaturityandLiquidityTransformationStep6Step7Step1Step2Step3Step4Step5"AssetFlows"17conduits,ABSissuers,orSIVs.
Commercialbanksinturnareownedbybankholdingcompanies(BHCs).
Mandel,Morgan,andWei(2012)provideadetailedanalysisofcommercialbanks'sponsorshipsofshadowbankingactivities.
ManyshadowbankingactivitiesareconductedundertheauspicesofBHCs.
Forexample,aBHCmightownawealthmanagementunitwithamoneymarketmutualfund,whichwewouldconsiderashadowbankinternaltotheBHC.
Anotherexampleistri-partyrepofundingbythebroker-dealersubsidiariesofBHCs.
AthirdexampleistheuseofABCPconduits,whichareoffbalancesheettotheBHC,butaresponsoredbythecommercialbanksubsidiaryoftheBHCviacreditandliquiditylines.
OnegaugeoftheextentofshadowbankingactivitybyBHCsistheirorganizationalcomplexity.
Whiletraditionalbankingisdoneinasingleentity,BHCstendtohavehundredsorthousandsofsubsidiaries,mostofwhichdonothavedirect,explicitaccesstopubliccreditandliquidityputs.
Forexample,Avraham,Selvaggi,andVickery(2012)documentthateachofthefivelargestBHCsintheU.
S.
hadover1,500subsidiariesin2012,withthelargestoneowningmorethan3,000.
Whilesomeofthesesubsidiariesareforeignbanks,mostofthemarenonbanksubsidiariesintheUnitedStates.
Themajorityofthesubsidiariesarefunds,trusts,andfinancialvehiclesthataretypicallyengaginginshadowbankingactivities.
Infact,Copeland(2012)showsthattheseshadowbankingactivitiesofbankholdingcompanieshavebeenincreasingovertimeandrepresentaquantitativelyimportantshareoftheholdingcompanies'totalearnings.
CetorelliandPeristiani(2012)investigatetheroleofBHCsinassetsecuritizations.
TheyassessquantitativelythedegreetowhichcommercialbanksareinvolvedinABS,CMBS,andCDOissuance,andinservicingandunderwritingsecuritizations.
FornonagencyABS,theBHCmarketshareisbetween35and75percentforunderwriting,issuance,andservicingandcloseto100percentfortrustservices.
Incontrast,forprivate-labelmortgages,themarketshareofBHCshasincreaseddramaticallyoverthepasttwentyyearstoover60percentforunderwriting,servicing,andissuance.
ThissuggeststhatthepresenceofBHCsinshadowbankingactivitiesrelatingdirectlytosecuritizationsissubstantial.
BordandSantos(2012)studytheroleofbanksintheoriginate-to-distributemodelofcreditintermediation.
TheybasetheirstudyondatafromtheSharedNationalCreditProgram(SNC).
Thedataarecollectedbythenationalsupervisorybankingagencies(FederalReserve,theFDIC,ortheOCC),whichtrackcreditheldbyfederallysupervisedinstitutions.
Unlikeanyformofpubliclyavailabledata,theSNCallowsBordandSantostotracktheownershipofloansbyvariousinstitutionsovertime.
Theauthorsdocumentthatmorethan75percentofsyndicatedcreditlinesareboughtbysyndicateparticipantbanksandthattheystaywiththosebanksafterthreeyears.
Theshareoftermloansownedbysyndicatebankshasfallenfromaround75percentinthemid-1990stoaround30percentinthemid-2010s.
Fortermloans,shadowbankingorganizationshavethusemergedasmoreandmoreimportantinvestorsoverthepasttwentyyears.
Buyersoftermloansthatareparticularlyimportantareinvestmentmanagersandcollateralizedloanobligations(CLOs).
BordandSantosconcludethattheshareoftermloanssoldtotheshadowbankingsystemamountedtolessthan10percentin1993androsetoover30percentby2007.
Whileloanoriginationsareconductedalmostexclusivelybycommercialbanks,theultimateownersoftermloansarethussplitamongbanksandshadowbanks.
18BesidesthesubsidiariesassociatedwithBHCinvolvementinsecuritizationactivities,thelargestnonbankBHCsubsidiariesconsistoffinancecompanies,broker-dealers,andwealthmanagementunitsincludingmutual,hedge,andmoneymarketmutualfunds.
Inmanyrespects,thefinancialcrisisof2007-09hasledtoafinancialsystemwheretheBHCsownalargershareofnonbanksubsidiariesthatconductshadowbankingactivities.
Forexample,thefivelargestindependentbroker-dealerspriortothecrisisallwereabsorbedby,ortransformedinto,BHCs.
Similarly,someofthelargestindependentissuers,originators,andservicersofprivate-labelmortgageswereabsorbedbyBHCs.
Sowhilethetwodecadesintherun-uptothefinancialcrisissawtheemergenceofashadowbankingsystemthatwaspartiallyindependentfromBHCs,thefinancialcrisisled,perhapsparadoxically,toamigrationofindependentshadowbankingactivityintoBHCs.
Cetorelli(2012)showsthat,asof2011,BHCscontrolledabout38percentoftheassetsofthelargestinsurancecompanies,41percentoftotalmoneymarketmutualfundassets,and93percentoftheassetsofthelargestbrokersanddealers.
Moreover,verylittlesecuritieslendingandrelatedcashcollateralreinvestmentstakeplacewithouttheservicesprovidedbythemaincustodianbanks.
ThetrendtowardconsolidationofshadowbankinginBHCssincethecrisisis,however,counteractedbyapowerfulforce:theenhancedprudentialstandardsofBHCs.
TightercapitalandliquidityrequirementswillarguablyleadtoanincreasedincentiveforsomeformsofcreditintermediationtomigrateoutofBHCsandintotheshadowbankingsystem.
Thistrendhasbeenobservedinproprietarytrading,whichhaslargelymigratedfromtheBHCstoindependentlyrunhedgefundsinanticipationoftheVolckerRule(seeDuffie(2012)foranacademicassessmentoftherule).
Furthermore,theCLOmarketcontinuestothriveandisatleastpartiallyindependentfromBHCs.
Abroadermovementofsecuritization-relatedactivityfromBHCstoshadowbankinginstitutionscanbeexpectedovertime.
C)ExternalandIndependentShadowBankingExternalshadowbankingentitiesareregulatedinstitutionsthatareindependentofanyentitieswithdirect,explicitgovernmentbackstopsandthatconductshadowbankingactivities.
Examplesofsuchinstitutionsarestand-alonebroker-dealers;independentwealthmanagementfirmsthatrunmoneymarketfunds;credithedgefunds;andfinancecompaniesthatareaffiliatedwithindustrialcompaniessuchastheautoloansubsidiariesofcarmanufacturers.
Independentshadowbanksareinstitutionsindependentofthegovernmentsafetynet;theyaresetuptoengageexclusivelyincertainshadowbankingactivities.
Incontrasttotheinstitutionsoftheexternalshadowbankingsystem,independentshadowbanksdonothavenon-shadowbankingactivitiesastheirprimarybusiness,butratherspecializeonlyinshadowbanking.
ExamplesofindependentlyrunshadowbanksarenonbankaffiliatedSIVs,stand-alonemoneymarketmutualfunds,independentCDOsandCLOs,andthemajorityofABSandprivate-labelRMBSandCMBS.
Thefifthandsixthstepsofthecreditintermediationchainrelyheavilyonprivatecreditriskrepositoriestoperformoriginate-to-distributesecuritizations.
Privateriskrepositoriesspecializein19providingcredittransformationservicesintheshadowbankingsystemandincludemortgageinsurers,monolineinsurers,certainsubsidiariesoflarge,diversifiedinsurancecompanies,credithedgefunds,andcreditderivativeproductcompanies.
Theseentities,asinvestorsinthejuniorequityandmezzaninetranchesofloanpools,allprovideriskcapitaltotheshadowbankingsystem,therebysupportingcreditextensiontotherealeconomy.
Differentcreditriskrepositoriescorrespondtospecificstagesoftheshadowcreditintermediationprocess.
Assuch,mortgageinsurersspecializeininsuringorwrappingwholemortgageloans;monolineinsurersspecializeinwrappingABStranches(ortheloansbackingaspecificABS'stranches);andlarge,diversifiedinsurancecompanies,credithedgefunds,andcreditderivativeproductcompaniesspecializeintakingontherisksofABSCDOtranchesthroughCDS.
Therearealsooverlaps,withsomemonolineswrappingbothABSandABSCDOs,forexample.
Effectively,thevariousformsofcreditputoptionsprovidedbyprivateriskrepositoriesabsorbtailriskfromloanpools,turningtheenhancedsecuritiesintosecuritiesthatarefreefromcreditrisk(atleastfrominvestors'perceptionspriortothecrisis).
Thisinturnmeansthatanyliabilitythatissuedagainsttheseassetsisperceivedtobefreeofcreditriskaswell,justasifitisFDIC-insured.
Theperceivedcredit-risk-freenatureoftraditionalbanks'andshadowbanks'liabilitiesstemsfromtwoverydifferentsources.
Inthecaseoftraditionalbanks'insuredliabilities(deposits),thecreditqualityisdrivenbythecounterparty—theU.
S.
taxpayer.
Asaresult,insureddepositorsinvestlesseffortintoexaminingabank'screditworthinessbeforedepositingmoneythantheywouldiftheywereuninsured.
Inthecaseofshadowbanks'liabilities(repoorABCP,forexample),perceivedcreditqualityisdrivenbythe"credit-riskfree"natureofcollateralthatbacksshadowbankliabilities,asitisoftenenhancedbyprivatecreditriskrepositories.
Thecreditputsprovidedbyprivatecreditriskrepositoriesarealternativestothecredittransformationperformedby1)thecredit-risk-basedcalibrationofadvanceratesandattachmentpointsonloanpoolsbackingtop-ratedABCPandABStranches,respectively;2)thecredit-risk-basedcalibrationofhaircutsoncollateralbackingrepotransactions;3)thecapitalnotessupportingLPFCs'andSIVs'portfoliosofassets;and4)thepoolingandrepackagingofnon-AAA-ratedtermABSintoABSCDOs.
ThecreditputsofprivatecreditriskrepositoriesarealsosimilarinfunctiontothewrapsprovidedbyFannieMaeandFreddieMaconconformingmortgagepools.
Justasthesegovernment-sponsored,publiccreditriskrepositories"borrowed"theAAA-ratingofthefederalgovernmentandextendedittopoolsofmortgageloans(turningthemintocredit-risk-freerateproducts),theprivatecreditriskrepositorieswereeffectively"borrowing"theAAAratingoftheirparent.
D)Government-SponsoredShadowBankingInmanyways,themodernshadowbankingsystemoriginatedinthegovernmentsector.
Securitizationwasfirstconductedbygovernment-sponsoredenterprises(GSE),whicharecomprisedoftheFHLBsystem(1932),FannieMae(1938),andFreddieMac(1970).
TheGSEshavedramaticallyimpactedthewayinwhichbanksarefundedandthewayinwhichtheyconductcredittransformation:TheFHLBswerethefirstprovidersoftermwarehousingofloans,andFannieMaeandFreddieMacpioneeredtheoriginate-to-distributemodelofsecuritizedcreditintermediation.
20Likebanks,theGSEsfundtheirloanandsecuritiesportfolioswithamaturitymismatch.
Unlikebanks,however,theGSEsarefundednotthroughdeposits,butthroughcapitalmarkets,wheretheyissueshort-andlong-termagencydebtsecurities.
Theseagencydebtsecuritiesareboughtbymoneymarketinvestorsandrealmoneyinvestorssuchasfixed-incomemutualfunds.
ThefundingfunctionsperformedbytheGSEsonbehalfofbanksandthewayinwhichGSEsarefundedarethemodelsforwholesalefundingmarkets.
TheGSEsuseseveralsecuritizationtechniques.
TheyusetermloanwarehousingservicesprovidedbytheFHLBs.
TheyalsousecreditrisktransferandtransformationthroughcreditinsuranceprovidedbyFannieMaeandFreddieMac.
SecuritizationfunctionsareprovidedbyFannieMaeandFreddieMac.
MaturitytransformationisconductedontheGSEs'balancesheetsthroughretainedportfolios.
ThesesecuritizationtechniquesfirstusedbytheGSEswereadoptedandimitatedbybanksandnonbankstogeneratethenongovernmentalshadowbankingsystem.
Theadaptationofthesetechniquesgaverisetothesecuritization-based,originate-to-distributecreditintermediationprocess.
4)WhyDoesShadowCreditIntermediationNeedtoBeRegulatedDuringthefinancialcrisisof2007-09,theshadowbankingsystemcollapsed.
ABCPconduitsexperiencedaseriesofruns.
Ofthefivemajorinvestmentbanks,onefailed,twowereacquiredbybanks,andtwoweretransformedintobankholdingcompanies.
Securitizationactivitytotallystopped,andmanyshadowbankinginstitutionssuchasSIVsandCDOsallbutdisappeared.
Thecollapseofshadowbankinginstitutionsandshadowbankingactivitiesoccurredbothontheassetandliabilitysides.
Ontheassetside,themainissuesweretheunderwritingstandards.
Ontheliabilityside,themainissueswererelatedtothefragilityofwholesalefunding.
Asaresultofthesefragilities,thegovernmentsectorsetupavarietyofbackstopsfortheshadowbankingsystem.
Thesebackstopsconsistedofbothliquidityfacilitiesandsolvencyguarantees.
Theywerecreatedbecauseofthepotentialforshadowbankdistresstospillovertootherinstitutionsanddamagetherealeconomy.
A)AssetQualityBecausetheyaretailoredtotakeadvantageofmispricedtailrisk,shadowbankinginstitutionsaccumulateassetsthatareparticularlysensitivetotailevents.
Atadeeplevel,thequestionbecomes,howcanthemispricingoftailriskexistinaworldwithfullyrationalactorsShouldn'tfinancialmarketparticipantsbeabletocalculatetailriskprobabilities,implicitguarantees,andvarioustailriskenhancementsAndshouldn'tthesecalculationsleadtotheproperassessmentoftailriskTheliteraturehasprovidedtwodistinct,complementaryanswers.
Thefirstreliesonthebehavioralexplanationof"neglectedrisk.
"Thesecondreliesoninformationopacityinarationalworld.
Wewilldiscusseachoftheseexplanationsinturn.
21Evidencefrompsychologyandbehavioralfinancearguesthatmarketparticipantsarefundamentallybiasedagainsttherationalassessmentoftailrisk.
Gennaioli,Shleifer,andVishny(2012a)developatheoryofindividualdecisionmakingbasedonthebehavioralevidence,positingthatactorsneglectrisk.
Inalaterpaper,Gennaioli,Shleifer,andVishny(2012b)applythistheorytotheeconomicsoftheshadowbankingsystem.
Theymodelaworldwhereinvestorssystematicallyignoretheworststateoftheworld,generatingoverinvestmentandoverpricingduringtheboomandexcessivecollapseofrealactivityandthefinancialsectorduringthebust.
Theirtheoryispossiblythemostparsimoniousnarrativeoftheboomandbustoftheshadowbankingsystem.
Infact,muchempiricalevidenceisconsistentwithsuchatheory.
Creditratingagenciesmodeledonlysmallornodeclinesinaggregatehousingprices,andinvestorsinsecuritizedproductsoftendidnotunderstandtheamountofriskexposurethatwasembeddedintheproducts.
Meanwhile,thepricesoftailriskfarintothefuture,faroutofthemoneyoptionsrelatingtomortgagecredit,weresurprisinglycheap.
Anearlypaperwarningofthefinancialsystem'sexposuretosuchtailriskwaspresentedbyRajan(2005),whopointedtopreciselythisphenomenonbyaskingwhetherfinancialinnovationhadmadetheworldriskier.
Neglectedrisksareonewaytointerpretthewidelyperceivedrisk-freenatureofhighlyratedstructuredcreditproducts,suchastheAAAtranchesofABS.
Coval,Jurek,andStafford(2009)pointoutthattheseAAAtranchesbehavelikecatastrophebondsthatloadonasystemicriskstate.
Insuchasystemicriskstate,assetsbecomemuchmorecorrelatedthaninnormaltimes.
Theunderestimationofcorrelationenabledfinancialinstitutionstoholdinsufficientamountsofliquidityandcapitalagainsttheputsthatunderpinnedthestabilityoftheshadowbankingsystem,whichmadetheseputsundulycheaptosell.
Asinvestorstendtooverestimatethevalueofprivatecreditandliquidityenhancementpurchasedthroughtheseputs,theresultisanexcesssupplyofcheapcredit.
Adrian,Moench,andShin(2009)documenttheclosecorrespondencebetweenthepricingofriskandthefluctuationsofshadowbankandbroker-dealerbalancesheets.
Timesoflow-riskpremiatendtobeassociatedwithexpandingbalancesheets—infact,intermediarybalance-sheetdevelopmentspredictthepricingofriskacrossmanyassetclasses.
Neglectedriskcanmanifestitselfthroughover-relianceoncreditratingsbyinvestors.
Forexample,Ashcraftetal.
(2011)documentthatsubprimeMBSpricesaremoresensitivetoratingsthanexpostperformance,suggestingthatfundingisexcessivelysensitivetocreditratingsrelativetoinformationalcontent.
Dang,Gorton,andHolmstrm(2009)presentanalternativetheorywhere,inaworldwithfullyrationalmarketparticipants,assetsarehighlyexposedtotailrisk.
Theirsisatheoryofinformationopacitythatcanserveasarationalizationofcreditproblemsfortheshadowbankingsystem.
Accordingtothistheory,debtcontractsareoptimalbecausetheygenerateopacity.
Opacity,inturn,minimizesadverseselectionandprovidestheleastpossibleincentivestocollectinformation.
Thisinsightjustifiesthegrowthofrelativelyopaquesecuritizedproductsintherun-uptothecrisis.
MortgagesandloanswerepackagedintoMBSandABSandfundedbyCDOs,SIVs,andMMMFsthathadrelativelylittleinformationabouttheunderlyingcreditquality.
However,Dang,Gorton,andHolmstrmshowthatsystemicriskisexacerbatedonceabadshockhitsinformationallyopaque,debt-fundedeconomies.
Theintuitionisthatabadshockleadstoanincreaseinprivateinformationcollection,whichexacerbatesthe22incorporationofadverseinformationinmarketprices.
Asaresult,adverseselectionstartstoaccumulateassystemiccrisesdeepen.
TheabovetheorycomplementstheexplanationbyGennaioli,Shleifer,andVishny(GSV)discussedearlier.
WhileDang,Gorton,andHolmstrm(DGH)emphasizeadverseselectionasanamplificationmechanism,GSVemphasizeawarenessofrisk.
InGSV,theriskinessoftheworststateoftheworldissimplyneglected,andthisneglectisbasedonbehavioralarguments.
Incontrast,inDGH'smodel,theopacityoffinancialcontractsingoodtimesisanequilibriumoutcomethatmaximizestheliquidityoffinancialcontracts.
Thecommonalitybetweenthetwotheoriesisthattheseverityoffinancialcrisisisneglected,eitherrationallyorbehaviorally.
Asaresult,thetailriskembeddedindebtsecuritiesisunderpricedfromanexpostpointofview.
InbothDGHandGSV,theassetsthatareaccumulatedduringtheboomexperiencelargeassetpricedeclinesduringtimesofcrisis.
Suchtheoriesofneglectedriskthusprovidearationalizationfortheaccumulationofriskexposuretothehousingmarketthatwasthemajoraggregateriskoftheshadowbankingsystem.
InthetheoriesofDGHandGSV,securitiessuchasABSandCDOsthatobscuretheunderlyingcreditrisksarisenaturally.
Suchsecurities,inturn,generatelargelossesintimesofcrisis.
SuchexplanationsoftheboomandbustcycleofsecuritizedcreditproductsneedtobecomplementedwiththeadditionalinsightsfromtheoriesdiscussedinSection2.
Thosealternativetheoriesprovideadditionalreasonsthatgiverisetothesystematicmispricingoftailrisk.
Inparticular,theimplicitorindirectaccesstogovernmentbackstopsviacreditandliquidityputsfrombankholdingcompaniesandinsurancecompaniesleadstotheunderpricingoftailriskandtheexcessivebuildupofsystemicrisk.
Theagencyconflictsarisingatvariousstagesoftheshadowbankingsystem,andparticularlythemisalignedincentivesofcreditratingagencies,willleadtoexcessiverisktakingingoodtimes,withassociatedexcessivecreditlossesintimesofcrisis.
Finally,thefailuretoprovideadequatefinancialdisclosurebasedonaccountingrulesgeneratesanadditionalmarketfailurethatfacilitatesexcessiverisktakingwiththeassociatedlargelossesindownturns.
B)FundingFragilityThefinancialfrictionsthatleadtoexcessiverisktakingandexacerbatedcreditlossesduringdownturnsalsointeractwiththefragilityoffunding.
Perdefinition,fundingsourcesforshadowbankingactivitiesareuninsuredandthusrunnable.
Inmanyways,thefragilityofshadowbanksduetotherun-abilityofliabilitiesresemblesthebankingsystemofthe19thcentury,priortothecreationoftheFederalReserveandtheFDIC.
Duringthattime,bankrunswerecommon,andtheyoftenhadsevereconsequencesfortherealeconomy.
Theshadowbankingsystem'svulnerabilitytorunsbearsresemblancetobankrunsasmodeledbyDiamondandDybvig(1983).
Shadowbanksaresubjecttorunsbecauseassetshavelongermaturitiesthanliabilitiesandtendtobelessliquidaswell.
Whilethefundamentalreasonforcommercialbankrunsisthesequentialservicingconstraint,forshadowbankstheeffectiveconstraintisthepresenceof23firesaleexternalities.
Inarun,shadowbankingentitieshavetosellassetsatadiscount,whichdepressesmarketpricing.
Thisprovidesincentivestowithdrawfunding—beforeothershadowbankingdepositorsarrive.
However,theanalogybetweenbankrunsandshadowbankrunsgoesonlysofar.
Thereasonisthatshadowbankingentitiesdonotofferdemanddeposits,butinsteadobtainfundinginwholesalemoneymarketssuchascommercialpaperorrepo.
Martin,Skeie,andvonThadden(2011)provideamodelforaruninrepomarketsthattakestheempiricalfactsoftheBearStearnsandLehmancrisesasastartingpoint.
Intheirmodel,repoborrowersfaceconstraintsduetothescarcityofcollateralandtheliquidityofcollateral.
Undersufficientlyadverseconditions,self-fulfillingrunscanoccur.
Themodelfocusesinparticularonthedifferencesbetweenthetri-partyrepomarketandthebilateralrepomarket(seeAdrian,Begalle,Copeland,andMartin(2013)foranoverviewofbothmarkets).
Arguably,runsoccurredinbothmarkets,buttheywereofverydifferentnatures.
Whiletheruninthebilateralmarketwascharacterizedbyasharpincreaseinhaircuts(asdocumentedbyGortonandMetrick(2012)),theruninthetri-partyrepomarketmaterializedasasimplewithdrawaloffundingwitharatherlimitedimpactonthelevelofhaircuts(seeCopeland,Martin,andWalker(2011)).
RunsintheABCPmarketwereequallycharacterizedbyawithdrawaloffunding(seeCovitz,Liang,andSuarez(2012)).
Fundingfragilityofshadowbankinginstitutionscanalsobeinterpretedastheresultoftheleveragecyclesofmarket-basedfinancialinstitutions.
Suchleveragecyclesrefertoequilibriumoutcomes,whereassetvaluesandbalancesheetcapacityofintermediariesaredeterminedendogenously.
Thefrictioninmodelsofleveragecyclesisduetothefundingconstraintsofintermediaries,whichreflecttheincentiveproblemsdiscussedearlier.
TheoriesofintermediaryleveragecycleshavebeenproposedbyFostelandGeanakoplos(2008),BrunnermeierandPedersen(2009),BrunnermeierandSannikov(2011),GarleanuandPedersen(2011),andAdrianandBoyarchenko(2012).
Suchtheoriesofleveragecycleshavethecommonalitythatintermediariesaresubjecttocollateralconstraints,asisthecaseforrepoandABCPfunding.
Thetightnessofthecollateralconstraintsdependsontheunderlyingriskofassets,theliquidityofassets,andthecollateralvalues.
Aseconomicconditionsdeteriorate,theleveragecycleactsasanamplificationmechanismtounderlyingshocks.
AdrianandBoyarchenko(2012)showthattheirtheoryofintermediaryleveragecycleshasstrongempiricalsupport.
Intermediarybalancesheetsexhibitstronglyprocyclicalleverage,meaningthatleverageexpandsinbooms.
Thisprocyclicalbehaviorofleverageisahallmarkofshadowbanking,asdocumentedbyAdrianandShin(2009).
Shadowbankleveragetendstobehighwhenbalancesheetsarelargeandcreditintermediationisexpanding.
Furthermore,equityiscountercyclical,bothinthetheoryandinthedata,asintermediariestendtoholdaslittleequityaspossibleduringbooms,butareforcedtoraiseequityduringdownturnswhenthemarketriskincreases.
AdrianandBoyarchenko(2012)alsodocumentthecloselinkbetweenintermediarybalancesheetsandassetprices.
Overtime,expandingleveragetendstocoincidewithcompressedriskpremiaandinflatedassetprices.
Inbusts,riskpremiawiden,generatingassetpricebusts.
Inaddition,marketvolatilityiscountercyclical.
Asaresult,thefundingofintermediariestendstocollapseduringtimesofcrisis.
24Theadvantageofthegeneralequilibriumtheoriesofleveragecyclesisthattheyallowwelfareanalysis,thustyingfundingfragilityinfinancialcrisestopossiblepolicyinterventions.
Ingeneral,themarketequilibriumisnotwelfareoptimizing,andpoliciesthatmitigatethecyclecanenhancewelfare.
Inthecontextofshadowbanking,thisimpliesthattherun-upofshadowbankingjustbeforethecrisisanditscollapseduringthecrisisareexacerbatingthefinancialcycleinawaythatdoesnotenhancewelfare.
C)LiquidityBackstopsWhilethenextsectiondiscussesstructuralchangestotheregulatoryenvironmentthataimatreducingthefragilityoftheshadowbankingsystem,andultimatelyatreducingtheamplitudeoftheleveragecycle,thissubsectionfocusesonexpostpoliciesthatweredeployedduringthefinancialcrisisof2007-09tostabilizethecollapseofshadowbanking.
Thiscollapsereflectedproblemsontheassetandliabilityside—inotherwords,bothcreditproblemsandliquidityproblems.
WhiletheFederalReserveinitiatedprogramsthatprimarilyaimedatliquidityinjections,theU.
S.
Treasury'sprogramssoughttoresolvecreditproblems.
Thefinancialcrisisstartedin2007withthecollapseofABCPconduitsandSIVs.
Themajorityofthoseconduitsweresingle-sellerconduitsthatwereundertheumbrellaofparticularBHCs.
AstheassetqualityoftheABCPconduitsandSIVsdeteriorated,moneymarketinvestorswithdrewfunding,forcingthesponsoringBHCstoseekothersourcesoffunding.
Asaresult,assetsfromtheconduitsandSIVswerere-intermediatedandfundedinunsecuredmarkets,suchastheLibormarket.
Asaresult,fundingintheLibormarketbecamedisrupted,andtheFederalReserveinitiatedtwoprogramstoaddressthefundingliquidityshortage.
TheTermAuctionFacility(TAF)providedtermfundingforcommercialbanks,effectivelyreplacingthetermfundingthatevaporatedintheABCPmarket(seeArmantier,Krieger,andMcAndrews(2008)foradescriptionoftheTAF).
Inaddition,foreignexchangeswapsprovidedtermfundingforforeignbanksthatdidnothaveaccesstotheTAFviatheFed'sdiscountwindow.
InstitutionsthatheldU.
S.
dollarassetsthatcouldnolongerbefundedintheABCPmarketwerethusabletoobtainfundinginforeigncurrencies,andswapintodollarfundingattheforeigncentralbankbyusingtheFed'sforeignexchangeswaps.
Thenextdysfunctionduringthefinancialcrisisoccurredintherepomarket.
Inearly2008,haircutsintheDVPrepomarketstartedtoincreasesubstantially,leadingtoforceddeleveragingofmanyfixed-incomeandcredithedgefunds.
Forexample,Carlyle'sCCCfund,whichinvestedinagencymortgagesandwasfundedintheDVPrepomarket,hadtodeclarebankruptcyinFebruary2008.
Subsequently,therepofundingshortagealsoimpactedthetri-partyrepomarket.
IntheweekofMarch11,2008,BearStearnswasnolongerabletoobtaintri-partyrepofunding.
Inordertopreventthesefundingdifficultiesfromspreadingtootherinstitutions,theFederalReserveintroducedthePrimaryDealerCreditFacility(PDCF)onMarch15,2008(seeAdrian,Burke,andMcAndrews(2009)foradetailedexpositionofthefacility).
ThePDCFallowedprimarydealerstoobtainfundingfromtheFedandthuseffectivelyallowedthedealersectortojoindepositoryinstitutionsinhavingaccesstolast-resortlending.
Inaddition,the25TermSecurityLendingFacility(TSLF)alloweddealerstoexchangeagencymortgagecollateralagainstTreasurycollateral(seeFleming,Hrung,andKeane(2010)fordetailontheTSLF).
FollowingLehman'sbankruptcyonSeptember15,2008,moneymarketsexperiencedarun,leadingtofundingshortagesofABCP,CP,andrepoissuers.
TheFederalReserveintroducedtwofacilitiestoaddressthesemoneymarketdislocations:theAsset-BackedCommercialPaperMoneyMarketMutualFundLiquidityFacility(AMLF)andtheCommercialPaperFundingFacility(CPFF).
TheCPFFofferedafundingsourcetocommercialpaperissuersthatreplacedmoneymarketfundingintheaftermathofLehman.
TheCPFFwasconstructedtobeaself-liquidatingfacilityduetotermsthatmadeitattractiveduringthefinancialcrisiswhenspreadswereunusuallylargebutthatwereuneconomicalinamorenormalspreadenvironment.
(Adrian,Kimbrough,andMarchioni(2011)describetheCPFFingreaterdetail.
)TheTermAsset-BackedSecuritiesLoanFacility(TALF)wascreatedtohelpmarketparticipantsmeetthecreditneedsofhouseholdsandsmallbusinessesbysupportingtheissuanceofasset-backedsecurities(ABS)collateralizedbyautoloans,studentloans,creditcardloans,equipmentloans,floorplanloans,insurancepremiumfinanceloans,loansguaranteedbytheSmallBusinessAdministration,residentialmortgageservicingadvances,orcommercialmortgageloans.
Ashcraft,Malz,andPozsar(2012)describethefacilityindetail.
Theseliquidityfacilitieshavethecommonalityofexpandingthecentralbank'slendingoflastresorttoinstitutionsoftheshadowbankingsystemthatdonothavedirect,explicitaccesstopublicliquiditybackstops.
Thefragilityofshadowbanksduetovulnerableassetsandliabilitiesmakesthemvulnerabletoexcessivecollapseintimesofadversefinancialcycles.
Lendingoflastresortaimstoinsulaterealeconomicactivityfromsuchdisorderlycollapses.
Whilelendingoflastresortisanecessaryactionexpost,oncecrisesmaterialize,theanticipationofsuchactioncanresultindistortedrisk-takingincentives.
Asaresult,structuralreforms,reviewedinthenextsection,aimatmitigatingincentivesforexcessiverisktakingintheshadowbankingsystemexante.
5)HowShouldShadowCreditIntermediationBeRegulatedAdrianandAshcraft(2012)reviewrecentregulatorychangestotheshadowbankingsystemindetail.
Weprovideashortoverviewofthosereformefforts,focusingonthreeareas:1)reformsrelatingtomoneymarkets,2)implicationsofbankingregulationfortheshadowbankingsystem,and3)reformsofsecuritizationandcreditratings.
Itshouldbenotedthatallthesereformeffortsareunderwayatthistime.
26A)MoneyMarketReforms:ABCP,Repo,andMoneyMarketMutualFundsReformsRelatingtoABCPConduitsInJune2009,theFinancialAccountingStandardsBoard(FASB)announcedtheStatementsofFinancialAccountingStandards(FAS)166and167,amendingexistingaccountingrulesforconsolidationofsecuritizationtransactions.
Sponsorsofsecuritizationtransactionshavegenerallyinterpretedthisnewguidanceasrequiringaccountingconsolidationintheeventthatafirst-losspositionandloanservicingareretainedbythesponsorforsecuritizationtransactions.
Followingrevisionstotheaccountingrules,theU.
S.
bankingagenciesclarifiedinSeptember2009thatdepositoryinstitutionswouldhavetoholdregulatorycapitalagainstconsolidatedsecuritizationtransactionsandABCPconduits.
Themovementofassetsontothebalancesheetwillresultinanincreaseincapitalrequirementsundertheminimumleverageratio,anincreaseinrisk-weightedassetsandcapitalrequirementsgiventheinabilityofbankstousea10percentcreditconversionfactorforliquidityguarantees,andarequirementthatbanksprovisionforlossesonloansheldinconsolidatedconduitsandsecuritizationtrusts.
Thecloselinkbetweenregulatorycapitalandaccountingtreatmenthaseliminatedthescopeforusingsecuritizationofloansservicedbythesponsortoreducecapitalrequirements.
Furthermore,Section331oftheDodd-FrankActrequiresFDICassessmentsonconsolidatedassetsminustangibleequityoflargebanksratherthanthehistoricalpracticeofcountingonlydepositliabilities.
Theconsolidationofconduitsontobankbalancesheetsmeansthatbankswillpayassessmentsontheseliabilities,makingconduitsponsorshipmoreexpensive.
ReformsRelatingtoTri-partyRepoReformsinthetri-partyrepomarketareongoing.
Animportantfrictioninthetri-partyrepomarketisthedependenceofmarketparticipantsonintradaycreditofthecustodianbanks.
In2009,anindustrytaskforcesponsoredbytheFederalReserveBankofNewYorkwascreatedwiththeaimofreducingmarketparticipants'dependenceonintradaycredit.
Thetaskforcehasshortenedthewindowofthedailyunwind,movingitfrom8:30inthemorningto3:30intheafternoon.
However,between3:30andthesettlementofallrepos,thedealersarestilldependentonthecreditoftheclearingbanks.
Anothermajorsourceofsystemicriskinthetri-partyrepomarketisvulnerabilitytothedefaultofamajordealer.
Suchaneventexposesthatclearingbanktocounterpartycreditrisk.
Moreover,itleadstoapotentiallydestabilizingtransferofriskacrossmarketparticipantsanddirectlyimpactsthedealers'clientswhoarenolongerabletoobtainleveragethroughthedealerinquestion.
Thevulnerabilityofshort-termfundingmarketswithrespecttosingleinstitutionsisamajorconcernforthestabilityofthesefundingmarkets.
Thetri-partyrepotaskforcehasnotbeensuccessfulinidentifyingasolutiontotheproblemofhowmoneymarketfundinvestorswouldbeabletoliquidatecollateralintheeventalargebroker-dealerbecameinsolvent.
Aslongasthetri-partyrepomarketacceptsasignificantamountofcollateralotherthanU.
S.
Treasuryandagencysecurities(suchasprivate-labelABSandcorporatebonds),thetri-partymarketwillremainpronetorunsandconstituteasourceofsystemicrisk.
Themajorbroker-dealersthatarethemostimportantborrowersinthetri-partyrepomarkethavebecomesubjecttotighterregulation.
Inparticular,oneoftheconsequencesofthefinancialcrisishas27beenthattwooftheformerlyfivemajorinvestmentbankshavebeentransformedintobankholdingcompaniesandtwohavemergedwithbankholdingcompanies.
Thefifthdealer,LehmanBrothers,declaredbankruptcy,anditsdealersubsidiarywasacquiredbyforeignbanks.
Asaresult,alloftheformerlymajorindependentinvestmentbanksarenowregulatedonaconsolidatedbasisbytheFederalReserveandwillbesubjecttothereformedBaselcapitalandliquiditystandards.
Inaddition,theDodd-FrankActinstitutedenhancedprudentialstandardsforlargebankholdingcompaniesandthedesignationof"systemicallyimportantnonbankfinancialinstitutions.
"Furthermore,theOrderlyLiquidationAuthorityprovidestheFDICwiththeauthoritytoactasreceiverfortheresolutionofnonbankfinancialinstitutions(includingbankholdingcompanies)forwhichasystemicriskdeterminationhasbeenmade.
AcurrentlyopenquestionconcernstheregulationofthemajorU.
S.
broker-dealersownedbyforeignbankingorganizations.
ReformsRelatingtoMMMFsMMMFshaveundergonesomereformsincethefinancialcrisisof2007-09.
Inparticular,theSEChasputnewrestrictionson2a-7fundstolimitriskandmaturitytransformationandrelianceonratings.
However,theserestrictionsdonotaddressthekeyfrictionthatexistsinthemarket,whichisimplicitsupportforastablenetassetvalue(NAV)byplansponsorsandtheofficialsectorthroughhistoricalexperience.
TheMMMFrulesasamendedin2010alsoincreasethefunds'incentivestolendforshorttenorsanddecreasetheirincentivestolookthroughtothecollateral.
TheSECrulesincentMMMFstoactasunsecuredratherthansecuredinvestors—whichisaproblemfromafinancialstabilitypointofview.
However,thesereformscontinuetoleaveMMMFsasasourceofsystemicrisk.
ThesusceptibilityofMMMFstorunsisillustratedintheaftermathofLehman'sfailureinSeptember2008whenastand-aloneMMMF—theReserveFund—brokethebuck.
Inthefollowingweeks,institutionalinvestorsbroadlywithdrewfromMMMFs,thusforcingmassiveliquidationsofMMMFassets,whichledtodownwardpressureonprices,andinstancesoffundsbreakingthebuck.
Wheninvestorsrunonmoneymarketfunds,theyareforcedtoliquidateassets,puttingdownwardpressureonthemark-to-marketvaluesofthemoneymarketmutualfundassetsandpotentiallyleadingtomorefundsbreakingthebuck.
ThemarketfrictionthatmakesMMMFsunstableisthusthestableNAVthatgivesrisetofiresaledynamicsthatoccurwheninvestorswithdrawinvestments.
AsaresultofthestableNAVrule,investorsofMMMFseffectivelytreatthefundslikedemanddepositaccounts.
Infact,manyMMMFsmarketthefundsasalternativestodemanddeposits.
However,MMMFshavenoexplicitbackstopthatwouldprotectthemagainstdeclinesinassetvalues.
MMMFsdorelyonimplicitdiscretionarysupportbysponsors.
However,inafinancialcrisis,investorscannotnecessarilycountonthesponsorstoprovidesupport.
ThesecondfrictionwewouldhighlightinthecontextofMMMFsisthustheimplicitguaranteeprovidedbythefunds'sponsors.
OneoftheproposalsforfurtherreformsofMMMFsistoabandonthestableNAVruleandoperateMMMFswithfloatingNAV.
Moneyfundswouldthenmarktheirassetvaluestomarketatalltimes.
Thiswouldremovesomeoftheincentivesforinvestorstorun.
However,changingmoneyfundsfromstabletofloatingNAVswouldnotremoveallincentivesofinvestorstorun.
Inthepresenceofsomeilliquidityintheassetmarket,earlywithdrawaloffundscanleadtotemporaryunder-valuationsofassetsand28provideincentivesforearlywithdrawal.
Thereare,indeed,someinstancesofrunsincertainEuropeancountriesthathavemoneymarketfundswithfloatingNAVs.
TheSECiscurrentlyconsideringarangeofreformoptions,andtheFinancialStabilityOversightCouncil(FSOC)hashighlightedtheneedforMMMFreformsinitsannualreportsof2011and2012.
Ingeneral,thesewereintendedtoaddressthefactthatMMMFshaveanumberofcharacteristics—includingastableNAV,redemptionupondemand,andextremelyrisk-averseinvestors—thatinteracttomaketheseentitiesvulnerabletoruns.
Severaloftheseproposalsentailthecreationofliquidityandcapitalbuffers.
Theformerprovideadditionalnear-cashassetstodealwithredemptions,whilethelatterenhancesthelossabsorptioncapacityavailabletodealwithacreditevent.
Broadlyspeaking,twokindsofbufferscanbesetup:exanteandexpost.
Onetypeofexantebufferistocreateaprivateemergencyliquidityfacility,capitalreserve,orinsurance.
RegulatedfixedNAVfundswouldbenefitfromanexantebuffer,butbeforcedtopaythecost.
Anotherapproachtoanexantebufferisforindividualfundstosetasideresourcesinadvancetoabsorblossesshouldtheyoccur,servingthesamepurposeascapitalreservesintraditionalbanks.
Asanalternative,theInvestmentCompanyInstitutehasproposedaprivate-sector"liquiditybank"thatwouldprovideabackstopbutmightitselfbenefitfromaccesstoofficialliquidity.
Anexpostbufferdoesnotrequireanyresourcestobesetasideinadvance,butiscreatedbytakingstepstoensurethatinvestorsabsorblosseswhentheyoccurandthattheycannotfleeandleavethelossesbehind.
Inparticular,suchmeasuresaredesignedtoforestallinvestorsredeemingsharesataNAVofonedollaronceacrediteventorliquidityeventhasbegun.
AvariableNAVmaybehelpfulinthisregard,becauseitcould,ifproperlycomputed,adjustrapidlyinresponsetolossesorliquidityshocks.
However,thiswouldbeafundamentalchangeinthenatureofMMMFs.
TheSquamLakeGroup(2011)putforwardaproposalforMMMFstohavetwoshareclasses.
TheseniortranchewouldbeastablenetassetvaluefundthatwouldbebackedbyaliquiditybufferamountingtoxpercentofthecurrentNAV.
Theliquiditybuffercouldbeimplementedinatleastfourdifferentways.
Mostrecently,McCabe(2012)hasinvestigatedthefeasibilityofrequiringatwo-sharesystemforMMMFs(labeledAandBshares).
McCabecalibratesthereturnstothetranchesunderrealisticscenariosaboutassetreturnsandrunrisk.
Theadvantageofthistwo-shareproposalisthatitwouldpreservethestableNAVfeatureofmoneymarketfundsundermuchmoreseverecircumstancesthaniscurrentlythecase,whileloweringthereturnstoinvestorsintothestableNAVsharesonlyslightly.
However,itshouldbenotedthatthetwo-shareproposaldoesnotfullyprotectfundsagainstrunsinallstatesoftheworld.
Foranyrealisticcapitalrequirement,therearesometaileventsthatwillinducethefundtohavetounwind.
McCabe,Cipriani,Holscher,andMartin(2012)haveproposedthatMMMFsbemadesubjecttoa"minimumbalanceatrisk"(MBR).
TheMBRwouldbeasmallfractionofeachMMMFinvestor'sbalancedemarcatedtoabsorblossesifthefundisliquidated.
ThisfeatureaccountsforthecreditriskthatMMMFshold.
Furthermore,redemptionsoftheMBRwouldbedelayedbythirtydays,thusaccountingfortheilliquidityofMMMFassetsintimesofcrisisandreducingfiresaleincentives.
LargeredemptionswouldsubordinatepartoftheMBR,creatingadisincentivetoredeemifthefundislikelytohavelossesandthusreducingincentivesforinvestorstorun.
29B)BankingRegulationReforms:CapitalandLiquidityCapitalRequirementsforSecuritizationExposuresInFebruary2011,regulatorsannouncedplannedchangestothetreatmentofsecuritizationexposuresheldbybanksinthetradingbook.
Ingeneral,assetsheldinthetradingbookfacelowercapitalchargesthanthoseinthebankingbookgiventhestatedintentoftheinstitutiontoactivelytrade,andthepresumptionofregulatorswasthattheinstitutionwillbeabletoexitthepositionbeforeincurringcreditlosses.
However,thebehaviorofbanksduringtherecentfinancialcrisissuggestedthattheseinstitutionswereunwillingtotradeoutofpositions,giventhelargedeclineinpricesrelativetoprojectedlosses.
TheproposedrevisionstotheMarketRiskAmendmentofBaselIIrecognizethisbehaviorandrequirebankstoholdcapitalagainstsecuritizationexposuresinthetradingbookasiftheywereinthebankingbook,eliminatingbanks'abilitytoholdlesscapitalagainsttheseexposures.
FDICSafeHarborInSeptember2010,theFDICapprovedrevisionstoitssafeharborfromrepudiationpowersinreceivership.
Inparticular,asreceiverofafailedbank,theFDIChastheauthoritytorepudiatecontracts,whichcouldpossiblyincludethesaleofassetstoabankruptcy-remotetrustaspartofabank-sponsoredsecuritizationtransaction.
Historically,theFDICcreatedasafeharborfromuseofthisauthoritytiedtotheaccountingtreatmentofthetransaction.
However,theaforementionedchangestoFAS166/167impliedthatmanysecuritizationtransactionswouldnowbeconsolidatedonabank'sbalancesheet,implyingthatinvestorswouldnolongerbenefitfromtheexistingsafeharbor.
Inthenewsafeharbor,theFDICrequiresbank-sponsoredsecuritizationstomeetminimalstandardsforcapitalstructure;disclosurerequirementstobealignedwiththeSEC'sproposedrevisionstoRegulationAB;anddocumentation,compensation,andriskretentiontobealignedwiththeinter-agencyimplementationofDodd-Frank941.
Therulehasmorestringentrequirementsforbank-sponsoredRMBStransactions,includingtheneedfora5percentcashreservefortwelvemonthstofundrepresentationsandwarrantiesandarequirementthatcompensationtoratingagenciesbebasedinpartontheperformanceoftheunderlyingtransactions.
ThestatedmotivationforusingthesafeharborinthisfashionistoprotecttheFDICasguarantorofbankdepositsfromthebank'sinvestmentinsecuritizationtransactions.
Asthescopeoftheruleappliesonlytobankssponsoringsecuritizationtransactions,itispossiblethat,whenbinding,itwillshiftsecuritizationactivitytothenonbanksector.
BankLiquidityRegulationInDecember2010,theBaselCommitteeproposednewliquidityrequirementsforbanks.
Inparticular,inadditiontocapitalrequirements,bankswouldhavetomeettwoliquiditystandards:aliquiditycoverageratio(LCR)andanetstablefundingratio(NSFR).
TheLCRisintendedtopromoteshort-termresilienceofabank'sliquidityriskprofilebyensuringthatithassufficienthigh-qualityliquidassetstosurviveasignificantstressscenariolastingforonemonth.
Inparticular,thebankisrequiredtoholdunencumberedhigh-qualityliquidassetsinanamountnolessthan100percentoftotalnetcashoutflowsoverthenextthirtydaysinastressscenario.
TheNSFRisintendedtopromoteresilienceover30alongertimehorizonbycreatingadditionalincentivesforbankstofundtheiractivitieswithmorestablesourcesonanongoingbasis.
IntheNSFRrequirement,stablefundingisdefinedas"theportionofthosetypesandamountsofequityandliabilityfinancingexpectedtobereliablesourcesoffundsoveraone-yeartimehorizonunderconditionsofextendedstress.
"Theamountofrequiredstablefundingisafunctionoftheliquiditycharacteristicsoftheinstitution'sfinancialexposures.
Collectively,theseliquidityrulesareexpectedtohaveanimpactonthecostsofprovidingliquidityguaranteestoABCPconduits,asbankswillnowberequiredtoholdanadequatelevelofunencumberedhigh-qualityliquidassetsfordrawsonlinesunderlyingtheexposuresintheconduits,aswellasanyABCPwithamaturityofthirtydaysorless.
Moreover,newproposedliquidityrequirementsforbankscouldmakebackuplinesmoreexpensivebyrequiringanadequatelevelofliquidassetstomeetstressliquidityneedsforathirty-daytimehorizon.
C)CreditMarketReforms:SecuritizationandCreditRatingsFDICSafeHarborInApril2010,theSECproposedrevisionstoRegulationABthatprovideguidanceonrequireddisclosurebysponsorsofsecuritizationtransactions.
TheproposalbytheSECwaslargelyconfirmedinSection942oftheDodd-FrankAct.
Theseruleswerere-proposedinApril2011,partlyinresponsetotheSection932Arequirementtoremovereferencestocreditratingsandpartlyinresponsetocommentsontheoriginalproposal.
Themotivationforrevisionstotheruleistheconclusionthatinvestorsdidnothaveadequateinformationortimetoconductduediligenceonnewissuesecuritizationtransactions,thatmarketparticipantsover-reliedoncreditratings,andthatincentivemisalignmentexistsbetweensponsorsandinvestors.
RiskRetentionInApril2011,regulatorsjointlyproposedrulesimplementingSection941ofDodd-Frank,requiringthatsponsorsretainmeaningfulriskofsecuritizationtransactions.
Intheproposal,thesponsorofasecuritizationtransactionisrequiredtoholdatleast5percentofaneligibleformofriskretentionmeasuredusingparvalue.
Eligibleformsofriskretentiongenerallyincludeverticalretention,wherethesponsorretainsafractionofeverytranche;horizontalretention,wherethesponsorretainsafirst-lossposition;andaspecificcombinationofthetwoformsofequalsize,referredtoasL-shaped.
Thesponsorisnotpermittedtosellorhedgetheretainedinterestforthelifeofthetransactionandisnotpermittedtopledgefornonrecoursefinancing.
TheproposedruleprovidesforexemptionsfromriskretentionforsecuritizationssponsoredbyU.
S.
governmentagencies,forgovernment-sponsoredenterprisesaslongastheyareinreceivership,andforqualifiedloanpoolsthatmeetstrictunderwritingrequirements.
TheproposedrulepermitsthesponsorofaCMBStransactiontosellahorizontaltranchetoaB-pieceinvestorthatre-underwriteseveryloaninthetransactionandpermitsthesponsorofanABCPtransactiontorecognizeriskretentionbytheoriginatorintheunderlyingreceivablesbeingfinancedintheconduit.
31CreditRatingAgenciesThe2010Dodd-FrankActincludesarangeofprovisionsintendedtoimproveratingagencyincentivesandperformance.
UnderSections7and11oftheSecuritiesActof1933,whenanissuerincludesstatementsinaprospectusfromexpertslikelawyersoraccountants,theprospectusmustalsoincludeconsenttoliabilityfromtheexpert.
WhileRule436(g)historicallyexemptedcreditratingagenciesfromthisrequirement,thisexemptionwasremovedbyDodd-Frank.
However,theexemptionhasnevergoneintoeffectbecausetheSECissuedano-actionletter,acknowledgingrefusalbythecreditratingagenciestoconsenttoexpertliability,threateningtobringnewissuancetoahalt.
Asofthiswriting,therepealof436(g)isstillnotinforce.
Dodd-FrankmadeamendmentstoRule17g-5inordertoprovideinvestorswithmoreviewsonthecreditworthinessofstructuredfinanceproductsandtoimprovethequalityofratingsbylimitingratingshopping.
Inparticular,theseamendmentsrequirearatingagencyhiredbyanissuertodisclosetheratingassignmentandobtainrepresentationfromthearrangerthattheywillprovideinformationtobothhiredandcertifiednon-hirednationallyrecognizedstatisticalratingorganizations(NRSROs).
Inpractice,thisprovisionhasnotbeenusedtoproducemanyshadowratings,givenconcernsbyNRSROsaboutlegalliabilitytheywouldfacefromdisclosingthisinformationwithoutexplicitpermissionoftheissuer.
TheCreditRatingAgencyReformActof2006(RatingAgencyAct)mandatedthattheSECestablisharegistrationandoversightprogramforNRSROs.
WhiletheSECwasgivenformaloversightauthorityfortheratingagenciesforthefirsttime,thelawexpresslyprohibitedregulationofthesubstanceofcreditratingsortheproceduresandmethodologies.
InJune2007,theSECadoptednewrulesestablishingaregulatoryprogramforNRSROs.
TheserulesrequireNRSROstohavewrittenpoliciesandprocedurestopreventthemisuseofmaterialnonpublicinformationandtomanagecertainconflictsofinterest.
TheyalsodisallowcertainotherconflictsofinterestoutrightandprohibitNRSROsfromengagingincertainunfair,coercive,orabusivepractices.
6)ConclusionInthispaper,wedefineshadowcreditintermediationtobetheintermediationofcreditwithoutthedirectorexplicitsupportoftheU.
S.
taxpayer.
Overthepastsixtyyears,thefractionofcreditfundedusingcommercialbanks'maturitytransformationhasdeclinedsignificantly,asmarket-basedcreditintermediationhasincreasedinimportance.
Thedeclineofcreditintermediationfinancedthroughtraditionalbankshasbeenoffsetonlyinpartbytheriseofshadowbanks.
Weexplorethreemotivationsforthegrowingimportanceofshadowbanking.
First,wehighlightfinancialinnovationinthecompositionoftheaggregatemoneysupply.
Second,wediscusstheincentivesoffinancialinstitutionstoavoidtaxes,accountingrules,orcapitalrequirements.
Third,wereviewthepresenceofagencyproblemsinfinancialmarkets,whichcreateperverseincentives.
32Weprovideatopologyforunderstandingdifferentpartsoftheshadowintermediationprocessaswellasdifferentsectorsthatengageinshadowbanking.
Wefurthermorearticulatethemarketfailuresthatultimatelyjustifytheneedforregulationofthissector.
Inparticular,wediscusshowcomplexitycanresultinneglectedriskbyinvestors,permittingthebuildupofsystemicrisks.
Moreover,wenotethewell-documentedexternalitiesassociatedwithrunsoninstitutionsinvolvedinmaturitytransformation.
Wealsodocumenthowthepublicsectorhasprovidedlending-of-last-resortfacilitiesduringthefinancialcrisisinordertoshieldrealeconomicactivityfromarunontheshadowbankingsystem.
Finally,weprovideanoverviewofnewrulestargetingshadowbankingandhighlighttheunevenimpacttheywillhaveonthelikelysizeoftheshadowbankingsectorinthefuture.
Whilechangestoaccountingandcapitalrequirementswillreduceincentivesbybankstoengageintypesofarbitrageactivitiesatthecoreofthefinancialcrisis—thatis,ABCPandsecuritizationactivity—significantincreasesintheoveralllevelandrisksensitivityofcapitalwillprovidestrongincentivesforcreditintermediationtobefundedoutsideofthebankingsystem.
LiteratureAcharyaV,SchnablP,SuarezG.
2010.
SecuritizationWithoutRiskTransfer.
NBERWorkingPaper15730.
AcharyaV,RichardsonM,NieuwerburghSV,WhiteLJ.
2011.
GuaranteedtoFail:FannieMae,FreddieMac,andtheDebacleofMortgageFinance.
PrincetonUniversityPress.
AdrianT,AshcraftAB.
2012.
ShadowBankRegulation.
AnnualReviewofFinancialEconomics.
Forthcoming.
AdrianT,NBoyarchenko.
2012.
IntermediaryLeverageCyclesandFinancialStability.
FederalReserveBankofNewYorkStaffReports567.
AdrianT,BegalleB,CopelandA,MartinA.
2013.
RepoandSecuritiesLending.
InQuantifyingSystemicRiskMeasurement:NBERResearchConferenceReportSeries,ed.
JGHaubrich,AWLo.
UniversityofChicagoPress.
Forthcoming.
AdrianT,BurkeC,McAndrewsJ.
2009.
TheFederalReserve'sPrimaryDealerCreditFacility.
FederalReserveBankofNewYorkCurrentIssuesinEconomicsandFinance15(4):1—12.
AdrianT,FlemingM.
2005.
WhatFinancingDataRevealAboutDealerLeverage.
FederalReserveBankofNewYorkCurrentIssuesinEconomicsandFinance11(3):1—7.
AdrianT,MoenchE,ShinHS.
2009.
FinancialIntermediation,AssetPrices,andMacroeconomicDynamics.
FederalReserveBankofNewYorkStaffReport422.
33AdrianT,KimbroughK,MarchioniD.
2011.
TheFederalReserve'sCommercialPaperFundingFacility.
FederalReserveBankofNewYorkEconomicPolicyReview17(1):25—39.
AdrianT,ShinHS.
2009.
TheShadowBankingSystem:ImplicationsforFinancialRegulation.
BanquedeFranceFinancialStabilityReview13:1–10.
AdrianT,ShinHS.
2010a.
LiquidityandLeverage.
JournalofFinancialIntermediation19(3):418–37.
ArmantierO,KriegerS,McAndrewsJ.
2008.
TheFederalReserve'sTermAuctionFacility.
FederalReserveBankofNewYorkCurrentIssuesinEconomicsandFinance14(5).
AshcraftA,Goldsmith-PinkhamP,HullP,VickeryJ.
2011.
CreditRatingsandSecurityPricesintheSubprimeMBSMarket.
AmericanEconomicReview101(3):115–19AshcraftAB,SchuermannT.
2008.
UnderstandingtheSecuritizationofSubprimeMortgageCredit.
FoundationsandTrendsinFinance2(3):191—309.
AshcraftAB,MalzA,PozsarZ.
2012.
TheFederalReserve'sTermAsset-BackedSecuritiesLoanFacility.
FederalReserveBankofNewYorkEconomicPolicyReview.
Forthcoming.
AvrahamD,SelvaggiP,VickeryJ.
2012.
AStructuralViewofBankHoldingCompanies.
FederalReserveBankofNewYorkEconomicPolicyReview18(2):65—82.
Bord,V,SantosJC.
2012.
TheRiseoftheOriginate-to-DistributeModelandtheRoleofBanksinFinancialIntermediation.
FederalReserveBankofNewYorkEconomicPolicyReview18(2):21—34.
Brunnermeier,MK,LHPedersen.
2009.
MarketLiquidityandFundingLiquidity.
ReviewofFinancialStudies22(6):2201—2238.
Brunnermeier,MK,YSannikov.
2012.
AMacroeconomicModelwithaFinancialSector.
WorkingPaper.
PrincetonUniversity.
CetorelliN.
2012.
APrincipleforForward-LookingMonitoringofFinancialIntermediation:FollowtheBanks!
FederalReserveBankofNewYorkLibertyStreetEconomicsBlog,July23,2012.
Availableathttp://libertystreeteconomics.
newyorkfed.
org/2012/07/a-principle-for-forward-looking-monitoring-of-financial-intermediation-follow-the-banks.
html.
CetorelliN,PeristianiS.
2012.
TheRoleofBanksinAssetSecuritization.
FederalReserveBankofNewYorkEconomicPolicyReview18(2):47—64.
CohenA.
2011.
RatingshoppingintheCMBSmarket.
PresentedatRegulationofSystemicRisk,Washington,DC,Sep.
15—16.
CopelandA,MartinA,WalkerM.
2011.
RepoRuns:EvidencefromtheTri-partyRepoMarket.
FederalReserveBankofNewYorkStaffReport506.
34CopelandA.
2012.
EvolutionandHeterogeneityamongLargerBankHoldingCompanies:1994to2010.
FederalReserveBankofNewYorkEconomicPolicyReview18(2),pp.
83—93.
CovalJ,JurekJ,StaffordE.
2009.
TheEconomicsofStructuredFinance.
JournalofEconomicPerspectives23(1):3—25.
CovitzD,LiangN,SuarezG.
2012.
TheEvolutionofaFinancialCrisis:PanicintheAsset-backedCommercialPaperMarket.
JournalofFinance.
ForthcomingDangTV,GortonG,HolmstrmB.
2009.
OpacityandtheOptimalityofDebtforLiquidityProvision.
Yale/MITWorkingPaper.
DiamondD,DybvigP.
1983.
BankRuns,DepositInsurance,andLiquidity.
JournalofPoliticalEconomy91:401–19.
Duffie,D.
2012.
MarketMakingUndertheProposedVolckerRule.
WorkingPaper.
StanfordUniversity.
FinancialAccountingStandardsBoard.
2012.
BriefingDocument:FASBStatement166and167.
Availableathttp://www.
fasb.
org/cs/ContentServerc=FASBContent_C&pagename=FASB%2FFASBContent_C%2FNewsPage&cid=1176155633483.
FinancialStabilityBoard.
2011.
ShadowBanking:StrengtheningOversightandRegulation.
Availableathttp://www.
financialstabilityboard.
org/publications/r_111027a.
pdf.
FinancialStabilityBoard.
2012.
ProgressReporttotheG20onStrengtheningtheOversightandRegulationofShadowBanking.
Availableathttp://www.
financialstabilityboard.
org/publications/r_120420c.
pdf.
Fleming,M,GarbadeK.
2003.
TheRepurchaseAgreementRefined:GCFRepo.
FederalReserveBankofNewYorkCurrentIssuesinEconomicsandFinance9(6):1—7.
Fleming,M,HrungWB,andKeaneFM.
2010.
RepoMarketEffectsoftheTermSecuritiesLendingFacility.
AmericanEconomicReview:PapersandProceedings100,pp.
591—6.
FostelA,GeanakoplosJ.
2008.
LeverageCyclesandtheAnxiousEconomy.
AmericanEconomicReview98(4),1211—1244.
FrameWSandWhiteLJ.
2005.
FussingandFumingoverFannieandFreddie:HowMuchSmoke,HowMuchFireJournalofEconomicPerspectives19(2),159—184.
GarleanuN,LHPedersen.
2011.
Margin-BasedAssetPricingandDeviationsfromtheLawofOnePrice,ReviewofFinancialStudies,24(6),1980—2022.
Gennaioli,N,ShleiferA,andVishnyR.
2012a.
NeglectedRisks,FinancialInnovation,andFinancialFragility.
JournalofFinancialEconomics.
Forthcoming.
35GennaioliN,ShleiferA,VishnyR.
2012b.
AModelofShadowBanking.
JournalofFinance.
Forthcoming.
GortonG.
1985.
ClearinghousesandtheOriginofCentralBankingintheUnitedStates.
JournalofEconomicHistory4(2)pp.
277-283.
GortonG,MetrickA.
2011.
RegulatingtheShadowBankingSystem.
BrookingsPaperonEconomicActivity261—312.
GortonG,MetrickA.
2012.
SecuritizedBankingandtheRunonRepo.
JournalofFinancialEconomics104:425—51.
HolmstrmB,TiroleJ.
1998.
PrivateandPublicSupplyofLiquidity.
JournalofPoliticalEconomy106(1),pp.
1-40.
JaremskiM.
2010.
FreeBankFailures:RiskyBondsVersusUndiversifiedPortfolios.
JournalofMoney,Credit,andBanking42(8),pp.
1565—1587.
KacperczykM,SchnablP.
2011.
DoesOrganizationalFormAffectRiskTakingEvidencefromMoneyMarketMutualFunds.
NewYorkUniversityWorkingPaper.
KeysB,MukherjeeT,SeruA,VigV.
2010.
DidSecuritizationLeadtoLaxScreeningEvidencefromSubprimeLoans.
QuarterlyJournalofEconomics125(1),pp.
307—62.
KrishnamurthyA,NagelS,OrlovD.
2011.
SizingUpRepo.
NBER/CEPR/Stanford/NorthwestWorkingPaper.
LevitinA,WachterS.
2011.
ExplainingtheHousingBubble.
ResearchPaper10-15,InstituteforLawandEconomics,UniversityofPennsylvania.
MandelB,MorganD,WeiC.
2012.
TheRoleofBankCreditEnhancementsinSecuritization.
FederalReserveBankofNewYorkEconomicPolicyReview18(2),pp.
35—46.
MartinA,SkeieD,vonThaddenE.
2011.
RepoRuns.
FederalReserveBankofNewYorkStaffReport444.
MathisJ,McAndrewsJ,RochetJC.
2009.
Ratingtheraters:ArereputationconcernspowerfulenoughtodisciplineratingagenciesJournalofMonetaryEconomics57(5):657—74.
McCabePE,CiprianiM,HolscherM,MartinA.
2012.
TheMinimumBalanceatRisk:AProposaltoMitigatetheSystemicRisksPosedbyMoneyMarketFunds.
FederalReserveBankofNewYorkStaffReport564.
McCulley,P.
2007.
TetonReflections.
PIMCOGlobalCentralBankFocus.
MertonRC.
1977.
AnAnalyticalDerivationoftheCostofDepositInsuranceandLoanGuarantees.
JournalofBankingandFinance1,pp.
3—11.
36MertonRC,BodieZ.
1993.
DepositInsuranceReform:AFunctionalApproach.
Carnegie-RochesterConferenceSeriesPublicPolicy38,pp.
1—34.
PassmoreW,SherlundSM,BurgessG.
2005.
TheEffectofHousingGovernment-SponsoredEnterprisesonMortgageRates.
RealEstateEconomics33(3),pp.
427–463.
Pozsar,Zoltan.
2008.
TheRiseandFalloftheShadowBankingSystem.
Availableathttp://www.
economy.
com/sbs.
PozsarZ,AdrianT,AshcraftAB,BoeskyH.
2010.
ShadowBanking.
FederalReserveBankofNewYorkStaffReport458.
Availableathttp://papers.
ssrn.
com/sol3/papers.
cfmabstract_id=1645337.
RajanR.
2005.
HasFinancialDevelopmentMadetheWorldRiskierProceedingsoftheFederalReserveBankofKansasCityEconomicsSymposium,pp.
313—69.
Ricks,M.
2010.
ShadowBankingandFinancialRegulation.
ColumbiaLawandEconomicsWorkingPaper370.
Schwarcz,S.
2012.
RegulatingShadowBanking.
BostonUniversityReviewofBankingandFinancialLaw.
SquamLakeGroup.
2011.
ReformingMoneyMarketFunds.
SquamLakeGroupStaffReport.
SteinJ.
2010.
Securitization,ShadowBanking,andFinancialFragility.
Daedalus139(4),pp.
41—51.
SunderamA.
2012.
MoneyCreationandtheShadowBankingSystem.
HarvardBusinessSchoolWorkingPaper.
TuckerP.
2010.
Shadowbanking,financingmarketsandfinancialstability.
RemarksbyMrPaulTucker,DeputyGovernorforFinancialStabilityattheBankofEngland,ataBernieGeraldCantor(BGC)PartnersSeminar,London,21January2010.
WermersR.
2011.
RunsonMoneyMarketMutualFunds.
UniversityofMarylandWorkingPaper.
WiggersT,AshcraftAB.
2012.
DefaultsandLossesonCommercialRealEstateBondsduringtheGreatDepressionEra.
FederalReserveBankofNewYorkStaffReport544.
XiaH,StroblG.
2012.
TheIssuer-PaysRatingModelandRatingsInflation:EvidencefromCorporateCreditRatings.
SSRNWorkingPaper.