关闭

关闭

封号提示

内容

首页 CDO market implosion and the pricing of subprim…

CDO market implosion and the pricing of subprime.pdf

CDO market implosion and the pr…

上传者: fengqiyaoye 2013-10-10 评分 0 0 0 0 0 0 暂无简介 简介 举报

简介:本文档为《CDO market implosion and the pricing of subprimepdf》,可适用于工程科技领域,主题内容包含CDOmarketimplosionandthepricingyBuiKenHallrfax,Vdeterminantsofcreditspread符等。

CDOmarketimplosionandthepricingyBuiKenHallrfax,Vdeterminantsofcreditspreads,GMMresultsindicatethattheemergenceandrapidcapitalizationofthesubprimebackedCDOmarketwasassociatedwithasignificanttighteningofsubprimeMBS–TreasuryyieldspreadsResultsofVARandotherrobustnesstestsserverities,viatherepackagingofrelativelyilliquidsubordinatedassetsintoderivativeCDOsecuritiesInastunningmarketaboutface,newissuanceofCDOsceasedinearly,inthewakeofimplosionandwholesalerepricingofcreditriskinthecapitalmarketsThesurgeinissuanceofsubprimebackedCDOscoincidedwithamarkedtighteninginsubprimeMBS–Treasuryspreads,suggestingsomemeasurableeffectofderivativesecuritizationonthe$seefrontmatterElsevierIncAllrightsreservedTheauthorsaregratefultoMichaelBrennan,JohnCotter,MarkFlannery,FrancisLongstaff,TimRiddiough,AvanidharSubrahmanyamandWalterTorousforexcellentcommentsandtoMinyeZhang,DaweiTianandHuanghaiLiforoutstandingresearchassistanceTheauthorsalsothankseminarparticipantsattheFederalReserveBoard,UCLA,GeorgeWashingtonUniversity,andtheAEAsessiononthesubprimecrisisforhelpfulcommentsAllerrorsareourresponsibilityCorrespondingauthorFax:Emailaddresses:ydengnusedusg(YDeng),stuartgabrielandersonuclaedu(SAGabriel),asandergmuedu(ABSanders)JournalofHousingEconomics()–ContentslistsavailableatScienceDirectJournalofHousijournalhomepage:wwweldoi:jjheThesubprimecrisisandthehousingbubbleanditsbursthavehadprofoundrepercussionsontheworldeconomyInparticular,thedownsideofthehousingbubblehasproduced(coupledwithrisingunemployment)increasedforeclosuresandfurtherdownwardpressuresonhousingpricesOneofthecontributingfactorstothehousingbubblewastheexpansionofcredittosubprimeborrowersandtheintroductionoflownodocumentationloanproductsweretypicallysoldtosecondarymarketparticipantsratherthanbeingkeptonbankbalancesheetsInordertounderstandthefinancialcrisis,itisimportanttounderstandthecollateralizeddebtobligation(CDO)marketwhichallowedthesubprimeandAltAmortgagemarketsgrowdramaticallyinsizeTheglobalmarketforcollateralizeddebtobligations(CDOs)witnessedexplosivegrowthoverthe–period,asthestockofglobalissuanceexpandedfrom$billiontoalmost$trillionCDOissuanceimportantlysupportedthemarketforsubprimemortgagebackedsecuqJELclassifications:RGGGKeywords:CollateralizeddebtobligationsSubprimecrisisYieldspreadsonmortgagebackedsecuritiesIntroductiontocorroboratethefindingsDynamicsimulationbasedontheimpulseresponsefunctionestimatesindicatessubstantialsubprimeMBSspreadwideninginthewakeoftherecentimplosionintheCDOmarketResearchfindingssuggesttheimportanceofsupplydemandshocksassociatedwithinnovationsinderivativesecuritiesmarketstothepricingofsecuritizedsubprimedebtElsevierIncAllrightsreservedTheloanstosubprimeborrowersandthe‘‘AltA’’loansArticlehistory:AvailableonlineNovemberWeevaluatetheeffectsofCDOissuanceonthepricingofsubprimeresidentialmortgagebackedsecuritiesUponcontrollingformortgageoptionvaluesandotherwellestablishedmortgagebackedsecuritiesqYonghengDenga,,StuartAGabrielb,AnthonaInstituteofRealEstateStudies,NationalUniversityofSingapore,HengMbUCLA,AndersonSchoolofManagement,WestwoodPlaza,EntrepreneurscSchoolofManagement,GeorgeMasonUniversity,UniversityDrive,FaiarticleinfoabstractofsubprimeSanderscgTerrace,#,Singapore,SingaporeSuite,LosAngeles,CA,USAA,USAngEconomicsseviercomlocatejhecofsubprimeMBS–TreasuryspreadsBy,however,theunderlyingcollateralformanyCDOs,subprimeMBS,begantoexperiencesharpincreasesindelinquenciesanddefaultsInthewakeofsharperosionincollateralperformance,spreadsbegantowidenmarkedlyinbothsubprimeMBSandCDOsandtheissuanceofCDOsbegantodeclineBy,issuanceofCDOsbackedbysubprimeMBShadceasedTheriseandfalloftheCDOmarketprovidesanexcellentTableResidentialmortgagedealsinABSCDOsVintageSubprimeAltASecondsPrimeTotalNumberofdealsbyvintageandmortgageloantypeTotalSource:UBS,‘‘MortgageandABSCDOLosses,’’December,YDengetalJournalofHousingEconomics()–supplydemandbalanceandpricingofmortgagebackedsecuritiesInandinthewakeoftheimplosionintheCDOmarket,spreadsonmortgagebackedsecuritieswidenedmarkedlyCDOsarefinancialstructureswherebyasetofassetsareheldinatrustformedasaSpecialPurposeVehicle(SPV)AseriesoftrancheswithdifferentexposurestotherisksoftheunderlyingassetsareissuedbythetrustACDOcashflowstructure,forexample,allocatestheinterestandprincipalpaymentsoftheunderlyingcollateralpoolofdebtinstrumentstotheCDOtranchesWhiletherearemanyvariations,acashCDO,forexample,isaseniorsubordinatedstructurewheretheseniorCDOdebttranchesarepaidfirst,thenthemezzanineandlowersubordinatednotesAnyremainingcashflowisavailabletoequityIntheCDOstructure,asetofassets(suchascorporatebonds,CMBS,orresidentialmortgagebackedsecurities)canbepackagedintoclaimsandsoldtoinvestorsAsshowninTable,subprimemortgageassetbackedsecurities(ABS)overwhelminglydominatedothermortgageproducts(suchasprimeandsecondmortgages)asthecollateralforCDOsAsshowninFig,issuanceofsubprimeassetbackedCDOsmovedupfivefoldduringthefirsthalfofthedecadefromabout$billionintoinexcessof$billioninCoincidentally,spreadstoTreasuryonsubprimeMBSbackedCDOsnarrowedsubstantiallyIndeed,asevidencedinFig,yieldspreadsofsubprimeMBStrendeddownfromahighinexcessofbasispointsintoabouttwofifthsthatlevelyearslaterInterestingly,asshowninFig,thatmarkedtrendingdowninsubprimeMBStoTreasuryspreadsoccurredevenassubprimeMBSissuancewasrapidlyexpandingIncreaseddemandforsubprimeMBSproduct,forpurposesofderivativesecuritizationviatheCDOvehicle,mayhaveresulted,allthingsequal,insometighteningWhiletherulesfordistributingthecashflowsoftheCDObonds’underlyingcollateralarerelativelystraightforward,thevaluationofthedebtandequitytranchescanbecomplicatedThereasonforthiscomplexityisthatCDOsoftencontainalargeportfolioofassetsandcreditexposureswithdiverseriskprofiles(forexample,default,lossandrecoverypatterns)ascollateralandtherearemanypossiblecapitalstructuresfromwhichdebtandequityareissuedUnlikethosemortgageobligationswhereprepaymentriskisthedominantfocus,inCDOvaluationtheprimaryfocusistypicallyonthecreditriskAnumberofpractitionerorientedpapersprovideexcellentdiscussionsoftheCDOmarket(eg,GoodmanandFabozzi,Lietal,RoyandShelton,Tavakoli,)laboratorytostudyspreadsoncreditsensitivesecuritiesFollowingpreviousstudiesoncreditspreads(see,forexample,CollinDufresneetal,CollinDefresneandGoldstein,Chenetal,),weexaminetherelationshipbetweenthespreadsontheunderlyingsubprimecollateralandtheissuancevolumeofCDOsInsodoing,webuildonthepriorliteratureinevaluatingthedeterminantsofyieldsspreadsonmortgagebackedsecuritiesMuchoftheresearchonCDOshasfocusedonpricingofthosederivativesecurities(see,forexample,DuffieandGarleanu,HullandWhite,Noh,Longstaffetal,EricssonandRenault,LongstaffandRajan,PanandSingleton,)Brennanetal()evaluatetheeffectsoftranchingofdebtcontractsonthepricesatwhichdebtsecuritiescanbemarketedInaddition,therehasbeenconsiderableresearchonarelatedtopic:correlateddefaultamongunderlyingassets(see,forexample,Duffee,Duffieetal,Giesecke,Dasetal,GieseckeandGoldberg,Indiscussionofthecreditcrisis,Gorton()assessesCDOstructureandissuanceInparticular,GortonidentifiesratingrelatedarbitrageasaprimarymotivationforCDOissuanceFurther,theopportunityfornegativebasistradesmayhavebeenimportanttoCDOpurchaseofsubprimemortgagebackedbondsAlthoughrecentstudieshavefocusedonCDOpricingandissuance,littleresearchhasattemptedtoassessthedirectimpactofCDOmarketevolutiononspreadsintheunderlyingcollateralAlternativehypothesescanbeputforthregardingtheeffectsoftheemergenceoftheCDOmarketonthepricingofsubprimeresidentialassetbackedsecuritiesOntheonehand,totheextentCDOsconferefficiencybenefitsasmarketcompletinginvestmentvehicles,institutionaldemandforCDOsandrelatedderiveddemandforassetbackedsecuritiesmayincreaseAccordingtothismarketcompletionhypothesis,CDOsmayhaveservedtoenhanceefficiencyintheassetbackedsecuritiesmarkets,viathepoolingandretranchingoflessliquidandlowerratedsubprimeMBSintoderivativesecuritiesthatweremoretradableFurther,aliquiditypremiumassociatedwithseniorCDOtranchesmayhavereducedthecostofraisingcapitalthroughtranchesbelowthecostofacquiringtheGorton()suggeststhatby,yieldspreadsonsubordinatedsubprimeMBStrancheswereelevatedrelativeofotherstructuredproductsofsimilarratings,providinganopportunitytoarbitragetheratingsbetweentheratingsonsubprimeMBSandontheCDOtranchesSuchnegativebasistradescouldoccurtotheextentyieldsonCDOtranchesexceedrequiredpaymentforcreditriskprotection(intheCDSmarket)OriginatorsofstructuredproductshadanincentivetoengageinratingsrelatedarbitragetotheextentspreadsonsubprimeMBStranchesexceededthoseonsimilarlyratedCDOtranchesnce($bn)YieldSpreads(BPS))ssubYDengetalJournalofHousingEconomics()–assetpool(see,forexample,Greenbaum,DeMarzoandDuffie,Demarzo,),AssuminglessthanfullyelasticsupplyofsubprimeMBSinthecapitalmarkets,CDOrelatedrelatedincrementstosubprimeMBSdemandwouldresultinsomecontractioninmortgagebackedsecurityspreads,allthingsequalOntheotherhand,capitalizationandgrowthoftheCDOmarketcouldservetoelicitasupplyresponse,intheformofincreasedsalesofsubprimesecuritiesintherealestateassetbackedsecuritiesmarketsThosesalesmightderivefromopportunitiesforimprovedpriceexecutionandorenhancedoutletsforliquificationofinvestorresidentialmortgagebackedsecurities(RMBS)portfoliosAssuggestedbyGreenbaumandThakor(),theadventofnewsecuritiesmarketsallowsfinancialintermediariestoremoveassetsfromtheirbalancesheetsandreduceothercostsofholdingdebtAllthingsequal,thissupplyshifthypothesiswouldservetodepresssubprimeassetbackedsecuritypricesandconcomitantlyresultinwiderMBS–SubprimeYieldSpread(BPSFigYieldspreadvTreasuryspreadsIndeed,priorresearchhassuggestedtheimportanceofsectorspecificsupplydemandimbalancesinthedeterminationofyieldspreadsCollinDufresneetal()showDeMarzoandDuffie()andDeMarzo()buildliquiditybasedmodelsoftranchingInthosemodels,thesecurityissuermaypossessprivateinformationregardingsecuritypayoffthatmaycauseilliquidityHowever,theseniortranches(lowrisktranches)arelesssensitivetothe(CDO)issuer’sprivateinformation,andthusmayenjoygreaterliquiditythantheunderlyingcollateralOntheotherhand,DeMarzo()showsthatforaninformedintermediary,purepoolingandsalesofassetsfromthepoolisinferiortosellingassetsseparatelyThisisbecauseassetpoolingeliminatestheintermediary’soptionregardinghowaggressivelytomarketeachassetandthuscanreducethepayoffThisiscalledthe‘‘informationdestructioneffect’’However,therecanbeanoffsetting‘‘riskdiversificationeffect’’ofpoolingandtranching–inthattheintermediarycancreatelowerriskderivativesecuritiesfromtheassetpool,andsuchsecuritiesarelesssensitivetotheintermediary’sprivateinformationandaccordinglycanbemoreattractivelypricedtotheinvestorInthecaseofCDOs,gainsfromriskdiversificationwereexpectedtoexceedlossesfrominformationdestruction,suchthatonnetpoolingandtranchingfacilitatedhighergainsthanindividualassetsalesthatmonthlycreditspreadchangesoncorporatebondsarelargelydrivenbysupplydemandshocksthatareindependentofbothcreditriskfactorsandstandardproxiesforliquidityDuffieandSingleton(),controllingforcreditriskandliquidityfactors,findthatswapmarketsupplydemandshocksdriveunexplainedchangesinswapratesSimilarly,supplydemandshocksappearimportanttoanalysesofGinnieMaeyieldspreads(Boudoukhetal,)Evidencealsopointstotheroleofsupplydemandimbalancesinthedeterminationofotherresidentialmortgageyieldspreads(see,forexample,Bradleyetal,Lehnertetal,)Alternatively,accordingtoasimpleshellgamehypothesis,elevateddemandforsubprimebackedMBSmighthavebeendrivenbyinaccurateCDOunderwritingorratings,misinformation,misrepresentations,orlackoffullunderstandingoftherisksoftheCDOvehicleSomecombinationofthosefactorscouldservetoexplaintighterspreadstoTreasuryonsubprimeMBSAnalternativeIssuaSubprimeCDOIssuance($bn)primeCDOissuanceasymmetricinformationhypothesissuggeststhatCDOissuersmaypossessprivateasymmetricinformationregardingassetreturns,andengageinrelatedpricediscriminationviatranchingtomaximizeprofits(Oldfield()orusesuchpoolingandtranchingmethodologiestodiversityrisk(DeMarzo,)Finally,onemightpositaproductionefficiencyhypothesis,wherebyspecializationandverticaldisintegrationofsuchfunctionsasMBSsecuritization,ownershipandservicing,wouldservetoenhanceproductionefficiencyandinsodoingdecreaseMBSproductioncostsThishypothesisderivesfromwellknownworkbyGreenbaum()andOldfield()arguesthattranchingmayallowsecurityissuerstofurtherenhancereturnsviapricediscriminationAssumingthatthedemandfunctionsforvariousderivativeproductsareimperfectlypriceelastic,Oldfield()explainsthatthesecurityissuerseeksprivateinformationaboutinvestordemandviathesecuritydesignandsalesprocess,andusesthatinformationtosegmentthemarketandpricediscriminateamongdifferentsetsofcustomersInthecaseofCDOs,thepricediscriminationcouldbefacilitatedviatherebundlingofthepoolandthesellingthedifferenttranchesatdifferentpricesandproxiesforcreditriskinthemacroeconomy,importantlyaffectthemagnitudeofmortgagesecurity–TreasuryspreadsFurther,subprimeMBStoTreasuryspreadsevissublnCDOissuancedoesnotGrangercauseDSubprimespread*DSubprimespreaddoesnotGrangercauselnCDOissuance*IndicatesrejectionofthehypothesisatthelevelYDengetalJournalofHousingEconomics()–HessandSmith()whichpositsgainsfromspecializationofactivityineachstepoftheintermediationandsecuritizationfunctionHerewewouldsimilarlyanticipatesomecontractioninMBS–TreasuryspreadsInthispaper,weundertakeempiricalassessmentoftheeffectsofCDOissuanceonthepricingofsubprimeMBSWeemployvariousempiricalspecificationstoidentifyCDOissuanceeffectsFirstly,toaddresscausality,weundertakeGrangerCausalityWaldTestsResultsofGrangerTests(seeTable)supportthehypothesisthatissuanceofsubprimebackedCDOsGrangercausechangesinsubprime–Treasuryspreads,butnotviceversaInordertodisentangletheeconomicrelationshipbetweensubprime–TreasuryspreadsandissuanceofsubprimebackedCDOYieldSpreads(BPS)SubprimeYieldSpread(BPS)FigYieldspreadvalongwithotherriskfactors,weundertakeparametricestimationofsubprimeMBSpricingdeterminantsusingGMMmodelswithNewey–WestKernelerrorcorrectionThosemodelsspecifyandcontrolforwellestablishedoptionsembeddedinMBS–TreasuryspreadsFinally,tofurtherexplorethedynamicdependencybetweensubprime–Treasuryyieldspreadsandtheunderlyingtimeseriesvariables,weestimateastructuralvectorautoregression(VAR)modelwithchangesinsubprime–TreasuryyieldspreadsandissuanceofsubprimebackedCDOsasendogenousvariablesTheestimatedimpulseresponsesfromtheVARmodelarethenusedtodynamicallysimulatetheeffectsofimplosionofderivativeCDOmarketsonthepricingofsubprimemortgagebackedsecuritiesResultsofGMMmodelestimationsuggestthatfactorsassociatedwiththeterminationrisksoftheunderlyingsubprimeresidentialmortgagecontracts,includinginterestratevolatility,thetermstructureofinterestrates,TablereportstheGrangerCausalityWaldTeststatisticsThepvalue<allowsrejectionofthehypothesisthatthenaturallogofissuanceofsubprimebackedCDOsdoesnotGrangerCausechangesinsubprime–TreasuryspreadsIssuacne($bn)SubprimeMBSIssuance($bn)primeMBSissuanceTableGrangerCausalityWaldTesthypothesisFstatisticPvaluedenceonlylimitedsensitivitytoreturnsandrelatedreturnvolatilityamongalternativeassetclasses,notablyincludingthoseofequitymarketsFinally,holdingconstantthosefactors,researchindicatesthattheemergenceoftheCDOmarketwasassociatedwithasignificantcontractioninsubprimemortgagesecuritytoTreasuryyieldspreadsVARestimationshowsthatCDOissuancelaggedmonthsexertssignificant,negativeeffectsonsubprimeMBS–TreasuryspreadsAlso,asafurtherindicationoftheabsenceofreversecausality,noneofthelaggedcoefficientsforchangesinsubprimeMBS–TreasuryspreadsarestatisticallysignificantintheestimatedimpulseresponseofCDOissuanceWethenusetheestimatedimpulseresponsefunctionsfromtheVARstosimulatechangesinsubprimemortgagebackedsecuritypricingassociatedwithcumulativenegativeshocksintheissuanceofsubprimebackedCDOsResultsindicatethataonestandarddeviationnegativeshocktolnCDOissuanceoveraperiodof,,andmonths,resultsinawideningofsubprimeMBS–Treasuryyieldspreadsbybps,bps,andbps,respectivelyFurther,findingssuggestthataonestandarddeviationpositiveshockinsubprimeMBS–Treasuryspreadslastingfor–months

职业精品

精彩专题

上传我的资料

热门资料

资料评价:

/ 13
所需积分:0 立即下载

意见
反馈

返回
顶部

Q