关闭

关闭

封号提示

内容

首页 Property delisting, housing cycle and pricing b…

Property delisting, housing cycle and pricing bias.pdf

Property delisting, housing cyc…

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《Property delisting, housing cycle and pricing biaspdf》,可适用于高等教育领域,主题内容包含dpGlades,USCanadRaforricesroperctsudownShonotgEvenoutagSipricemaybeabiased符等。

dpGlades,USCanadRaforricesroperctsudownShonotgEvenoutagSipricemaybeabiasedindicatoroftherealmarketconditionsThedegreeofsuchbiasisdependentonthefrequencyofdelistingInhotmarketwheresalesarequick$seefrontmatterElsevierIncAllrightsreservedCorrespondingauthorEmailaddresses:pchengfauedu(PCheng),llincobilanmsstateedu(ZLin),yingchunLiufsaulavalca(YLiu)Obviously,theimplicitassumptionisthatalllistedpropertiesarereasonablypricedInpractice,thisisensuredtoalargeextentbybrokerswhopresumablytendtorefusetolistanypropertieswithunrealisticpriceexpectationsinthefirstplace,especiallyincoldmarketTherefore,propertieslistedandthendelistedrepresenttrue(shadow)supply,whilepropertieswhoseownershaveunrealistichighpriceexpectationsarenotconsideredastruesupplyJournalofHousingEconomics()–ContentslistsavailableatScienceDirectJournalofHousingEconomicsjournalhomepage:wwwelseviercomlocatejhecdoi:jjhehousesarepartofthetruesupply(orshadowsupply)inthemarket,givendemand,thetrueequilibriumprice(ortheshadowprice)maybesignificantlylowerthantheobservedmarketpricesbasedonthesoldhousesIntuitively,ifallthehouseslistedforsalemustremainonthemarkettobesold,competitionamongsellerswilllikelybeintensified,thustheeventualpricesonaveragewillbeevenlower,andtheexpectedtimeonmarket(theshadowtimeonmarket)willbelongerthanthoseobservedonlyfromsuccessfulsalesInotherwords,theobservedmarkettypicallybasedononlytransactionprices,theyarelikelytosufferfromsuchabiasSimplyput,transactionbasedhousingindicestendtooverstatethetrueequilibriumpriceandunderstatethepricedeclinefrompreviousperiods,thusmakingthehousingmarketappeartobeinbetterconditionthanitreallyisduringdowncyclesThisproblemisintuitivelywellunderstoodButhowcanwecorrectsuchpricingbiasandfindoutthetruemarketconditionTheoreticalandempiricalsolutionstothisproblemhavesignificantpolicyimplicationsbecausetheKeywords:PropertydelistingHousingcycleHousepricesAwidespreadphenomenoninaestatemarket(suchastherecentUthathouseslistedforsaleoftendotendedperiodoftimeonmarketthemarepulledoffthemarketwithThisisknownaspropertydelistincycleoftherealusingmarket)isetsoldafterextually,manyofcceptableoffersncethedelistedanddelistingisrare,thebiasisnegligibleButincoldmarket,likeintherecentyears,thebiascanbesevereinthefaceoframpantpropertydelistingAnimportantimplicationofsuchbiasisthatitmaycausebiasedhousingpriceindicesSincemajorhousingmarketindices,suchasthewidelywatchedSPCaseShillerHomePriceIndices,areJELclassifications:GGmakerstoformlessbiasedviewsofthetruestateofthehousingmarket,especiallyduringthedowncyclesElsevierIncAllrightsreservedPropertydelisting,housingcycleanPingChenga,ZhenguoLinb,,YingchunLiucaDepartmentofFinance,CollegeofBusiness,FloridaAtlanticUniversity,bDepartmentofFinanceandEconomics,MississippiStateUniversity,MScDepartmentofFinance,InsuranceandRealEstate,LavalUniversity,Quebec,articleinfoArticlehistory:ReceivedJulyAvailableonlineFebruaryabstractThispaperprovidesvabletransactionpnumbersoflistedptoidentifyandcorrericingbiasRoad,BocaRaton,FL,USAAaGVAmalanalysisonawellknownissueofthehousingmarket–obserareabiasedindicationofthetruemarketconditionifsignificanttiesaredelistedwithoutsaleWeprovideaclosedformformulachpricingbiasThemodelcanhelpmarketparticipantsandpolicydegreeoftheindexbias,besidescausingpotentialmisjudgmentamongmarketparticipants,islikelytobemostsevereduringtheworstmarkettimes,whenaccurateassessmentofthetruemarketconditionismostneededforappropriatepolicyresponsesThepurposeofthisstudy,therefore,istoseekgeneralunderstandingoftheimpactofpropertydelistingonpropertyprice,andtodevelopanforsuchasuccessfulbidtoarrive(ie,timeonmarket)shouldfollowanexponentialdistributionwithparameterofgkThishasrecentlybeenempiricallyconfirmedbythefindingsofBondetal(),inwhichUKdataareusedAtP*tpOObservableBUnobservabletPInfact,ouressentialresultswouldholdunderawidevarietyofmorecomplexdistributionfunctionassumptionsIncertaincircumstances,itiswellrecognizedthatpotentialbuyerscangetintoa‘‘biddingwar’’inwhichtheybidapriceabovetheaskingpricehowever,thishappensrarelyandonlywhenthemarketisexceptionally‘‘hot’’orapropertyisdramaticallyunderpricedBasedonthedatafromtheNationalAssociationofRealtors,GreenandVandell()findthatsuchasituationoccursinaboutfivepercentoftransactionsStatisticsprinciplesstatethattheexponentialdistributionoccursnaturallywhendescribingthelengthsoftheinterarrivaltimesinahomogeneousPoissonprocessPChengetalJournalofHousingEconomics()–analyticalframeworkthatenablesustodeterminetheshadowpriceandtheshadowtimeonmarketwithobservedtransactionanddelistinginformationUsingempiricaldata,wedemonstratethat,whenthereisrampantpropertydelisting(asreportedinthemedia),thetruemarketconditionislikelytobesignificantlyworsethanthemajorhousingindicessuggestAmodeloftherealestatetransactionprocessInthissection,weexaminethematchingprocessoftherealestatetransactionThisprocessischaracterizedbysequentialsearchandbargainingDuringthesearchprocess,asellerreceivesoffersovertimefromastreamofbuyerswhoseofferpricesandtimingofarrivalarestochasticinnatureThebuyersmakebidsbasedontheinformationacquiredfromtheirsearchEachtimeabuyermakesanofferthesellerevaluatesthecostsandbenefitsofwaitingforapotentiallybetteroffer,anddecideswhethertoacceptthecurrentbidIfthebidisrejected,thesearchcontinuesAcommonstoppingruleforthisprocessistoassumethatthesellerwillacceptthefirstbidabovethereservationprice,andtorejectallbidsbelowitThereservationpriceisaffectedbytheseller’sholdingcostsduetohisheruniquepersonalandfinancialconstraintsGenerallyspeaking,sellerswhoaremoreconstrainedtendtosetlowerreservationpricesinordertoincreasetheprobabilityofsalewithinaplannedorexpectedtimeframeToformallymodelsuchaprocess,wemakethefollowingassumptions:a)Thedistributionofbuyers’arrivalsWeassumethebuyers’stochasticarrivalfollowsaPoissonprocesswithratekThisassumptioniswidelyadoptedbypreviousstudiesincludingSirmansetal(),Arnold(),Gloweretal(),Miceli(),Chengetal(),amongothersItisalsowidelyusedinstandardsearchmodelsinlaboreconomicsb)ThedistributionofbiddingpricesConsiderasellerwhoplacesapropertyonthemarketattimeandsellsitattimetAsshowninFig,assumethedistributionofbiddingpricesisuniformlydistributedoverPPwithdensityfunctionf(Pbid),wherePandParetheminimumandmaximumbidprices,respectively,andptistheseller’sreservationpriceBythestoppingrule,aselleronlyacceptsanofferthatisatleastashighasthereservationpriceThatSee,forexample,TheNewYorkTimesreport:‘‘HomeSalesPutSharesInaTailspin’’,Jul,,Page,ByMichaelMGrynbaumandEricDashEarlystudiesinlaboreconomicsliteratureoftenrelyonthisassumption(egStigler(),Whipple()andBarron()),andithasbeenextensivelyappliedtotherealestatemarketsincethes(egYinger(),Read(),QuanandQuigley(),Yavas(),Arnold(),LinandVandell(),LinandLiu())is,thedistributionoftransactionpricesisatruncateddistributionofbiddingpricesThedegreeofthetruncationdependsonthereservationpriceThehigherthereservationprice,thehigherthelikelytransactionprice,butthesmallertherangeoftransactionpricevariationsTheassumptionofbiddingpricesbeinguniformlydistributedisanotherwidelyadoptedassumptioninearlystudiesincludingRead(),Yavas(),Sirmansetal(),Arnold(),andChengetal()Fortechnicalsimplicity,weadoptthesameassumptionSincenobuyerwillnormallybidsabovethelistingprice,PcanberegardedasthelistingpriceInaddition,weshouldpointoutthattheunderlyingvalueofthepropertyiscloselyrelatedtobothholdingperiodandmarketconditionsTosimplifynotation,weintentionallyomitthesubscriptsdenotingmarketstateandholdingperiodHence,thebuyers’bidpriceisdistributedasfðPbidÞðPPÞPbidPPotherwise(ðÞGiventhetwoassumptionsabove,itisobviousthattherearetworandomprocessesinthedeterminationoftheprobabilityofarealestatesaleThefirstisthepotentialbuyers’stochasticarrival,whichisassumedtofollowaPoissondistributionwithconstantratekThesecondistheprobabilityoftheofferingpricebeingabovetheseller’sreservationprice,PpPPgivenkLetgPpPP,thearrivalrateofasuccessfulbidisthenthejointprobabilityofthetwostochasticprocesses,iegkTherefore,thetimeittakestFigRealestatebiddingpricesandtransactionpricesTDLgkðdÞðÞtransactions(T)ratherthantheshadowtimeonmarket(TDL),thesellershavetocompeteharderUndertheassumptionofhomogeneouspropertyandmarketparticipants,theimplicationisthatsellerswillhavetolowerPChengetalJournalofHousingEconomics()–Eqs()and()suggestTDLTdðÞSinceTisobservablefromtransactions,thetrue(orshadow)timeonmarketTDLcanbecomputedgivendelistingpercentagedAlso,since(d),Eq()suggeststhat,inthepresenceofdelisting,theshadowTOMisalwayslongerthanwhatcanbeobservedsolelyfromthesoldproperties,andthatahigherpercentageofdelisting(oftenoccursinweakrealestatemarket)impliesabiggerdifferencebetweenthetwoFromthepriceperspective,ifallthepropertiesweretobesoldwithinthesametimeframeasthoseobservedtoinvestigateanumberofpossibleassumptionsaboutthedistributionoftimestosale,suchasthenormal,chisquare,gammaandWeibulldistributionsBondetal()findsthattheexponentialdistributionexplainsthedatabetterthantheothersForexponentiallydistributedtimeonmarket,themean,ortheexpectedTOMofthesale,issimplyTEðTOMÞgkðÞItisimportanttonoticethat,giventhebiddingpricedistribution,gisdeterminedbytheseller’sreservationpricep⁄,whichinturnisaffectedbytheseller’sconstraintInotherwords,gcanbeconsideredasanindicatoroftheseller’sconstraintCorrectionofpricingbiasinthehomogeneousmarketBasedonthemodelabove,wenowstudytheeffectofpropertydelistingonrealestatepricereturnandvolatilityWebeginwithasimplesettingofahomogeneousrealestatemarket,inwhichweassumethereisonlyonetypeofpropertiesandallmarketparticipantshavehomogeneousexpectationsThisisausefulsimplificationifoneagreeswiththattheheterogeneousrealestatemarketcanbeconsideredastoconsistofmanyrelativelyhomogeneoussegmentsInthissectionwefocusonthewithinsegmentanalysisInthelatersection,wewillextendthemodeltointersegmentsanalysisinthecontextofheterogeneoushousingmarketAspreviouslydiscussed,ifthesupposedlydelistedpropertiesweretoremainonthemarketuntilasuccessfulsale,theobservedaveragetimeonmarketofallsoldpropertieswouldbelongerThisnotioncanbedemonstratedformallySupposethatthereisapoolofhousingsupplyinthemarket,letd(d)denotethepercentageofpropertiestobedelistedinthepool,sothepercentageofthepropertiesremainingonthemarketis(d)Theprobabilityofasuccessfulsalenowcanbeviewedasthejointprobabilityofthebuyer’sarrival(k),theofferingpricebeingabovethereservationprice(g),andthehouse’sremainingonthemarket(d)SimilartoEq(),theexpectedTOMuponasuccessfulsaleshouldbe,theirreservationpricetoP⁄<p⁄inordertoincreasetheirchanceofsuccessfulsaletogPPPPFromEq()and(),letTDLgkðdÞTgk,wecanconcludethat(g⁄)mustsatisfy:ggdðÞItisobviousthat,sinceg<g⁄,gindicatedbythesuccessfulsaleswilloverstatethereservationprice(thustheexpectedsalesprice)ifmorepropertiesarelikelytobedelisted(ied>)Thedegreeoftheoverestimationispositivelyrelatedtod–fromnegligibleinhotmarketwhenallpropertiesaresoldquicklytoveryhighincoldmarketwhenasignificantnumberofpropertiesarepulledoffthemarketNowwemovetoexaminetheimpactofdelistingonrealestatereturnandriskAsillustratedinFig,eachbiddingprice~Pbidreceivedbythesellerimpliesatotalreturn~Rbidoveraholdingperiods,whichcanbeexpressedas~Rbid~PbidPPðÞwhere~PbidisuniformlydistributedoverPPandPistheoriginalpurchasepriceSimilarly,thetotalreturnuponsaleforasellerwithreservationpricep⁄orgcanbeexpressedas~Rg~PgPPðÞwhere~PgisuniformlydistributedoverpPFromEq(),theexpectedtotalreturnofbiddingpricesisuMsE~PbidPP"#ðPþPÞPðÞAndthevolatility(orrisk)ofsuchreturnisrMsVar~PbidPP"#ðPPÞPðÞwhere(uMsrMs)characterizethedistributionofbiddingpricesfrompotentialbuyers(orbuyers’valuation)Similarly,theexpectedtotalreturnandvolatilityuponasuccessfulsaleintheabsenceofdelistingcanbecomputedas,usE~PgPP~PPpi"pþPPðÞrsVar~PgPP~PPpi"ðPpÞPðÞSincegPpPP,wehavegpPPpðÞThiscanalsobeseenbynotinggPPPPandgPpPP,whereP⁄<p⁄Tosimplifynotation,weintentionallyomitthesubscriptdenotingtheholdingperiodinpricesffiApplicationoftheformulasinEq()and()requiresknowledgeofthepercentageofdelisting(d)Empirically,dnariosandshowstheimpactofdelistingontheaveragePChengetalJournalofHousingEconomics()–GivenðPþPÞPðPþpÞPffiffiffippPPpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðPpÞPsðÞandffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðPPÞPsþpPPpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiðPpÞPsðÞusingEqs()–(),wecanthereforerewriteEqs()and()asuMsusffiffiffipðgÞrsðÞrMsgrsðÞEqs()and()illustratehowthebuyers’valuationcanbeinferredfromtheobservedtransactioniftherearenopropertiestobedelistedAtthepresenceofdelisting,inordertoachievetheobservedTOMinEq(),thesellerneedstosetlowerreservationpricesP⁄whereg⁄shouldsatisfyEq()andP⁄<p⁄SupposeuDLsandrDLsaretheexpectedreturnandvolatilitywheng⁄satisfiesEq()Thebuyer’svaluationofthesameproperty(uMsrMs)canbeinferredinthesimilarfashion:uMsuDLsffiffiffipðgÞrDLsðÞrMsgrDLsðÞFromEqs()–(),wecanhaveusffiffiffipðgÞrsuDLsffiffiffipðgÞrDLsðÞgrsgrDLsðÞSimplifyingEqs()and()byusingEq()yieldstherelationshipbetweentheobservedtransactionsindicatedby(us,rs)andtheshadowreturnandvolatilitydescribedby(uDL,s,rDL,s):uDLsusffiffiffipddrsðÞrDLsdrsðÞNotethatifd=,thatis,nopropertyisdelisted,therewillbenopricingbiasandtheobservedreturnandvolatilityareinfactunbiasedestimatesoftheexpectedreturnandvolatilityButifd>,itcanbeeasilyshownthat,uDLsus<andðuDLsusÞd<ðÞrDLsrs>andðrDLsrsÞd>ðÞEqs()and()suggestthat,inthepresenceofdelisting,theobservedreturnbasedontransactedpropertiesisupwardbiased,whiletheobservedvolatility(risk)isdownwardbiasedSuchbiasesaregreaterifalargerpercentageofpropertiesonthemarketaredelistedindexreturnandvolatilityduringdownmarketperiodsTheshadowreturnandvolatilityarecomputedbasedonEqs()and()foreachassumeddelistingpercentageAsshownabove,iftheshadowreturnandvolatilityindicatethetruemarketconditions,itisclearthattheobservedreturnsignificantlyoverstatesthetruemarketreturnandunderstatesitsvolatilitywhendelistingisnotnegligible(ieormore)Ifoneacceptsthatpropertydelistingiswidespreadindownmarketssuchastheonewearecurrentlyexperiencing,theshadowreturnandIncomparison,duringtheupmarketperiods,theaverageannual(yearoveryear)indexreturnandvolatilityareand,respectivelyforaparticulartimeperiod(ieaquarter)canbeestimatedfromthetotalnumberofpropertieslistedforsaleatthebeginningofthequarterandthenpulledoffthemarketwithoutsaleduringthatquarterForresidentialproperties,thisinformation,thoughnotoftenreported,canbeextractedfromtheMultipleListingService(MLS)Thisstudyhighlightstheusefulnessofthedelistinginformationandcallsforsystematictrackingandrecordingdelistingpercentageovertime,sincethisinformationissocriticaltounderstandthetruehousingmarketconditionsNumericalexampleNowwepresentanumericalexampletodemonstratetheapplicationofEqs()and()toprovidesomepracticalsenseastothemagnitudesofthepricingbiasesduetopropertydelistingTobeginwith,itisnecessarytostressthattheformulascanbeappliedtoanyrelevantdatainspecificmarkets,aslongasthedelistinginformationisavailableForconvenience,weusethedatafromoneconvenientsource:theOFHEORepeattransactionHomePriceIndexfortheperiodfromtoWeusetheOFHEOindextocomputethequarterlyUShomepricegrowthfortheperiodofanalysisFigpresentsahistoricperspectiveoftheUShousingmarketviatheOFHEOHomepriceindicesDuringthepastyearsorso,thehousingmarketexhibitsclearcyclicalpatternsInnominalterms,thehousingmarketneverexperiencednegativeyearoveryeargrowthrateuntilveryrecentlyInrealterms(inflationadjusted),negativegrowthratehappenedduringtheearlys,theearlys,and,ofcourse,themostrecenttwoyearsofandDuringthesedownmarketperiods,itisexpectedthatsalesareslowandtimeonmarketisrelativelylongPropertydelisting,therefore,ismorelikelyduringtheseperiodsFromtheindexdata,wecanobservetheaverageannual(yearoveryear)indexreturnandstandarddeviation(volatility)overthesedownmarketperiods

职业精品

精彩专题

上传我的资料

热门资料

资料评价:

/ 6
所需积分:1 立即下载

意见
反馈

返回
顶部

Q