In this note we represent the well known discrete time stochastic volatility (SV) model with a leverage effect and the SV model of Jacquier, Polson and Rossi (JPR) (2002) using Gaussian nonlinear state space forms with uncorrelated measurement and transiti
study,Andersen,ChungandSorensen(1999)comparedthe nitesampleperformancesofvariousmethodsforestimatingthelognormalSVmodelandfoundthatMCMCisthemoste cienttool.Their ndingisnotsurprisingsinceMCMCprovidesafullylikelihood-basedinference.
MotivatedbytheempiricalevidencethatthebasicSVmodelcanbetoorestrictiveformany nancialtimeseries,Jacquier,PolsonandRossi(2002)(JPRhereafter)extendtheirearlierworktoanalyzeaSVmodelwhichgeneralizesthebasicSVmodelintwodi erentdimensions,onebyreplacingtheGaussianinnovationbyafat-taileddistribu-tionofinnovationsandonebyincorporatingtheso-calledleveragee ect.TheSVmodelwithaleveragee ect(whichisalsotermedtheasymmetricSV(A-SVhereafter)modelinHarveyandShephard(1996))isparticularlyimportantfromthe nanceperspectiveandisconnecteddirectlytothecontinuoustimeSVmodelswidelyusedinthe nanceliteratureonoptionpricing;seeforexampleHullandWhite(1987),Wiggins(1987),andChesneyandScott(1989).Sincetheparameterwhichcapturestheleveragee ectisofcriticalimportance,JPRstudythesamplingpropertiesoftheBayesMCMCesti-matorusingMonteCarloexperimentsand ndlittlelossinprecisionfromaddingtheleverageparameterintothebasicmodel.Theyalso tthemodeltomanyreal nancialtimeseriessequencesand ndoverwhelmingevidenceofastrongleveragee ectinmost nancialtimeseriesconsidered.
Althoughwe ndtheproposedMCMCalgorithmandsimulationandempiricalre-sultsinJPRveryinteresting,theirspeci cationoftheA-SVmodelleavesusabitpuz-zled.First,theirspeci cationdoesnotcorrespondtothewellknowncontinuoustimeA-SVmodelcommonlyusedinthe nanceliteratureandhenceitislessusefulfromtheoretical(suchasoptionpricing)viewpoints.Infact,theirspeci cationisnotevenconsistentwiththee cientmarkethypothesisbecausethemodelisnotamartingalesequence,asnotedinHarveyandShephard(1996).Second,we ndtheirspeci cationisempiricallyinferiortotheconventionalspeci cationwhenSP500dataareused.Third,sincetheirspeci cationisoflittletheoreticalrelevanceaswellasoflimitedempiricalimportance,thesamplingpropertiespresentedintheirpaperarenotpracticallyveryuseful.
Onepurposeofthisnoteistoclarifythepuzzle.Toachievethisobjective,wederiveaGaussiannonlinearstatespacerepresentationoftheconventionallyspeci edA-SVmodel.Usingthenewrepresentationweshowthattheconventionalspeci cationcancapturetheleveragee ect,butthesameargumentisnotnecessarilytrueforthespeci cationinJPR.Wethen tbothmodelstoacommonlyusedstockindexbasedontheallpurposeBayesiansoftwarepackageBUGS,asdescribedinMeyerandYu(2000)andshowthatJPR’sspeci cationisinferior,judgedbyBayesianstatisticalcriteria.Theremainderofthenoteisorganizedasfollows.Section2derivesthestatespacerepresentationofbothmodelsandexplainswhytheleveragee ectmaynotbewarrantedinJPR’sspeci cation.Section3 tsbothmodelstoanSP500index.InSection4,wepresentthesamplingpropertiesoftheconventionallyspeci edA-SVmodel.Section5concludes.
MCMC Methods for Estimat... 暂无评价 11页 免费 Volatility in Financial ....Asymmetry and Leverage in Stochastic Volatility Models-An Exposition 隐藏>> Asym...
MCMC Methods for Estimat... 暂无评价 11页 免费 Volatility in Financial ....Asymmetry and Leverage in Stochastic Volatility Models-An Exposition 隐藏>> Asym...
热门文档