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
overJPR’sA-SVmodelwhichis2.44×1014.Theevidencestronglyfavorstheconven-tionalspeci cationagainstJPR’sspeci cation.ThesameconclusionisalsodrawnfromthecomparisonoftheDICvalues.2
Analternativewayforcomparingthesetwomodelsistonestboththemodelsintoasinglemodel.Todoso,considerthefollowingspeci cation, X=σtut,t(4.7) lnσ22=α+φlnσ+σv,t+1tvt+1
with
ut
vt+1 iid~N 00 , 1
ρ1ρ11 and ut
vt iid~N 00 , 1
ρ2ρ21 .
Inthismodelweallowcorrelationatbothtimelags,butwithdi erentdegreesofcorrelation.Ifρ1=0,wehaveJPR’sA-SVmodel.Ifρ2=0,wehavetheconventionallyusedA-SVmodel.
TomakeuseoftheallpurposeBayesiansoftwarepackageBUGS,weobtainthefollowingstateandobservationequationsfortheencompassedmodel:
22ht+1|ht,α,φ,σv~N(α+φht,σv), ht/2e2(ρ2(ht α φht 1)+ρ1(ht+1 α φht)),Xt|ht+1,ht,ht 1,α,φ,σv,ρ1,ρ2,~Nσv 2eht(1 ρ21 ρ2).
2Weadoptthesamepriordistributionsforφandσvasbefore.Forρ1andρ2weassume
auniformpriorwithsupportbetween-1and1.Table2reportstheestimationresults,includingtheposteriormeans,standarddeviations,95%Bayescon denceintervalsforalltheparametersandtheharmonicmeanestimatesoflogmarginallikelihood.Theposteriormeanofρ1is-0.2939whiletheposteriormeanofρ2is-0.2140.Theycomparetotheposteriormeanof-0.3179intheconventionalA-SVmodelandtheposteriormeanof-0.2599inJPR’sA-SVmodel.The95%posteriorcredibilityintervalforρ1is[ 0.4490, 0.1443]whichindicatesthepresenceofasigni cantnegativecorrelationbetweenutandvt+1.The95%posteriorcredibilityintervalforρ2is[ 0.3619, 0.0828]whichsuggestssomebutweakerevidenceofnegativecorrelationbetweenutandvt.ThemarginallikelihoodvaluesfromtheencompassedmodelandJPR’smodeldi ersby2.03×1011.Theevidencestronglyfavorstheencompassedspeci cationagainstJPR’sspeci cation.Ontheotherhand,themarginallikelihoodvaluesfromtheconventionalspeci cationandtheencompassedmodeldi ersby1200.03whichfavorstheconventionalspeci cationagainsttheencompassedspeci cation.TheoverallrankingofthreemodelsistheconventionalA-SVmodelcomes rst,followedbythetheencompassedA-SVmodelandthenbyJPR’sA-SVmodel.
2OnehastochoosethemodelwiththesmallestvalueofDIC.
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