文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > MCMC Methods for Estimating Stochastic Volatility Models with Leverage Effects Comments on
侵权投诉

MCMC Methods for Estimating Stochastic Volatility Models with Leverage Effects Comments on

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

2StateSpaceRepresentationofA-SVModels

Inthe nanceliteratureonoptionpricing,theA-SVmodelisoftenformulatedintermsofstochasticdi erentialequations.ThewidelyusedlognormalA-SVmodelspeci esthefollowingequationsforthelogarithmicassetprices(t)andthecorrespondingvolatilityσ2(t), ds(t)=σ(t)dB1(t),(2.1) dlnσ2(t)=α+βlnσ2(t)dt+σdB(t),v2

whereB1(t)andB2(t)aretwoBrownianmotions,corr(dB1(t),dB2(t))=ρands(t)=lnS(t)withS(t)beingtheassetprice.Whenρ<0wehavetheleveragee ect.Thisnegativecorrelationimpliesthatanegativeshocktothereturnincreasesthedebt-equityratioofa rmandsoincreasestheriskinessofthe rminsubsequentperiods(seee.g.Black(1976)).

Intheempiricalliteraturetheabovemodelisoftendiscretizedtofacilitateestima-tion.Forinstance,theEuler-MaruyamaapproximationleadstothefollowingdiscretetimeA-SVmodel: X=σtut,t(2.2) lnσ22=α+φlnσ+σv,t+1tvt+1

whereXt=st+1 stisacontinuouslycompoundedreturn,ut=B1(t+1) B1(t),vt+1=B2(t+1) B2(t),φ=1+β.Hence,utandvtareindependentiidN(0,1)andcorr(ut,vt+1)=ρ.ThismodelisestimatedbyquasimaximumlikelihoodinHarveyandShephard(1996)andbyMCMCinMeyerandYu(2000).

Comparingequation(2.2)withequation(8)inJPR,wenoteasmallbutimpor-tantdi erence.Insteadofassumingcorr(ut,vt+1)=ρ,JPRadoptthespeci cationofcorr(ut,vt)=ρ.Tofullyunderstandthedi erenceandalsotheirlinkagetotheleveragee ect,itisconvenienttoadopttheGaussiannonlinearstatespaceformwithuncorrelatedmeasurementandtransitionequationerrorterms.Todothis,denote wt+1≡(vt+1 ρut)/andrewriteequation(2.2)as

X=σtut,t(2.3) 1 lnσ22=α+φlnσ+ρσσX+σw,t+1tvttvt+1

wherewtisiidN(0,1)andcorr(ut,wt+1)=0.

22Obviouslylnσt+1isalinearfunctionofXtand lnσt+1/ Xt=ρσv/σt.Therefore,

ifρ<0andholdingeverythingelseconstant,afallinthestockreturn(ieXt<0)leads

22toanincreaseoflnσt+1andhenceσt+1.

Usingthesameapproach,werewriteequation(8)inJPRinthefollowingGaussian

上一页第4页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email