资源描述:
《LONG MEMORY STOCHASTIC VOLATILITY IN OPTION PRICING》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、May3,200514:8WSPC-104-IJTAFSPI-J07100301InternationalJournalofTheoreticalandAppliedFinanceVol.8,No.3(2005)381–392cWorldScientificPublishingCompanyLONGMEMORYSTOCHASTICVOLATILITYINOPTIONPRICINGSERGEIFEDOTOV∗andABBYTANSchoolofMathematics,TheUniversityofManchesterM601QD,
2、UK∗sergei.fedotov@manchester.ac.ukReceived16March2004Accepted21September2004Theaimofthispaperistopresentastochasticmodelthataccountsfortheeffectsofalong-memoryinvolatilityonoptionpricing.ThestartingpointisthestochasticBlack–Scholesequationinvolvingvolatilitywithlong-r
3、angedependence.WedefinethestochasticoptionpriceasasumofclassicalBlack–Scholespriceandrandomdeviationdescribingtheriskfromtherandomvolatility.Byusingthefactthattheoptionpriceandrandomvolatilitychangeondifferenttimescales,wederivetheasymptoticequationforthisdeviationinvo
4、lvingfractionalBrownianmotion.Thesolutiontothisequationallowsustofindthepricingbandsforoptions.Keywords:Longmemory;stochasticvolatility;optionpricing.1.IntroductionOverthelastfewyears,self-similarityandlong-rangedependencehavebecomeimportantconceptsinanalyzingthefinanc
5、ialtimeseries[24,26].Thereisstrongevidencethatthereturn,rt,haslittleornoautocorrelation,whereasitssquare,r2,orabsolutereturn,
6、r
7、,exhibitnoticeableautocorrelation[6].ThisphenomenonttcanbedescribedbytheARCH(p)model[14]oritsGARCH(p,q)extension[4].However,theexponentiald
8、ecayforλ=cov(r2,r2)isbelievedtobetoofasttostt+sdescribecorrectlythepersistentdependencebetweentheseriesobservationsasthetimelagincreases.Itturnsout[3,27]thatthemodelswithhyperbolicdecaywhichhaveslowlydecayingcovariancesprovidebetterfittingtofinancialtimeseries.Thechara
9、cteristicfeatureofthesemodelsisthattheircovarianceλshasthepowerlawdecays2d−1(010、ay3,200514:8WSPC-104-IJTAFSPI-J07100301382S.Fedotov&A.Tanthecorrelationsdecayveryslowlytozero.Letusnotethattheseriesissaidtohavesho