上海市场股票收益决定因素的实证研究

华中科技大学硕士学位论文上海市场股票收益决定因素的实证研究姓名：蒋俊锋申请学位级别：硕士专业：企业管理指导教师：夏新平20050421IBETACAPM20BETA199812401(CF/P)B/M1999712002630BETASIZEB/MB/MB/MCF/PCAPMB/MIIABSTRACTEarlyempiricalstudiesshowedthatexpectedstockreturnscanbedecidedbybetaandcanbeforecastbytheCAPMmodel.However,inthelast20yearsstudiesfoundthatexpectedstockreturnscannotbedecidedbybetaaloneandothercharacteristicvariablescanbeaddedtoexplainexpectedstockreturns,suchassizeandB/M.Usingasampleof401stockslistedonShanghaiStockExchangebeforeDecember,1998,thispaperaddspastsalesgrowthandCF/PtothestudyandappliesdifferentmethodstocalculateB/M.Thispaperanalyzesthecross-sectionaldeterminantsofexpectedstockreturnsbybothone-factormodelandmulti-factormodels.AndthispaperfindsthatBETAdoesnothaveexplanatorypowertoexpectedstockreturnsacrossallthemodels;SizeandB/Mcanalwayssignificantlyexplainexpectedstockreturns,theformerisnegativelyrelatedtoexpectedstockreturnsandthelatterispositivelyrelatedtoexpectedstockreturns,butSizehavenotexplanatorypowerwhenassociatedwithB/M;allotherfactors,suchaspastsalesgrowthandCF/P,don’thaveexplanatorypower;whensizeiscontrolled,theproportionofnegotiablesharestototalsharesalsohavesignificantexplanatorypower.ThispaperdiscussesthereasonwhythedeterminantssuchassizeandB/Mhaveexplanatorypowerafteraseriesofempiricalanalysis.AndthepaperalsousestheviewpointsofbehavioralfinancetoexplaintheanomaliesthatexpectedstockreturnsdeviatetheCAPM.Thepaperincludes5parts:introduction,summaries,empiricalparts(2parts)andtheconclusionoftheempiricalanalysis.KeyWords:Expectedstockreturn,SizeeffectBook-to-marketratio(B/M),Multi-factormodel111.110WTOQFII(Sharpe)(Lintner)(Black)CAPM(SLB)BETABETABETABETACAPMCAPMAnomalies(Data-snooping)CAPMCAPM2CAPMFamaFrenchCAPMFamaFrench(1992)Book-to-marketratioB/MBETAB/MBETA19411990BETA19631990FamaFrench(19931995)1.212CAPMBETA3BETA31(Salesgrowth)(CF/P)2(2001)(B/P)(B/M)ChenXiong(2002)70%-80%70%80%MEB/M(2001)422.1CAPM2.1.1CAPMCAPMSLBCAPMMarkowitz195219561959MarkowitzMarkowitzMarkowitz60SharpeLintnerMarkowitzMarkowitz519641965CAPMCAPM2.1.2CAPMRiimRfRiβiiεCAPMifmifiRRRRεβ+−+=)(2-1CAPMββicov(iR,mR)/var(mR)cov(iR,mR)ivar(mR)iβ1iβ1iβ1β0Markowitz6βCAPM2.1.3CAPM1)CAPMBlack,JensenScholesFamaMacbethBlack,JensenScholes197219311965CAPMBJSβ0β0FamaMacbeth197319351968CAPM1β2β2031935196840242-2402ppppRεσγβγγγε++++=32210(2-2)Rpβpβ2pββσεpβpεpγ0γ1γ2γ34025γ0γ1γ2γ3tβp7σεpCAPMCAPMCAPMCAPM2)2070CAPMCAPMCAPMβCAPMCAPMBasu(1977)Earnings-priceratioE/PE/PCAPME/PCAPMBanz1981Sizeeffect,CAPMBook-to-marketratio,B/M,Stattman(1980)Rosenberg,ReidLanstein1985B/MB/MB/MBasu(1983)BETAE/PBhandari(1988)(Leverage)βCAPMAnomalies8Ball(1978)CAPMCAPMEarnings-priceratioDebtequityratioB/MCAPM1992FamaFrench1992FamaFrench1992NYSEAMEXNASDAQ19621990BETAB/M,BETA,BETA,BETA19411990,BETA,19631990,B/MFamaFrench(1992)CAPM2-3FamaFrench(1996)2-4Rpt=pMttb1γεα++1,2−=∑tjpJjjtZγ+ptε,p=1,……,P,t=1,……,T.(2-3)itftMtiMiftitRRRRεβα+−+=−)((2-4)Z9FamaFrench(1992)(Reinganum(1981),Stambaugh(1982),LakonishokShapiro(1986))CAPMFamaFrench(1992)CAPMCAPMBarberLyon(1997)FamaFrench(1992),(Datasnooping),,;Kothari,ShankenSloan(1995)FamaFrench(1992)(sortingtechnique),(Datasnooping),Out-of-sampleChan,Hamao,andLakonishok(1991)B/MCapaul,Rowley,andSharpe(1993)B/MFamaFrench(1998)CAPM12CAPMDavis,FamaFrench(2002)1926B/MCAPMLakonishok,ShleiferVishny(1994),FamaFrench(1995)B/MB/MB/ME/PB/MB/MDeBondt10Thaler(1987),Lakonishok,ShleiferVishny(1994)Haugen(1995)CAPMFamaFrench(1993)Merton1973ICAPMB/MFamaFrench(19931996)Three-factormodelB/MB/MCAPMFamaFrench(1998)13CAPMFamaFrenchCAPMValueeffectB/MCAPMCAPMCAPMCAPMFamaFrench(2003)CAPMCAPMCAPMFamaFrench2.2Merton(1973)ICAPMIntertemporalcapitalassetpricingmodelCAPMICAPMCAPM11ICAPMt-1ICAPMt1t2tCAPMICAPMICAPMMMVMMVMMVFamaFrench(19931996)CAPMICAPM)()(])([)(HMLESMBERRERREihisfMiMfiβββ++−=−2-5SMBHMLB/MB/MRiRfRMRfSMBHML(2-5)ittihtisftMtiMiftitHMLSMBRRRRεβββα+++−+=−)(2-6iαi0FamaFrench(19931996)B/MSMBHMLCAPM2-6iαLoughranRitter(1995),MitchellStafford(2000)Carhart(1997)12CAPMFamaFrench1993CAPMCAPM,CAPMJedadeeshTitman1993momentumeffect312B/MCAPMCarhart1997MomentumfactorCAPMFrankelLee1998Dechow,HuttonSloan1999Piotroski2000B/MCAPM13CampbellShiller(1989),Vuolteenaho(2002)B/MB/MB/M,2.3BETA(2000)CAPMBETA(2000)(2000)1994919989ABCAPM2001(BETAB/M)ABETABETAB/MBETAB/M20032003(2001)14FamaFrench1992BETA20011533.1199812401A1999712002630149STPT199812BETA4080BETA40401()3.2(1)CAPMCAPMtiR,1,,1,,,−−+−=tititititiPDPPR3-1tiP,it1,−tiPit1tiD,It(2)(BETA)(3)(SIZE)SIZE=ln(ME)MEME=(180%)(3-2)16(4)(B/M)(5)(LEVERAGE)(6)(E/P)(7)(LTBL)(8)(CF/P):(9)(RSG)n-11999-2002n7n16n66303.33.3.1βtifmitiRRR,,)(εβα+−+=(3-3)3-3tiR,itmRfR()ti,εmRt11/)(−−−=tttmINDEXINDEXINDEXR(3-4)Scholes-Williams(SW,1977)iβˆS&W-17iβˆ010101021ˆ+−++++=miiiibbbbβ(3-5))(),(1,01mtmttmmRVarRRCovb++=)(),(0mtmtitiRVarRRCovb=)(),(1,01mtmttiiRVarRRCovb++=)(),(1,01mtmttiiRVarRRCovb−−=mtRtitRtiBETAβswβ3.3.2ttRiittWCRεα++=0,(3-6)Ct=t,1t,2t,kW=iSIZEiVkTt12149k128WCVkk149149149Bt,k181,11,kBt,k=(3-7)149,1149,k149T1944.14.1.14.1NRi-mean*401-0.005740.008790.002130.002200.00225401-0.3