基于KMV模型的我国上市公司信用风险研究

381020085MATHEMATICSINPRACTICEANDTHEORYVol.38　No.10　May,　2008　KMV王建稳1,　梁彦军2(1.　100144)(2.,　100144):　KMV,,19992006、、、.,,,,,.,2003—2006,.300,,.,.:　;;KMV;1　　　:2007-12-10:(20051D0500207);(km200710009011)　　,.、、、,.,,、.、、,、,,KMV..,KMV,.,,,,KMV,.KMV,,.,KMV,,,,.1.1　KMV,McQuown[1](1993),,,.Vasicek[2](1995)108,EDF,EDF.MattHewburBatIrinaKorablev[3](2002)KMV,KMVEDF,、EDFEDF.PeterCrodbieJeffBohn[4](2003)KMV,EDF、.《》(2004),KMV.Stefan.M.Denzler[5](2006)KMVEDF.KMVKMV.KMV,[6](2002)《》“”[7](2002)《》“”.[8](2003)KMV,.、[9](2003)KMVeVeE.[10](2005),EDF,.[11](2006)KMV,Logistic、Fisher,.1.2　,,19992006,.,..2　KMV2.1　KMVKMV,,,.,.VD,;DDVDP,,,;,EDF,.,Black-Scholes:E=VN(d1)-De-rtN(d2)=f(V,eV,r,D,f)(2.1)4710,:KMV　　(2.1),E,V,D,eV,r,f,N().,d1=ln(V/D)+(r+e2V/2)feVf(2.2)d2=ln(V/D)+(r-e2V/2)feVf=d1-eVf(2.3)　　(2.1),,,KMVeAeE.eE=VN(d1)EeV=g(V,eV,r,D,f)(2.4)(2.1)(2.4),,VeV.,(2.5):DD=V-DVeV(2.5)　　(2.5)D.(DD),.,EDF.EDF=N(-DD)　　,KMV,.2.2　1),.,,.,,,,,.,,,,,.,:P()=×(1+)=×2)eE.:ri=lnPiPi-1(3.1),Pi,Pi-1MatLab,GARCH,ed.48　　　　　　　38eE=ed×250(3.2)　　3),;,.,.KMV,,.4),,f=1..3　3.1　,(1)19992006(2)(3)(4)GARCH,3.1.3.1　(:)1999200020012002200320042005200663081991710531102117312671332133012318651003106911511218132213371105GARCH:1999112007523.wind.3.2　,,,,,,.,,50,,50.199920063.23.2　(:)19992000200120022003200420052006166129112125114119106773752719211515516618019992006(3.1),3.1,19992006(2002),4910,:KMVT,(3.3),5%,20012002,.19992002,.,.20032006,,.3.1　3.3　T199920002001200220032004200520062.112.282.332.432.212.042.051.942.412.482.442.372.322.252.212.07T0.300.200.11-0.060.110.210.160.13T-5.79-2.72-1.630.81-1.98-4.44-3.31-1.99P0.000.010.110.420.050.000.000.0516612911212511411910677375271921151551661803.3　3002005,,?2004300,.2006300,,T,T,3.4,5%,.T(3.5),5%,300,.50　　　　　　　383.4　KS-Z0.9070.742P()0.3840.6413.5　TTP2.14122.1017-0.03953.0630.0020.29850.32220.0237-5.420.0003.4　、、、,.2006,232,1190.0.941,,3.6.3.6.=1.572-2.015*+0.364*+0.003*　　,,.,,1%,2.015%.3.6　Betatp(Constant)1.572.008198.427.000-2.015.018-1.116-110.498.000.364.016.24322.914.000.003.001.0172.195.028　　:4　　　,,,,,.,,,.300,,,.,5110,:KMV,1%,2.015%.:[1]　JohnAndrewMcQuown,ACommentonMarketvs.Accounting-basedMeasuresofDefaultRisk[R].WhitePaper,Moody′sKMV.Septemper,1993.[2]　OldrichAlfonsVasicek.EDF(tm)CreditMeasureandCorporateBondPricing,1995,[3]　MatthewKurbat,IrinaKorablev.MethodologyforTestingtheLeveloftheEDF(tm)CreditMeasure2003,[4]　PeterCrosbie,JeffreyRBohn.ModelingDefaultRisk[R].WhitePaper,Moody′sKMV,RevisedDecember18,2003.[5]　StefanMDenzler.FinancialModelingGroupConveriumLtd,2006,GeneralGuisan-Quai26,8022Zu　　··rich,Switzerland.[6]　.[J].,2002,19(3):80-82.[7]　,.[J].,2002,22(4):14-16.[8]　,.[J].,2002,23(8):48-53.[9]　,,,.KMV[J].,2003,16(3):30-33.[10]　.EDF[J].,2005,19(3):151-154.[11]　.KMV[J].,2006,25(5):593-601.AnEmpiricalStudyoftheCreditRiskofListedCorporationinChinaBasedontheKMVModelWANGJian-wen1,　LIANGYan-jun2(1.CollegeofScienceNorthChinaUniversityofTechnology,Beijing100144,China)(2.CollegeofEconomyandManagementNorthChinaUniversityofTechnology,Beijing100144,China)Abstract:　ByimprovingmeasurementofequityvaluebasedonKMVmodel,thepapercalculatestheDistancetoDefault、companyvalue、equityvalueoflistedCompaniesinChinafrom1999to2006,etc.Accordingtotheanalysis,thesizeoflistedcompanyaffectedcreditrisk:Thebiggerthescaleoflistedcompany,thelowerthecreditriskwillbe;TheDistancetoDefaultshowedadownwardtrendfrom2003to2006.Itmeansthecreditriskincreasedinrecentyears.ByContrasting300listedcompaniesinShanghaiandShenzhenstockmarketbeforeandafterEquityDivisionReform,wediscoverthatcreditsituationhassignificantlydifferencebeforeandafterEquityDivisionReform.ListedcompanieshavehighercreditriskafterEquityDivisionReform.ThesensitivityanalysisshowsthatthevolatilityofsharesvalueismostsensitivetoDistancetoDefault.Keywords:　listedcompany;creditrisk;KMVmodel;distancetodefault52　　　　　　　38