基于信贷风险管理的财务危机预警系统

华中科技大学硕士学位论文基于信贷风险管理的财务危机预警系统姓名：庞卫宏申请学位级别：硕士专业：企业管理指导教师：陈君宁20060429IAltman20022006T-3;;IIAbstractThecreditriskisoneoftheoldestrisksinthefinancialmarketandisalsothecoreinthemanagementofcommercialbank.Therearelotsofcorporatebefoundupanddyeeveryday.Thefoundation,developmentanddyingofcorporationisboundupwithcommercialbank,soithasbecomethefocusoftheoryanddemonstrationthathowtofindthesignoffinancialcrisis.Thetheoryoffinancialcrisisbeintroducedintothecreditriskmanagementofcommercialbankisveryimportantinthetheoryandreality.Thecreditriskmanagersofcommercialbankfocusontheprofitabilityandabilityofrepayment,especiallyrepaymentinshort-term.Byusingtheearlywarningsystemoffinancialcrisisthemessageofcorporation’sfinancialstatementcansendtothecreditmanagersintime,andbringdownthecreditriskutmost.SinceAltmanappliedthemultiplelinearregressionmodeltothepredictionofcorporatefinancialrisk,theearlywarningmodeloffinancialcrisishasbeenwidelyusedinthefieldofcorporatecreditrankingandloanranking.Theearlywarningmodeloffinancialcrisisincludessinglevariablemodel,multiplelinearregressionmodel,logisticregressionmodelandANNmodeletc.Onthebasisofillustrationoftheoryofearlywarningoffinancialcrisis,thispapercomparestheadaptability,advantageandshortcomingofeachmodel.Usingfinancialdataof“specialtreat”corporationsinChinesestockmarketwechooselineprobabilitymodelandlogisticregressionmodelasthecreditriskearlywarningmodel.Thelogisticregressionmodelisexcellentthanthelineprobabilitymodel,soitshouldbethefirstchoiceincreditriskmanagementofcommercialbank.Thefinallineprobabilitymodelandlogisticregressionmodelaremadeupofcashflowvariables.Fromthiswecanknowthatthecorporation’scashflowincludemoreinformationaboutitsfinancialcondition,andthemodelbasedonthecashflowhashighercorrectratio.Keywords:FinancialCrisis;CreditRisk;LogisticRegression_____111.11.1.11997·[1]801994[2]WTO[3]2[4]1.1.21.1.3SPSS31.21.2.1[5]Carmichael[6]Froster[7]Ross[8][9][10]4,[11-13]1.2.2[14]()Ross*STST*ST,ST522.1[15][16][17]62.2[18]2.2.1=+=+[19]1272.2.2[20]122.2.312832.2.4[21]12;32-192.311[22]240%80%[23]3[24]207050050010212200536341133.1Fitzpatrick[25]19SmithWinakor/[26]Beaver[27]1954196479/90%88%Beaver199827ST2712199519974[11][28]3.23.2.11ZAltman[29-30]1968Z3333221-5554321999.0006.0033.0014.0012.0XXXXXZ++++=3.11X=/2X=/133X=/4X=/5X=/AltmanZZ1.811.81Z2.99Z3.0ZAltman1977ZZZETA//7196919755358ZETA70%91%ZETA1968AltmanZ2ZZ310[31]6543215.23.27.16.11.2XXXXXXZ+---+=3.21X2X3X4X5X6XZZ100Z10Z03F[32]ZF543214961.00302.09271.11704.01091.11774.0XXXXXF+++++-=3.31X2X3X4X5X0.0274,F0.0274,F0.0274143.2.2LogisticSS01[33-34]()()∑∑+++=iiiiiXXPbabaexp1exp3.4∑+=⎟⎟⎠⎞⎜⎜⎝⎛-=iiiiiXPPLnYba13.5XiiaibiPiOhlsom[35]1970-19761052058Ohlsom9496.12[13]200170705216FisherLogistic153.2.3BPArtificialNeuralNetworkANN501985RumelhartBPMinsky3-1Barbro[36]Swales[37]()()()()3-116[38]966462050%80%[39]BP1206090.8%90%3.2.4AndreasTrigeorgis[40]B-S19831994139LogisticLogistic3.3[27]79//ST[41-44][45-46]CEVCashEarningValueCAVCashAddedValue17=+++=FEWCIFEWCI0.10.1FEWCI0-0.1FEWCI0-0.3FEWCI-0.11844.1200461988194.24.2.1CharacterCapacityCapitalCollateralCondition5W5P5WWhoWhyWhenWhatHow5PPersonalPurposePaymentProtectionPerspective“5C”“5C”“5P”,,,,4-1CharacterCapacityCapitalCollateralCondition5C204.2.2[47]EdwardI.Altman1968ZZ-scoremodel1977ZETAZETAcreditriskmodelZ4.2.3J.P.1994VaR(RiskMetrics)1997(CreditMetrics)CreditRisk+KMVEDFKMVMckinseyCreditPortfolioView199944.321[48]1232255.11200620025-123222345.25.2.1*STSTSTST*ST12,34567STT-3TST245-1STSTT-1T-2TT-1T-225T-1T-2ST2002-20062131055-1200620063112005252004202003282002315.2.2ST123200320032105-221014862199819995-25.2.3LPM(Logistic)265.3Wind5.3.1123455.3.222,55-3275-3x1(ROE)=/x2(ROA)=/x3/x4=/x5=/x6=/x7=/x8=/x9=/x10=/x11=/x12=()/x13=/x14=/x15=/x16=/x17=/x18=/x19=/x20=(-)/x21=(-)/x22=(-)/1285235.45.4.1747422Z[][]2/1001101//NSNSMMZ+-=(5.1)1M1S1N0M0S0NZ5-45-51ST222-0.03650.10393ZZ295-4Z305-5Z5.4.21235-460%—70%85%315-245-4X2X6X7X8X9X10X11X13X14X16X20115.4.3SPSS11F0.050.105-65-65-2321LPMY=0.667-0.045X2+0.217X7-0.659X9(5.2)YX2ROAX7X925-6LPM0.05F9.450.0013TOLVIFTOLj=1-R2j=1/VIFj5.3R2Xjk-1TOLjVIFj105-6VIFj10R247474T-33100.50.50.55-7745067.5%5675.6%71.6%5-75.5335.5.1Kolmogorov-Smimov(K-SSPSSK-S()()XFxSnD0max-=5.4()xSn()XF05-8148148148148148148148148345-9SPSSDs0.0514822K-S5-85%[49]ST35Mann-WhitneyUSPSS5-90.05150.05X1X2X3X6X7X8X9X10X14X15X16X17125.5.2SPSS1485-1015-10Waldp0.055-103625-11-2LoglikelihoodNagelkerkeRSquareNagelkerkeR2NagelkerkeR2=R2/R2max5.5R2max=1[L(0)]2/NNL(0)0-11SPSS5-115-1235-10()()9729721422.41622.12216.08652.0exp11422.41622.12216.08652.0expXXXXXXPi-+-+-+-=5.60.5P0.537STY0.574ST177424ST,28%72%5.5.315.6P235.6625-13319772.6%3831ST101166.1%5-133966.112100%3100%4046.2(1)210148(2)T3T4T520022006(3)4142200641343[1]BestRW.TheRoleofDefaultRiskinDeterminingtheMarketReactiontoDebtAnnouncement.TheFinancialReview,1997,(2):87-105[2]..1999.41-55[3]..2001121-24[4].[].2004[5]EdmisterRO.AnEmpiricalTestofFinancialRatioAnalysisforSmallBusinessFailurePrediction.JournalofFinancialandQuantitativeAnalysis,1972,(3):28-45[6]Camichae