信用评分模型在我国企业贷款评估中的应用研究
摘要
本文以信用风险评价问题为核心，通过采用多元判别分析模型、
Probit和Logit模型、概率神经网络模型以及数据包络分析对我国企业
的信用风险度进行信用评分模型的建立。在研究方法上，本文采用理论
研究和实证分析并重的方式。在问题研究上，着重定量分析的运用，从
而得到既有理论依据同时也具有现实可操作性的解决方法。
本文首先对现有信用风险评估方法－专家评分法，信用评分模型，
现代信用风险度量模型进行了综合论述。接下来讨论了各种方法的优缺
点，以及各种方法在中国目前的可行性。然后介绍了信用评分模型建立
的主要方法，多元判别分析和二元Probit、Logit模型，并且引入新的
方法：概率神经网络和数据包络分析技术，对这两种方法在信用风险评
估中应用的可行性进行论述。最后将以上四种方法应用于我国上市公司
贷款信用风险度的评估，采用主成分分析法进行指标选取。
本文综合了财务管理、计量经济学和信息技术相关内容，运用多种
方法建立企业信用风险度和多维财务指标之间的量化关系。通过实证结
果表明，建立的信用评分模型具有较好的预测效果。
The main purpose of this thesis is to find out how to quantify credit
risk which appeared in loan appraisal. Multi-Discriminate analysis、Probit
and Logit models、Probabilistic Neural Networks and Data Envelopment
Analysis are all adopted to develop credit scoring models. To find the
best and practical method, a quantified analysis and experimental study
are emphasized..
First existing methods were introduced, namely specialist’s score
method, credit score model, and modern credit risk models. Then
advantages and disadvantages of each method were discussed and the
adaptability in China is presented. Some popular techniques for
developing credit scoring models such as multiple-discriminate analysis,
probit and logit models were also introduced. In addition, probabilistic
neural networks and data envelopment analysis techniques for credit
scoring are presented. Lastly four techniques were applied to
measurement of credit risk for listed companies in China, in which main
factor method was adopted to select input variables. Experimental results
show credit scoring models developed here have good forecasting ability.
Probit和Logit模型、概率神经网络模型以及数据包络分析对我国企业
的信用风险度进行信用评分模型的建立。在研究方法上，本文采用理论
研究和实证分析并重的方式。在问题研究上，着重定量分析的运用，从
而得到既有理论依据同时也具有现实可操作性的解决方法。
本文首先对现有信用风险评估方法－专家评分法，信用评分模型，
现代信用风险度量模型进行了综合论述。接下来讨论了各种方法的优缺
点，以及各种方法在中国目前的可行性。然后介绍了信用评分模型建立
的主要方法，多元判别分析和二元Probit、Logit模型，并且引入新的
方法：概率神经网络和数据包络分析技术，对这两种方法在信用风险评
估中应用的可行性进行论述。最后将以上四种方法应用于我国上市公司
贷款信用风险度的评估，采用主成分分析法进行指标选取。
本文综合了财务管理、计量经济学和信息技术相关内容，运用多种
方法建立企业信用风险度和多维财务指标之间的量化关系。通过实证结
果表明，建立的信用评分模型具有较好的预测效果。
The main purpose of this thesis is to find out how to quantify credit
risk which appeared in loan appraisal. Multi-Discriminate analysis、Probit
and Logit models、Probabilistic Neural Networks and Data Envelopment
Analysis are all adopted to develop credit scoring models. To find the
best and practical method, a quantified analysis and experimental study
are emphasized..
First existing methods were introduced, namely specialist’s score
method, credit score model, and modern credit risk models. Then
advantages and disadvantages of each method were discussed and the
adaptability in China is presented. Some popular techniques for
developing credit scoring models such as multiple-discriminate analysis,
probit and logit models were also introduced. In addition, probabilistic
neural networks and data envelopment analysis techniques for credit
scoring are presented. Lastly four techniques were applied to
measurement of credit risk for listed companies in China, in which main
factor method was adopted to select input variables. Experimental results
show credit scoring models developed here have good forecasting ability.
引文
[1]何五星. 银行风险论，北京：中国计划出版社，1995，4～12.
[2]安东尼?桑德森，信用风险度量－风险估值的新方法与其他范式，北京：机械
工业出版社，2001.45～47.
[3]约翰?考埃特，爱德华?爱德曼，《演进着的信用风险管理》，北京：机械工
业出版社，2001.23～24.
[4]Caouette,J.,ALtman,E.,Narayanan,P.,Managing credit risk: The ironic challenge in
the next decade. John Wiley and Sons,Inc.,1998，33～35
[5]Altman E I. Financial ratios,discriminant analysis and the prediction of corporate
bankruptcy. Journal of Finance,1968,(9)：15～19.
[6]王春峰，万海晖，张维，基于神经网络技术商业银行信用风险评估，系统工
程理论与实践，1999，（9）：26～32.
[7]Emad W.Saad, Danil V.Prokhorov and Donald C.Wunsch. Comparative study of
stock trend prediction using time delay, recurrent and Probabilistic Neural Networks.
Neural Networks ,1998，(9):1456-1470.
[8]Moody's Quantitative Risk's Public Firm Risk Model. Available: www.moodysqra.
com.
[9]Kiesel R, Perraudin W. and Taylor A, CreditMetrics, Technical Document, JP
Morgan,1997
[10]Tom Wilde, Credit Suisse Financial Products, “CreditRisk: Credit management
Framework” October 1997, New York/London.
55
参考文献
[11]T.Wilson, Portfolio credit risk, Risk, 1987,(9)10.
[12]Bessis J., Risk management in banking.Chichester: John Wiley and Sons,
1998,76～87.
[13]Basle Committee on Banking Supervision. Credit risk modeling.Currrent
practices and Applications. Bastle: Bastle Committee Publication, 1999.
[14]Treacy WF,Carey M. Credit risk rating at large US banks, The Journal of Banking
and Finance, 2000(24):167-201.
[15]English WB,Nelson WR,Bank risk rating of business loans,Bord of Governors of
the Federal Reserve System. Finance and Economics Discussion Series ,1998:37～45.
[16]Bertels K,Jacques JM, Neuberg L, Gatot L. Qualitative firm performance
evaluation: linear discriminant analysis and neural network models. European Journal
of Operational Research ,1999(115):608~15
[17]何小群,现代统计分析方法与应用,北京:中国人民大学出版社，1999:79～
89.
[18]Mark Zmijewski, Methodological Issues Related to the Estimation of Financial
Distress Prediction Model，Journal of Accounting Research, 1984,67~78.
[19]Altman,E.I.,T.K.N.Baidya,L.M.R.Dias, Accessing Potential Financial Problems
for Firms in Brazil,Working paper 125, NYU Salomon Center, 1977(9):125~130.
[20]Sanvicente,A.S.,and F.L.C.Bader, Filing for Financial Reorganization in Brazil:
Event Prediction with Accounting and Financial variables and the Information content
56
参考文献
of the Filing Announcement.,Working paper, Sao Paulo University, Sao Paulo, Brazil,
1996(7):144~155.
[21]Jensen,Herbert,Using neural networks for credit scoring, Managerial
Finance,1992,18(6):15~26.
[22]P.Coats, L.Fant, Recognizing financial distress patterns using a neural network
tool, Financial Mangement, vol.22,pp.142~155,1993.
[23]E.Altman,G.Marco,F.Varetto,Corporate distress diagnosis:Comparisons using
linear discriminant analysis and neural networks,Journal of Banking and Finance,
1994(18):505~529.
[24]David West.,Neural network credit scoring models,Computers and Operations
Research,2000(27):1131~1152.
[25]Reshmi Malhotra,D.K.Malhotra.Differentisting between good credits and bad
credits using neuro-fuzzy systems,Computing,artifical intelligence and information
technology,2002(136):190~211.
[26]Specht,Donald,Probabilistic Neural Networks,Neural Network,1990(3): 109~118.
[27]Masters,T. Advanced Algorithm for Neural NetworkWiley, New York,1995.
[28]Mark T.Leung,Hazem Daouk,An-Sing Chen.Forecasting stock indices: a
comparision of classification and level estimation models,International Journal of
Forecasting,2000(16):173~190.
[29]Bessis J.Risk management in banking.Chichester:John Wiley and Sons,1998:77～
57
参考文献
128.
[30]Thomas G. Calderon,John J.Cheh.A roadmap for future neural networks research
in auditing and risk assessment.International Journal Accounting Information
Systems,2002(3):203~236.
[31]魏权龄,数据包络分析（DEA),科学通报，2000（9）:24～29.
[32]马占新,唐焕文,宏观经济发展状况综合评价的DEA方法,系统工程, 2003
（2）：56～60.
[33]宗刚，马宁,工业企业竞争力的 DEA 评价研究,数量经济技术经济研究，2002
（11）:79～86.
[34]薛峰,杨德礼,评价银行经营与管理综合效益的DEA模型,数量经济技术经济
研究，1998（5）：124～128.
[35]樊宏,基于DEA模型的我国证券公司评价方法以及应用,数量经济技术经济研
究，2002（4）:44～50
[36]李光金,评价相对效率的投入－产出型DEA,管理科学学报,2001（4）:57～62.
[37]李光金,陈刚,综合 DEA 模型的理论研究,系统工程理论方法应用,2000
（3）:66～72.
[38]安希忠,林秀梅,实用多元统计分析, 长春:吉林科学技术出版社,1992,57～
64.
[39]张新民,企业财务报表分析，2001，29～56.
[40]张宜华,精通 SPSS,北京:清华大学出版社，2001,12～99.
[2]安东尼?桑德森，信用风险度量－风险估值的新方法与其他范式，北京：机械
工业出版社，2001.45～47.
[3]约翰?考埃特，爱德华?爱德曼，《演进着的信用风险管理》，北京：机械工
业出版社，2001.23～24.
[4]Caouette,J.,ALtman,E.,Narayanan,P.,Managing credit risk: The ironic challenge in
the next decade. John Wiley and Sons,Inc.,1998，33～35
[5]Altman E I. Financial ratios,discriminant analysis and the prediction of corporate
bankruptcy. Journal of Finance,1968,(9)：15～19.
[6]王春峰，万海晖，张维，基于神经网络技术商业银行信用风险评估，系统工
程理论与实践，1999，（9）：26～32.
[7]Emad W.Saad, Danil V.Prokhorov and Donald C.Wunsch. Comparative study of
stock trend prediction using time delay, recurrent and Probabilistic Neural Networks.
Neural Networks ,1998，(9):1456-1470.
[8]Moody's Quantitative Risk's Public Firm Risk Model. Available: www.moodysqra.
com.
[9]Kiesel R, Perraudin W. and Taylor A, CreditMetrics, Technical Document, JP
Morgan,1997
[10]Tom Wilde, Credit Suisse Financial Products, “CreditRisk: Credit management
Framework” October 1997, New York/London.
55
参考文献
[11]T.Wilson, Portfolio credit risk, Risk, 1987,(9)10.
[12]Bessis J., Risk management in banking.Chichester: John Wiley and Sons,
1998,76～87.
[13]Basle Committee on Banking Supervision. Credit risk modeling.Currrent
practices and Applications. Bastle: Bastle Committee Publication, 1999.
[14]Treacy WF,Carey M. Credit risk rating at large US banks, The Journal of Banking
and Finance, 2000(24):167-201.
[15]English WB,Nelson WR,Bank risk rating of business loans,Bord of Governors of
the Federal Reserve System. Finance and Economics Discussion Series ,1998:37～45.
[16]Bertels K,Jacques JM, Neuberg L, Gatot L. Qualitative firm performance
evaluation: linear discriminant analysis and neural network models. European Journal
of Operational Research ,1999(115):608~15
[17]何小群,现代统计分析方法与应用,北京:中国人民大学出版社，1999:79～
89.
[18]Mark Zmijewski, Methodological Issues Related to the Estimation of Financial
Distress Prediction Model，Journal of Accounting Research, 1984,67~78.
[19]Altman,E.I.,T.K.N.Baidya,L.M.R.Dias, Accessing Potential Financial Problems
for Firms in Brazil,Working paper 125, NYU Salomon Center, 1977(9):125~130.
[20]Sanvicente,A.S.,and F.L.C.Bader, Filing for Financial Reorganization in Brazil:
Event Prediction with Accounting and Financial variables and the Information content
56
参考文献
of the Filing Announcement.,Working paper, Sao Paulo University, Sao Paulo, Brazil,
1996(7):144~155.
[21]Jensen,Herbert,Using neural networks for credit scoring, Managerial
Finance,1992,18(6):15~26.
[22]P.Coats, L.Fant, Recognizing financial distress patterns using a neural network
tool, Financial Mangement, vol.22,pp.142~155,1993.
[23]E.Altman,G.Marco,F.Varetto,Corporate distress diagnosis:Comparisons using
linear discriminant analysis and neural networks,Journal of Banking and Finance,
1994(18):505~529.
[24]David West.,Neural network credit scoring models,Computers and Operations
Research,2000(27):1131~1152.
[25]Reshmi Malhotra,D.K.Malhotra.Differentisting between good credits and bad
credits using neuro-fuzzy systems,Computing,artifical intelligence and information
technology,2002(136):190~211.
[26]Specht,Donald,Probabilistic Neural Networks,Neural Network,1990(3): 109~118.
[27]Masters,T. Advanced Algorithm for Neural NetworkWiley, New York,1995.
[28]Mark T.Leung,Hazem Daouk,An-Sing Chen.Forecasting stock indices: a
comparision of classification and level estimation models,International Journal of
Forecasting,2000(16):173~190.
[29]Bessis J.Risk management in banking.Chichester:John Wiley and Sons,1998:77～
57
参考文献
128.
[30]Thomas G. Calderon,John J.Cheh.A roadmap for future neural networks research
in auditing and risk assessment.International Journal Accounting Information
Systems,2002(3):203~236.
[31]魏权龄,数据包络分析（DEA),科学通报，2000（9）:24～29.
[32]马占新,唐焕文,宏观经济发展状况综合评价的DEA方法,系统工程, 2003
（2）：56～60.
[33]宗刚，马宁,工业企业竞争力的 DEA 评价研究,数量经济技术经济研究，2002
（11）:79～86.
[34]薛峰,杨德礼,评价银行经营与管理综合效益的DEA模型,数量经济技术经济
研究，1998（5）：124～128.
[35]樊宏,基于DEA模型的我国证券公司评价方法以及应用,数量经济技术经济研
究，2002（4）:44～50
[36]李光金,评价相对效率的投入－产出型DEA,管理科学学报,2001（4）:57～62.
[37]李光金,陈刚,综合 DEA 模型的理论研究,系统工程理论方法应用,2000
（3）:66～72.
[38]安希忠,林秀梅,实用多元统计分析, 长春:吉林科学技术出版社,1992,57～
64.
[39]张新民,企业财务报表分析，2001，29～56.
[40]张宜华,精通 SPSS,北京:清华大学出版社，2001,12～99.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |