Prognostic personal credit risk model considering censored information

Credit scoring is one of the most successful applications of quantitative analysis that helps organizations decide whether or not to grant credit to consumers who apply to them. However, standard credit risk models based on binary classifying approaches appear to have missed several important time-varying factors and censoring information. This paper looks at the extensions of the survival analysis model to analyze personal credit risk. Survival analysis has mainly been used in the clinical domain, which can handle the above issues. This paper investigates the ability of the survival-based approach to predict the probability of personal default. The proposed method can give a prediction of 'time' as well as 'probability' of personal default. We develop a survival-based credit risk model and assess the relative importance of different variables in predicting default. Standard binary classifying models are also developed for assessing a new way in the context of classifying power. These models are applied to personal credit card accounts dataset. According to the experiment results, survival-based credit risk modeling is a more useful approach for classifying risky customers than others. The survival-based approach is a useful alternative and a complement in view of personal credit risk.

[1]  J. Kalbfleisch,et al.  The Statistical Analysis of Failure Time Data , 1980 .

[2]  Bart Baesens,et al.  Bayesian neural network learning for repeat purchase modelling in direct marketing , 2002, Eur. J. Oper. Res..

[3]  Kevin Keasey,et al.  Multilogit approach to predicting corporate failure—Further analysis and the issue of signal consistency , 1990 .

[4]  Eric Rosenberg,et al.  Quantitative Methods in Credit Management: A Survey , 1994, Oper. Res..

[5]  Ian A. Cooper,et al.  Default risk and derivative products , 1996 .

[6]  John Banasik,et al.  Not if but when will borrowers default , 1999, J. Oper. Res. Soc..

[7]  David W. Hosmer,et al.  Applied Logistic Regression , 1991 .

[8]  Laurence L. George,et al.  The Statistical Analysis of Failure Time Data , 2003, Technometrics.

[9]  W. R. Lane,et al.  An application of the cox proportional hazards model to bank failure , 1986 .

[10]  C. de Villiers,et al.  The nature of end-user relationships in the development of electronic commerce applications , 2000, SIGCPR '00.

[11]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[12]  R. Gill,et al.  Cox's regression model for counting processes: a large sample study : (preprint) , 1982 .

[13]  D. Cox,et al.  Analysis of Survival Data. , 1986 .

[14]  Maria Stepanova,et al.  Survival Analysis Methods for Personal Loan Data , 2002, Oper. Res..

[15]  Jonathan N. Crook,et al.  Credit Scoring and Its Applications , 2002, SIAM monographs on mathematical modeling and computation.

[16]  Johan A. K. Suykens,et al.  Benchmarking state-of-the-art classification algorithms for credit scoring , 2003, J. Oper. Res. Soc..

[17]  V. Srinivasan,et al.  Credit Granting: A Comparative Analysis of Classification Procedures , 1987 .

[18]  Robert A. Jarrow,et al.  The intersection of market and credit risk , 2000 .

[19]  D.,et al.  Regression Models and Life-Tables , 2022 .

[20]  Erkki K. Laitinen,et al.  Survival analysis as a tool for company failure prediction , 1991 .

[21]  Anlong Li.  Three Essays on Contingent Claims Pricing , 1992 .

[22]  Bart BaesensRudy.  Using Neural Network Rule Extraction and Decision Tables for Credit-Risk Evaluation , 2003 .