Reject Inference in Credit Score Model Based on Semi-parametric Method

›› 2018, Vol. 30 ›› Issue (10): 40-48.

Previous Articles Next Articles

Reject Inference in Credit Score Model Based on Semi-parametric Method

Xia Liyu^1,2, He Xiaoqun²

1. State Grid Energy Research Institute, Beijing 102209;
2. Center for Applied Statistics, Renmin University of China, Beijing 100872

Received:2016-06-17 Online:2018-10-28 Published:2018-10-23

Abstract

Abstract:

Reject inference can be considered a specific case of non-ignorable missing data analysis. It can deal with the biased estimation of model parameters caused by sample bias in credit score model resulting from absence of the credit quality of rejected applicants. Based on the semi-parametric regression model of mean function with non-ignorable missing response proposed in Kim and Yu (2011), we propose an iterative semi-parametric method to infer credit-quality data with non-ignorable missing mechanism. Simulation studies in 5 missing scenarios are implemented and empirical studies with Australian dataset and bank of C dataset are conducted. The results demonstrate that our method is a relatively conservative approach that can effectively identify the "bad" applicants and reduces credit risk faced by financial institutions.

Key words: credit score model, reject inference, semi-parametric estimation, non-ignorable missing

Xia Liyu, He Xiaoqun. Reject Inference in Credit Score Model Based on Semi-parametric Method[J]. , 2018, 30(10): 40-48.

References

[1] Rubin D. B. Inference and Missing Data[J]. Biometrika, 1976,63(3):581-592
[2] Boyes W., Hoffman D., Low S. An Econometric Analysis of the Bank Credit Scoring Problem[J]. Journal of Econometric, 1989,40(1):3-14
[3] Greene W. Sample Selection in Credit-scoring Models[J]. Japan and the World Economy, 1998,10(3):299-316
[4] Jacobson T., Roszbach K. F. Bank Lending Policy, Credit Scoring and Value at Risk[J]. Journal of Banking Finance, 2003,27(4):615-633
[5] Crook J., Banasik J. Does Reject Inference Really Improve the Performance of Application Scoring Models[J]. Journal of Banking and Finance, 2004,28(4):857-874
[6] Banasik J., Crook J. Reject Inference, Augmentation and Sample Selection[J]. European Journal of Operational Research, 2007,183(3):1582-1594
[7] Banasik J., Crook J. Reject Inference in Survival Analysis by Augmentation[J]. Journal of the Operational Research Society, 2010,61(3):473-485
[8] Heckman J. J. Sample Selection Bias as a Specification Error[J]. Econometrica, 1979,47(1):153-161
[9] 张景肖,魏秋萍,姜玉霞,等. 基于两阶段思想处理拒绝推断的信用评分模型[J]. 数理统计与管理, 2012,31(6):1049-1060
[10] Sohn S. Y., Shin H. W. Reject Inference in Credit Operations Based on Survival Analysis[J]. Expert Systems with Applications, 2006,31(1):26-29
[11] 魏秋萍,张景肖,张波. 基于核函数法进行拒绝推断的信用评分模型[J]. 统计与决策, 2012,12(12):4-8
[12] Chen G. G., Astebro T. Bound and Collapse Bayesian Reject Inference for Credit Scoring[J]. Journal of the Operational Research Society, 2012,63(10):1374-1387
[13] Kim J. K., Yu L. C. A Semi-parametric Estimation of Mean Functionals with Non-ignorable Missing Data[J]. Journal of the American Statistical Association, 2011,106(493):157-165
[14] Zhao P. Y., Tang M. L., Tang N. S. Robust Estimation of Distribution Functions and Quantiles with Non-ignorable Missing Data[J]. The Canadian Journal of Statistics, 2013,41(4):575-595

Reject Inference in Credit Score Model Based on Semi-parametric Method

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 1

Recommended Articles

Metrics

Comments