Money, finances and credit

A practical approach to validation of credit scoring models

Economic Annals-XXI

Economic Annals-ХХI: Volume 141, Issue 5-6, Pages: 64-67

Citation information:
Lukashevich, N. (2014). A practical approach to validation of credit scoring models. Economic Annals-XXI, 5-6, 64-67. https://ea21journal.world/index.php/ea-v141-16/

View Full TextDownload PDF

Nikita Lukashevich
PhD (Economics),
Associate Professor,
Business and Commerce Department,
St. Petersburg State Polytechnical University
29 Polytechnicheskaya Str., St. Petersburg, 195251, Russia
lukashevich@kafedrapik.ru

A practical approach to validation of credit scoring models

Abstract. Introduction. The implementation of the Third Basel Accord raises many technical and methodological issues regarding the development and validation of credit risk models and makes these issues much more important. Bank regulators will pay more and more attention to testing model validation processes in order to examine the predictive accuracy of banks’ credit scoring models. Lenders therefore need to develop and apply the approaches for operational monitoring of predictive accuracy and modifying the cut-off value as one of the most important parameters of credit scoring models. The author poses the receiver operating characteristic (ROC) curve technique that can be successfully used to validate credit scoring models. The purpose of the research is testing the application of ROC curve technique for estimating the validity and predictive accuracy of credit scoring models. Results. The main parameters of credit scoring models validation were summarized. The possible criteria for determining the acceptable cut-off value for credit scoring models are presented. The approbation of the ROC curve technique is given by comparing two logistic models developed on the factual statistical data. Conclusions. The ROC curve technique can be applied successfully to estimate validity and compare credit risk models. The research gives recommendations of how to apply the proposed technique in the validation process. The area for the further research can be the consideration of the ROC curve technique in terms of the economic benefits and losses from the true and false classified credit applications.

Keywords: Credit Risk; Credit Scoring; Logistic Regression; Validity; ROC Curve

JEL Classification: C52; E42; E58; G21

References

  1. Hand, D. J., & Henley, W. E. (1997). Statistical classification methods in consumer credit scoring: a review. Journal of the Royal Statistical Society: Series A (Statistics in Society), 160(3), 523-541.
  2. Abdou, H. A., & Pointon, J. (2011). Credit scoring, statistical techniques and evaluation criteria: a review of the literature. Intelligent Systems in Accounting, Finance and Management, 18(2), 59-88.
  3. Thomas, L. C. (2000). A survey of credit and behavioural scoring: forecasting financial risk of lending to consumers. International Journal of Forecasting, 16(2), 149-172.
  4. Yingxu, Y. (2007). Adaptive credit scoring with kernel learning methods. European Journal of Operational Research, 183(3), 1521-1536.
  5. Yap, B. W., Ong, S. H., & Husain, N. H. M. (2011). Using data mining to improve assessment of credit worthiness via credit scoring models. Expert Systems with Applications, 38(10), 13274-13283.
  6. Tsai, C.-F., & Wu, J.-W. (2008). Using neural network ensembles for bankruptcy prediction and credit scoring. Expert Systems with Applications, 34(4), 2639-2649.
  7. Hand, D. J., & Till, R. J. (2001). A simple generalization of the area under the ROC curve to multiple class classification problems. Machine Learning, 45(2), 171-186.
  8. Ben-David, A., & Frank, E. (2009). Accuracy of machine learning models versus «hand crafted» expert systems – A credit scoring case study. Expert Systems with Applications, 36(3, Part 1), 5264-5271.
  9. Seidenfeld, T. (1985). Calibration, coherence, and scoring rules. Philosophy of Science, 52, 274-294.
  10. Egan, J. P. (1975). Signal detection theory and ROC analysis. New York: Acad. Press.
  11. Swets, J. A. (1988). Measuring the accuracy of diagnostic systems. Science, 240, 1285-1292.
  12. Drummond, C., & Holte, R. (2004). What ROC curves can’t do (and cost curves can). Proceedings of ROC Analysis in Artificial Intelligence (pp. 19-26). Valencia, Spain.
  13. Stein, R. M. (2005). The relationship between default prediction and lending profits: Integrating roc analysis and loan pricing. Journal of Banking and Finance, 29, 1213-1236.
  14. Fawcett, T. (2004). ROC Graphs: Notes and Practical Considerations for Researchers. Kluwer Academic Publishers, Netherlands.
    Retrieved from http://home.comcast.net/tom.fawcett/public_html/papers/ROC101.pdf
  15. Youden, W. J. (1950). Index for rating diagnostic tests. Cancer, 3, 32-35.

Received 24.03.2014

View Full TextDownload PDF
In, , , ,

More Papers