Credit scoring and decision making in Egyptian public sector banks
Author:
Abdou Hussein A.,Pointon John
Abstract
PurposeThe main aims of this paper are: first, to investigate how decisions are currently made within the Egyptian public sector environment; and, second, to determine whether the decision making can be significantly improved through the use of credit scoring models. A subsidiary aim is to analyze the impact of different proportions of sub‐samples of accepted credit applicants on both efficient decision making and the optimal choice of credit scoring techniques.Design/methodology/approachFollowing an investigative phase to identify relevant variables in the sector, the research proceeds to an evaluative phase, in which an analysis is undertaken of real data sets (comprising 1,262 applicants), provided by the commercial public sector banks in Egypt. Two types of neural nets are used, and correspondingly two types of conventional techniques are applied. The use of two evaluative measures/criteria: average correct classification (ACC) rate and estimated misclassification cost (EMC) under different misclassification cost (MC) ratios are investigated.FindingsThe currently used approach is based on personal judgement. Statistical scoring techniques are shown to provide more efficient classification results than the currently used judgemental techniques. Furthermore, neural net models give better ACC rates, but the optimal choice of techniques depends on the MC ratio. The probabilistic neural net (PNN) is preferred for a lower cost ratio, whilst the multiple discriminant analysis (MDA) is the preferred choice for a higher ratio. Thus, there is a role for MDA as well as neural nets. There is evidence of statistically significant differences between advanced scoring models and conventional models.Research limitations/implicationsFuture research could investigate the use of further evaluative measures, such as the area under the ROC curve and GINI coefficient techniques and more statistical techniques, such as genetic and fuzzy programming. The plan is to enlarge the data set.Practical implicationsThere is a huge financial benefit from applying these scoring models to Egyptian public sector banks, for at present only judgemental techniques are being applied in credit evaluation processes. Hence, these techniques can be introduced to support the bank credit decision makers.Originality/valueThie paper reveals a set of key variables culturally relevant to the Egyptian environment, and provides an evaluation of personal loans in the Egyptian public sector banking environment, in which (to the best of the author's knowledge) no other authors have studied the use of sophisticated statistical credit scoring techniques.
Subject
Finance,Business, Management and Accounting (miscellaneous)
Reference37 articles.
1. Altman, E.I. (1968), “Financial ratios, discriminant analysis and the prediction of corporate bankruptcy”, Journal of Finance, Vol. XXIII No. 4, pp. 589‐609. 2. Altunbas, Y., Goddard, J. and Molyneux, P. (1999), “Technical change in banking”, Economics Letters, Vol. 64 No. 2, pp. 215‐21. 3. Baesens, B., Gestel, T.V., Stepanova, M., Van den Poel, D. and Vanthienen, J. (2005), “Neural network survival analysis for personal loan data”, Journal of the Operational Research Society, Vol. 56 No. 9, pp. 1089‐98. 4. Baesens, B., Gestel, T.V., Viaene, S., Stepanova, M., Suykens, J. and Vanthienen, J. (2003), “Benchmarking state‐of‐the‐art classification algorithms for credit scoring”, Journal of the Operational Research Society, Vol. 54 No. 6, pp. 627‐35. 5. Banasik, J., Crook, J. and Thomas, L. (2001), “Scoring by usage”, Journal of the Operational Research Society, Vol. 52 No. 9, pp. 997‐1006.
Cited by
18 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
|
|