Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

---

With the continuous development and changes in the credit industry, credit products play a more important role in the economy. This has led institutions to expand the role of technology in their credit management processes. Credit scoring is a method used to estimate the probability that a loan applicant or existing borrower will default or become delinquent. There are two types of methods used for scoring: Traditional statistics models like Probit and Logistic regression and Data Mining models such as Classification and Regression Trees (CART). In spite of popularity in applying Logit model in credit assessment of applicants, it is attempted to present another method which is theoretically and empirically superior to Logit model. It is also tried to study the capability and accuracy of this method in comparison with Logit model. In this paper, we have examined the performance of different models in credit scoring on real data of a bank and the two approaches above are compared as well. After building a model using Logistic regression; we have built a model using classification and regression trees. Our aim is to emphasize on the specification of CART and testing its capability and comparing its accuracy with the Logit model. The results reveal the accuracy of CART through a bootstrap simulation. Finally it is suggested that classification and regression trees method could be used in credit scoring process instead of Logit model.

Keywords: Credit scoring, Credit risk

Full-Text [PDF 241 kb]   (3487 Downloads)    

Add your comments about this article : Your username or Email:

---