Business intelligence for delinquency risk management via Cox regression

Abstract

The recent economic downturn has made many delinquent credit card customers, which in sequence reduced profit margins as well as sales of the retail companies. This study focuses on customers who have recovered from credit delinquency to analyze repayment patterns of delinquents. A Cox regression model is designed as a credit-predicting model to handle with credit card debtors. The model predicts the expected time for credit recovery from delinquency. The model's prediction accuracy is compared with that of other known algorithms. © 2010 Springer-Verlag Berlin Heidelberg.