The impact of sample bias on consumer credit scoring performance and profitability

This article seeks to gain insight into the influence of sample bias in a consumer credit scoring model. Considering the vital implications on revenues and costs concerned with the issuing and repayment of commercial credit, predictive performance of the model is crucial, and sample bias has been suggested to pose a sizeable threat to profitability due to its implications on either population drainage or biased estimates. Whereas in previous research, different techniques of reducing sample bias have been proposed and deployed, the debate around the impact of sample bias itself has predominantly been held on a theoretical level. The dataset that was used in this study, however, provides the opportunity to investigate the issue in an empirical setting. Based on the data of a mail-order company offering short term consumer credit to their consumers, we show that (i) given a certain sample size, sample bias has a significant effect on consumer credit-scoring performance and profitability, (ii) its effect is composed of the inclusion of rejected orders in the scoring model, and the inclusion of these orders into the variable-selection process, (iii) the impact of the effect of sample bias on consumer credit scoring performance and profitability is limited and (iv) in consumer credit scoring, by merely increasing the sample size of the biased sample, the impact of sample bias can likely be reduced. Hence, we conclude that the possible impact of any reduction of sample bias is modest in a consumer credit scoring model, and that attention might optimally be focused on other factors leading to improved consumer credit scoring.