Evaluation de Techniques de Traitement des Refusés pour l'Octroi de Crédit

Abstract

We present the problem of "Reject Inference" for credit acceptance. Because of the current legal framework (Basel II), credit institutions need to industrialize their processes for credit acceptance, including Reject Inference. We present here a methodology to compare various techniques of Reject Inference and show that it is necessary, in the absence of real theoretical results, to be able to produce and compare models adapted to available data (selection of "best" model conditionnaly on data). We describe some simulations run on a small data set to illustrate the approach and some strategies for choosing the control group, which is the only valid approach to Reject Inference.