Which Curve Fits Best: Fitting ROC Curve Models to Empirical Credit-Scoring Data

In the practice of credit-risk management, models for receiver operating characteristic (ROC) curves are helpful in describing shape an ROC curve, estimating discriminatory power a scorecard, and generating without underlying data. The primary purpose this study is to review curve proposed literature, primarily biostatistics, fit them actual credit-scoring data order determine which could be used credit-risk-management practice. We list several theoretical describe context. model includes binormal, bigamma, bibeta, bilogistic, power, bifractal curves. then tested against empirical from publicly available presentations papers, as well European retail lending institutions. Except all presented quite well. However, based on results other favourable properties, it suggested that binormal preferred choice modelling