Statistical models for the Basel II internal ratings-based approach to measuring credit risk of retail products

The Basel II Accord is a financial risk management standard recently adopted by many financial institutions and regulators around the world.The general spirit of the accord is to develop a systematic approach to evaluating and controlling risks based on timely data and their analysis and interpretation.The interface between statistical modeling and the financial application is of pivotal importance in the development of the internal ratings-based (IRB) approach recommended by the Basel II Accord.This article reviews the IRB requirements and develops new empirical Bayes models for modeling probability of default and loss given default, which are the key ingredients in the IRB approach to credit risk analysis of retail exposures.