The Analytics of Risk Model Validation
A collection of papers in this new and developing area of research, this book focuses on the quantitative side of model validation and covers model validation in credit, market, and operational risk.
TABLE OF CONTENTS The Analytics of Risk Model Validation
Chapter 3: The Validity of Credit Risk Model Validation Methods
George Christodoulakis [ ] and Stephen Satchell [ ]
Abstract
In this chapter, we discuss the nature, properties and pitfalls of a number of credit risk model validation methods. We focus on metrics of discriminatory power between sick and healthy loans, their association and their properties as random variables, which may lead to pitfalls in model validation processes. We conclude with a discussion of bootstrap and credit-rating combinations.
[*] The views expressed in this paper are those of the authors and should in no part be attributed to the Bank of Greece.
[ ] Manchester Business School and Bank of Greece, Manchester, UK
[ ] Trinity College and Faculty of Economics, University of Cambridge, Cambridge, UK
1. Introduction
The development of various types of credit risk models has its origins in the pricing of assets and has been further strengthened by the Basel Capital Accord, which allows for the determination of capital adequacy of credit institutions using internal rating models. This process has led the financial institutions as well as their supervisors to develop not only internal models but also validation methods to assess the quality and adequacy of those models. The need for credible assessment stems from the fact that using low-quality models could lead to sub-optimal allocation of capital as well as ineffective management of risks. The assessment of credit risk model adequacy is usually based on the use of statistical metrics of discriminatory power between risk classes, often referred as model validation, as well as on the...
Copyright Elsevier Ltd. 2008 under license agreement with Books24x7
