Evaluation of Performance Measures for Probability of Default and Related Credit Risk Assessment Models
Abstract
The performance of Probability of Default (PD) models and other credit classification methods such as ratings is often measured by assessing how well the method differentiates entities that eventually default within a given time frame from those that do not default. A standard approach is to apply the theory of Receiver Operating Characteristic (ROC) curves taken from signal detection theory. This talk considers the effectiveness of ROC-like performance measures with regard to characteristics of the data sample including size, distribution and correlation. The speaker presents a new approach to the problem derived from the distributional properties of the data sample.
About the speaker
Dr. William Morokoff received his PhD in Mathematics from New York University, where he specialized in Monte Carlo methods and numerical analysis. He had been senior member of the credit research group at Moody’s KMV and vice president of quantitative market risk management at Goldman Sachs. He is currently Managing Director at Standard & Poor’s and heads the Quantitative Analytics and Research Group.
Dr. Morokoff is responsible for leading the development and application of quantitative methodologies for all of Standard & Poor’s Ratings Services. In partnership with Structured Finance Ratings, Corporate Ratings, Global Risk Management and Economics and Research, his team is also responsible for research support of the quantitative models and criteria used in Standard & Poor’s credit assessment products and services. He has worked extensively in credit and market risk modeling, with a research focus on numerical analysis for portfolio risk management problems.
