Applying the zero-adjusted inverse Gaussian model to predict probability of default and exposure at default for a credit card portfolio

Rafael Rodrigues Troiani

Abstract

ABSTRACT

Although the literature about measuring probability of default (PD) in retail credit portfolios is vast, the same thing cannot be said about measuring exposure at default (EAD). This paper aims to contribute toward plugging this gap, presenting the estimation of both PD and EAD of a credit card portfolio using a model based on the zero-adjusted inverse Gaussian distribution, which is a mixed discrete-continuous distribution with a mass probability for the value zero and a continuous right-skewed distribution for positive values. The paper also suggests a method for measuring accuracy in loss-prediction models by adjusting the Gini coefficient.