Efficient calculation of expected shortfall contributions in large credit portfolios

Michael Kalkbrener, Anna Kennedy, Monika Popp

Abstract

ABSTRACT

In the framework of a standard structural credit portfolio model, we investigate the Monte-Carlo-based estimation of capital allocation according to expected shortfall. We develop and analyze several variance reduction techniques based on importance sampling, analytic approximations of portfolio loss distributions and a semi-analytical allocation technique. The focus of the paper is on the application of these techniques to large credit portfolios used in economic capital calculations. Our results show that the inherent numerical problems of expected shortfall allocation can be overcome and, as a consequence, economic capital allocation according to expected shortfall is a viable option for financial institutions.