The Calibration of Market Risk Measures During Period of Economic Downturn: Market Risks and Measures

Abstract

This study explores the calibration of market risk measures during period of economic downturn. This calibration is done in two frameworks: firstly individual profit and loss distribution is modelled using two different types of extreme value distribution namely: the generalized extreme value (GEV) distribution, and the generalized Pareto distribution (GPD). The resulting shape parameters are all positive indicating that these distributions can in fact capture the negative skewness and excess kurtosis of the profit and loss (PL we consider one elliptical copula (student t copula) and one Archimedean copula (Gumbel copula). Using two stock market indices we compute what we refer to as EVT based mark risk measures and the copula based market risk measures for both the left and right tails of the P&L distribution. Our results suggest that copula based risk measures are more reliable in predicting the behavior of market risks during period of economic downturn.