Is the filtered historical simulation method adequate to forecast the expected shortfall ? An assessment based on the risk map

Abstract

The Global Financial Crisis prompted the Basel Committee on Banking Supervision to call for new measures to address risks that had not yet been handled. One of these requirements is the replacement of the Value at Risk by the Expected Shortfall, which will help financial institutions to capture tail risks and capital adequacy in periods of severe market stress. This change has the effect of positioning the monitoring of market risk no longer on a certain quantile of the Profit & Loss distribution but on the anticipation of losses beyond the Value at Risk. In addition to this change, some theoretical issues have been identified, such as the unavailability of simple tools to backtest the Expected Shortfall forecasts. In this thesis, the main objective will be to verify whether the Filtered Historical Simulation approach (Barone-Adesi et al., 2002) can be used to predict the Expected Shortfall. First, a GARCH model will be used to estimate the Value at Risk. Based on this estimated Value at Risk, the Expected shortfall will then be forecasted. Finally, the Risk Map tool (Colletaz et al., 2013) will determine whether or not to validate the use of this model.