Eric Girardin: eric.girardin[at]univ-amu.fr
Christelle Lecourt: christelle.lecourt[at]univ-amu.fr
Jean-François Carpantier: jean-francois.carpantier[at]univ-amu.fr
We perform a large–scale empirical study to compare the forecasting performance of single–regime and Markov–switching GARCH (MSGARCH) models from a risk management perspective. We find that MSGARCH models yield more accurate Value–at–Risk and left–tail distribution forecasts than their single–regime counterpart. Also, our results indicate that accounting for parameter uncertainty improves left–tail predictions, independently of the inclusion of the Markov–switching mechanism.