Empirical Performance of Regime Switching Model to VaR

Value at Risk (VaR) has become the standard measure of market risk employed by financial institutions for both internal and regulatory purposes. In asset allocation and portfolio risk management, we need a good model that can capture the joint dynamics of the assets under consideration. Financial asset returns often exhibit time-varying conditional distributions, departures from normality, and persistence in volatility. Higher moments of financial returns distribution are especially important in applications in which the tail probability is of major interest or concern. Therefore, a good data-generating process for financial returns should allow for both temporal dependence and stochastic variation in the moments. We examine whether a regime switching (RS) model is viable for measuring risk. Specifically, we evaluate the performance of VaR under RS and some well-known alternatives