Revisiting the accuracy of standard VaR methods for risk assessment: Using the Copula–EVT multidimensional approach for stock markets in the MENA region

The aim of this study is twofold. First, it aims to show how to overcome some of the shortcomings of the standard risk measurement methods using value-at-risk (VaR) in the context of extreme events and propose an alternative empirical method. Second, it aims to fill the research gap in analysis of risk in the Middle East and North Africa (MENA) region. In this regard, for six daily stock indices in the MENA region, from January 3, 2005 to December 31, 2014, we employ a vine copula-based generalized autoregressive conditional heteroskedastic (GARCH) method and the extreme value theory (EVT) to model the dependence between the marginal distributions of returns and forecast the VaR. Based on backtesting, we assess the efficiency of the standard risk measurement models from the following families—the exponentially weighted moving average, the historical simulation, and the GARCH. By implementing the GARCH-EVT-C-vine method in the MENA region, we find that, empirically, standard methods overestimate (underestimate) the violation ratio, implying an underestimation (overestimation) of the risk, and therefore a misallocation of the capital covering the risk. The method also provides better VaR estimates for the MENA stock markets than that of the standard methods. We also verify that the greater the openness of the capital accounts and the flexibility of the exchange rate regimes, the greater will be the conditional dependence of the MENA countries on the developed markets. Finally, the empirical method we propose has an important implication for the MENA countries in that it can be adopted by these countries to forecast VaR effectively.