Robust Portfolio Selection with Regime Switching R-Vine Copulas

This paper develops a robust portfolio optimization model based on regime switching R-Vine copulas, where regime switching R-Vine copulas capture asymmetric dependence and regime switching in financial markets. We consider the uncertainty in hidden economic states and define WSCVaR as CVaR in the worst mixture regime to capture extreme risk in times of crisis. We compare our model with the models without regime switching or robustness in an empirical study of thirteen market indices. The results show that our model can select assets with smaller lower tail dependence and outperforms other models in dynamic investments. In particular during periods of crisis, our model performs most robustly in terms of the highest return and smallest risk. This also shows that regime switching R-Vine copulas can capture extreme tail dependence in times of crisis and WSCVaR with respect to regime switchig copulas can assess extreme risk in times of crisis adequately