The purpose of this paper is to investigate whether a dynamic Value at Risk model and high frequency realized volatility models can improve the accuracy of 1-day ahead VaR forecasting beyond the performance of frequently used models. As such, this paper constructs 60 conditional volatility...