Forecasting VaR with Real-Time Realized EGARCH Model

Abstract

In this paper, we propose a Real-Time Realized EGARCH model for forecasting the Value-at-Risk (VaR) in the Chinese stock market, utilizing a skewed t-distribution. This model introduces a novel approach by incorporateing high frequency intraday information and current return information into daily VaR forecast. We conduct an empirical analysis using two main indexes of Chinese stock market (Shanghai Stock Exchange Composite Index and Shenzhen Stock Exchange Component Index), and backtesting approach as well as the model robustness test prove that the VaR forecasts of Real-Time Realized EGARCH model generally gain an advantage over the GARCH model, the Real-Time GARCH model and the Realized EGARCH model. Our empirical findings highlight the value of incorporating both the high-frequency intraday information and current return information for forecasting the VaR of Chinese stock market.