Study on CVaR forecasts based on weighted realized volatility

Value-at-Risk (VaR), as a a risk measure, has been widely accepted all over the world. However, VaR is not the best risk measure. VaR is not sub-additive. Moreover, it doesnpsilat indicate the size of the potential loss. Conditional Value-at-Risk (CVaR) is the most attractive coherent risk measure and has been studied by many authors. In this paper, we study on CVaR calculations. In addition, we study the issue of volatility forecasting for CVaR calculations by using weighted realized volatility. Weighted realized volatility is a non- parametric measure of volatility and can be modeled and forecasted with usual time series models. Furthermore, weighted realized volatility is based on high frequency financial data and can fully take advantage of the intraday information. Finally, we do empirical research in Chinese stock market.