Volatility forecasting using intraday information with the CARR models for the China stock markets

This study mainly focuses on utilizing the range-based CARR and CARRX models to explore the non-trading effects during the lunch break and overnight periods in volatility changes forecasting for the two stock markets in China. We find that the price difference in absolute values between the close price in day t-1 and the open price in day t are able to predict the volatility change in the morning trading session significantly. The price range during the lunch break and the trading volume change percentage in the morning session appear positive correlation with the volatility in the afternoon trading session.