Improving volatility forecasting based on Chinese volatility index information: Evidence from CSI 300 index and futures markets

This paper investigates whether iVX, the newly launched implied volatility index in China contains incremental information about volatility forecasting. We use high frequency data of Chinese CSI 300 stock index and futures to calculate realized volatility, then estimate various constant coefficients and time-varying coefficients HAR models (TVC-HAR), and finally adopt one-step and smooth multi-step rolling forecasting methods to evaluate the forecasting errors. Our analysis confirms that iVX does have significant influence to the realized volatility forecasting. Both the in-sample and out-of-sample forecasting errors indicate that iVX plays a crucial role to volatility forecasting, combining both continuous volatility, jump volatility and iVX information leads to best performance. TVC-HAR models outperform HAR models for multi-step ahead forecasting while with iVX as regressor perform best for one-step ahead forecasting. TVC-HAR models with iVX as driven factor is more suitable for index while models with time as the driven factor perform better for futures. MCS test further confirms the superiority of the selected models in volatility forecasting. Our study is important for financial market risk management and the healthy development of derivatives market in China.