The role of high-frequency data in volatility forecasting: evidence from the China stock market

This research investigates the role of high-frequency data in volatility forecasting China stock market by particularly feeding different frequency return series directly into a large number GARCH versions. The contributions this are as follows. 1) We provide clear evidence to support that superiority traditional time models remains taking advantage data. 2) incorporate distribution assumptions capture stylized facts result shows that: application substantially influence accuracy forecasting, higher is series, better forecasts provided; non-normal distributions such skewed student-t and generalized error more capable at reproducing both intraday daily than normal distribution; 3) estimated 5-min returns not only outperforms other alternatives, but also considerably beats RV-based HAR ARFIMA forecasting.