Using ARFIMA Model to Calculate and Forecast Realized Volatility of High Frequency Stock Market Index Data
The forecast precision of realized volatility can be affected by both measurement error and market microstructure error when we analyze volatility using high frequency data. This paper adopts the method of second moving average to balance these two errors and establishes ARFIMA model to study the distribution characteristics of realized volatility based on high frequency data of hushen300, its parameters are estimated applying estimation of distribution algorithm. Finally, the superiority of ARFIMA model in volatility forecast is proved by comparing the performances of ARFIMA model and GARCH model.
Keywordshigh frequency data realized volatility optimal sampling frequency ARFIMA model
Unable to display preview. Download preview PDF.
- 3.Oomen Roel, C.A: Modelling realized variance when returns are series correlated (December 23, 2002), http://www.iue.it/Personal/Researchers/oomen/oomen02b
- 5.Huang, H., Chen, L.: The high frequency estimation and characteristic analysis of Chinese stock market volatility. Economic Research Journal (2), 75–94 (2003)Google Scholar
- 10.Jin, X., Yao, J., Zhuang, X.: The study of ARFIMA model and forecasting effect based on fractional order difference. Mathematical Statistics and Management (9), 896–907 (2007)Google Scholar