Empirical Study On Volatility Of Nonferrous Metal Futures Based On HAR Model

With the vigorous development of Chinese capital market and the constant popularity of high-frequency data,the volatility of financial markets is also increasing.At the same time,effective management and accurate forecasting of financial market risks have become particularly important.However,volatility has become an important indicator for accurately measuring and forecasting financial markets,and it is also an important indicator for investors to accurately predict investment risks.The realized volatility model uses high-frequency data to model the volatility,so that the amount of information used is relatively sufficient and there is no need to consider parameters and other advantages.Therefore,the modeling of the realized volatility used high frequency data has been promoted.Unlike the traditional volatility model,the modeling of realized volatility uses high-frequency data.It compensates for the inadequacy of some information by the traditional modeling method and is simpler than the traditional model in calculating the model and estimating the parameters.Therefore,it has been widely used.Based on the HAR model,this paper models and predicts the volatility of Chinese non-ferrous metals futures market,and it uses high-frequency data to provide new insight into the time-varying volatility of realized volatility and leverage effects.This paper also extends the HAR model and conducts the empirical research on the high-frequency data of copper and aluminum futures of the Shanghai Futures Exchange.The result shows the long-term memory LHAR-RV-CJ and HAR-RV-NT models captures that the dynamic correlation of realized fluctuations is most consistent with time-varying volatility.