Can Option Trading Volume Forecast Volatility?

Using daily trading volume of Shanghai Stock Exchange 50 ETF option and Classing the option’s delta as the moneyness of option,the paper calculates the volume percentages of out-of-the-money(OTM),at-the-money(ATM),in-the-money(ITM)options and investigates the forecasting ability of each kind of option trading on the realized volatility of underlying ETF.The results show that OTM option trading can significantly promote the predicting power of HAR and HAR-CJ models.Further findings show that,the information of OTM option mainly comes from the deep-out-of-the-money(DOTM)option,while the non-DOTM option has weaker performance in forecasting volatility.Comparing with implied volatility,DOTM option has better performance in in-sample fitting and out-ofsample forecasting volatility.Among the DOTM options,call option has more information than put option,and the call option trading negatively relates to future volatility while the put option trading positively relates to future volatility.These results indicate that informed traders prefer to using DOTM option,and the opposite correlations of DOTM call and put option with future volatility are consistent with leverage effect.This paper uses a series of robustness tests to compare implied volatility,replace volatility model and volatility proxy variables,divide different sample groups and forecast groups,and finds that the trading volume of DOTM is very robust to the prediction and impact of future volatility.Therefore our findings can be applied in improving current volatility forecasting models,and can also be helpful in optimizing governors’ risk monitoring systems.