Research On Option Volatility Based On Realized Volatility Model

With the rapid development of China’s financial market,the volatility of the stock market has been widely concerned by investors.It reflects the operation of the stock market.At the same time,the volatility is also closely related to asset pricing,risk management and hedging.Therefore,it is of great significance to build a reasonable model to predict the volatility of the stock market more accurately.On the one hand,scholars have begun to study behavioral finance,and the research combining stock volatility and investor sentiment has become a hot topic.And because investors in China’s stock market tend to invest in the short term,sentiment has a greater impact,which leads to the deviation between the option price of financial assets or securities portfolio and the actual price,so this paper considers adding investor sentiment as an explanatory variable to the volatility prediction model;On the other hand,most of the volatility prediction models built by previous scholars are based on historical volatility data for empirical research.A large number of documents show that the implied volatility based on the option market can provide more information for predicting the volatility of the stock market.Therefore,this paper considers adding the China ETF volatility index as the representative of the implied volatility to the volatility prediction model as an explanatory variable,Compare the improvement effect of investor sentiment and China ETF volatility index on volatility prediction,and find a more accurate model for volatility prediction.Based on the above research background,this paper selects six indicators,namely,the net capital inflow(NIF)of the Shanghai 50 Index constituent stocks,the transaction amount(VT)of the Shanghai 50 Index constituent stocks,the turnover rate(TR),the margin trading balance(SMT),the closing price(C)of the Shanghai 50 Index,and the average price-earnings ratio(PE)of the Shanghai 50 Index constituent stocks,as the proxy variables of the comprehensive index of investor sentiment,The investor sentiment comprehensive index(SENT)is constructed by principal component analysis.For the volatility research,this paper selects the 5-minute high-frequency data of Shanghai Stock Exchange 50 ETF option as the research object.From the perspective of long-term memory,this paper constructs the HAR-RV model and the expansion model,and adds investor sentiment and China ETF Volatility Index(VXFXI)to the HAR-RV model to build the expansion model.At the same time,considering the jump phenomenon of the realized volatility series,this paper also adds the jump term as an explanatory variable to the HAR-RV model.Carry out out-of-sample and intra-sample prediction for the above constructed model,compare the improvement effect of investor sentiment and China ETF volatility index on volatility prediction with the loss function method,and explore the realized volatility model suitable for China’s national conditions.From the perspective of volatility agglomeration,this paper constructs the Realized-GARCH model and the expansion model.In order to further study the improvement effect of investor sentiment and China ETF volatility index on volatility prediction,the two variables are added to the mean equation and the variance equation of Realized-GARCH respectively to study the improvement effect of investor sentiment and China ETF volatility index on volatility prediction.Considering that the realized volatility data does not have the normal distribution characteristics and has the phenomenon of peak and thick tail,this paper extends the Realized-GARCH model and the extended model,extends the residual from the normal distribution to the standard t distribution,the partial t distribution and the generalized error GED distribution,and conducts a more detailed empirical study of volatility.At the same time,Va R is used for risk estimation.On the one hand,it compares the contribution of investor sentiment and China ETF volatility index to volatility prediction,and on the other hand,it describes the characteristics of stock market earnings volatility.Through empirical research,it is found that both investor sentiment and China ETF volatility index can improve the prediction performance of the volatility model from the perspective of long-term memory and volatility aggregation,but investor sentiment has a greater impact on the prediction performance of the model,and the volatility prediction performance will be further improved after adding both to the HAR family model at the same time,It is proved that the volatility prediction model built in this paper has good practicability in reality.