Prediction Research On Implied Volatility Surface Of SSE 50 ETF Option

With the growth of the economy,various financial instruments have also flourished.Among them,the representative option products have emerged in the financial market and have begun to play their role of risk management and arbitrage speculation.Options can be divided into stock options,index options,etc.The underlying assets can be stocks,securities indices,futures,and bonds.In China,the options market appeared relatively late,and the SSE 50ETF options were officially listed on February 9,2015.After more than 4 years of development,more and more investors are aware of the advantages of option products,and have started to use SSE 50ETF options in asset management.Hedge institutions have also begun to use them for risk diversification and asset management.As SSE 50ETF options trading becomes more and more active,in order to promote the healthy development of China's capital market,this article chooses to study the SSE 50ETF options market.In the financial market,if the prices of financial products can be effectively predicted,investors and related institutions can establish trading strategies based on them,perform accurate asset management,and market makers can also perform risk management.However,because options are complex derivatives,it is very difficult to directly price and predict them.Therefore,the option price can be studied indirectly by studying the implied volatility(IV)corresponding to the option price.If the implied volatility can be predicted more accurately,and then the option pricing model is used,the options can be correctly priced and effective risk management can be performed.At the same time,because the fluctuation of the option market price is related to the trading behavior of the option market participants,the transaction volume information will help predict the implied volatility,and the prediction error can be further reduced by the transaction volume.In order to formulate correct and effective asset management and risk management strategies,it is necessary to anticipate the price changes of options,which also requires more accurate predictions of implied volatility.This article uses the two-step analysis framework to model the implied volatility surface of the SSE 50ETF options.In the first step,this article uses the implied volatility calculated by the option pricing model and fits a direct deterministic IVS model daily.In this deterministic IVS model,the dependent variable is the implied volatility,and the independent variable is the adjusted degree and remaining period after the period adjustment.In this way,different daily implied volatility surfaces are obtained during the sample period.In the second step,a time series model is established for the coefficients estimated in the first step to fit the dynamic changes of the coefficients.In this way,through the above two steps,the dynamic process of the implied volatility surface can be characterized.The conclusions of this paper are as follows:(1)The research results of this paper find that the dynamic model of the implied volatility surface of the SSE 50ETF option based on the improved two-step method can better fit the dynamic characteristics of the implied volatility surface of the SSE 50ETF option.In particular,the relative root mean square error(RRMSE)of the call option sample can be reduced to 6.62%.And with the development of the options market,forecasting errors will decrease significantly.Investors and hedging institutions can refer to this model for risk management of assets and design more precise trading strategies.(2)Fitting results It is found that good results are obtained when the call and put options are fitted separately.For example,in the process of constructing the daily implied volatility surface,the average fitting goodness can reach about 90%,which is much higher than the Taiwan market and even slightly higher than 85%of mature European and American countries.This result not only means that the method used in this paper can well fit the cross-section data.At the same time,the existence of this unique phenomenon may be caused by the existence of a large amount of information transactions in China's options market.Therefore,in order to promote the healthy and equitable development of the options market and reduce information asymmetry,the relevant regulatory authorities should improve the regulatory system and strengthen supervision.(3)The study found that although the root-mean-square error was not significantly reduced after adding the exogenous variable of trading volume,when the sample period was divided by trading volume,with the development of the options market,the forecast error of the period with large trading volume Significant reduction,indicating that transaction volume will have an impact on the forecast of implied volatility.This also shows that China's financial market development is still immature and there is still much room for improvement.Therefore,China should vigorously promote the development of the options market,expand trading volume and varieties,and relevant institutions should establish a broad and diversified customer base.To create a good investment atmosphere,to achieve realism and improve market efficiency.The innovations of this paper are as follows:First,Hentschel(2003)pointed out that there are universal measurement errors in implicit volatility,which will cause problems such as heteroscedasticity and autocorrelation.However,this is often ignored in previous research.This article improves the previous estimation of OLS by using GLS to estimate the cross section,and strives to make the model's prediction results more accurate.Second,China's options market has just developed and market supervision is incomplete.The presence of information traders will cause differences in the underlying implied volatility.In the process of constructing the daily implied volatility surface,this paper finds that the prediction and fit of bullish and bearish respectively can greatly improve the effect.Therefore,this article implements separate forecasts based on the characteristics of the domestic option market.Third,because the options market in China has just emerged,if the trading volume is directly introduced into the dynamic model of the factor as an exogenous variable,the prediction effect is not obvious.Therefore,this article divides the sample into two periods according to the size of the transaction volume,and found that the transaction volume can indeed significantly reduce the prediction error.