Research On Option Pricing Based On Variable Volatility

As an important class of financial derivatives,options are widely used in areas such as risk hedging and hedging.In mature financial markets abroad,options are heavily used by risk managers to hedge the risk of asset price fluctuations.In China,due to the lack of corresponding risk hedging instruments,investors often suffer heavy losses in the market volatility.In order to meet the needs of investors to avoid risks,the Shanghai Stock Exchange launched China's first on-the-spot option,the SSE 50 ETF opton,on February 9,2015.Since the listing of options for 4 years,the transaction has been active and has become a powerfful tool for investors to manage portfolio risk.When investors trade options,the most important concern is the option pricing problem.This is because the accuracy of the pricing will affect the investor's subsequent trading strategy construction and portfolio position hedging.The volatility of basic financial assets is a key parameter in option pricing.Therefore,correctly estimating and predicting the volatility of basic financial assets is very important for the pricing of options.In order to better analyze the inpact of changes in the volatility of basic financial assets on option pricing,the research in this paper is carried out through the following two parts.The first part studies the modeling problem of variable volatility.Variable volatility mainly includes historical volatility and realized(high frequency)volatility.These two types of volatility have their own strengths in revealing information on changes in the return on financial assets.Historical volatility can better capture the fluctuation of financial assets in the past period of time,but its shortcoming is that it can oniy reflect the fluctuation ofasset prices during the day,and will lose a lot of intraday trading information.The realized volatility is calculated based on the intraday high-frequency income data which can reflect the variation characteristics ofasset price fluctuations in 5 minutes,1 minute or even shorter time.However,the current mainstream volatility model(HAR-RV model)only considers the endogeneity of higb-frequency volatility,ignoring the impact of external information shocks,and there will still be some bias in volatility estimation and prediction.Therefore,the first part of the thesis also focuses on how to introduce the influence of external information impact in the current mainstream volatility model,and constructs a symbolic hopping high frequency volatility model based on external information impact(HAR-V-RV-Jump model).The model not only considers the common influence of endogenous factors and external information impact on high-frequency volatility,but also considers the asymmetric effect of multi-information impact.The high-frequency transaction data of CSI 300 and CSI 500 index were selected as the research samples,and the prediction ability of HAR-V-RV-Jump model was evaluated by rolling time window prediction and SPA test.The results show that the model can According to the type of external information impact,the high-frequency volatility is more accurately predicted,and its prediction ability is significantly better than the existing HAR-RV-Jump modelIn the second part of the thesis,the static and dynamic binary tree simulation method is used to study the option pricing problem based on historical volatility and realized volatility.The SSE 50ETF in-the-money call option and out-the-money put option are selected as research samples and compared.Option pricing model based on historical volatility(Constant volatility,GARCH(1,1)process and EGARCH(1,1)process)and "option pricing model based on realized volatility(HARGL process)"The prediction accuracy shows that the realized volatility option pricing can better capture the volatility variation characteristics of the option sample,and the simulation and prediction of the option price is obviously better than the historical volatility-based option pricing.