| The issue of option pricing has always been one of the core research topics of modern financial engineering.In the study of pricing of financial derivatives,the study of option pricing model is one of the most widely used one.The characteristics of leptokurtosis and fat-tailed and heteroscedasticity often occurs to financial time series data.For these financial data,the first proposed GARCH model can solve the heteroscedasticity of volatility,but it is defective when fitting data,considering its characteristics of thick tail and biasing.Based on this,a mixed normal heteroscedasticity model is proposed.This model not only retains the excellent characteristics of normal distribution but also solves the problems of leptokurtosis and fat-tailed to some extent,and can properly improve the defects of normal distribution.This paper first introduces the mixed normal heteroscedasticity model and some basic features.According to the process of asset price return of the mixed normal heteroscedasticity model,the Nadom-Nikodym derivative is used to derive the risk neutrality measure needed in option pricing,and the change in the model under this measure.Then,the Gibbs sampling strategy of Bayesian method is used to estimate the historical data of the daily yield of SSE 50 ETF,so as to determine the appropriate pricing model,and empirical analysis is conducted afterwards.The empirical results show that the mixed normal heteroscedasticity model can not only effectively estimate the parameters of the model when fitting the time series data of the rate of return,but also has a small pricing error when applied to the SSE 5OETF option in China. |
