Research On Hedging Of CSI 300 Stock Index Futures Based On GAS Model

In recent years,the "Black Swan" incident has occurred frequently.Under the influence of the superposition of factors such as the COVID-19 and the Russia Ukraine war,the volatility of domestic and foreign stock markets has increased significantly,and the risks faced by investors in the stock market have increased.As a derivative financial instrument,stock index futures can hedge the systematic risk of the stock market.This paper uses the Shanghai and Shenzhen 300 stock index futures to hedge the stock price of the domestic stock market has important practical significance.As a modeling method of time-varying parameters,the Generalized Autoregressive Score(GAS)model has simple calculation,and data driven time-varying parameters update the distribution function,which can better adapt to the nonlinear structure of data.In recent years,it has been applied to the study of financial risk measurement.Using GAS model to discuss the hedging of Shanghai and Shenzhen 300 stock index futures can further improve the relevant research on hedging in theory and practice.In this paper,Harvest CSI 300 ETF and Huatai Berry CSI 300 ETF are selected to establish an ETF portfolio as the spot assets to track the CSI 300 index.First,the dual GAS(1,1)model is used to study the CSI 300 spot hedging problem.Considering that in the case of abnormal market fluctuations,the current yield of Shanghai and Shenzhen300 may have nonlinear correlation,and the copula function can well describe the nonlinear correlation of financial time series,this paper also constructs the copula GAS(1,1)model.In order to evaluate the advantages and disadvantages of the above models,this paper compares the classical OLS model and DCC-GARCH model,and compares the hedging performance of each model from two perspectives:(1)calculate the hedging performance under the principle of risk minimization;(2)The hedging performance is calculated under the principle of utility maximization,and its evaluation index is negatively correlated with the variance of return rate and positively correlated with the mean value.In addition,this paper adds a risk aversion coefficient to reflect the risk aversion of different hedgers.Through the analysis of the daily yield trend chart and correlation coefficient of CSI 300 futures and cash,it is found that the futures and cash series have strong correlation,obvious left bias,thick tail and asymmetry characteristics,and do not obey normal distribution.The empirical research shows that: first,from the analysis of hedging performance of risk minimization and utility maximization,the application of binary GAS(1,1)model and copula GAS(1,1)model to the hedging research of Shanghai and Shenzhen 300 stock index futures has obvious advantages;Using the binary GAS(1,1)-t model for hedging can most effectively reduce the risk of hedging the CSI 300 spot market,and the HE value(risk minimization index)when hedging the ETF portfolio between regions outside the sample is 97.9268.Second,Gumbel copula GAS(1,1)model and Clayton copula GAS(1,1)model can well depict the upper tail correlation and lower tail correlation characteristics of financial time series,and also show good hedging effect.Third,the best copula GAS(1,1)models corresponding to different risk aversion degrees are different.When the risk aversion coefficient is small( < 2),the performance of Gaussian copula GAS(1,1)and t copula GAS(1,1)models is better;When the risk aversion coefficient is large( ≥ 2),the performance of Clayton copula GAS(1,1)model is better.Fourth,compared with a single ETF,building an ETF portfolio for hedging can more effectively reduce the risk of holding spot assets.