Research On Cross-species Arbitrage Of Chinese Stock Index Futures Based On Cointegration-GARCH Model

Statistical arbitrage is a market-neutral strategy based on mean reversibility of spreads,which does not depend on the judgment of market trends and can produce low risk,low volatility and stable returns.This trading strategy has long been matured into investments in the international financial futures market.However,compared with the international financial futures market,the application of statistical arbitrage in the domestic financial futures market began on April 16,2010,when the China Financial Futures Exchange launched its first stock index futures product,the SCI 300 stock index futures.It mainly focused on cross-period statistical arbitrage and current statistical arbitrage of stock index futures,and the application of cross-species statistical arbitrage of stock index futures is very rare.On April 16,2015,the launch of SSE50 stock index futures and CSI500 stock index futures not only enriched China's stock index futures varieties,but also made it possible to use stock index futures for cross-species statistical arbitrage.Then,does there exist cross-species statistical arbitrage opportunities among the three major stock index futures in China? In the presence of arbitrage opportunities,how should investors conduct cross-variety arbitrage transactions? Therefore,this paper will take the CSI300 stock index futures,SSE50 stock index futures and CSI 500 stock index futures as the research object to research on the cross-species arbitrage trading of China's stock index futures,in order to design a statistical arbitrage strategy that can be applied to the stock index futures market and provide investors with more choices.There are many specific operational strategies for statistical arbitrage.The most widely used statistical arbitrage strategy is based on cointegration.There are two important prerequisites for the co-integration statistical arbitrage method.First,arbitrage is possible only when there is a high correlation between trading assets;second,only when there is a long-term stable equilibrium relationship between trading assets,that is,co-integration relationship,there will be arbitrage opportunities.However,the cointegration statistical arbitrage method has a shortcoming: when using the variance to establish the arbitrage trading mechanism,the constant variance is used,and the ARCH effect of the residual of the cointegration equation is not considered.Therefore,based on the cointegration theory,this paper introduces the GARCH model to describe the conditional heteroscedasticity of the residuals in the cointegration regression equation,and uses the cointegration-GARCH model to study the crossspecies arbitrage of Chinese stock index futures.Therefore,this paper first analyzes the correlation between the two of the three major stock index futures.Secondly,it conducts ADF unit root test and cointegration test on the logarithmic price series of the three major stock index futures,and screens out the arbitrage that can carry out cross-species trading.then,the ARCH effect test is performed on the residual of the cointegration regression equation,and the cointegration-GARCH cross-species arbitrage trading model is established.The data in the sample is used to estimate the parameters of the cointegration-GARCH model.Finally,the arbitrage trading rules such as trading positions,exit and entry signals and stop-loss signals are designed,and a complete set of cross-species arbitrage trading strategies is constructed.Then,the validity of the arbitrage trading strategies is tested with the data in and out of the sample,and the results of both inside and outside the sample are obtained.This paper mainly has the following research findings:First,according to the correlation analysis and co-integration test results of Chinese stock index futures,it is found that there are high correlations between the two of the three major stock index futures in China's stock index futures market,and there is a long-term equilibrium relationship,that is the cointegration relationship,which can carry out cross-species statistical arbitrage.Secondly,according to the internal and external transaction results of the statistical arbitrage trading strategy based on the cointegration-GARCH model,it is found that this statistical arbitrage strategy can be applied to the cross-species arbitrage of China's stock index futures,and it can obtain stable and considerable benefits.Thirdly,by comparing the intra-sample and out-of-sample transactions of the three portfolios,it is found that the arbitrage portfolio of the SSE 50 stock index futures and the CSI 500 stock index futures has the best performance,with the largest annualized rate of return and the smallest rebound;the performance of the Shanghai and Shenzhen 300 stock index futures and the SSE 50 stock index futures arbitrage combination is second;the combination of the Shanghai and Shenzhen 300 stock index futures and the CSI 500 stock index futures has the worst performance,and the profitability and transaction robustness are lower than the other two combinations.When investors choose the cross-species arbitrage portfolio of stock index futures,they can give priority to the arbitrage combination of SSE 50 stock index futures and CSI 500 stock index futures.In summary,the results of the sample internal and external transaction verify the validity of the statistical arbitrage strategy based on the cointegration-GARCH model,which can provide an important reference for domestic stock index futures investors.