Arbitrage Research In China's Futures Market Base On GARCH Model And Threshold Autoregressive Model

China's HS 300 stock index futures contract officially launched at the Shanghai Futures Exchange on April 16,2010,marking the formal formation of a risk-aversion mechanism for the stock index futures market in China.With the advent of Internet+ and the era of big data,the trading volume and transaction volume of China's stock index futures contracts have continued to rise,and the market scale has continuously expanded.These have brought new opportunities and challenges to the application of statistical arbitrage in the futures market.Therefore,it is of great practical significance to apply statistical arbitrage on China's futures market.This article based on GARCH model and threshold autoregressive model for futures contract arbitrage research,its main research idea is firstly to process futures time series data pretreatment,analysis of Granger causality and cointegration relationship between futures contracts.Then,we test whether the residuals of the cointegration model have the ARCH effect,and construct the GARCH model to describe the heteroscedasticity of the residual series.Then,the SETAR model in the threshold autoregressive model is established based on the standardized residual sequence obtained from the GARCH model.Finally,using the SETAR model's threshold,the related statistical arbitrage trading strategy is set up.Based on the mean-reversion theory,statistical arbitrage is performed on futures contracts.Through empirical research,this paper has obtained the following conclusions:1.From the results of descriptive statistics and relationship studies,we can see that there is a high correlation between China's silver futures contract.They have Granger causality,and there is a long-term equilibrium relationship.An error correction model is established and the error correction coefficient is 0.9127.2.From the establishment of the model and parameter estimation,the GARCH model is established to eliminate the ARCH effect,and the threshold autoregressive model is constructed based on the standardized residual sequence of the GARCH model.The upper and lower thresholds are 0.1371 and-0.9333.Thus,the arbitrage interval is determined and the upper and lower thresholds are set as arbitrage warning signals.3.From the statistical arbitrage effect of the model,arbitrage opportunities and arbitrage returns are very good.The arbitrage trading strategy is determined based on the combination position,the warning signal,the risk determination and the stop loss signal of the set arbitrage transaction.The positive arbitrage opportunities are greater than the negative arbitrage opportunities on inside and outside the sample.The total profit is large on inside and outside the sample.In general,the risk of futures market arbitrage trading strategies established is relatively low by using the GARCH model and threshold autoregressive model.The small losses that occur are the fees incurred when opening a position,the arbitrage opportunities are relatively high,the arbitrage gains are good,and there is a very good arbitrage effect.4.The arbitrage research applied to China's futures market based on GARCH model and threshold autoregressive model is feasible.The standardized residual sequence of the GARCH model established by the silver futures contracts ag1806 and ag1812 of the China's futures market that can be described by a three-mechanism SETAR model,the fitting effect of the model is higher than that of the linear model and the SETAR model of the two mechanisms.5.According to the threshold of the threshold autoregressive model,the arbitrage range of the futures market is determined,the arbitrage trading strategy of the futures market is formulated,and statistical arbitrage is performed on the silver futures contracts ag1806 and ag1812 of the China's futures market.This provides a new research idea for statistical arbitrage research in China's futures market.