Research On Futures Cross-varietal Arbitrage Based On Threshold Co-integration Model

The futures market is an important part of the financial system and plays an important role in the development of the national economy.In recent years,China’s futures market has developed well,expanded overseas investment channels,improved trading rules and regulations,and diversified futures trading types,and the market size is also growing.Cross-species arbitrage strategies are generally based on the co-integration relationship between the two futures targets.Based on the co-integration relationship,determine the trading signals and implement arbitrage.How to find more suitable trading signals has become the focus of most scholars’ research.In the previous literature,most cross-species arbitrage strategies are constructed based on the co-integration relationship between the targets.Based on this,this article introduces the concept of threshold co-integration,which describes the Nonlinear cointegration relationship,that is,there is a long-term equilibrium relationship between the two commodity futures price series and the threshold value of the residual series.Different from the traditional co-integrated arbitrage strategy,the threshold co-integrated arbitrage strategy constructs a threshold autoregressive model for the residuals to obtain the upper and lower thresholds as the critical value of the arbitrage transaction,that is,the trading signal.This article constructs a crossspecies arbitrage strategy for commodity futures based on a threshold co-integration model.By comparing with traditional co-integration arbitrage strategies,this article analyzes the effectiveness of the threshold co-integration model to optimize trading signals.The cross-species arbitrage research object selected in this paper is soybean oil futures and palm oil palm futures in the Dalian Commodity Futures Exchange.Based on the co-integration relationship between the two,a traditional cointegration arbitrage strategy is constructed as a control model for the study.Then,based on the residuals of the soybean oil and palm oil price series obtained from the cointegration test,a threshold effect test is performed on the residuals,and after passing the test,a three-system threshold autoregressive model(TAR)is established for the residuals.Limits,construct a cross-species arbitrage strategy based on a threshold co-integration model as a research model.Then carry out empirical analysis of arbitrage within and outside the sample for soybean oil and palm oil price data at different frequencies,compare the arbitrage effect of the control model and the research model,and test the effectiveness of the threshold co-integrated arbitrage strategy.The results show that the threshold co-integration arbitrage strategy has better profitability and risk management than the traditional co-integration arbitrage strategy in all frequency samples.This verifies the effectiveness of the threshold cointegration model for arbitrage trading signal optimization.It is also found in the study that comparing different frequency data,the threshold co-integrated arbitrage strategy has the best arbitrage effect under the 5-minute frequency price series of soybean oil and palm oil.