Comovement And Arbitrage Of Soybean,Soybean Meal And Soybean Oil Futures Prices

Exploring the interrelationships between futures in an industry chain is an important guideline for investors’ arbitrage behavior.As soya beans,soya oil and soya meal are part of the same industrial chain,they are highly correlated and there may be some comovement in futures prices.Price comovement between these three elements are a prerequisite for arbitrage research.With regard to previous arbitrage studies,most arbitrage studies have taken the cointegration relationship between the underlying as the entry point,because when the residual error deviates from the equilibrium value,there will be a long-term equilibrium between futures components.it triggers arbitrage behavior in the market and the residual tends to try to return to the equilibrium state This is said to give rise to the possibility of arbitrage.In this paper,the threshold autoregressive and ARCH models are introduced to optimize the conventional covariance arbitrage strategy.When the residuals are within the threshold,they do not immediately converge to the equilibrium value and tend to wander randomly,with no arbitrage opportunities.On the other hand,when the residuals are outside the threshold,they tend to converge rapidly to the equilibrium value,which is the arbitrage time.The innovation of this paper is to combine the threshold covariance and ARCH models,introduce them into futures arbitrage strategies and measure arbitrage performance for benchmarking purposes.The empirical part of the study is the main continuous futures contracts for Soya I,Soya meal and Soya oil,with data selected for hourly closing prices in the sample from February2021 to March 2022.First,empirical research on the price comovement of soybean,soybean meal and soybean oil futures was conducted,analysing the underlying futures price series one by one by means of homogeneity tests,cointegration tests,VAR model tests,variance decomposition,Granger causality analysis and error correction models,laying the foundation for subsequent arbitrage research.The traditional republican arbitrage strategy was then used as the control model for the study,and the threshold republican and threshold republicanARCH models were constructed as research models.Traditional republican strategies used±1 and ±2 standard deviations of the residuals as opening and closing price signals and stop loss signals to simulate arbitrage.Traditional republican strategies used k times the standard deviation of the residuals as a threshold,which was subjective.Furthermore,traditional cointegration models have the disadvantage that they assume a constant standard deviation of the residuals when designing an arbitrage strategy and do not consider the heterogeneity of the residuals.Therefore,in this paper,the ARCH model is introduced to optimize the conditional heteroscedasticity problem,the threshold covariance-ARCH model is constructed,the new equilibrium equation and residual series are used as benchmarks to construct arbitrage strategies,the thresholds that determine the performance of trading signals are obtained,and the arbitrage performance of three arbitrage strategies are compared.Finally,the threshold covariance-ARCH model is validated on out-of-sample data to test the effectiveness of the model in optimize traditional covariance arbitrage strategies.Empirical analysis shows that there is price comovement between soya,soya meal and soya oil,making an arbitrage strategy feasible.The arbitrage performance of the gated cointegration strategy shows that the gated cointegration model outperforms the traditional cointegration model in terms of both return and risk.The arbitrage performance of the gated cointegration-ARCH model outperforms both the traditional and gated cointegration models in terms of risk,as well as return,and testing results show that the model is an effective optimization of the traditional cointegration model The results show that it is an effective optimization of the traditional cointegration model.Finally,recommendations are made for futures trading institutions,investors and government regulators.