Research On Statistical Arbitrage Of Cross Commodity Futures Based On Threshold Cointegration

In the past ten years,the development of China's futures market is very rapid.But compared with the foreign futures market,we also belong to the early stage,and many advanced methods and theories are acclimatized.Also because there are not all rational investors in the market that causes unpredictable forecasting of volatility.And the futures market trading volume in our country has been in a large proportion,so it is very important to do the research on this.In the futures market,investors are often faced with the risk of huge losses in a short period of time,thus the statistical arbitrage is generated.The difference between statistical arbitrage and arbitrage methods in general is obvious.What we want to study is no longer a futures price,but two futures price difference.We selected two kinds of futures(buy one and sell one).when there is arbitrage opportunity in the future,we then have the reverse operation of them.This method has made the great development of the futures market,reducing the risk of investors,making our income predictable.And this method is especially for small and medium-sized investors.In the application of statistical arbitrage,futures arbitrage is the most widely used.But cross commodity arbitrage is simpler and has a long history.In this article,we selects cross commodity arbitrage to study.For the cross commodity arbitrage's objects of study,we often involve the common two kinds: one is copper industry chain,and the other is the soybean industry chain.We can also select two products that have a strong relationship,but this correlation is not very stable(may be interrupted for some reason).So we are more inclined to study industrial chain products.This paper is to study the statistical arbitrage of bean 1 and corn.For existing research,we usually apply co-integration arbitrage method to cross commodity arbitrage,but we find that threshold co-integration method can better explain economic some phenomena.So we try to apply this method to cross commodity arbitrage.This paper first combed the theory and method foundation of statistical arbitrage,threshold co-integration,and also introduced the methods for controlling risk.Then the model is built using threshold co-integration theory(to construct the TAR model and TVECM model).The TVECM model is selected for complexity,here we use the method of Bias factor.Then we estimated and tested the model.Finally we carried out empirical analysis.Here we use the actual data.On the one hand,we can verify the validity of the threshold co-integration method,on the other hand we can better guide the cross commodity futures arbitrage.So the study is of great significance on China's statistical arbitrage method.This method of threshold arbitrage has greatly reduced the risk of investors,but we know that when the risk is low,the relative income will be lower.Although high return and low risk are want things for all investors,this is often not possible.So we can expect the theory results.Obviously,this method is more suitable for risk averse investors.We only intend to give investors with a new investment way,and contribute to the follow-up research and development of the futures market.