Strategies And Analysis Of Statistical Arbitrage Of Futures Cross-Variety

The emergence of the agricultural product futures market is a good supplement to the spot market.Traders in the spot market can evade,diversify and transfer risks and guide their trading behavior by relying on the functions of hedging and price discovery in the futures market.A series of arbitrage modes extended from this,such as futures arbitrage,inter-commodity arbitrage,inter-period arbitrage,etc.,on the one hand,bring good returns to investors,on the other hand,it also promotes the stability of the price market of agricultural products and guides agricultural products.The price is back in the right direction.Soybean meal futures in my country were listed and traded on July 17,2000,and have been developing for more than 20 years.Rapeseed meal,an important substitute for soybean meal,has been on the market for nearly 10 years since it was listed on the futures market in 2012.Up to now,the trading volume of soybean meal futures in my country has ranked first in the global agricultural futures options for 10 consecutive years,and soybean meal futures have always been one of the most active trading varieties in the agricultural futures market.The soybean meal market is closely related to animal husbandry.The development of the soybean meal market affects the operating risks and production costs of breeding enterprises and manufacturers,thus affecting the supply and price stability of pork,and even closely related to the living standards and happiness index of residents.Therefore,It is of great significance to study the conversion arbitrage between soybean meal and its substitutes to help its value return to stability.The traditional arbitrage method depends on the investor’s own investment level and on the accuracy of the price difference and timing prediction.This prediction ability often depends on the investor’s experience and skills,which leads to the subjectivity,experience and limitations of traditional arbitrage decisions.The investment effect is also uneven.The emergence of statistical arbitrage eliminates and controls this subjectivity to a certain extent,because it determines whether it has a stable long-term relationship by retrospecting historical data,as well as the specific price difference embodiment of this relationship,and applies this relationship used to arbitrage.Of course,for the various risks in the fundamentals,it is also necessary to set stop loss signals and stop loss mechanisms to avoid and control them.A major feature of statistical arbitrage is mean reversion.As long as the asset attribute characteristics do not change for a long time and there is a long-term mean,then its price will fluctuate around this mean,and the fluctuation will continue to return to its mean.So for commodity futures,although the commodity futures contract itself may not have obvious mean-reversion characteristics,years of research have found that the spread between commodity contracts with high correlation often exhibits mean-reversion characteristics.Therefore,when carrying out crosscommodity arbitrage,it is necessary to have paired long and short positions.Whenever the spread deviates from a certain range of the long-term equilibrium spread,it is possible to open long and short positions,and when the spread returns to the mean value,then close the position in both directions,thereby to arbitrage.At present,the statistical arbitrage strategy based on cointegration is becomingmore and more mature and widely used in the academic and business circles at home and abroad.At the same time,the emergence of quantitative trading platforms can program statistical arbitrage strategies,and enter model parameters and trading signals in a programmatic manner to automatically track data and place orders,so that ordinary investors can also participate in statistical arbitrage transactions through reference.Based on the detailed introduction of the principle,method and process of statistical arbitrage based on co-integration theory,this paper uses co-integrationOLS,GARCH,EG ARCH and TGARCH models to carry out an empirical study on the cross-variety arbitrage scheme of soybean meal futures.Compared with the current The research on arbitrage strategies more comprehensively considers the heteroscedasticity and asymmetric characteristics of financial data time series,and compares their respective returns intuitively,helping investors to broaden their trading ideas and obtain higher returns.The development and stability of the market help to guide soybean meal prices to stabilize.This paper describes in detail the construction ideas,construction methods and specific steps of the soybean and rapeseed meal arbitrage strategy,and also visually presents the application results and comparative analysis of the arbitrage strategy,and draws the following research findings:First,according to the results of correlation analysis and cointegration test of soybean meal and rapeseed meal futures,it is found that soybean meal and rapeseed meal futures in my country’s commodity futures market have high correlation,and there is a long-term equilibrium relationship,that is,cointegration relationship,which can be used for cross-variety statistics.arbitrage.Second,according to the trading results,it is found that the statistical arbitrage trading strategy based on the TGARCH model can be applied to the cross-variety arbitrage of soybean meal and rapeseed meal futures.The GARCH model can bring more stable and considerable returns.Third,the arbitrage strategy is evaluated,through the backtesting test of the strategy,the results show that the TGARCH model has an annualized yield of 40.95%in the case of the maximum drawdown of 8.86%,while the TGARCH model has a short holding time,the average holding time is 16 days,so that its arbitrage efficiency is high and it can avoid the situation of needing stop loss and closing multiple times,the Sharpe ratio is 1.86,the transaction win rate is 86%,the risk reward per unit is higher,and the winning rate of each position opening is higher.The entire model has performed well over the past two years and is ideal as an investment vehicle.Due to the lack of ability,experience and knowledge,there are still many unsatisfactory points in this paper that deserve to be improved,such as selecting research data of other frequencies for investigation instead of simply specifying daily settlement data as the research basis,adopting dynamic trading signal design or trading signal parameters.Comparison,programmatic arbitrage trading,etc.If the above content is futher studied,more excellent results may be obtained.