Research On Statistical Arbitrage Strategy Of Commodity Futures

Commodity futures market has the characteristics of two-way trading and high leverage.Its risk is greater than that of stock market,but its return is also more considerable.Faced with the high risk of the futures market,many futures investors hope to find an investment strategy that can hedge most of the market risk while obtaining excess return.Statistical arbitrage strategy can meet the needs of these investors.Pairs trading is one of the statistical arbitrage strategies.In pairs trading,investors long one asset and short another assets that is highly related to it.Investors use the mean recovery characteristics of the spread between two assets to obtain return.In this thesis,We use commodity futures price data from 2015 to 2017 to research the investment strategy.Four pairs of commodity futures arbitrage portfolios are selected from the perspective of liquidity and realistic logic of commodity futures portfolio.Then,this thesis test these futures price data by cointegration and confirm the stationary of spread time series.In this thesis,we use the data of 2018 outside the sample to measure the strategy.The results show that most portfolios have the advantages of higher Sharp ratio and lower withdrawal.After combining several strategies with equal funds,it is found that the maximum withdrawal and annual volatility of the strategies are further reduced.Compared with single arbitrage strategy,portfolio strategy has less risk and higher Sharp ratio.