Statistical arbitrage in crude oil futures market

Abstract

In this study, we use crude oil futures contracts that differ in maturity to come up with a trading algorithm that takes forecasted convenience yields as a base. Because there is a liquidity constraint, we take 5 different futures contracts with a maximum maturity of 5 months from January 1985 to December 2012. By forecasting convenience yields of each contract day by day, we develop a profitable real life simulated trading strategy. The results indicate that there exist statistical arbitrage opportunity in crude oil futures market. After controlling for transaction cost and interest payments, our trading algorithm yields 19.11%, 16.37%, 15.45% and 11.92% annualized returns for the contracts 2-month, 3-month, 4-month and 5-month maturities respectively. In this respect it outperforms alternative trading strategies and generates statistical arbitrage.