Pairs Trading in Chinese Commodity Futures Markets: An Adaptive Cointegration Approach

Abstract

This study comprehensively examines pairs trading in Chinese commodity futures markets, which, although less researched, represents an important scenario for analysing commodity price behaviour. Based on a sample of daily future returns from 2006 to 2016, we propose a cointegration model that employs an adaptive learning process, and we show that our model yields an average annualised return of 26.94 percent before trading costs, using a closed‐loop strategy. Our results are robust to various tests, including parameter uncertainty, holding period constraints, trading period selection and trading costs.