Research On The Existence And Formation Mechanism Of Time Reversal Effect In Commodity Futures Market

Since the discovery of the cross-sectional reversal and momentum effects,they have been shown to exist in various financial markets.Existing literature has extensively studied their formation mechanisms from both traditional finance and behavioral finance perspectives.In contrast,the study of time-series reversal and momentum effects started later and there are relatively fewer related publications.Based on high-frequency data of the US S&P500 ETF,Gao et al.(2018)discovered the existence of intraday time-series momentum effect.Subsequently,time-series reversal and momentum effects based on high-frequency data have been found in more markets,enriching the theory of time-series reversal and momentum effects and providing guidance for the investment practice of highfrequency traders.Chinese commodity futures market is booming,with an increasing variety of futures products.At the same time,in order to improve pricing efficiency and align with international markets,some futures products in China have added night trading sessions.There are significant differences in market characteristics between night trading and daytime trading,but currently there is no literature studying the time-series reversal and momentum effects in China’s night trading and adjacent time periods.Based on high-frequency price data of 28 commodity futures products in China and referring to the method of Gao et al.(2018),this paper studies the existence and formation mechanisms of time-series reversal effects in China’s commodity futures market during night trading and adjacent time periods.In terms of the study of existence: this paper uses the predictive regression method to discover two types of time-series reversal effects during night trading;analyzes these two types of reversal effects out of sample and finds that they have good outof-sample predictive ability;and also finds that the reversal effects are more significant in high volatility market environments.Based on the time-series reversal effects,a timing strategy is constructed,which has higher returns and win rates than the passive "buy and hold" strategy.In terms of the study of formation mechanisms: this paper uses hypothesis testing to discover that the mechanisms of the two types of time-series reversal effects are the regression of prices to rational expected values after continuing overreaction and compensation for liquidity providers,respectively;and also discovers a "tug-of-war" caused by heterogeneous investor preferences in some futures products.