Pairs trading with fractional Ornstein-Uhlenbeck spread model

This paper proposes to model the spreads associated with financial asset prices by fractional Ornstein - Uhlenbeck (fOU) processes and constructs a pairs trading strategy based on the fOU spread model. It is shown that the Hurst parameter of the fOU process contains substantial information about the anti-persistence, or mean-reversion, characteristic of the spread over time. Consequently, the Hurst parameter is a key measure of mean-reversion for selecting the preferred candidates to trade. Adaptive methods for setting the optimal trading thresholds are proposed as well. The enhanced performance of the proposed trading model is demonstrated by comparison with existing models through simulation studies as well as an empirical analysis using high-frequency Chinese equity market data.