A hidden Markov model for statistical arbitrage in international crude oil futures markets

We study statistical arbitrage strategies in international crude oil futures markets. We analyze strategies that extend classical pairs trading strategies, considering two benchmark crude oil futures (Brent and WTI) together with the recently introduced Shanghai crude oil futures. We show that the time series of these three futures prices are cointegrated and we introduce a mean-reverting regime-switching process modulated by a hidden Markov chain to model the cointegration spread. By relying on this model and applying online filter-based parameter estimators, we implement and test several statistical arbitrage strategies. Our analysis shows that: (i) arbitrage strategies involving the recently introduced Shanghai futures can be profitable even under conservative levels of transaction costs; (ii) strategies based on our model outperform those relying solely on observed spread values; and (iii) incorporating three futures contracts enables the implementation of arbitrage strategies even in cases where pairwise cointegration is not detected.