A hidden Markov model for statistical arbitrage in international crude oil futures markets (2309.00875v2)

Abstract: In this work, we study statistical arbitrage strategies in international crude oil futures markets. We analyse strategies that extend classical pairs trading strategies, considering the two benchmark crude oil futures (Brent and WTI) together with the newly introduced Shanghai crude oil futures. We document that the time series of these three futures prices are cointegrated and we model the resulting cointegration spread by a mean-reverting regime-switching process modulated by a hidden Markov chain. By relying on our stochastic model and applying online filter-based parameter estimators, we implement and test a number of statistical arbitrage strategies. Our analysis reveals that statistical arbitrage strategies involving the Shanghai crude oil futures are profitable even under conservative levels of transaction costs and over different time periods. On the contrary, statistical arbitrage strategies involving the three traditional crude oil futures (Brent, WTI, Dubai) do not yield profitable investment opportunities. Our findings suggest that the Shanghai futures, which has already become the benchmark for the Chinese domestic crude oil market, can be a valuable asset for international investors.

• The paper introduces a Hidden Markov Model (HMM) framework for exploring statistical arbitrage opportunities in Brent, West Texas Intermediate, and Shanghai crude oil futures markets.
• The study models the cointegration spread among futures prices using an HMM that incorporates regime-switching to adapt to changing market conditions and estimate parameters.
• Empirical analysis demonstrates that including Shanghai futures enhances profitability, showing significant returns even with transaction costs due to rapid price adjustments.

Overview of a Hidden Markov Model for Statistical Arbitrage in Crude Oil Futures Markets

This paper employs a Hidden Markov Model (HMM) framework to explore statistical arbitrage strategies within the context of international crude oil futures markets. With a specific focus on three selected futures contracts—Brent, West Texas Intermediate (WTI), and the recently introduced Shanghai crude oil futures—the paper investigates the feasibility and profitability of statistical arbitrage through dynamic modeling techniques.

Methodological Approach

The authors establish evidence of the cointegration among the price series of these three futures contracts. Building on this foundation, they model the cointegration spread using a mean-reverting process governed by a Hidden Markov Chain. This methodological choice allows them to incorporate regime-switching behaviors to simulate different market conditions through a stochastic filtering approach alongside the Expectation Maximization (EM) algorithm for parameter estimation.

The HMM framework offers a robust method for adapting to changing market environments by dynamically estimating the parameters that characterize the relationship among the futures prices. This dynamic modeling is critical given the inherent volatility and complexity of futures markets, allowing investors to fine-tune their strategies in real-time.

Empirical Analysis

In their empirical analysis, the authors demonstrate that statistical arbitrage strategies incorporating the Shanghai futures achieve significant returns, even when accounting for substantial levels of transaction costs. They show that the return on investment improves when the trading strategies include all three futures contracts compared to using pairs of only two of these contracts.

Furthermore, the results emphasize the rapid price adjustment in the Shanghai futures, facilitating quicker reversion to the long-term equilibrium. This characteristic enhances the profitability of strategies that capitalize on temporary price discrepancies. The Shanghai futures' distinct role aligns with trends observed in new market entrants generally offering lucrative arbitrage opportunities before markets mature.

Practical and Theoretical Implications

The findings present practical implications for financial arbitrageurs seeking opportunities across global futures markets. These statistical arbitrage strategies harness the newly available Shanghai crude oil futures, presenting a significant edge over using traditional benchmarks like the Brent and WTI futures alone. The model's dynamic nature allows traders to adapt quickly to market changes, optimizing their position-taking decisions based on the most current market estimates.

From a theoretical standpoint, this paper advances the application of HMMs in financial markets, particularly in modeling cointegration with regime-switching mechanisms. It substantiates the potential of using sophisticated econometric models to exploit transient departures from long-term pricing equilibria effectively.

Future Prospects

The research lays a foundation for future explorations into derivative markets using regime-switching models. As financial products and market dynamics evolve, the potential for such models to capture transient arbitrage opportunities will intensify. Further research could integrate machine-learning techniques and artificial intelligence with the HMM framework to enhance predictive accuracy and adaptability in ever-changing market conditions.

In conclusion, this paper provides a comprehensive framework for evaluating statistical arbitrage opportunities in crude oil futures markets by leveraging the flexibility and predictive power of an HMM. The integration of the Shanghai futures represents a strategic move to exploit newly introduced market dynamics, presenting synthetic avenues for refined trading strategies amidst the global commodities landscape.