Optimizing Mean-Reversion Pairs with Ornstein-Uhlenbeck Models

Summary

This study develops a mean-reversion pairs strategy for US equities. Rather than selecting pairs through customary pair-trading practice, it uses maximum likelihood estimation to form static portfolios that fit an Ornstein-Uhlenbeck process. The estimated model parameters generate contrarian signals, while trading thresholds and the in-sample window length are selected through multiple tests.

The document reports nine promising pairs with Sharpe ratios above 1.9 in both in-sample and out-of-sample periods. It highlights CCI with HCP and CCI with Realty Income as examples, with the latter pair showing a higher reported out-of-sample Sharpe ratio. These results illustrate the proposed selection and signal process, but the evidence is limited to selected pairs and the stated evaluation periods. The summary provides no detail on transaction costs, execution assumptions, sample dates, or whether selection procedures could introduce bias, so the reported Sharpe ratios alone do not establish general or live-trading profitability.

Key ideas

• Maximum likelihood estimation is used to construct static equity pairs that fit an Ornstein-Uhlenbeck process.
• Model parameters generate contrarian signals for trading deviations from the fitted mean.
• Thresholds and the length of the estimation period are optimized through multiple tests.
• The study reports strong in-sample and out-of-sample Sharpe ratios for nine selected pairs.
• The excerpt does not specify transaction costs or execution assumptions.

Tags

# On the Profitability of Optimal Mean Reversion Trading Strategies


# On the Profitability of Optimal Mean Reversion Trading Strategies









We study the profitability of optimal mean reversion trading strategies in the US equity market. Different from regular pair trading practice, we apply maximum likelihood method to construct the optimal static pairs trading portfolio that best fits the Ornstein-Uhlenbeck process, and rigorously estimate the parameters. Therefore, we ensure that our portfolios match the mean-reverting process before trading. We then generate contrarian trading signals using the model parameters. We also optimize the thresholds and the length of in-sample period by multiple tests. In nine good pair examples, we can see that our pairs exhibit high Sharpe ratio (above 1.9) over the in-sample period and out-of-sample period. In particular, Crown Castle International Corp. (CCI) and HCP, Inc. (HCP) achieve a Sharpe ratio of 2.326 during in-sample period and a Sharpe ratio of 2.425 in out-of-sample test. Crown Castle International Corp. (CCI) and Realty Income Corporation (O) achieve a Sharpe ratio of 2.405 and 2.903 respectively during in-sample period and out-of-sample period.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.