Introduction/Hypothesis:
In this project, we've implemented a pairs trading strategy on Exxon Mobil (XOM) and Chevron (CVX), two major oil companies operating within the same industry. The principle behind pairs trading is the expectation of mean reversion when the spread between two correlated stocks diverges significantly. We've operationalized this by defining conditions based on the z-scores of the spread to trigger our long and short positions.
Analysis:
The chart presents the dynamic performance of our pairs trading strategy over time. From 2015 to 2020, the strategy struggled, with cumulative returns dipping as low as -50%. This could have been due to a variety of factors such as changes in the oil market, lack of mean reversion in the spread, or other macroeconomic variables affecting the correlation between the stocks.
However, in 2020, the outbreak of the COVID-19 pandemic had significant impacts across all sectors, including oil. The strategy's performance dramatically improved during this period, with cumulative returns skyrocketing to 250%. This sudden upswing may have been driven by increased mean reversion in the stock prices, potentially due to shared responses to the industry-specific factors or broader economic trends induced by the pandemic.
Despite a significant pullback to 100% and a subsequent bounce to 200% in 2021-2022, the strategy experienced another drawdown to 50% followed by a substantial rise to 300% in 2022-2023. The extreme volatility during this period coincided with geopolitical tensions, including Russia's invasion of Ukraine in 2022, which led to a significant surge in oil prices.
Conclusion:
The performance of the pairs trading strategy has been highly volatile and appears to be significantly influenced by major global events. The periods of high returns during the COVID-19 pandemic and following Russia's invasion of Ukraine suggest that the strategy can capitalize on market volatility induced by such events. However, the periods of losses underline the potential risks and the importance of careful risk management.
It would be beneficial to investigate these periods further to understand how these global events affected the correlation and mean reversion of the two oil stocks and consider if and how the strategy could be improved to mitigate risks during other periods.
Furthermore, it would be worthwhile to test the strategy on other pairs, adjust the z-score thresholds, or modify the lookback period for calculating the mean and standard deviation to see if the performance can be improved.
As always, all investment strategies carry risk, and past performance does not guarantee future results. Thorough backtesting, consideration of transaction costs, and portfolio diversification are essential for risk management.