Understanding Strategy Correlation
Mean reversion strategies exhibit distinct return correlations with other common trading styles. These correlations impact portfolio diversification and overall risk management. Traders must quantify these relationships.
Momentum strategies buy recent winners and sell recent losers. Value strategies buy undervalued assets and sell overvalued ones. Carry strategies exploit interest rate differentials or other cost-of-carry anomalies. Mean reversion, by contrast, profits from prices returning to an average.
Mean Reversion vs. Momentum Correlation
Mean reversion and momentum strategies often show negative correlation. Momentum thrives when trends persist. Mean reversion profits when trends reverse. Consider a stock, XYZ, trading at $100. A momentum strategy might buy XYZ if it rises to $105. A mean reversion strategy might short XYZ if it reaches $105, anticipating a return to $100.
During strong, sustained market trends, momentum strategies typically outperform mean reversion. For example, during the NASDAQ's dot-com bubble in 1999, momentum strategies buying technology stocks generated substantial returns. Mean reversion strategies attempting to short these overextended stocks faced significant drawdowns. Conversely, during periods of market chop or range-bound trading, mean reversion strategies often outperform momentum. From January 2018 to December 2019, the S&P 500 exhibited several sharp reversals. A simple mean reversion strategy buying the S&P 500 after a 2-standard-deviation drop and selling after a 2-standard-deviation rise could have captured profits while momentum strategies struggled with whipsaws.
A 2017 study by AQR Capital Management found a long-term correlation between mean reversion and momentum strategies of approximately -0.40 across various asset classes. This negative correlation provides diversification benefits. Combining mean reversion and momentum can reduce portfolio volatility compared to holding either strategy alone. A portfolio with 50% mean reversion and 50% momentum, each with a standard deviation of 10% and a correlation of -0.40, would have a portfolio standard deviation of 7.75%. This is lower than either individual strategy.
Mean Reversion vs. Value Correlation
Mean reversion and value strategies share some conceptual overlap. Both seek to exploit mispricings. However, their time horizons and triggering mechanisms differ. Value strategies typically operate on longer time horizons, identifying assets fundamentally undervalued based on metrics like price-to-earnings or price-to-book ratios. Mean reversion strategies operate on shorter to medium time horizons, reacting to price deviations from recent averages.
A stock trading at a P/E of 5 might be a value play. A stock that just dropped 5% in one day might trigger a mean reversion buy signal, regardless of its P/E. Sometimes, value stocks become mean reversion candidates after large, unjustified price drops. Other times, fundamentally sound companies can trade at high valuations for extended periods, making them poor mean reversion short candidates despite being "overvalued" by value metrics.
Consider Apple (AAPL) in 2016. Its P/E ratio hovered around 12-14, suggesting it was reasonably valued. A value investor might hold AAPL. If AAPL experienced a sudden 7% drop in a week due to minor news, a mean reversion trader might buy it, anticipating a bounce back to its short-term average. The correlation between mean reversion and value strategies is generally low, often near zero or slightly positive. This suggests they are largely uncorrelated return streams.
Mean Reversion vs. Carry Correlation
Carry strategies typically involve borrowing in a low-interest-rate currency and lending in a high-interest-rate currency, or selling low-volatility assets and buying high-volatility assets. These strategies profit from persistent differentials. Mean reversion strategies profit from temporary deviations.
The correlation between mean reversion and carry strategies depends heavily on the specific market and carry implementation. In currency markets, a carry trade might involve shorting JPY (low rate) and buying AUD (high rate). If the AUD/JPY pair experiences a sudden, large upward spike, a mean reversion strategy might short AUD/JPY, betting on a return to its average. This would be negatively correlated with the carry position if the spike was a continuation of the carry trend. However, if the AUD/JPY pair experiences a sudden, large downward spike, a mean reversion strategy might buy AUD/JPY. This would be positively correlated with the carry position if the spike was a reversal of the carry trend.
Generally, carry strategies perform well in stable, low-volatility environments. Mean reversion strategies can perform well in volatile, range-bound environments. During periods of risk-off sentiment, carry trades often unwind rapidly, leading to losses. In such environments, mean reversion strategies might profit from the sharp reversals. For example, during the 2008 financial crisis, many currency carry trades unwound, causing significant losses. Mean reversion strategies could have exploited the extreme swings. The correlation between mean reversion and carry strategies is often near zero or slightly negative, depending on the market cycle.
Practical Implications for Portfolio Construction
Understanding these correlations allows for more robust portfolio construction. Negative correlations reduce overall portfolio risk. Low or zero correlations provide diversification benefits.
A portfolio combining mean reversion, momentum, value, and carry strategies can achieve a more stable return profile. For example, a global macro fund might allocate 25% to each strategy. If momentum performs poorly during a market downturn, mean reversion might provide offsetting gains as markets whipsaw. If value strategies underperform during a growth-led rally, carry or momentum might compensate.
Traders should regularly calculate the rolling correlations between their chosen strategies. A 12-month rolling correlation provides insight into how these relationships change over time. If the correlation between two strategies becomes strongly positive, their diversification benefits diminish. Rebalancing or adjusting allocations may be necessary. For instance, if a trader observes the 12-month rolling correlation between their long-term mean reversion equity strategy and their short-term momentum equity strategy shifts from -0.4 to +0.2, they should investigate the cause and consider reducing exposure to one or both.
Practical Takeaway: Quantify the correlation of your mean reversion strategies with other strategies. Use a 12-month rolling correlation for major asset classes. Adjust portfolio allocations to maintain desired diversification levels.