Understanding Mean Reversion Returns
Mean reversion strategies profit when prices deviate from a statistical average. These strategies assume prices revert to a mean over time. This assumption shapes their return profile. Mean reversion strategies typically generate frequent, small profits. They occasionally incur large losses. This differs from trend-following, which seeks infrequent, large profits and frequent small losses.
Consider a simple mean reversion strategy on a stock. The strategy buys when the stock price falls two standard deviations below its 20-day moving average. It sells when the price returns to the moving average. This strategy aims to capture small price corrections. Each trade seeks a modest profit.
Skewness and Kurtosis in Mean Reversion
Mean reversion strategies often exhibit negative skewness. Negative skewness means the strategy has more frequent small gains and fewer, larger losses. This happens because the strategy consistently profits from small price corrections. However, when the market trends strongly against the position, the strategy can experience substantial drawdowns. These large losses pull the tail of the return distribution to the left.
For example, a mean reversion strategy trading the S&P 500 (SPY) from 2005 to 2015 might show a daily return distribution with a mean of 0.05%, a standard deviation of 0.8%, and a skewness of -0.7. A trend-following strategy over the same period might show positive skewness.
Kurtosis measures the "tailedness" of the return distribution. Mean reversion strategies often display high kurtosis, or "fat tails." This indicates a higher probability of extreme outcomes, both positive and negative, than a normal distribution. The frequent small gains contribute to a peak around the mean. The occasional large losses contribute to the fat tails.
Imagine a strategy trading a pair of stocks, like Coca-Cola (KO) and PepsiCo (PEP). The strategy shorts the outperforming stock and buys the underperforming stock when their price ratio deviates significantly. During calm markets, the ratio oscillates, generating small profits. During a market dislocation, one stock might decouple from the other for an extended period. This can lead to a substantial loss for the strategy. The probability of such an event is higher than a normal distribution would predict, contributing to high kurtosis.
Drawdown Characteristics
Mean reversion strategies are susceptible to significant drawdowns during strong trends. When a security or market segment trends persistently in one direction, the strategy will continuously take counter-trend positions. Each subsequent position will likely incur a loss. This accumulates into a substantial drawdown.
Consider a mean reversion strategy on crude oil futures (CL). The strategy buys when oil drops below its short-term average, expecting a bounce. From June 2014 to January 2016, crude oil prices fell from over $100 per barrel to below $30. A mean reversion strategy would have repeatedly bought into a falling market. It would have experienced a prolonged and deep drawdown. The strategy might have generated small profits during sideways periods. These profits would be dwarfed by losses during the sustained downtrend.
The maximum drawdown (MDD) for mean reversion strategies can be substantial. Traders must size positions carefully to manage this risk. A strategy with a typical daily profit of 0.1% might experience an MDD of 20% or more. This requires robust risk management.
Risk-Adjusted Returns
Evaluating mean reversion strategies requires metrics beyond raw returns. Sharpe ratio and Sortino ratio are important. The Sharpe ratio measures risk-adjusted return, considering total volatility.
Where $R_p$ is portfolio return, $R_f$ is risk-free rate, and $\sigma_p$ is portfolio standard deviation.
The Sortino ratio focuses on downside deviation, ignoring upside volatility. It is often more fitting for strategies with negative skewness.
A mean reversion strategy might have a lower Sharpe ratio than a trend-following strategy due to its higher volatility and negative skewness. However, its Sortino ratio might be competitive, especially if the strategy effectively limits large drawdowns.
For example, a mean reversion strategy on a basket of commodities might generate an annual return of 10% with a standard deviation of 15%. Assuming a 2% risk-free rate, its Sharpe ratio is $(0.10 - 0.02) / 0.15 = 0.53$. If its downside deviation is 10%, its Sortino ratio is $(0.10 - 0.02) / 0.10 = 0.80$. Comparing these ratios to other strategies provides a clearer picture of its risk-adjusted performance.
Traders often combine mean reversion with other strategies, like trend following, to diversify return profiles. This can smooth overall portfolio returns and improve risk-adjusted metrics. A portfolio combining strategies with negative and positive skewness can achieve a more symmetrical return distribution.