In a previous newsletter, I discussed momentum strategies. In this edition, I’ll explore mean-reverting strategies.
Mean reversion is a natural force observed in various areas of life, including sports performance, portfolio performance, volatility, asset prices, etc. In this issue, I specifically examine the mean reversion characteristics of individual stocks and indices.
Long-Run Variances of Trending and Mean-Reverting Assets
Trading strategies are often loosely divided into two categories: trend-following and mean-reverting. They’re designed to exploit the mean-reverting or trending properties of asset prices. Reference [1] provides a different perspective and approach for studying the mean-reverting and trending properties of assets. It compares the long-run variances of mean-reverting and trending assets to that of a random-walk process.
Findings
-The paper provides an alternative perspective on studying mean-reverting and trending properties of assets.
– Long-run variances of mean-reverting and trending assets are compared to a random-walk process. The paper highlights a probabilistic model for investment styles.
– Theoretical analysis indicates the variance’s direct dependence on the probability of consecutive directional movements.
– It suggests that variance may be reduced through mean reverting strategies, capturing instances of assets moving in opposing directions.
-The model is applied to US stock data. It is found that in 97 of the largest stocks, a regime of mean-reversion is prevalent.
-The paper demonstrated that relative to a random walk, the variance of these stocks is reduced due to this behavior.
-It concluded that most large-cap US stocks exhibit mean-reverting behavior.
-Mean-reverting asset prices are deemed more predictable than a random walk.
In short, the paper concluded that most large-cap US stocks are mean-reverting, and the mean reversion resulted in a reduction of the variances of the assets. This means that mean-reverting asset prices are more predictable as compared to a random walk. The opposite is true for trending assets: larger variances and less predictability.
Reference
[1] L. Middleton, J. Dodd, S. Rijavec, Trading styles and long-run variance of asset prices, arXiv:2109.08242
Mean-Reverting Trading Strategies Across Developed Markets
Reference [2] studies the mean reversion strategy of individual stocks across developed markets. It shows that the mean-reversion strategy is not profitable in all markets. However, when we apply filters for stock characteristics, the strategy becomes profitable.
Findings
-This study examined the reversal strategy in the five largest developed markets using portfolio analysis and the Fama–Macbeth (FM) regression method.
-Portfolio analysis revealed that the unconditional reversal strategy is persistent only in Germany and Japan.
-When applied to firms with higher expected liquidity provision costs, the reversal returns became stronger across all markets.
-The FM regression method provided the strongest support for the reversal strategy while accounting for key firm-related characteristics.
-Reversal returns were significantly linked to market volatility, indicating that they are more pronounced during periods of higher market liquidity costs.
-The lack of liquidity in smaller, high book-to-market, high volatility stocks contributes to their higher reversal effect.
-Small, high book-to-market ratio and volatile stocks exhibit a prominent reversal effect based on portfolio analysis.
-Traditional asset pricing models like CAPM, FF-3, and CF-4 fail to explain the observed reversal returns.
Reference
[2] Hilal Anwar Butt and Mohsin Sadaqat, When Is Reversal Strong? Evidence From Developed Markets, The Journal of Portfolio Management, June 2024
Closing Thoughts
We have examined the mean reversion characteristics of stocks and indices in both U.S. and international markets. Gaining insights into this dynamic can lead to better risk-adjusted returns for your portfolio.