Combining Momentum with Long-Term Reversal

Two of most documented anomalies in the asset pricing literature are the momentum effect and the long-term reversal effect. Momentum is typically defined as the last 12 months of returns excluding the most recent month (i.e., months 2–12) because it tends to show a reversal, which some have attributed to microstructure (trading) effects in which securities that have outperformed recently tend to continue to outperform. The long-term reversal effect is typically defined as the returns from month t – 13 to t – 60, in which securities that have outperformed for a long time tend to see a reversal in performance.

Adam Zaremba, Mehmet Umutlu, and Andreas Karathanasopoulos contribute to the literature on momentum and reversal with their study “Opposites Attract: Combining Alpha Momentum and Alpha Reversal in International Equity Markets,” which was published in the April 2020 issue of The Journal of Investing. The authors examined whether an integrated investment approach that combines the trading strategies of momentum and reversal would be a superior strategy than either individually.

They began by noting that researchers usually highlight the initial under-reaction that creates the trend, and the subsequent over-reaction that leads to the eventual return reversal. A trend usually starts with some catalyst, such as negative earnings release or positive macroeconomic news, to which investors under-react due to several behavioral mechanisms. Some investors anchor their views to historical prices and insufficiently adjust them to new information. Others are influenced by the so-called disposition effect, riding losers too long and selling winners too early. Finally, slow-moving capital or low analyst coverage may also lead to a delayed response of certain institutional investors. Once the trend is in place, some investors may decide to jump on the bandwagon because of feedback trading or herding, and the effect may be augmented by the similar behavior of analysts. Also, confirmation bias and representativeness may play some role. The eventual consequence of the overreaction phenomenon is that a series of abnormal profits (losses) may lead to a positive (negative) short-run abnormal return and a subsequent overvaluation (undervaluation). Importantly, the eventual consequence of the overreaction phenomenon is that a series of abnormal profits (losses) may lead to a positive (negative) short-run abnormal return and a subsequent overvaluation (undervaluation).

To test their hypothesis that a combined strategy would be superior to either of the two individual strategies the authors examined country and industry portfolios from 51 international equity markets covering the period January 1973 through October 2017. The momentum return was computed as the average monthly excess return on an index in months t − 12 to t – 1, and the long-run reversal return was computed as the average monthly excess return on an index in months t − 72 to t – 13 (one-year lagged five-year period). In the first pass, they calculated momentum and reversal signals. In the second pass, they proposed a blended alpha signal which is an average rank of momentum and reversal signals. They then applied an array of portfolio-based tests to examine the efficiency of the country- and industry-picking strategies based on past performance. Following is a summary of their findings:

• Combining the alpha momentum and reversal strategies into the blended signal produces a superior return predictive signal, outperforming the individual momentum and reversal strategies.
• The long-short equal-weighted quintile portfolio of countries (industries) delivers a three-factor (market beta, size, and value) model alpha of 1.16 percent (1.44 percent) per month with a Sharpe ratio of 0.86 (1.26), more than doubling (tripling) that of the market portfolio.
• Their findings were robust to many considerations including various asset pricing models, value-weighted and equal-weighted portfolios, different subperiods, as well as across large, medium, and small countries and industries.
• The strategy was also profitable after conservatively accounting for trading costs.
• Consistent with other research findings, the momentum and reversal phenomena are stronger across the small countries and industries than across large ones.
• The blended strategy results are not explained by the individual momentum and reversal anomalies.

These findings led the authors to conclude:

“The blended alpha strategy provides incremental value for investors who already pursue these strategies.”

They then offered the following observation:

“The concepts presented in this study could be theoretically extended to other asset classes. The momentum and reversal effects are remarkably pervasive and have been documented not only in equities, but also in government and corporate bonds, commodities, and currencies. Could the blended alpha strategy be successfully implemented also in these universes? This question remains open for future research.”

Conclusion

Investors can benefit from integrating the two strategies into their portfolio design. With that the following caution is offered. The authors showed that the combined strategies delivered superior results over the long term. However, all strategies go through long periods of underperformance. That means there will almost certainly periods when the integrated strategy will underperform. Discipline, patience and the ability to avoiding engaging in resulting⁽¹⁾ (judging the quality of a decision by the outcome instead of by the quality of the decision-making process) are the keys to successful investing.