What are the momentum and reversion patterns for stocks that have been consistent past winners or losers? In his June 2010 paper entitled “Does Bad Economic News Play a Greater Role in Shaping Investors’ Expectations than Good Economic News?”, Abdulaziz Alwathainani investigates the relationship between the consistency of past monthly stock performance and future returns. He defines consistent past winners (losers) as those ranking in the top (bottom) 30% of monthly returns for at least six of the last 12 months. He defines stocks ranking in the middle 40% for at least six of the last 12 months as a consistently moderate benchmark. Using monthly return and characteristics data for a broad sample of U.S. stocks spanning 1963-2007, he finds that:
- Over the entire sample period, consistent losers exhibit stronger average return momentum during the first year after portfolio formation and more pronounced and persistent average return reversal in years two through five than do consistent winners (see the chart below).
- During the first year after portfolio formation, the consistent winner (loser) portfolio outperforms (underperforms) the benchmark consistent moderate portfolio by about 0.47% (0.69%) per month.
- In years two through five after portfolio formation, the consistent winner (loser) portfolio marginally underperforms (substantially outperforms) the benchmark consistent moderate portfolio.
- Results generally suggest that investors overreact asymmetrically to past performance of consistent winners and losers over the intermediate term (one year) and subsequently correct asymmetrically over the long term (two to five years), with intermediate positive and long-run negative autocorrelations in returns distinct components of price evolution.
- The consistent loser portfolio exhibits a strong counter-tendency (gain) in January of the first year after formation. The bulk of long-term return reversal concentrates in January, particularly for the consistent loser portfolio.
- Results are robust to the Fama-French three-factor (market, size, book-to-market) model and the momentum factor. Specifically, the momentum factor does not subsume the ability of return consistency to predict future returns.
The following chart, constructed from data in the paper shows average monthly returns of equally weighted portfolios formed each month during January 1963 through December 2002 based on consistency of past stock returns and held without rebalancing for the next five years. Results show momentum and reversal effects, with consistent past losers exhibiting stronger intermediate-term momentum and stronger long-term reversal compared to the benchmark consistent moderate portfolio than do consistent past winners.
Results do not include any trading frictions involved in forming and liquidating portfolios.
In summary, evidence indicates that investors may be able to exploit consistency of past stock performance, independently of widely used momentum measures, via continuation over the intermediate term and reversal over the long term.
Frequently Asked Questions
What is a consistent past winner or loser stock in momentum investing?
A consistent past winner (loser) is a stock ranking in the top (bottom) 30% of monthly returns for at least six of the last 12 months. Stocks ranking in the middle 40% for at least six of the last 12 months form the consistently moderate benchmark. The measure focuses on how steadily a stock performed rather than on its cumulative past return.
Does consistency of past stock performance predict future returns?
Consistency of past stock performance does predict future returns in a study of U.S. stocks over 1963-2007. In the first year after formation, consistent winners outperformed the moderate benchmark by about 0.47% per month and consistent losers underperformed it by about 0.69% per month. In years two through five, consistent losers substantially outperformed the benchmark while consistent winners only marginally underperformed it. Results exclude trading frictions.
How is stock return consistency tested for momentum and reversal effects?
Return consistency is tested by ranking U.S. stocks each month on how often their monthly returns fell in the top 30%, bottom 30%, or middle 40% over the prior 12 months, requiring at least six months in a group. Equally weighted portfolios are formed monthly and held without rebalancing for five years, with returns compared to the consistently moderate benchmark. Results were checked against the Fama-French three-factor model and the momentum factor.
Is return consistency the same as the standard momentum factor?
Return consistency is not the same as the standard momentum factor, because the momentum factor does not subsume the ability of return consistency to predict future returns. The findings remain robust after controlling for the Fama-French three-factor model (market, size, book-to-market) and momentum. This suggests consistency of past performance carries information independent of widely used momentum measures.
What are the limitations of trading on past return consistency?
The reported return consistency results do not include trading frictions from forming and liquidating portfolios, so real-world profits may be lower. Much of the long-term reversal is concentrated in January, especially for consistent losers, which makes the effect seasonal rather than evenly spread. The evidence also comes from a single U.S. sample spanning 1963-2007, and the chart portfolios cover 1963-2002 formation dates.