Mean reversion
Where momentum bets on continuation, mean reversion bets on reversal: whatever recently ran extremely strong or weak tends to return toward the average. The effect shows up over very short and very long horizons, but barely over medium ones.
Mean reversion is the opposite bet from momentum: that a security which has moved far from its usual level tends to return there. Buy what's fallen sharply, sell what's risen sharply.
The long-term variant became known through economists Werner De Bondt and Richard Thaler. They showed in 1985 that stocks with the worst performance over three to five years clearly outperformed the former top performers over the following three to five years.
There's also a very short-term variant: stocks that fall sharply on one day often partly recover over the following days. This effect is heavily affected by high trading costs, since it rests on very short periods and requires frequent trading.
Worth noting is an observation that seems contradictory at first glance: over the medium term of three to twelve months, momentum tends to hold; over very short and very long terms, mean reversion tends to hold. The two patterns don't contradict each other, they simply apply on different time scales.
De Bondt and Thaler (1985) sorted stocks by their performance over the preceding three to five years and built a loser and a winner portfolio from that. Over the following three to five years, the loser portfolio outperformed the winner portfolio by several dozen cumulative percentage points, with the effect asymmetric: the reversal was considerably stronger among the losers than among the winners.
For the short-term variant, Nagel (2012) shows that individual stocks exhibit negative serial correlation at the daily and monthly level, which strengthens when market makers demand higher prices for supplying liquidity due to liquidity constraints. This variant is therefore often interpreted as compensation for supplying liquidity, rather than as a pure behavioral anomaly.
To explain the long-term effect, De Bondt and Thaler's overreaction hypothesis stands against a risk-based view, under which former losers are more often financially distressed companies with higher expected risk. Both explanations remain contested in the literature. Practically significant is that the short-term variant loses considerable strength after accounting for realistic trading costs in several studies, while the long-term variant requires lower turnover and is therefore less cost-sensitive.
Summary
- Mean reversion bets on a return to the average, momentum on continuation.
- The effect shows up over very short and very long periods, barely over medium ones.
- The short-term variant is especially cost-sensitive because of the frequent trading it requires.
Did you get it?
What did De Bondt and Thaler find in 1985?
That stocks with the worst performance over three to five years clearly outperformed the former top performers over the following three to five years.
How can the short-term reversal effect be alternatively explained?
As compensation for supplying liquidity, when market makers demand higher prices during constraints.
On which time scales do momentum and mean reversion seem to contradict each other without actually doing so?
Momentum tends to hold over the medium term of three to twelve months, mean reversion over very short and very long terms.
Sources and further reading
- De Bondt, W. F. M. and Thaler, R. H. (1985), Does the Stock Market Overreact?, Journal of Finance
- De Bondt, W. F. M. and Thaler, R. H. (1989), Anomalies: A Mean-Reverting Walk Down Wall Street, Journal of Economic Perspectives View source ↗
- Nagel, S. (2012), Evaporating Liquidity, Review of Financial Studies
Related
- Momentum and trend followingStrategy Types
- Why active funds usually underperformStage 0
- Grid-trading and DCA botsStrategy Types