Evidence of Predictable Behavior of Security Returns
Source: Jegadeesh (1990) · The Journal of Finance · DOI: 10.1111/j.1540-6261.1990.tb05110.x
TL;DR
Monthly stock returns exhibit strong negative first-order autocorrelation: last month's winners tend to be next month's losers and vice versa. A contrarian strategy buying prior-month losers and shorting prior-month winners generated large gross returns. This is the founding paper of short-term (one-month) reversal — and the reason momentum strategies skip the most recent month.
What anomaly it documents
Predictor: the prior one-month return.
Direction:negative — high last-month return predicts low next-month return (reversal), the opposite of intermediate-horizon momentum.
Horizon: one month. The effect is distinct from De Bondt-Thaler long-term (3–5 year) reversal and from Jegadeesh-Titman 12-month momentum; the autocorrelation structure of returns flips sign across horizons.
OSAP predictor: STreversal.
How to construct it
Sorting variable: prior-month total return.
Universe: US common stocks (1934–1987 in the paper); in practice apply price/liquidity filters because the effect is microstructure-driven.
Portfolio formation: monthly.
Long / short: long prior-month losers, short prior-month winners.
Holding period: one month, then fully rebalance.
Weighting: equal-weighted historically (the effect is much stronger in small, illiquid names).
Evidence and replication
Period
Gross return
Net of costs
Source
IS (1934–1987)
large monthly reversal profit
sharply reduced
this paper
OOS (post-1990)
persists gross, marginal net
—
post-publication
OSAP replication (STreversal)
clear gross
—
Chen & Zimmermann 2022
The gross profits are large and highly significant, driven by strong negative monthly serial correlation.
Net of transaction costs the strategy is largely or entirely unprofitable for most investors: monthly full-turnover trading of small, illiquid stocks incurs costs that consume the gross edge. The realistic interpretation is that short-term reversal is compensation for providing liquidity, earned by market makers and low-cost traders, not a free lunch.
Why it might work
Liquidity provision (leading view): uninformed order-flow shocks and price pressure push prices temporarily away from value; liquidity providers who take the other side earn the reversal as compensation (Nagel 2012 explicitly links short-term reversal returns to expected returns from liquidity provision, which spike in volatile markets).
Bid-ask bounce / microstructure: measured monthly returns contain mechanical reversal from quotes bouncing between bid and ask, inflating raw reversal estimates.
Overreaction to firm-specific information: investors overreact to recent idiosyncratic news, later corrected.
Limitations and risks
Turnover is brutal: 100% monthly turnover concentrated in the highest-cost names — the central reason paper profits rarely survive in practice.
Capacity: very limited; lives in small, illiquid stocks.
Microstructure contamination: part of the raw signal is bid-ask bounce, not tradeable alpha; skip-a-day/skip-a-week conventions reduce but complicate it.
Why momentum skips a month: because this one-month reversal would otherwise contaminate a momentum signal, the standard momentum construction measures t−12 to t−2.
No free full text: paywalled; see DOI.
Key references
Jegadeesh, N. (1990) — Evidence of Predictable Behavior of Security Returns — Journal of Finance — DOI: 10.1111/j.1540-6261.1990.tb05110.x
Lehmann, B. (1990) — Fads, Martingales, and Market Efficiency — Quarterly Journal of Economics
Nagel, S. (2012) — Evaporating Liquidity — Review of Financial Studies
Jegadeesh, N. & Titman, S. (1993) — Returns to Buying Winners and Selling Losers — Journal of Finance
Chen, A. & Zimmermann, T. (2022) — Open Source Cross-Sectional Asset Pricing — Critical Finance Review
