// STUDY RESULT · 1990-01 TO 2026-07
Short-Horizon Reversal
NO EDGEAt the 1-week to 1-month horizon, returns reverse: last period's losers beat last period's winners. Later work links the effect to liquidity provision.
| THE STRATEGY | $SPX BUY & HOLD | |
|---|---|---|
| Growth a year (CAGR) | -1.7% | 9.0% |
| Return per unit of risk (Sharpe) | 0.12 | 0.66 |
| Worst fall (max drawdown) | -85.4% | -52.6% |
// GROWTH OF $1, LOG SCALE
AMBER: THE STRATEGY · GREY: $SPX BUY & HOLD
Tested on years it never saw (from the split): -4.9% a year against 12.4% for the benchmark; Sharpe 0.15 in the training years, 0.07 after.
After 25 bps of trading costs: -6.1% a year. BELOW RANDOM
// DECADE BY DECADE
| ERA | STRATEGY | BENCHMARK |
|---|---|---|
| 1990s | 5.6% | 15.5% |
| 2000s | -7.6% | -2.6% |
| 2010s | 9.1% | 11.6% |
| 2020s | -17.1% | 14.1% |
// THE IDEA, AS PUBLISHED
Jegadeesh (1990); Lehmann (1990) · 1990
- Jegadeesh: rank on predicted monthly return (dominated by a negative one-month lag), long top decile, short bottom
- Lehmann: weekly, weights proportional to minus each stock's prior-week return relative to the market
- Hold one month (Jegadeesh) or one week (Lehmann); costs decide implementability
A real computation on the QuantGPT warehouse, survivorship-free, walk-forward. Gross returns, no costs except where stated. Past performance is not a promise. Findings, not advice. COMPUTED 2026-09-24 19:42 ON THE QUANTGPT WAREHOUSE: EVERY DELISTED NAME KEPT, FUNDAMENTALS AS FILED. NOT INVESTMENT ADVICE.