Source: George & Hwang (2004) · The Journal of Finance · DOI: 10.1111/j.1540-6261.2004.00695.x
TL;DR
A stock's price relative to its 52-week high explains much of momentum profits — and forecasts returns better than past returns. The 52-week-high strategy earns ~1.06% per month, versus ~0.38% for Jegadeesh–Titman and ~0.25% for Moskowitz–Grinblatt industry momentum; outside January it earns ~1.13%/mo. Crucially these forecasts do not reverse long-run, suggesting short-term momentum and long-term reversal are separate phenomena.
What anomaly it documents
Predictor: nearness to the 52-week high = current price / 52-week high price.
Direction: positive — stocks near their 52-week high earn higher future returns.
Shape: ~1.06%/mo, dominating JT (0.38%) and MG (0.25%) momentum; no long-run reversal.
OSAP predictor: High52.
How to construct it
Sorting variable: ratio of current price to the trailing 52-week high.
Universe: NYSE/AMEX/Nasdaq common stocks (1963–2001).
Portfolio formation: rank into deciles on the 52-week-high ratio.
Long / short: long near-high, short far-from-high.
Weighting: equal-weighted.
Rebalancing: monthly, ~6-month holding.
Evidence and replication
Period
Notes
Source
1963–2001
52-week-high ~1.06%/mo, dominates individual & industry momentum; no reversal
this paper
OOS (post-2004)
persists; anchoring-flavored momentum
post-publication
OSAP (High52)
positive, robust
Chen & Zimmermann 2022
Why it might work
Anchoring: traders treat the 52-week high as a reference and underreact to news pushing price through it.
No reversal: gains don't unwind — favoring underreaction/anchoring over overreaction.
Limitations and risks
Momentum-like risks: shares momentum's crash exposure and turnover.
Correlated with momentum: smaller marginal alpha in a momentum-aware book.
Key references
George, T. & Hwang, C. (2004) — The 52-Week High and Momentum Investing — Journal of Finance — DOI: 10.1111/j.1540-6261.2004.00695.x
Jegadeesh, N. & Titman, S. (1993) — Returns to Buying Winners and Selling Losers — JF
Provenance: verified/generated from the paper's full text.
