- 1George and Hwang (2004) showed that the ratio of current price to 52-week high predicts future returns with comparable or stronger signal strength than standard 6-to-12-month momentum.
- 2Stocks close to their 52-week high tend to continue outperforming; stocks far below their 52-week high tend to underperform.
- 3The authors argued that investors use the 52-week high as a psychological anchor, underreacting to news that should push prices through that reference point.
- 4The signal survived controls for industry momentum, size, book-to-market, and the standard Jegadeesh-Titman momentum factor — it carries genuinely independent information.
- 5Blank Capital Research uses nearness-to-high as a component within the 25% momentum weight of our composite rather than as a standalone screen.
#The Paper at a Glance
Title: The 52-Week High and Momentum Investing Authors: Thomas J. George, Chuan-Yang Hwang Published: Journal of Finance, 2004
If markets are efficient, reference points from past prices should be economically irrelevant. Yet practitioners have long paid attention to 52-week highs. Financial news wires tag them. Discretionary traders treat them as resistance or breakout signals. George and Hwang asked whether that folklore had any quantitative substance, and if so, whether it added anything beyond what standard momentum already captured. Their answer was yes on both counts, and their proposed mechanism was simple: anchoring bias. Investors use round numbers and salient past levels as reference points. When good news arrives near a 52-week high, they underreact because they resist pushing the price "above" the anchor. The underreaction is later corrected, generating predictable continuation returns.
#What the Paper Found
The Signal Construction
The authors ranked stocks on the ratio of current price to 52-week high:
nearness = price_t / max(price over trailing 52 weeks)Stocks were sorted into deciles each month. The top decile ("near-high") contained stocks trading closest to their 52-week high. The bottom decile ("far-from-high") contained stocks trading furthest below their 52-week high.
The Return Spread
Going long the near-high portfolio and short the far-from-high portfolio produced a significant long-short spread over holding horizons of 1 to 12 months. The magnitude was comparable to — and sometimes larger than — the classical Jegadeesh-Titman momentum spread at the same horizons.
Independent Information
The critical test was whether the 52-week-high signal added anything beyond standard momentum. George and Hwang ran horse-race regressions with both signals together. The 52-week-high signal retained statistically significant predictive power after controlling for past 6-month and 12-month returns, while standard momentum weakened in the presence of the 52-week-high variable. The two signals were correlated but not redundant.
Persistence and Weak Reversal
A notable feature of the 52-week-high signal is that its predictive returns do not appear to reverse at longer horizons in the same way as standard momentum does. In classical momentum, the 12-month winners-minus-losers portfolio tends to give back a meaningful fraction of the gain over the subsequent 2 to 5 years (the De Bondt-Thaler reversal). The 52-week-high signal shows meaningfully less of that reversion, which the authors interpret as evidence that its information is about sustained anchoring rather than transient overreaction.
Behavioral Interpretation
Why should nearness to a round-number reference level predict returns? The authors built on Barberis, Shleifer, and Vishny (1998) and on behavioral accounting evidence. Investors anchor on salient past prices. When fundamentals justify a higher price, investors demand an above-normal return before they will push prices through the reference level. That delayed response is a source of predictable continuation.
#The Math (Lite)
Define the 52-week high ratio for stock i at date t:
nearness_i,t = P_i,t / max(P_i,tau for tau in [t - 252, t])where P_i,t is the price and the maximum is taken over the trailing 252 trading days. By construction nearness lies in the interval (0, 1]. A value near 1 means the stock is at or near its 52-week high.
In a Fama-MacBeth style regression, the predictive specification is:
r_i,t+1 = alpha + beta_1 * nearness_i,t + beta_2 * momentum_i,t + beta_3 * size_i,t + beta_4 * book-to-market_i,t + epsilon_i,t+1where momentum_i,t is the past 6- or 12-month return. George and Hwang document that beta_1 is positive and statistically significant even when beta_2 is included in the regression.
In a portfolio context, one sorts stocks each month on nearness, buys the top decile, shorts the bottom decile, and holds for 1 to 6 months. The long-short return R_52wh has low-to-moderate correlation with the Jegadeesh-Titman long-short momentum return.
#How Blank Capital Research Uses This
The 52-week-high effect fits naturally into the Momentum component of our six-factor composite:
| Factor | Weight |
|---|---|
| Quality (profitability) | 30% |
| Momentum | 25% |
| Value | 15% |
| Investment | 10% |
| Stability | 10% |
| Short Interest | 10% |
We blend multiple momentum-related signals inside the 25% Momentum weight: 6- and 12-month total return, risk-adjusted momentum, residual momentum, and nearness to the 52-week high. Treating the 52-week-high signal as one input among several produces a more robust composite than picking a single momentum definition and riding it. Quality (30%) and Stability (10%) do the heavy lifting on avoiding the classic momentum pitfall — a stock near its 52-week high because of unsustainable hype or low-quality earnings will typically rank poorly on quality and thus receive a more measured composite score.
#Practitioner Watch-Outs
- Post-IPO and limited history. The 52-week-high ratio is not well-defined for stocks with less than a year of trading history. Many factor libraries handle this silently with look-back extensions that introduce subtle look-ahead issues.
- Price path matters. Two stocks with the same
nearnessratio can have very different paths to get there — one might be a steady climber, the other might have broken out on a single event. Adding a smoothness or drawdown filter helps. - Event-driven distortions. Mergers, spinoffs, and large one-day moves can move a stock instantly through its 52-week high in a way that does not reflect the anchoring dynamic. Clean the input series accordingly.
- Correlation with momentum. Even though the signals are not redundant, they overlap. Do not count the two as independent alpha sources when budgeting factor risk.
- Crowding. The 52-week-high metric is widely watched and easily visualized. Crowded trading can shorten the alpha half-life.
#See It in Action
#Further Reading
- George, T. J., and Hwang, C.-Y. (2004). "The 52-Week High and Momentum Investing." Journal of Finance.
- Jegadeesh, N., and Titman, S. (1993). "Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency." Journal of Finance.
- Barberis, N., Shleifer, A., and Vishny, R. (1998). "A Model of Investor Sentiment." Journal of Financial Economics.
- Grinblatt, M., and Keloharju, M. (2001). "What Makes Investors Trade?" Journal of Finance.
#Related Factor Explainers
More on the momentum literature.
The Capital Memo writes through factor papers Monday through Friday. The Learn library has every momentum walkthrough: Jegadeesh-Titman, Carhart, Asness, and the follow-ups worth reading.
Last updated · April 21, 2026