The Stock That's Closest to Its High Usually Keeps Going: US Data

Stocks near their 52-week high tend to keep outperforming. That's the core finding from George & Hwang (2004), one of the most replicated results in behavioral finance. We tested it on US stocks from 2000 to 2025 using a systematic quarterly screen. The strategy returned 12.78% annually vs 8.01% for the S&P 500, with one particularly striking feature: a down capture ratio of 65.0%.

Contents

1. Method
2. Why It Works: The Anchoring Hypothesis
3. Results
4. Annual Returns
5. The Three Phases
6. Phase 1: The dot-com era (2000-2007)
7. Phase 2: The GFC and recovery (2008-2018)
8. Phase 3: Momentum market and concentration risk (2019-2025)
9. The Down Capture Advantage
10. Limitations
11. Takeaway
12. Part of a Series
13. References

That number is the real story. Classic price momentum strategies typically show down capture above 100%, meaning they amplify losses in bear markets. This one did the opposite. In 2008, the strategy fell 30.2% while the S&P 500 fell 34.3%. In 2000, when the dot-com bubble burst and the index lost 10.5%, this strategy gained 48.5%.

Method

The signal is simple: compute how close each stock's current price is to its 52-week high, expressed as a ratio between 0 and 1. A stock trading exactly at its 52-week high has a proximity ratio of 1.0. A stock trading at half its annual peak has a ratio of 0.5. We buy the top 30 each quarter.

Financial data uses a 45-day look-back lag to prevent forward-looking bias. Market cap filtering uses the most recent annual report (key_metrics FY) to ensure point-in-time integrity.

Why It Works: The Anchoring Hypothesis

The academic explanation comes from George & Hwang's 2004 paper in the Journal of Finance (59(5), 2145-2176). Their argument: investors use the 52-week high as a psychological anchor. When a stock approaches that level, investors who missed the run-up hesitate to buy. They feel they've already "missed it." This creates systematic underreaction to good news near the high.

The result is a predictable drift. Stocks that breach their 52-week high tend to continue upward as the psychological ceiling breaks and late buyers pile in. Stocks far from their highs, by contrast, often stay depressed, sometimes for good reason and sometimes because sentiment has overshot.

This is different from 12-month price momentum. Classic momentum selects stocks with the best recent returns, which naturally includes stocks that have already run hard and are extended. Proximity to the 52-week high is more selective. A stock can have mediocre 12-month returns but be near its high if it fell and recovered. The signal captures a specific behavioral pattern rather than raw return persistence.

The key practical advantage: in bear markets, fewer stocks trade near their 52-week highs. The signal naturally generates fewer candidates, pushing the portfolio toward cash (the minimum is 10 stocks). This automatic de-risking explains the 65.0% down capture ratio.

Results

The strategy produced 12.78% annually over 25 years, turning $1 into $21.15. It captured 107.77% of the market's upside and only 65.0% of its downside. That combination, more up and less down, is what drives the CAGR advantage over time.

The Sharpe of 0.556 and Sortino of 0.949 are solid numbers. The Sortino being nearly double the Sharpe tells you that the volatility is mostly upside. Drawdowns are real but shorter and shallower than the benchmark.

One caveat on the stock count: the average of 25.5 stocks per period is slightly below the target of 30. This is a recent phenomenon driven by 2021-2025, when a small number of growth stocks dominated the 52-week high list and many had price gaps that made clean exits difficult. The earlier periods were closer to 30.

Annual Returns

The Three Phases

Phase 1: The dot-com era (2000-2007)

The strategy's strongest stretch. In 2000, while the S&P 500 dropped 10.5% as the dot-com bubble deflated, the strategy gained 48.5%. The mechanism was clean: dot-com stocks at their highs in 1999 crashed in 2000, so they dropped out of the proximity screen entirely. The portfolio rotated into value and industrial names that were near their highs for the right reasons.

From 2000 to 2007, the strategy compounded at a substantially higher rate than the index. Eight consecutive years of positive returns, with the only underperformance in years the index itself recovered strongly.

Phase 2: The GFC and recovery (2008-2018)

2008 was the defining test. Down 30.2% vs the index's 34.3%. Not unscathed, but meaningfully less damaged. The down capture effect worked as expected: as the market fell, fewer stocks held up near their highs, and the portfolio tilted defensive.

The recovery years were mixed. 2009 was a miss: the strategy gained 15.0% while the index rebounded 24.7%. This is the classic weakness of the proximity approach. Deep-value crash rebounds involve stocks that have fallen far from their highs. Those don't appear in a proximity screen. You miss the early stages of recovery.

But 2013-2018 was exceptional: seven years of consistent outperformance, including +39.8% in 2017 and an extraordinary +16.7% in 2018 when the index fell 5.2%.

Phase 3: Momentum market and concentration risk (2019-2025)

The recent period is the honest part of the story. 2019, 2021, and 2023 were significant misses. In 2019, the strategy returned +1.8% vs the index's +32.3%. In 2023, +5.4% vs +26.0%.

What happened: a narrow set of large-cap growth stocks, the Magnificent Seven and similar names, drove almost all of the index's gains. Many of these stocks were near their 52-week highs for extended periods, and we did hold them. But the index weights those names heavily, so when they outperformed, matching them required concentrated exposure we couldn't replicate with equal weighting across 25 stocks.

2020 was a counter-example: +34.3% vs +15.6%, powered by tech and pandemic winners near highs. The strategy can capture concentration risk when it's broad. When it's confined to 5-7 mega-caps, equal weighting dilutes it.

2024 was a return to form: +36.3% vs +25.3%, the second-best excess return in the dataset.

The Down Capture Advantage

The 65.0% down capture ratio deserves focus because it separates this strategy from most momentum approaches.

Standard 12-month price momentum strategies typically have down capture ratios above 100%, sometimes well above. This is because momentum portfolios, by construction, hold the recent winners. When markets turn, last period's winners often become this period's most overvalued names and they fall harder.

The 52-week high proximity signal sidesteps this. As stocks fall from their highs, their proximity ratios decline, and they drop out of the selection. The portfolio doesn't carry the baggage of recent momentum champions through drawdowns because the signal itself fades as prices deteriorate.

The result: in 2008, down 30.2% vs 34.3%. In 2002, down 15.5% vs 19.9%. In 2022, down 24.3% vs 19.0%. That last number is worth noting: 2022 was a rare period where the strategy fell more than the index. It wasn't the strategy's finest year.

Limitations

Equal weight vs index concentration. When market returns concentrate in a few mega-cap names, equal weighting across 25 stocks will trail a market-cap-weighted index. This explains 2019, 2021, and 2023.

No fundamental filter. The screen selects solely on price proximity. A stock can be near its 52-week high because it reported excellent earnings or because investors haven't yet sold after bad news. The signal doesn't distinguish. Adding a quality overlay (ROE threshold, debt screen) would reduce the false positives but also the stock count.

Thin universe at cycle turns. When markets recover from deep drawdowns, the best returns often come from stocks far from their highs. Those are exactly the stocks this strategy avoids. Missing the first year of a bull market is a consistent structural cost.

Transaction costs not included. Quarterly rebalancing with 25-30 stocks, equal weight, means meaningful turnover. With realistic transaction costs (0.1-0.3% one-way depending on market cap), the CAGR would be modestly lower. The relative advantage vs SPY would narrow slightly but not reverse.

Data dependency. The 252-day rolling high requires a full year of price history. New listings, spin-offs, and recently relisted stocks are excluded until they accumulate enough history. This is a minor effect on the US universe.

Takeaway

The 52-week high proximity strategy works on US stocks over a 25-year period. 12.78% CAGR vs 8.01% for the S&P 500. The mechanism is behavioral: investors anchor to the 52-week high as a ceiling, creating systematic underreaction that resolves in the direction of the signal.

The defining characteristic is the down capture ratio of 65.0%. This isn't a momentum strategy that amplifies crashes. It's closer to a trend-following approach that naturally reduces risk as markets deteriorate. Capturing 107.77% of the upside while absorbing only 65.0% of the downside is a durable edge.

The strategy isn't perfect. It misses recovery rallies. It lags when market returns concentrate in a handful of names. And 2022 was a rough year. But across 103 quarterly periods, the win rate was 64.1%, and the compounding advantage over SPY was significant.

Part of a Series

This is the US post in our 52-week high proximity global comparison. We tested the same signal across 18 exchanges:

• 52-Week High Proximity: 18 Markets Compared - full table, geographic analysis
• 52-Week High Proximity on Indian Stocks - 26.73% CAGR, the standout
• 52-Week High Proximity on Korean Stocks - 11.43% CAGR, 39.1% down capture
• 52-Week High Proximity on Thai Stocks - 12.55% CAGR

References

• George, T. J. & Hwang, C.-Y. (2004). "The 52-Week High and Momentum Investing." Journal of Finance, 59(5), 2145-2176.
• Jegadeesh, N. & Titman, S. (1993). "Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency." Journal of Finance, 48(1), 65-91.

Run the US 52-week high proximity screen: https://cetaresearch.com/data-explorer?q=BPrIuTI_-7

Data: Ceta Research (FMP financial data warehouse). Universe: NYSE + NASDAQ + AMEX. Quarterly rebalance, equal weight, 2000-2025.