Does the head and shoulders pattern predict a fall?
No. Since 1995, US stocks showing a head and shoulders top trailed the same day's median stock over the next month 50 times in 100, and over three months 49 times; a coin flip gives 50. Only 7 in 100 reached the pattern's drawn target before its stop. At no horizon does the pattern show a measurable edge, in either direction.
7 of 100 reached the drawn target before the stop
The paper
The full study: its data, method, robustness, limits and sources.
Opulence Alpha Research · Published Sep 25, 2026 · Data through Aug 21, 2026
Keywords: head and shoulders pattern, does the head and shoulders pattern work, head and shoulders success rate, head and shoulders backtest, head and shoulders top, neckline break·JEL classification: G11, G12, G14, C12, C58
- Universe
- The same for every study: 1,767 US common stocks in the 11 GICS sectors, 422 of them since delisted; S&P 500 members since 1996 plus large and mid-sized companies outside the index
- Period
- Every Wednesday from Jan 4, 1995 to Aug 19, 2026: 1,550 Wednesdays across 7,959 trading sessions
- Sample
- n = 1,764 stocks of the 1,767-stock universe; 1,361,444 formation detections, a median of 906 per Wednesday
- Outcome
- Return over the next 5, 21 and 63 trading sessions against the same day's median stock; whether the drawn target was reached
- Inference
- Date-clustered intervals; t on non-overlapping dates; proven only when |t| ≥ 3
1Introduction
A head and shoulders top is three peaks, the middle one highest, resting on a shared support line called the neckline. Chart readers take a close below the neckline as the end of an advance and project a fall about as deep as the head stands above the neckline. It is among the most cited reversal patterns in technical analysis. This study asks whether stocks showing the pattern go on to trail other stocks, and how often the projected target is reached.
2Data and method
2.1Sample design
Population and frame. The population is US common stocks listed on the NYSE and Nasdaq; funds, ETFs, trusts, preferred shares, warrants and units are excluded. The sampling frame is a fixed universe of 1,767 companies, drawn once when the platform was built and not re-sampled since, in two strata. Stratum 1 is a census of the S&P 500: every company in the index at any time since 1996 whose price history could be recovered, 1,027 companies of which 413 have since delisted; it holds 76% of the index's members in 1996 and at least 96% in every year from 2010. Stratum 2 is 740 large and mid-sized companies outside the index, selected in proportion to the market's sector weights from the stocks that passed a minimum share price of $15 and a minimum average daily trading value of $25 million; 9 of them have since delisted.
Sample. The unit of observation is a formation detection: one stock on one Wednesday on which the pattern detector recorded a formation. Weeks the platform's regime model flags as a likely change of market regime are left out, as are bars flagged as bad data. This study's sample is n = 1,764 stocks of the 1,767: 1,361,444 detections on 1,550 Wednesdays from Jan 4, 1995 to Aug 19, 2026, a median of 906 stocks with a formation per Wednesday. 217,945 detections (16%) come from the 420 companies that have since delisted. Every detection and every Wednesday carries equal weight.
Representativeness. Table 1 gives the sample by GICS sector beside the S&P Composite 1500: 7.9% of companies would have to change sector for the two to match exactly. By latest market value, 55% of the active companies are large (at least $10bn), 38% mid ($2–10bn) and 6% small. Because Stratum 2 was chosen from companies listed at construction, its history carries survivorship bias. Section 4 reads the result by market volatility.
Companies in this study's sample, of them those since delisted, the size of the active companies, the sample's share of stock-weeks and the sector's share of the S&P Composite 1500.
| Sector | Sample | Delisted | Large | Mid | Small | Stock-weeks | S&P 1500 |
|---|---|---|---|---|---|---|---|
| Information Technology | 283 | 67 | 122 | 76 | 18 | 13.5% | 12.7% |
| Financials | 255 | 59 | 114 | 74 | 8 | 14.7% | 17.2% |
| Industrials | 255 | 52 | 126 | 67 | 11 | 15.9% | 17.5% |
| Health Care | 237 | 47 | 92 | 81 | 16 | 12.3% | 10.9% |
| Consumer Discretionary | 209 | 39 | 69 | 87 | 12 | 12.4% | 12.9% |
| Energy | 105 | 30 | 39 | 31 | 5 | 6% | 4.7% |
| Consumer Staples | 101 | 38 | 40 | 17 | 5 | 5.9% | 4.9% |
| Materials | 95 | 34 | 35 | 24 | 2 | 5.6% | 5.1% |
| Real Estate | 84 | 10 | 40 | 32 | 2 | 5.9% | 6.9% |
| Communication Services | 80 | 31 | 28 | 13 | 5 | 3.9% | 3.3% |
| Utilities | 60 | 13 | 35 | 12 | 0 | 4.1% | 4% |
| All sectors | 1,764 | 420 | 740 | 514 | 84 | 100% | 100% |
Sample: n = 1,764 of the 1,767 companies. Size by latest market value for the 1,338 active companies with one: large ≥ $10bn, mid $2–10bn, small < $2bn. S&P 1500 shares count the constituents of the S&P 500, MidCap 400 and SmallCap 600 (1,506 companies, lists read Sep 25, 2026). Sectors are each company's current GICS sector.
2.2Signals
| Signal | Definition | Since | Wednesdays |
|---|---|---|---|
| Head and shoulders top | Three peaks, the middle one highest, on a shared neckline; bearish, scored on the Wednesday it is detected | Jan 4, 1995 | 1,550 |
2.3Measurement
Outcome. The return from the close on the detection's Wednesday to the close 5, 21 and 63 trading sessions later, in excess of the S&P 500, compared with the same day's median stock, so chance is 50 in 100 on every date; and whether the price reached the formation's drawn target before its stop.
Each detection is scored on its Wednesday: a bullish formation is a hit when the stock beat the same day's median stock, a bearish one when it trailed it. The detector also records whether the price reached the formation's drawn target before its stop (21 sessions for bullish formations, 42 for bearish).
The t-statistic uses non-overlapping dates only. A result is called proven when |t| is at least 3, it held in at least 60% of years and it is large enough to matter; with 1,470 tests across the studies, a looser bar would pass dozens by luck.
Full data and methods3Results
Figure 1 reads the pattern at every horizon, as the share of detections whose stock trailed the same day's median stock, which is the call a head and shoulders top makes. It did so 49 times in 100 over the next week, 50 over the next month and 49 over the next three months. The month's interval runs from 47.3 to 52.4; all three intervals contain 50, and none of the three readings is a measurable edge.
Share of detections that beat (bullish) or trailed (bearish) the same day's median stock. The line at 50 is chance; the grey band is the range chance produces for this sample (95%, clustered by date).
Share of detections that reached the drawn target before the stop, by the pattern's stage at detection.
The further away the drawn target, the less often it was reached.
Figure 2 asks the chartist's own question. Of every 100 detections, 7 reached the drawn downside target before the stop, within the 42-session window; those that did took a median of 23 sessions. Figure 3 splits the target by stage: 9 in 100 for tops whose price had already closed below the neckline, 3 in 100 for tops still forming.
A head and shoulders is usually pointed out after the fact, on charts where the fall is already visible, so the examples people remember are the ones that worked. Counted on every stock where the shape appeared, before the outcome was known, the stocks trailed other stocks about as often as a coin flip would predict.
4Robustness
Table 3 splits the month's reading by the market's volatility on the detection date; a positive edge there means the stocks trailed the median stock more often than chance, as the pattern calls. The edge stays close to zero in the low, normal and elevated regimes and falls below zero in the high regime. Only in the extreme regime does it lean the way the pattern calls, and that is the smallest slice of the sample, not tested on its own. Year by year, the pattern went the called way less than half the time in 18 of 32 years and more than half the time in the rest.
Edge over 50 in points at 21 sessions, by the market's volatility regime on the detection date.
| Volatility regime | Detections | Edge, points |
|---|---|---|
| Low | 4.3K | 0.42 |
| Normal | 7.4K | -0.34 |
| Elevated | 4.4K | -0.76 |
| High | 2.7K | -1.77 |
| Extreme | 1.1K | 4.72 |
5Limitations
- The formations are found by a detector working to fixed rules, not drawn by eye; a chartist may pick out different tops, and may discard some the detector counts.
- The target is graded within 42 sessions; a target reached later counts as a miss.
- Before costs. Averages exclude trading costs, taxes and market impact.
- Same-close timing. Returns start at the close the signal is computed from; a real trade would start later.
- Survivors among smaller companies. The non-index names were chosen from companies listed when the universe was built; no delisting returns are added.
6Conclusion
The head and shoulders top describes a fall that chart readers expect to continue. Measured on every stock where the detector found one, against the same day's median stock, the pattern's stocks trailed other stocks about as often as a coin flip would predict, over a week, a month and three months, and the drawn target was reached in only a small minority of cases. The shape is easy to recognise after the fact; in this data it carried no measurable edge over other stocks.
References
- Lo, A. W., Mamaysky, H. and Wang, J. (2000). Foundations of Technical Analysis: Computational Algorithms, Statistical Inference, and Empirical Implementation. Journal of Finance 55(4).
- Bulkowski, T. N. (2005). Encyclopedia of Chart Patterns, 2nd ed. Wiley.
- Benjamini, Y. and Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society, Series B 57(1).
- Grinold, R. C. and Kahn, R. N. (2000). Active Portfolio Management, 2nd ed. McGraw-Hill.
- Harvey, C. R., Liu, Y. and Zhu, H. (2016). … and the Cross-Section of Expected Returns. Review of Financial Studies 29(1).
- Shumway, T. (1997). The Delisting Bias in CRSP Data. Journal of Finance 52(1).
Appendix A. Questions readers ask
- Does the head and shoulders pattern work?
- Not as a forecast. Since 1995, stocks showing a head and shoulders top trailed the same day's median stock over the next month 50 times in 100, and over three months 49 times, against 50 for a coin flip. Neither reading can be told apart from chance.
- How often does a head and shoulders reach its price target?
- In this sample, 7 times in 100: the price reached the target drawn from the pattern before it hit the stop, within 42 sessions. Tops that had already broken the neckline reached it 9 times in 100.
- Why do so many traders believe in it?
- A head and shoulders is usually pointed out after the fact, on charts where the fall is already visible, so the examples people remember are the ones that worked. Counted on every stock where the shape appeared, before the outcome was known, the stocks trailed other stocks about as often as a coin flip would predict.
- How was the head and shoulders measured?
- Every Wednesday from Jan 4, 1995 to Aug 19, 2026, a pattern detector scanned the same universe of 1,767 US stocks every study uses. Each head and shoulders top it found was compared with the same day's median stock over the next 5, 21 and 63 trading sessions, and checked for whether the price reached the drawn target before the stop.
Data availability and citation
Every figure in this paper is quoted from one frozen snapshot, published as JSON with the sample description, the robustness results and the test counts. The same snapshot feeds the research console, so the two cannot disagree.
Opulence Alpha Research (2026). Does the head and shoulders pattern predict a fall? Opulence Alpha Studies, Sep 25, 2026. https://opulencealpha.ai/studies/head-and-shoulders
Research, not advice. A measured tendency across hundreds of stocks is a nudge for any one of them, never a forecast of its price.