Head and Shoulders Pattern: How to Read It
How a head and shoulders top forms, how to measure its target, and three real 2022–2024 cases: one hit, one fell short, one never triggered.
Contents
The head and shoulders is the reversal pattern most people can name, and one of the most over-spotted. Almost any three bumps on a chart can be squinted into one. What makes it useful is not the drawing but the sequence of events it describes, and the fact that it tells you exactly which price has to break before it counts.
What the shape is saying
Head and shoulders topping pattern with a left shoulder, higher head, right shoulder, and neckline. A breakdown below the neckline projects a downside target equal to the head-to-neckline distance.
- Left shoulder: an ordinary high in an uptrend, followed by a pullback.
- Head: buyers push to a new high. The pullback that follows goes back to roughly the same level as the first one. The trend is still intact, but the higher high did not lead anywhere.
- Right shoulder: the next rally stalls below the head. That is the first lower high, and the first sign that buyers are losing the argument.
- Neckline: a line through the two troughs. A close below it is the confirmation: the pullback lows that held twice have now failed.
- Measured move: the distance from the head down to the neckline, subtracted from the neckline. It gives a rough idea of scale, not a destination.
Amazon, summer 2024
Amazon's chart from June to August 2024 is a clear example, though not a perfect drawing. The left shoulder topped at 188.35 on June 12. The stock then ran to a broad, flat head around 200, peaking at 201.20 on July 8. The right shoulder reached 190.60 on August 1, within about 1.2% of the left one. The two troughs between them were 181.22 on June 17 and 176.80 on July 25, so the neckline sloped slightly down. Using the lower trough:
| Point | Date | Price |
|---|---|---|
| Left shoulder high | June 12 | 188.35 |
| First trough | June 17 | 181.22 |
| Head high | July 8 | 201.20 |
| Second trough (neckline) | July 25 | 176.80 |
| Right shoulder high | August 1 | 190.60 |
| Head to neckline | 201.20 − 176.80 | 24.40 |
| Measured target | 176.80 − 24.40 | 152.40 |
Here is the part that textbook diagrams leave out. Amazon reported earnings after the close on August 1. On August 2 it opened at 166.75, 10.05 below the neckline, and closed at 167.90 on 3.45 times average volume. Nobody got to sell at the neckline. Anyone holding with a plan to exit "on a close below 176.80" got out roughly 5.7% below that level. The next session the stock traded down to 151.61, 0.79 below the measured target, before bouncing.
The pattern was right about direction and about scale, and it still would have hurt someone relying on a neat exit at the line. It is also worth knowing what happened later: Amazon closed back above the head's 201.20 on November 6, 2024. A completed top is a statement about the next few weeks, not the next year.
Three heads, three outcomes
One success proves little. Here are two more tops from the same dataset, found with the same rules (shoulders within a few percent of each other, head at least 5% above both).
| Stock | Head | Neckline | Target | What happened |
|---|---|---|---|---|
| AMZN 2024 | 201.20 (Jul 8) | 176.80 | 152.40 | Gapped through the neckline on Aug 2; low of 151.61 on Aug 5 |
| MSFT 2022 | 294.18 (Aug 15) | 245.94 | 197.70 | Closed at 245.38 on Sep 15; fell to 213.43 by Nov 4 (−13.0%) but never reached the target |
| AAPL 2024 | 237.23 (Jul 15) | 206.59 | 175.95 | Traded below the neckline intraday but never closed there; closed at 259.02 on Dec 26 |
The Microsoft case is the most typical outcome: the break was real and the decline was substantial, but the measured move overshot. Apple is the most instructive. Its left shoulder (220.20), head (237.23) and right shoulder (225.60) look just as convincing on a chart. The lowest close after the right shoulder was 207.23 on August 6, just 0.6 above the neckline. Anyone who sold because the shape "looked like" a head and shoulders missed a 25% rise from that low close to the December high.
The measured move is a ruler, not a promise
The target formula is simple arithmetic, which is exactly why it gets more respect than it deserves. Use it to judge whether a trade has room: if the target is 12% away and the nearest support zone is 3% away, the first stop is probably the support, not the target. And if your stop above the right shoulder is further away than the target is below the neckline, the trade does not pay for its risk, however clean the pattern looks.
The inverse version
Flip the whole thing and you get an inverse head and shoulders at the end of a downtrend: a low, a lower low, a higher low, and a neckline across the two rally highs. A close above the neckline completes it. The same caveats apply, and volume on the breakout matters more here, because a rally without participation fails more easily than a decline does.
Other chart patterns in one table
Most classic chart patterns work the same way: a shape made of swing highs and lows, a trigger line, and a measured move. The trigger is always the part to watch.
| Pattern | Usual bias | Trigger | Common failure |
|---|---|---|---|
| Double top | Bearish | Close below the trough between the two highs | Price holds the trough and breaks to new highs |
| Double bottom | Bullish | Close above the peak between the two lows | Third test breaks the lows |
| Ascending triangle | Bullish | Close above the flat top | Rising lows give way first |
| Descending triangle | Bearish | Close below the flat bottom | Short squeeze through the falling highs |
| Symmetrical triangle | Neutral | Close outside either line | Breakout reverses within days |
| Bull flag | Bullish | Close above the flag's upper line | Flag drifts too deep, over half the pole |
| Rising wedge | Bearish | Close below the lower line | Keeps grinding up with shrinking ranges |
A plan around the neckline
- Draw the neckline only after the right shoulder has formed. Before that you are guessing.
- Wait for a daily close below it. An intraday poke, like Apple's in August 2024, is not a break.
- Check the calendar. An earnings date inside the right shoulder means the break may come as a gap, as it did for Amazon.
- Put the invalidation above the right shoulder high, and size the position so that distance is an amount you accept losing. The risk management guide shows the arithmetic.
- Treat a retest of the neckline from below as information: if price closes back above it, the pattern has failed and you step aside.
Questions readers ask
How reliable is the head and shoulders pattern?
It depends heavily on whether you wait for the neckline break. Of the three 2022–2024 examples above, the two that closed below the neckline both fell meaningfully (one reached its target, one fell 13% but stopped short). The one that never broke went on to rise 25%.
How do you calculate the head and shoulders target?
Measure from the head's high down to the neckline and subtract that distance from the neckline. For Amazon in 2024: 201.20 − 176.80 = 24.40, and 176.80 − 24.40 = 152.40. Treat the result as a rough scale, not a price that must be reached.
What if the neckline is sloped?
Draw it through the two troughs as they are, or use the lower trough as a conservative horizontal line, as this guide did for Amazon. A downward-sloping neckline gives a later, lower trigger; be consistent about which method you use.
Sources
- StockCharts ChartSchool — Head and Shoulders Top
- StockCharts ChartSchool — Chart Patterns
- TradingView — How to Read Chart Patterns
Charts and worked examples use daily prices from the One Candle Ahead historical dataset (split-adjusted). Past price behaviour does not predict future results.
Written and maintained by the Studio Solum Editorial Team. How we write and review guides