Head and shoulders pattern: neckline, target and reliability

The head and shoulders is the best-known reversal pattern in chart reading. It is three rallies in a row in which the middle one goes highest: a left shoulder, a higher head, and a right shoulder that fails to reach the head. A line under the two dips between them, the neckline, is the level that matters. When price closes below it, the pattern is said to be complete, and the uptrend that came before it is in doubt.

Its mirror image, the inverse head and shoulders (also called a head and shoulders bottom), forms at the end of a decline and signals a possible turn up. Both versions come with a built-in price objective, the measured move, which is why traders like them: the pattern tells you where you are wrong and roughly how far the move might go.

This guide explains how to identify both versions, draw the neckline, calculate the target, and judge the volume and context. It then looks at three real occurrences on US stocks, including one that failed, and a mechanical test of every head and shoulders our rules found over two years on 60 large US stocks and ETFs.

Anatomy of a head and shoulders

A head and shoulders top has five turning points and one line:

1. Left shoulder. Price rallies to a peak as part of an existing uptrend, then pulls back. Nothing about it looks unusual at the time.
2. First reaction low. The pullback ends and buyers return.
3. Head. The rally carries to a new high, above the left shoulder, then fails and falls back to roughly where the first pullback ended.
4. Second reaction low. This dip ends near the first one. The two lows define the neckline.
5. Right shoulder. A third rally fails below the head, often near the height of the left shoulder. It is the first lower high in the sequence: buyers could not even match the previous peak.
6. Neckline break. Price closes below the line through the two reaction lows. This close is the signal. Before it, the shape is only a possible head and shoulders.

The shoulders do not have to be the same height, and the neckline does not have to be horizontal. It can slope up or down. What matters is the sequence: a higher high (the head) followed by a lower high (the right shoulder), and then a break of the support that held twice.

The inverse pattern reverses every element. Three troughs form during a decline, the middle one lowest. The neckline runs across the two rally highs between them, and the signal is a close above it.

Why the shape means something

Each part records a shift in who is in control. During the left shoulder and head, buyers are still strong enough to make new highs. The head's decline is the first sign of trouble: it gives back the whole rally to the head. The right shoulder is the decisive tell, because the next rally runs out of buyers below the old high. When the neckline breaks, everyone who bought in the pattern's range is now holding a loss, and some of them sell, which can feed the decline.

Edwards and Magee, who wrote the classic description in 1948, also stressed volume. In a textbook top, trading volume is heavy on the left shoulder, lighter on the head even though price is higher, and lighter still on the right shoulder. Rising prices on shrinking volume suggest fewer participants are driving the advance. For the inverse pattern, a breakout on clearly rising volume is considered more important, because a rally needs new buying, while a decline can happen simply because buyers step away. Finabase charts show a volume pane under price by default, so you can check this on any symbol.

How to calculate the measured-move target

The pattern's height is the vertical distance from the head to the neckline, measured at the head. Project that distance from the point where price breaks the neckline.

Two cautions. First, the target is a minimum objective in the classical literature, not an expected outcome. Price often stops short of it, and sometimes overshoots. Second, a tall head relative to the stock's price makes for an ambitious target. The longer the projected move, the less often it is reached before something else happens.

Three real examples, one of which failed

The examples below were found by the mechanical rules described in the test section, run on daily bars. The shaded column marks the session the outcome was decided.

Comcast, 2026: a completed top. After a 28% rally from a $21.28 low on July 24, 2026, CMCSA made a left shoulder at $24.88 (July 29), a head at $27.35 (August 25) and a right shoulder at $26.25 (September 9). The neckline through the July 31 low ($23.28) and the September 9 low ($24.19) sloped upward. On September 16 the stock closed at $23.73, below the neckline's $24.36. The head stood $3.50 above the neckline, so the target was about $20.86, 12% below the breakout close. It traded there on October 8, sixteen sessions later.

JPMorgan Chase, 2026: a textbook shape that failed. JPM peaked at $318.80 on December 9, 2025, reached a head of $337.25 on January 5, 2026, and formed a right shoulder at $326.40 on February 9. The neckline was almost flat near $297. On March 5 the stock closed at $293.55, below it, which projected a target near $254.76, 13% lower. The decline stalled at $279.10 on March 12, less than 5% below the breakout. Over the next three months the stock recovered, and on June 16 it closed at $331.14, above the right shoulder. A short seller who had waited for the neckline break and used the right shoulder as the stop would have lost at least 11%.

Exxon Mobil, 2026: an inverse head and shoulders. XOM made a left shoulder low of $144.71 on May 29, 2026, a head at $134.95 on June 25 and a right shoulder at $149.09 on August 4. The neckline through the June 3 high ($154.91) and the July 29 high ($159.07) sloped up. The stock closed at $161.46 on August 17, above the neckline, for a target of about $182.10. The path was not clean: by August 28 price had fallen back to $155.32, below the neckline, which is a classic retest. It held above the right shoulder, then climbed and reached the target on October 23, 48 sessions after the breakout.

The three cases show the range of outcomes. One reached its target quickly, one went nowhere and then reversed, and one worked only for a trader willing to sit through a dip back into the pattern.

How often did it work? A test on 60 large US stocks

Published statistics on chart patterns vary with the definition used, the market and the period. Thomas Bulkowski's Encyclopedia of Chart Patterns is the largest public catalog, and an academic study by Lo, Mamaysky and Wang (2000) detected head and shoulders patterns with a smoothing algorithm and found they carried some information about later returns. We ran our own test so the numbers on this page can be checked.

The rules. Swings were marked with a 5% ZigZag on daily highs and lows. A head and shoulders needed a head at least 2% beyond both shoulders, shoulders within 6% of each other, at least 25 sessions from shoulder to shoulder, a neckline tilted by no more than 5%, and each shoulder rising at least 40% of the head's height above the neckline. The breakout was the first close through the neckline within 40 sessions of the right shoulder. A case counted as a hit if the target traded before any close back beyond the right shoulder (the failure), within 80 sessions.

The benchmark. A target and a stop on either side of the entry create odds of their own. If prices moved randomly with no drift, the chance of reaching the target first would be the distance to the stop divided by the sum of both distances. We averaged that figure over the same cases, so the comparison is fair to the pattern's geometry.

The pattern was rare and did not beat chance. Our rules found only 28 completed patterns in about 30,000 stock-days. Fewer than a third reached their target before failing, below the benchmark in both directions. Nine of the 15 tops saw price close back above the right shoulder first.

Mistakes to avoid

• Acting before the neckline breaks. Many apparent heads and shoulders never complete: the right shoulder turns into a new high and the uptrend continues. The break is the signal.
• Forcing the pattern. If you have to squint, it probably is not there. The head should clearly stand out and the shoulders should be recognizably balanced in time and height.
• Ignoring the prior trend. A top needs an advance to reverse, and a bottom needs a decline. The same three-peak shape in a sideways range carries little meaning.
• Treating the target as a promise. Set expectations from the stop, not from the target. If the risk to the right shoulder is $10 and the target is $12 away, the trade is roughly even before you account for how often targets are missed.
• Moving the stop after a retest. A pullback to the neckline after a break is common, as the XOM example shows. Deciding in advance where the idea is wrong prevents you from rewriting the rules in the middle of the trade.

Head and shoulders compared with similar patterns

A triple top in which the middle peak happens to be a little higher is, in practice, a head and shoulders. Different traders label the same chart differently, which is one reason published statistics disagree.

How Finabase helps you find and test the pattern

Finabase does not label chart patterns automatically. Pattern recognition depends on judgment calls (how big a swing must be, how level the shoulders), and we prefer to give you the tools to apply your own rules:

• ZigZag study. Add the ZigZag to any chart and set the deviation, for example 5%, to mark the swing highs and lows our test used. The CMCSA chart with a 5% ZigZag shows the three peaks of the 2026 top. The ZigZag indicator: filtering swings, and why the last leg repaints guide explains why its last leg can change.
• Drawing tools. Each symbol's chart has a trend line tool for a sloped neckline and a horizontal line for a flat one or for the target. Signed-in users' drawings are saved per symbol, so you can come back after the breakout and see how it played out.
• Volume pane. The chart shows volume under price by default, so you can check whether volume faded through the shoulders and expanded on the break.
• Scanner. The Scanner lists stocks making 52-week highs or lows, gapping, or crossing moving averages. A stock that just failed to make a new high after a long rise is a natural place to look for a developing top.

Frequently asked questions

Is the head and shoulders bullish or bearish?

A head and shoulders top is bearish: it suggests an uptrend is ending. The inverse head and shoulders is bullish: it suggests a downtrend is ending. Neither is confirmed until price closes beyond the neckline.

What if the neckline slopes?

Draw it through the two reaction points anyway and measure the target from the neckline's value at the breakout. Many practitioners view a neckline that slopes against the pattern (down for a top, up for a bottom) as a sign of more weakness or strength, but the break itself is what matters.

How long does a head and shoulders take to form?

There is no fixed duration. On daily charts the formations most traders watch span a few weeks to several months. Our test required at least 25 sessions from one shoulder to the other.

What is a throwback or pullback to the neckline?

It is a return to the broken neckline after the breakout, as XOM did in August 2026. Some traders wait for it to enter at a better price. The risk is that it never comes, or that the return becomes a full failure.

Summary

The head and shoulders turns a familiar story into a rule: a market that makes a higher high and then a lower high, and finally breaks the support that held twice, has lost its trend. The neckline gives a precise trigger, the right shoulder a precise invalidation point, and the height a rough objective. On real data, though, the targets were reached less often than the geometry alone would predict, and textbook shapes like JPMorgan's in 2026 failed outright. Use the pattern to frame a decision with defined risk, check the trend, volume and broader market, and judge it by its failures as much as its successes.