Quick answer
A head and shoulders pattern is a three-peak reversal formation: a first peak (the left shoulder), a higher peak (the head), and a third peak roughly matching the first (the right shoulder), all sitting above a support line called the neckline. The pattern isn’t confirmed until price actually closes below the neckline on increased volume, not when the third peak simply appears. Acting before that close is the single most common mistake beginners make with this pattern.
| Part | What it is | What to watch |
|---|---|---|
| Left shoulder | First peak, previous trend still intact | Often the highest volume of the three peaks |
| Head | Higher peak, the final push of the old trend | Volume often exceeds the left shoulder |
| Right shoulder | Third peak, roughly matches the left shoulder’s height | Volume is typically the weakest of the three |
| Neckline | Support line connecting the two pullback lows | The pattern isn’t confirmed until this breaks on a closing basis |
The head and shoulders pattern is one of the oldest named formations in technical analysis, first documented in detail in the 1930s and popularized further by Robert Edwards and John Magee’s landmark 1948 book on technical analysis of stock trends. Nearly a century later, it’s still one of the most misused. Most beginners learn to spot the three peaks and stop there, treating the shape itself as the signal. The shape is only half the pattern. The other half, the neckline, the volume behind each peak, and what happens after the break, is where the actual edge lives, and it’s the part most quick explanations skip.
What the pattern actually looks like
Picture a stock in an established uptrend. It rallies to a peak, pulls back, that’s the left shoulder. It rallies again, this time to a new, higher peak, pulls back again, that’s the head. It rallies a third time, but this attempt falls short of the head, roughly matching the height of the left shoulder instead, then pulls back once more, that’s the right shoulder.
Connect the two pullback lows, the one between the left shoulder and the head, and the one between the head and the right shoulder, and you get the neckline. That line is the actual support the whole pattern rests on. Everything above it is just the three peaks. What happens at and after the neckline is what turns this from an interesting shape into an actual trading signal.
The mirror version, a head and shoulders bottom, works the same way upside down: a low, a lower low, a third low roughly matching the first, with a neckline now acting as resistance overhead instead of support underneath. Everything covered below applies to both, just flipped.
Volume: the confirming signal most beginners skip
The shape alone isn’t enough. The volume behind each peak tells you whether the pattern has real weight behind it.
Confirmed or not yet?
1 / 4Each chart shows a right shoulder forming near the neckline. Has the reversal actually confirmed, or not yet?
| Peak | Typical volume pattern | What it suggests |
|---|---|---|
| Left shoulder | Often the heaviest of the three | Strong buying still active, trend not yet in question |
| Head | Frequently exceeds the left shoulder, though not always | The final push, sometimes on real strength, sometimes on speculative excess |
| Right shoulder | Usually the lightest of the three | Buying interest fading, fewer participants willing to chase a new high |
| Neckline break | Should show a noticeable increase over the days before it | Confirms the reversal has real participation, not just a quiet drift lower |
That declining volume across the three peaks, heaviest on the left shoulder, lighter on the head relative to the size of the move, lightest on the right shoulder, is one of the more reliable tells that a genuine reversal is forming rather than a normal pullback within a continuing uptrend. A right shoulder that forms on volume just as heavy as the left shoulder is a weaker version of the pattern and deserves more skepticism before acting on it.
The neckline is the vital line, not the third peak
This is the part worth repeating until it sticks: seeing the right shoulder form is not a sell signal. It's a reason to start paying closer attention, nothing more.
The temptation is obvious. The pattern looks complete, the third peak has clearly fallen short of the head, and jumping in early feels like getting ahead of the crowd. But price pulling back from the right shoulder can just as easily stop at the neckline and turn back up, continuing the original uptrend, as it can break through. Acting on the shape alone, before the neckline actually breaks, means risking a trade on a pattern that was never technically completed.
The actual signal is a closing break below the neckline, ideally accompanied by a visible increase in volume compared to the days before it. That combination, a real close below the line plus volume confirming it, is what separates a completed head and shoulders reversal from three peaks that happen to look like one.
Ascending vs. descending neckline
Not every neckline is flat, and the slope changes both how the pattern behaves and how much confidence to place in it.
| Neckline type | What it means | Behavior after the break |
|---|---|---|
| Descending (drooping) | Right pullback low sits below the left pullback low | Often a sign of fast-developing weakness; the decline can accelerate quickly once the neckline breaks, sometimes with less volume confirmation than usual |
| Horizontal | Both pullback lows sit at roughly the same level | The most textbook version, generally reliable |
| Ascending | Right pullback low sits above the left pullback low | Usually the clearest, most dependable version, the eventual decline tends to unfold with more time to act, and volume on the break is typically more convincing |
A descending neckline is the trickiest version to trade well. The break can happen on relatively light volume and then accelerate rapidly once the move gets underway, giving less time to react than the more common horizontal or ascending versions. An ascending neckline head and shoulders pattern, by contrast, is generally considered the more dependable setup of the two, since it forms when buyers are still managing to push each pullback low slightly higher even as the overall pattern signals exhaustion, a genuine tug-of-war rather than a one-sided rout. Recognizing this variation matters more than most quick explanations of the pattern let on.
What shoulder width tells you
The relative width, meaning how long each shoulder takes to form, carries information too.
A wide left shoulder that takes longer to form than the right often suggests larger holders were already distributing their position while buying interest was still strong, meaning the eventual rally to the right shoulder tends to be weaker and narrower, since a lot of the available supply has already changed hands into weaker holders.
A wide right shoulder that takes longer to form than the left often suggests larger holders held on longer, only beginning to distribute after the head, with the right shoulder representing a genuine, extended fight between the last buyers and increasingly determined sellers before the trend finally gives way.
Neither version invalidates the pattern. Both are still head and shoulders reversals. The width just adds context about how the decline is likely to behave once it starts.
There's a related, more precisely documented finding worth knowing too. Thomas Bulkowski, a private investor and researcher who has published statistical studies covering thousands of chart pattern examples, found that shoulder height symmetry, not width, actually correlates with performance in a counterintuitive direction: patterns with a distinctly higher left shoulder tend to show larger average declines than patterns with even, textbook-symmetrical shoulders, which came out weakest of the three configurations in his data. In other words, a genuinely asymmetric head and shoulders pattern, one where the two shoulders clearly don't match in height, is not a flaw to be suspicious of. The picture-perfect, symmetrical version taught in most quick guides isn't the best-performing one.
Does it actually work? What the honest numbers say
Search around for how reliable this pattern actually is and you'll find wildly different numbers: some pages claim a 96-99% success rate, others cite failure rates anywhere from roughly 20% to well over 50% depending on the market and time period studied. That range itself is the honest answer, this isn't a pattern with one clean, agreed-upon statistic, and treating any single number as settled fact is a bigger mistake than not knowing the number at all.
A few things worth knowing about where these numbers actually come from:
- The often-repeated 96-99% figure traces back to a single older study that gets copied from page to page without anyone re-verifying it. Treat any number that clean with real skepticism.
- Bulkowski's own published research, based on a large sample of tracked chart patterns, has reported average post-breakout declines in the range of roughly 20-25% for confirmed tops, with failure rates that vary noticeably depending on the broader market environment at the time, generally worse in strong bull markets, when downside reversals have to fight the prevailing trend.
- Some academic research has found the pattern carries real predictive signal in currency markets but weaker, less consistent evidence in individual stocks, which matters if you're applying this specifically to equities rather than forex.
The honest takeaway: this is a genuinely useful pattern with real, if inconsistent, statistical support behind it, not a coin flip and not a guarantee either. Anyone quoting a single precise win rate without naming where it came from is telling you less than they're implying.
Failure modes worth knowing before you trade one
A well-formed head and shoulders pattern still fails to produce a sustained move a meaningful share of the time. Knowing the common ways it fails matters as much as knowing how it's supposed to work.
The neckline retest that keeps going. Price breaks the neckline, drifts back up to retest it from below as expected, but instead of rolling back over, it keeps climbing back above the old right shoulder high. This is the clearest failure signature, and it's exactly what the stop-above-the-right-shoulder placement covered below is designed to catch early.
The gap-through. News or an earnings surprise can send price gapping straight through the neckline with no orderly test at all. The pattern's directional call may still turn out correct, but the clean, low-risk entry the setup normally offers isn't available, and chasing it after the gap changes the risk profile of the trade considerably.
The pattern that never had a real trend to reverse. Three peaks of similar height sitting inside a directionless, choppy range aren't a head and shoulders, there's no established uptrend for the pattern to actually be reversing. This is less a failure of the pattern than a misidentification of it in the first place, and it's the single most common reason beginners report the pattern "not working."
The measuring rule, and its real reliability
Once the neckline breaks, there's a rough rule of thumb for how far the move might carry: measure the vertical distance in price from the head down to the neckline, then project that same distance downward from the point where the neckline broke. That projected level is a reasonable first target, not a guarantee.
Worth being honest about this rule's actual track record: it works reasonably well in a meaningful share of cases, and misses by a wide margin in plenty of others, sometimes stopping well short of the projected target, sometimes blowing straight through it with no pause at all. Treat it as a rough planning tool for where a first partial profit might make sense, not as a precise price target to bet the whole position on.
A walkthrough, start to finish
Here's a hypothetical head and shoulders top on a daily chart, followed candle by candle.
| Day | Price and volume | What it means |
|---|---|---|
| 1-8 | Stock rallies from $60 to a peak of $68, pulls back to $63 on fading volume | Left shoulder forms |
| 9-16 | Stock rallies again to a new high of $72, pulls back to $62.50 on heavy reversal-day volume | Head forms; pullback low sits just below the first pullback low, an early hint of a descending neckline |
| 17-23 | Stock rallies a third time, stalls at $67.50, well short of $72 | Right shoulder forms, on the lightest volume of the three rallies |
| 24 | Price drifts down toward $63 | Approaching the neckline, no signal yet |
| 25 | Closes at $61.80, below the neckline, on volume nearly double the recent average | Neckline confirmed broken, the reversal signal |
| 26-30 | Stock continues lower to $57, roughly matching the head-to-neckline distance | Measuring rule plays out reasonably closely in this case |
The trade isn't day 23, when the right shoulder becomes visually obvious. It's day 25, when the close actually confirms below the neckline with volume behind it. A trader who sold on day 23 would have been sitting through two more days of uncertainty with no confirmation the pattern would actually complete.
The inverse pattern, given equal treatment
Head and shoulders bottoms get mentioned as an afterthought far too often, despite working through identical logic. Here's the same walkthrough, flipped, on a hypothetical stock coming out of a downtrend.
| Day | Price and volume | What it means |
|---|---|---|
| 1-8 | Stock declines from $45 to a low of $38, bounces to $42 on fading volume | Left shoulder forms |
| 9-16 | Stock falls again to a new low of $34, bounces to $42.50 on heavy reversal-day volume | Head forms; bounce high sits just above the first bounce high, an early hint of an ascending neckline |
| 17-23 | Stock falls a third time, holds at $37, well above $34 | Right shoulder forms, typically on lighter volume than the head |
| 24 | Price rallies toward $42 | Approaching the neckline, no signal yet |
| 25 | Closes at $43.10, above the neckline, on volume nearly double the recent average | Neckline confirmed broken, the reversal signal |
| 26-30 | Stock continues higher to $50, roughly matching the head-to-neckline distance | Measuring rule plays out reasonably closely in this case |
Everything that applies to the top applies here in reverse: the neckline is what confirms it, not the visual shape of the right shoulder, and the stop-loss logic simply flips to just below the right shoulder's low instead of above a high.
A documented historical example
The pattern isn't a modern invention. Some of the clearest, most frequently cited examples in early technical analysis literature come from the 1929 market top, when a wave of major industrial stocks formed textbook reversals within weeks of each other.
Republic Steel's 1929 peak is one of the most commonly referenced. The stock formed a left shoulder in late August, a head in mid-September at a marginally higher level, and a right shoulder in early October that fell short of the head, a fairly standard picture. What makes it a useful teaching example is the neckline: it drooped, with the pullback low after the head sitting below the pullback low after the left shoulder, the descending variation covered above. True to that pattern's typical behavior, once the neckline finally broke, the decline that followed was sharp and sustained, part of a downtrend that continued for roughly three years.
The lesson worth taking from a century-old example like this isn't that the pattern is magic, it's that the same mechanics, the declining volume across the three peaks, the significance of a drooping versus horizontal neckline, the danger of acting before confirmation, have held up as a description of crowd behavior for as long as liquid, chart-able markets have existed. The tool is old. That's a point in its favor, not against it, given how many trading fads from any given decade quietly disappear by the next one.
How to actually trade it: entry, stop, and target
Putting the pieces together into an actual trade plan:
Entry. Wait for a daily close below the neckline, not an intraday touch. A candle that dips below the line and closes back above it hasn't confirmed anything yet. If the neckline is sloped, use its value at the specific point where price is currently testing it, not the level from weeks earlier.
Stop loss. A common placement is just above the high of the right shoulder. If price rallies back above that level after a neckline break, the pattern has likely failed, and holding on hoping it resumes the original forecast is usually the more expensive mistake compared to taking a small, defined loss.
Target. Use the measuring rule covered above as a first reference point, the head-to-neckline distance projected downward from the breakout point, while remembering its real-world track record is inconsistent. Many traders take partial profits as price approaches that level and let the remainder run with a trailing stop, rather than treating it as a hard exit.
This structure, confirmed entry, a stop tied to a real invalidation point rather than an arbitrary percentage, and a target treated as a guide rather than gospel, applies just as well to the head and shoulders bottom, simply inverted.
Common mistakes
Acting on the shape before the neckline breaks. This is the single most common error. The right shoulder forming is a reason to watch closely, not a reason to trade.
Ignoring volume entirely. A right shoulder on heavy volume, or a neckline break on light volume, are both weaker versions of the signal than the textbook picture, and deserve more caution than a picture-perfect example would.
Treating the measuring rule as a precise target. It's a rough guide for planning, not a number to bet an entire position's exit on. Plenty of genuine head and shoulders reversals blow past it or stop well short.
Forcing the pattern onto a chart that doesn't really have one. Three peaks of roughly similar height in a genuinely choppy, directionless market isn't automatically a head and shoulders. The pattern means the most after a clear, established trend, since it's describing that trend running out of steam.
Missing the reverse version. Head and shoulders bottoms are just as valid as tops, and the same neckline and volume rules apply, just flipped. Beginners scanning only for tops miss half the pattern's usefulness.
Frequently asked questions
Is a head and shoulders pattern bullish or bearish?
A standard head and shoulders top is bearish, it signals a likely reversal of an existing uptrend once the neckline breaks. The inverse version, a head and shoulders bottom, is bullish, signaling a likely reversal of an existing downtrend. Neither direction is confirmed by the shape alone, in both cases, the signal only counts once price actually closes through the neckline.
How reliable is the head and shoulders pattern?
More reliable than many chart patterns when the full picture, shape, declining volume across the peaks, and a confirmed neckline break, is present. Less reliable when traders act on the shape alone before the neckline actually breaks, which is where most of the pattern's poor reputation among skeptics actually comes from.
What's the difference between a head and shoulders top and a double top?
A double top has two roughly equal peaks with one pullback between them. A head and shoulders top has three peaks, with the middle one higher than the other two. The confirmation logic, waiting for a support line to break rather than acting on the shape alone, applies to both.
Can the pattern fail even after the neckline breaks?
Yes. A neckline break can occasionally be followed by a sharp reversal back above the line, sometimes called a failed pattern or a bull trap in this context. This is part of why position sizing and a stop above the right shoulder matter, no confirmation signal is guaranteed.
Does this pattern work on any timeframe?
The mechanics apply broadly, from intraday charts to weekly charts, though the significance scales with the timeframe. A head and shoulders forming over several months on a weekly chart carries more weight than one forming over a few hours on a 5-minute chart.
How long does a head and shoulders pattern typically take to form?
There's no fixed duration. Larger, more significant reversals tend to take longer to build, weeks to months on a daily chart, while smaller patterns can complete in days. As a rough guide, the more time and price range a pattern consumes in forming, the more significant the resulting move tends to be.
Putting it together
The shape gets all the attention because it's the easiest part to recognize at a glance. The neckline, the volume behind each peak, and the honest limits of the measuring rule are what separate someone who's memorized a picture from someone who can actually trade the pattern with any consistency.
Next time you spot what looks like a head and shoulders forming, resist the urge to act on the right shoulder alone. Mark the neckline, watch the volume on the approach, and wait for an actual close through it before treating the reversal as real.
About the Author — Jamaluddin K.A.
Jamaluddin is the founder of The First Time Investor, a US and UK-focused personal finance and investing education site. He covers stock market fundamentals, risk management, and trading strategy for people who want to invest confidently without a finance degree.
Disclosure: This article is for educational purposes only and does not constitute financial advice. Trading on margin involves significant risk, including the potential loss of more money than initially invested. Consult a licensed financial advisor before making investment decisions.