Head and Shoulders — How to Spot and Trade the Reversal

Head and shoulders (H&S) is probably the most recognizable pattern in technical analysis. Three peaks, the middle one highest, a neckline — every TA course starts with this drawing. The problem is that most material stops at the drawing. We'll go further: we'll show what Thomas Bulkowski's measurements across thousands of stock-market cases say about H&S, when the pattern actually matters, what a sensible entry looks like, and why "perfect" symmetry isn't actually a virtue.

Short version: H&S is one of the most effective top-reversal patterns, and in a bear market its stats get downright indecent — an average decline of 29% with just 1% failures. But only if you wait for confirmation.

How to Recognize the Head and Shoulders Pattern

H&S is an uptrend-reversal pattern. It's made up of three peaks:

• Left shoulder — a peak within an ongoing uptrend, followed by a pullback.
• Head — the next, HIGHER peak. The market prints a new high, but demand starts to fade.
• Right shoulder — a third peak, clearly LOWER than the head, usually near the level of the left shoulder. This is the first hard evidence that buyers can no longer push price to new highs.

We connect the troughs between the peaks with a line — the neckline. It can be horizontal, descending, or ascending.

Validity conditions worth taking seriously:

1. There must be an uptrend to reverse. An H&S drawn in consolidation, or after a short, shallow rise, has nothing to reverse — and as you'll see in the stats, the size of the prior move matters.
2. Proportions. The head clearly above the shoulders; the shoulders similar to each other in height and in distance from the head. But — careful — don't overdo the search for an ideal shape (more on that shortly).
3. Volume classically peaks on the left shoulder or the head and clearly fades on the right shoulder. Weakening volume on the right shoulder is the signature of exhausting demand.
4. Confirmation. The pattern only becomes valid once price closes below the neckline. Before that, you have three bumps on the chart, not a pattern. This is the most commonly ignored condition — and the most common source of losses "on H&S."

What the Numbers Say (Bulkowski, Encyclopedia of Chart Patterns)

Thomas Bulkowski studied chart patterns across tens of thousands of cases from US stocks (daily timeframe) and published the results in the "Encyclopedia of Chart Patterns." Before we get to the numbers, four terms he uses — without them the stats float in a vacuum:

• Break-even failure rate — how often, after confirmation, price failed to move even 5% in the breakout direction. In other words, how often the pattern "didn't fire."
• Average move (average decline/rise) — the average distance from the breakout price to the extreme of the move (the so-called ultimate low/high), measured on idealized, perfect trades. It's inflated relative to what a real trader would actually capture — treat it as a comparison of pattern strength, not a profit promise.
• Target % — how often price reaches the target set by the measure rule (more on that rule in the trading section).
• Throwback / pullback — price returning to the breakout level within 30 days. Throwback applies to upward breakouts, pullback to downward ones. With H&S we care about the pullback: price breaks the neckline, drops, then comes back to "say goodbye" to the neckline from below — and only then continues down. This behavior is common and has two consequences: it gives you a second entry chance, but patterns with a pullback statistically perform worse afterward.

Now the specifics for H&S:

Two things in this table deserve comment.

First — bear markets. An average decline of 29% with 1% failures is among the best statistics in the entire catalog of bearish patterns. An H&S confirmed in an ongoing bear market is a case where the pattern trades WITH the current, not against it. Practical takeaway: the same pattern has different value depending on the surrounding market environment.

Second — symmetry is overrated. Bulkowski's research on symmetry showed that the more asymmetric the H&S (different distances of the shoulders from the head), the BETTER the result after the breakout. Bulkowski himself summed it up with the phrase "ugly patterns work best." On top of that, a higher left shoulder outperforms equal shoulders. So if you're rejecting setups because "the shoulders aren't perfectly equal" — you're rejecting the statistically better cases. (Fun fact: with the inverse H&S it's exactly the opposite — there, symmetry helps.)

The caveat we repeat with every pattern: these are stats from US stocks on the daily timeframe, with a bull-market bias in the sample. On BTC, ETH or indices the pattern mechanics work the same way, but the specific numbers may differ — nobody has reliably measured them there on a sample this size.

How to Trade H&S

Entry. Base variant: short on a candle close below the neckline (with a descending neckline, Bulkowski says to watch for a break of the line itself; with an ascending one — the low of the right-shoulder trough). Second variant, often with a better risk-reward: wait for a pullback to the neckline from below and enter when price gets rejected. Compromise: half a position on the break, half on the pullback — the pullback might not come.

Stop loss. Logical placements: above the right shoulder's peak (conservative) or above the last local high before the break (more aggressive). A stop above the head is usually too far away — it wrecks the risk-reward ratio to the point that you're better off skipping the setup.

Target — the measure rule. Measure the pattern's height: from the top of the head straight down to the neckline. Subtract that height from the neckline break point — that's the full target. Bulkowski recommends being realistic about it: multiply the height by the target-achievement percentage before subtracting it, since only some patterns reach the full range. In practice it's also worth checking whether strong support sits along the way to the target — if so, that's the first real target.

What to avoid.

• Entering before confirmation. Three bumps without a close below the neckline is not a pattern.
• An H&S after a tiny rise — price needs something to fall from. A small move before the pattern means a small move after it.
• Ignoring context: an H&S against a strong higher-order uptrend is swimming upstream.

Example from crypto: on BTC's daily chart, H&S patterns have preceded several major local tops, but just as often "textbook" setups on the H1 or H4 got broken to the upside instead. The lower the timeframe, the more noise — and the less these statistics have to do with reality, since Bulkowski's measurements refer to the daily chart.

Busted H&S — When the Pattern Fails

A "busted" pattern is one that, after confirmation, moves less than 10% in the breakout direction, then reverses and closes on the opposite side of the pattern. For H&S: price breaks the neckline, drops a token amount, comes back, and closes ABOVE the top of the pattern (the head).

Why does this matter? Because a broken bearish pattern is often fuel for the bulls — everyone who shorted the neckline break has to bail, and their stops above the pattern turn into demand. Bulkowski treats busted patterns as a separate signal class in the third edition of the Encyclopedia; for H&S's cousins, double tops, he measured that the move after a busted pattern can be larger than the move from a "successful" one.

In practice: if your H&S short got stopped out because price came back above the pattern, don't treat that purely as a loss. It's information. A close above the head after a failed downside break signals demand strength and, for some traders, stands as a standalone long setup.

And a myth to bust before we close: "H&S is the surest pattern, it almost always works." No. H&S is good — especially in a bear market — but we're still talking about a statistical edge, not a certainty. Some patterns will fail, some will stall halfway to target, and throwbacks and pullbacks will regularly stop out overly tight stops. The pattern only gives you an edge when risk management stands behind it.

FAQ

Does the head and shoulders pattern work on crypto? Bulkowski's statistics come from US stocks on the daily timeframe, so on BTC or ETH treat them as an approximation, not a guarantee. The pattern's mechanics are universal — demand exhaustion looks similar on every market — but the specific numbers (22%, 29%, 1% failures) were measured elsewhere.

When is the pattern confirmed? When price closes below the neckline. Before that it's just three bumps. Shorting "because the right shoulder is already visible" is guessing — the statistics apply to CONFIRMED patterns.

Do the shoulders have to be symmetrical? No. Measurements show that asymmetric H&S patterns perform better after the breakout than perfectly even ones, and a higher left shoulder is statistically the best variant (-25% versus -19% for equal shoulders). The textbook ideal is not the measured optimum.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero