Skip to main content

Head and shoulders, double top/bottom: reversal patterns

The two most emblematic reversal patterns. Learn to recognize them, validate them, and avoid being trapped by false positives.

Among the dozens of chart patterns catalogued in technical analysis books, two stand out for their reliability: the head and shoulders and the double top (or double bottom). These are reversal patterns — they appear at the end of a trend and signal that a change of direction is likely. Properly identified, they're among the most exploitable setups in TA. Poorly identified, they're the number one source of false reversal signals.

Head and shoulders (H&S)

Anatomy

Head and shoulders consists of three successive peaks, with the second (the head) higher than the other two (the shoulders), which are roughly at the same level. The horizontal (or nearly horizontal) line connecting the two troughs between the peaks is called the neckline.

Neckline L. Shoulder Head R. Shoulder Projected target

After the second shoulder forms, price breaks the neckline downward. This is the signal of the end of the bullish trend and potentially the beginning of a significant decline.

The inverted version — inverted head and shoulders — appears at the end of a downtrend: two troughs (shoulders), a deeper trough in the middle (head), and a neckline that breaks upward.

Why it works

Head and shoulders tells a coherent psychological story. In an uptrend:

  1. Left shoulder: classic bullish push, everything's fine.
  2. The trough: normal pullback.
  3. The head: new push that exceeds the left shoulder — the trend is "even stronger".
  4. The second trough: pullback, but this time it returns only to the neckline. There are fewer buyers.
  5. Right shoulder: new attempt to push, but it fails to exceed the head. Momentum is gone.
  6. Neckline break: buyers have capitulated, sellers take control.

This sequence reflects a progressive weakening of buying pressure. This is what makes the pattern predictive.

Validation and pitfalls

A head and shoulders is only valid after the neckline breaks. As long as the neckline holds, it's just a "potential pattern" — and many never break at all (the market resumes its bullish trend). Never trade a head and shoulders before confirmed break.

Criteria for a reliable head and shoulders:

  • Clear bullish trend before the pattern (no H&S in a range);
  • Three distinct, well-formed peaks, not a confusing mass;
  • Reasonable symmetry: the shoulders are roughly at the same level (not exactly, but close);
  • Horizontal or slightly sloping neckline (very sloped versions are less reliable);
  • Neckline break with volume;
  • Failed retest (bonus): the neckline acts as resistance after the break.

Target projection

The "theoretical" target for head and shoulders is obtained by measuring the vertical distance between the head and neckline, then projecting it downward from the breakout point (or upward for the inverted version).

This target is reached in roughly 60-70% of cases when the pattern is well-formed. It's not a certainty — it's a statistical expectation that justifies a realistic intermediate target.

Double top

Anatomy

Simpler than head and shoulders: two peaks at roughly the same level, with a trough between them. The horizontal line passing through the trough is sometimes also called the neckline.

The sequence:

  1. Price rises and forms a peak.
  2. Pullback.
  3. Price rises again, but stops at the same level as the first peak.
  4. Pullback again toward the previous trough level.
  5. Neckline break downward → double top validation.

Double bottom is the inverse: two troughs at the same level at the end of a downtrend, followed by a break above the intermediate resistance.

Why it works

Same logic as head and shoulders but simpler: price has tried twice to exceed a level and failed twice. The buyers supporting the trend progressively exhaust themselves. When the intermediate trough gives way, it's capitulation.

Validation criteria

A good double top has:

  • Two clear peaks at the same level (±1-2%, no more);
  • A sufficiently deep trough between them (at least 3-5%);
  • Clear bullish trend before;
  • Confirmed break of the trough with volume;
  • Ideally, an RSI or MACD divergence on the second peak (sign of momentum exhaustion).

The more of these criteria are met, the more reliable the pattern.

Projected target

Classic target: distance between the peaks and the trough, projected downward from the breakout point. Same logic as head and shoulders, same order of magnitude of reliability (60-70%).

Common pitfalls

Pitfall #1: seeing patterns everywhere. Once you know head and shoulders, you start seeing it on every chart. Except most are "almost head and shoulders" that lack the validation criteria. Be strict: as long as 5 criteria aren't met, it's not the right pattern.

Pitfall #2: trading before the break. "I see the pattern forming, I'll get ahead of it." Bad idea. In many cases, the pattern never breaks the neckline — price resumes its bullish trend, and your counter-trend position gets crushed.

Pitfall #3: ignoring the higher trend. A head and shoulders on 1h in a strong bullish trend on 1D has a good chance of being just consolidation. Reversal patterns work better when they align with a dominant trend above.

Pitfall #4: forgetting volume. A break without volume is suspicious. Statistically, patterns that break with massive volume go much further than those that break on "background noise".

Pitfall #5: waiting too long. After the break, there's often a retest of the neckline — that's the good entry point. If you wait for too long confirmation (5 candles, 10 candles), you enter too late and your risk/reward becomes mediocre.

False breakouts

A significant proportion of head and shoulders or double top breaks are false breakouts: price breaks briefly, then returns into the pattern. A few ways to filter them:

  • Wait for the close of a candle clearly on the other side (not just a wick).
  • Check the volume of the breakout candle: low volume = caution.
  • Use the retest: entering on a successful retest rather than on the break itself filters out many false signals.

Accept that some bad trades slip through your filter — it's inevitable. The important thing is to limit their size with a properly placed stop, not to seek a perfect filter.

In DYOR

DYOR automatically detects some of these patterns on the detailed coin view. Always verify visually — automatic detection is based on heuristics, it's never perfect. Your eye is often better than any algorithm at judging if a pattern is "clean".

To go further

Related reads