Head and shoulders pattern: neckline, confirmation and measured move

Recognize the structure before naming it

A head-and-shoulders top is a possible reversal pattern with three peaks that can form after an advance. The center high is the head; the outer highs are the shoulders, and the intervening lows define a neckline.

Shoulder size can vary, and the neckline can slope with the two reaction lows. Use the swing sequence to separate this candidate from similar shapes:

The inverse head and shoulders is the upside counterpart: three reaction lows, with the middle low deepest, and a neckline drawn through the intervening highs. Its structure is similar, but its expected direction is opposite. This page focuses on the top formation.

Define the neckline and confirmation rule

Draw the neckline through the two reaction lows that separate the left shoulder, head and right shoulder. Extend that line to the right. If it slopes, its value changes with time: measure the line at the candle where you evaluate the break rather than reusing the first trough’s price as a fixed level.

A chart can trade below the neckline intrabar and close back above it. Decide whether your rule is an intrabar breach, a candle close below, or a close followed by a failed retest from underneath. A close-based rule waits longer and can produce a worse entry price; an intrabar rule can trigger on a move that is quickly reversed. Neither is universally superior. Compare them on the same market, interval, sample and execution assumptions.

A reclaim of the neckline after a downside break is evidence against immediate continuation on that timeframe. A close above the right shoulder or head is a stronger structural invalidation, although a trader’s actual stop may need to be closer to control risk. Distinguish pattern invalidation from a stop: the former describes the setup; the latter is the exit level specified by a trading plan.

Volume can describe how active the move is, but it does not confirm the pattern by itself. Classic descriptions often look for stronger activity on the left shoulder and weaker activity on the right, with expanding activity on a downside break. In crypto, visible volume usually belongs to a particular spot or derivatives venue and can vary by feed. Check the market and volume definition; do not treat a volume shape as a requirement or a probability.

Work a measured-move example

Use a simple synthetic example. Suppose the head reaches 110, and the neckline directly below the head is at 100. The vertical pattern height is 110 − 100 = 10. At a later point the neckline is at 100 and price closes below it at 99. Projecting the 10-unit height down from the neckline at the break gives a rough measured level near 90; projecting from the actual 99 close gives 89. State which convention your method uses and keep it consistent. These levels are arithmetic illustrations, not forecasts.

Before acting on a projection, compare it with visible support, prior lows, volatility and the distance to invalidation. If a nearby support area sits at 94, a target near 90 may be less reachable than the textbook geometry suggests. A measured move does not say how often price reaches the target, how long the move takes or whether an entry can survive an intervening rally.

An example plan can be described without presenting it as advice: wait for the selected break rule, decide whether to enter at the close or after a retest, place an invalidation level based on the structure, and calculate size from the distance between entry and that exit. Account for fees, spread, slippage and—on perpetual contracts—funding. A wider stop with the same position size increases the potential loss; a distant measured target does not offset that risk automatically. See the crypto stop-loss guide for the position-sizing calculation.

What the published statistics do—and do not—say

Study scope: the figures in this section come from ThePatternSite’s bull-market results for historical stocks. Bull market describes the source’s market regime; it does not mean that every individual pattern or breakout was bullish. These figures do not combine all market regimes.

The current ThePatternSite page for head-and-shoulders tops reports more than 2,800 historical “perfect trades,” a 19% break-even failure rate, an average decline of 16%, a 68% pullback rate and 51% meeting its price target. The site defines break-even failure as not moving at least 5% in the expected direction after breakout. The 16% is a sample average decline; the 51% is the separate measure-rule target endpoint.

These values describe historical stock patterns and the site's study definitions. The summary is not a crypto-specific sample, not limited to a stated cryptocurrency timeframe, and not net of a trader’s entry, stop, costs or leverage. A 19% break-even failure rate therefore cannot be rewritten as an 81% chance to make money. Price might fall 5% and then reverse before reaching a planned target. In another case, it might reach the target only after first triggering the trader’s stop.

For your own review, record every candidate using the same detection and confirmation rule, including those that fail. Specify instrument, exchange, candle interval, date range, entry, stop, target, evaluation horizon, fees and funding. Record separately whether price moved a chosen amount in the expected direction, reached the target or hit the stop first. Also record the largest move against the position (maximum adverse excursion) and the actual return after costs. A collection of selected screenshots cannot estimate those rates.

altFINS lists 82% for head and shoulders and defines guide-wide success as hitting a forecast target within one pattern length. This differs from the 81% of historical stock cases that moved at least 5% after breakout; the guide does not state the sample size or study period for head and shoulders.

Common failures and practical tradeoffs

The right shoulder never breaks the neckline. The pattern is not confirmed by its appearance alone. Price can rise through the right shoulder and invalidate the bearish interpretation. Avoid treating an unfinished shape as a completed signal.

A brief break is quickly reclaimed. A wick or candle close under the line can be followed by a move back above it. Decide beforehand whether a retest is part of the rule, and record how many missed moves or false breaks that extra condition creates.

The pattern points into nearby support. The geometry may imply a large measured move, but a prior swing low or another observed support level can interrupt it. Map the surrounding chart before assigning a target.

A small chart pattern conflicts with the broader market. A 5-minute head and shoulders inside a strong 4-hour advance may be a short pullback rather than a durable trend reversal. Higher intervals can add context, but their candles summarize overlapping trades and are not independent confirmation. Use completed higher-timeframe bars when your rule relies on their close; the multi-timeframe guide explains the tradeoff.

The break-even statistic does not solve these decisions. The best confirmation rule depends on the trading horizon and the cost of false signals versus late entries. A pattern label organizes observations; it does not make the uncertainty disappear.

Frequently asked questions