Failed Head and Shoulders Pattern: Why It Happens and What the Data Shows

A failed head and shoulders pattern is a head and shoulders top that never completes its bearish reversal: instead of a break below the neckline, price turns back up and closes above the right shoulder. The sellers who shorted the pattern are trapped, and classical chart analysis reads the move that follows as a bullish signal.

This guide explains how the failed pattern forms, why it happens, how traders identify and trade it, and what our research shows for trading it. We scanned full price history on stocks, crypto, forex, ETFs, indices and commodities with one fixed rule for head and shoulders patterns that failed this way, and compared what came next with random dates on the same assets. The short answer: the failure is common enough to measure, and on its own it did not beat chance in any market or timeframe we tested.

What is a head and shoulders pattern?

The head and shoulders pattern is one of the classic chart patterns of technical analysis. It forms after an uptrend and is considered a bearish reversal: the rise loses momentum and the market turns down.

The four parts of the pattern

• Left shoulder: a first top, then a pullback to a first low.
• Head: the highest point of the pattern, then a second pullback to the lowest area of the neckline.
• Right shoulder: the last of the three peaks, lower than the head, roughly level with the left one.
• Neckline: the line drawn through the two lows (troughs) between the peaks. It acts as support.

The top is confirmed when price closes below that line. The classic measured objective projects the height of the head below the breakout level. The mirror image after a downtrend is the inverse head and shoulders, a bullish reversal that forms after a downtrend.

How our bench defines the pattern

Our test bench uses fixed rules: after a rise, a head above two shoulders of similar height, with a limit on how unequal the two sides may be in time. It counts two levels of strictness: a textbook pattern, with the head at least 25% of its height above both shoulders and the two shoulders within 20% of the head height of each other, and a standard level, with 10% and 30%. Most occurrences come from the standard level: 5,382 of the 5,567 daily failures, and 3,064 of the 3,184 inverse ones. The pattern is confirmed by a close below the neckline. On the daily chart, 61.0% of the patterns it found were confirmed this way, a confirmation rate that reminds us many tops never complete.

What is a failed head and shoulders pattern?

In its simplest form, a failed head and shoulders pattern is a complete head and shoulders top whose expected outcome does not happen. Technical analysis books and articles describe two versions of the failed pattern.

Failure before the breakout: price clears the right shoulder

The pattern is complete, every trader can see the three peaks, but price never ends a session below the two troughs. It rallies instead and closes above the right shoulder, which acted as resistance. The lower top is gone, and the top is canceled. This is the version our bench measures: the signal is recorded at that close, as a bullish one.

Failure after the breakdown: price reclaims the neckline

Price makes a break below the neckline, short sellers enter, and then price turns around and gets back above the broken level. This is the failed breakout, or bear trap. We have not measured this second version yet: the numbers on this page cover the first one only.

The psychology behind the failed breakout

The head and shoulders is well known, so it invites early positioning: selling the last rally, or placing a short order just under the neckline with a protective order above the last top. When price rises through that resistance instead, those orders are triggered and the sellers buy back. That forced buying can add fuel to the rally. The story is plausible; whether it creates a measurable edge is a question we test below.

Why do head and shoulders patterns fail?

No chart pattern will work every time, and the reasons for a failure become clear only afterwards. Four causes are worth checking.

Lack of volume confirmation at the breakout

Classic chart reading wants rising volume on the break of the neckline to confirm the top. In that reading, a break on low volume suggests few sellers are committed, and price can recover quickly. Our numbers do not settle it: on hourly crypto, 3,736 failures printed on volume above 1.5 times the average and 2,126 on normal volume, with close results in a descriptive split (see the reliability section).

Market context: trading against the primary trend

Trading a head and shoulders top inside a strong long-term uptrend is a bet against the main move. Of the 7,050 hourly failures we recorded, 6,510 formed above the 200-period moving average: failures happen where the larger trend is still up.

Stop runs around the pattern's levels

In the classical reading, the levels around a well-known pattern attract orders: exits above the last top, sell orders under the base. In that reading, a move through one of those levels triggers many of these orders at once, which can produce a quick spike that reverses straight away; we have not measured how often it happens.

The impact of major news events

An earnings report, a central bank decision or a large crypto listing can move price more than any chart pattern; we have not tested news days. A head and shoulders is a picture of past supply and demand; it does not know about tomorrow's news.

How to identify a failed head and shoulders pattern in real time

A failed pattern is only certain after the fact, but a few key points help you find it early.

First: the pattern was complete

Second: price ends above the right shoulder

Third: the failed retest of the neckline

In the second version, look for a break below the neckline and then a return above it. A retest from below that fails to stay under it is the first warning for short sellers.

Divergence indicators: RSI and MACD

Momentum indicators can be read at the head, the second of the three peaks. In classical reading, a bearish divergence (the RSI or MACD makes a lower high while price makes a higher high) supports the top. If momentum also makes a higher high there, the top is weaker. We have not tested these signals.

Candlestick formations at the support level

Bullish candles at that level, such as a hammer or a bullish engulfing candle, are read in classical analysis as buyers defending that level. They are warning signals, not proof, and we have not tested them here.

The failed inverse head and shoulders pattern

A related setup works in reverse. A failed inverse head and shoulders is an inverse head and shoulders bottom, formed after a downtrend, where price ends below the low of its last shoulder before getting above the neckline. The expected bullish move does not happen, buyers are trapped, and the pattern points down. Our bench measures the failed inverse head and shoulders with the mirror rule. Inverse patterns and their failures follow the same logic as the top: the inverse version simply flips every level, and its results are in the table below.

How to trade a failed head and shoulders pattern

The classic trading strategy buys the failed pattern: a long entry when price closes above the right shoulder, a stop loss below the neckline, an objective above the head. Here is the trade in detail, and what it did on our bench.

Entry tactics: catching the reversal move

The entry is the close above the right shoulder, the third peak. An aggressive entry buys as soon as it happens. A conservative one waits for a pullback toward the old resistance, which now acts as support, but misses the fastest rallies.

Managing risk: where to place your stop loss

The logical exit is below the neckline: if price falls through it, the head and shoulders top is valid after all. The problem is how far away it is. In the three daily crypto cases written up below, that level sat between 7.8% and 45.6% below the entry. A level that far away forces a very small position, so the money at risk stays the same. A tighter stop under the old top is another choice; we have not tested that version.

Setting profit targets for a failed pattern

One target adds the height of the head above the breakout level, the mirror of the measured move. Our simulation uses an objective at twice the risk (a 2:1 ratio) and ends the trade after 100 candles if neither level is hit. Because that distance is so large, the objective is far away: on the hourly chart, half of the positions never got further than 0.75 R in their favor.

Successful vs. failed head and shoulders patterns

Volume profiles: comparing the two scenarios

In the textbook top, volume is high on the left shoulder, lower on the head and lighter still on the third high, then expands on the breakout below the neckline. In a failure, the expansion comes on the way up, through the old resistance. Our breakdowns, descriptive only, do not show that this volume makes the move up clearly more reliable.

Price action characteristics after the failure

Failed patterns can be slow. On the daily timeframe (5,478 trades), winners took a median of 50 sessions to reach the target and losers 28 sessions to hit the stop, and 40.0% of trades reached neither level and were closed at the end of the 100-day window, tying up money for months.

Is the failed head and shoulders pattern reliable?

No. On daily and hourly charts alike, buying a failed head and shoulders did no better than buying on random dates of the same assets. Here are the main results of our statistical analysis.

How to read the table

Compare each pair of columns. On the daily chart, 53.7% of failed tops were followed by a higher price 10 days later, against 54.3% of random days on the same assets. The excess return over the rest of the market after 10 days was −0.21%, with a 95% confidence interval from −0.39% to −0.03%: zero sits just outside, on the wrong side, so the failures rose slightly less than their market. The statistics section rates the row Weak, below chance: the gap passes a simple test but not our correction for multiple testing. The 2:1 setup reached its target in 21.0% of cases, against 22.3% for the same entry, stop and target distances replayed on random dates, rated no clear edge. Expectancy is in R, the initial risk, before fees and slippage.

On the hourly timeframe, where we recorded the most failures, the failure moved up less often than random hours: 47.2% against 48.8%, with an expectancy of +0.03 R against +0.03 R for the random version. The statistics section still rates the row No clear edge: the gap does not pass our tests, so it is not an established effect. Hourly forex, with 1,188 cases, reads no clear edge too, and the failed inverse version also lagged random hours in raw numbers, rated no clear edge: 47.1% moved down against 50.6%. The daily sample is large: 5,567 cases since 1971, on 1,557 assets.

Statistical performance across different asset classes

These splits are descriptive and not corrected for multiple testing:

• Height: on the hourly timeframe, tall patterns (above the median) moved up 46.4% of the time with a 2:1 hit rate of 18.9%, against 48.0% and 20.2% for short ones.
• Volume: on hourly crypto, the 3,736 failures on volume above 1.5 times the average moved up 46.3% of the time (expectancy +0.05 R), against 47.2% (+0.02 R) for the 2,126 on normal volume. The gap is small, and neither group stands out.
• Stability: the hourly excess return was −0.07% in the first half of the history and −0.04% in the second.

How rare is the failure?

Under our strict rule, less common than the top itself: one occurrence every 1,523 eligible daily candles, against one every 1,145 for the confirmed head and shoulders top. Some weekly rows, on smaller markets, rest on fewer than 30 occurrences, which is too few for a verdict; we do not draw conclusions from them. The full method is in our methodology.

What earlier research found

Automated detection of this top is not new. Andrew Lo, Harry Mamaysky and Jiang Wang, in "Foundations of Technical Analysis" (Journal of Finance, 2000), detected the head and shoulders and other shapes by algorithm on US stocks and found that they carry some information about later returns. They did not isolate failures. Our bench adds three things: six asset classes up to 2026, a comparison of every setup with the same setup on random dates, and a correction for the thousands of tests we run.

Failed head and shoulders case studies

Here are three daily crypto cases, with their actual outcome. None of the three reached its target.

Case study: a failure that collapsed

ETC/USDT formed a head and shoulders top on its daily chart, with a head at 39.67 on March 5, 2024 and a right shoulder at 34.96 on March 29, 2024. On April 8, 2024, it closed at 35.65, above that shoulder, before any close below the neckline. The stop below the neckline was at 28.45, 20.2% under the entry, and the 2:1 target at 50.04. Price never traded above the signal day's high: it fell to 26.78 on April 12 and hit the stop, a full 1 R loss, and was 26.7% lower 10 days after the signal.

Case study: stopped within eight sessions

ADA/USDT put in its right shoulder at 0.3807 on May 17, 2023, below a head at 0.4617, and closed above it at 0.3828 on May 28, 2023. The stop sat at 0.3530, 7.8% under the entry, with a 2:1 target at 0.4424. Price never got above 0.3863. It hit the stop on June 5, 2023: a full 1 R loss, and was 15.8% lower 10 days after the signal.

Case study: the slow drift

XRP/USDT closed above its right shoulder (2.8054, February 16, 2025) at 2.9396 on March 2, 2025, below a head at 3.40. The neckline was so far below that the stop sat at 1.598, 45.6% under the entry, and the 2:1 target at 5.623. XRP was 23.8% lower 10 days later, yet the stop was never hit: the trade was closed after 100 days, in June 2025, for a result of −0.47 R. A wide stop limits the damage per unit of risk, but it also makes the target almost unreachable.

Common mistakes to avoid

• Entering short before the break. Until the breakout, the top is only a potential reversal. Acting early is the error the confirmation rule prevents.
• Treating every rally after a top as a failure. A bounce from the support is not a close above the right shoulder.
• Sizing on the candle, not on the stop. The stop can be far away, up to 45.6% below the entry in our three daily crypto cases; size the position on that gap.
• Trusting volume blindly. In our hourly crypto sample, high volume on the failure did not come with clearly stronger results.
• Ignoring the trading costs. Our expectancy figures are before fees and slippage; long holding periods add financing costs on leveraged trading accounts.
• Remembering only the squeezes. The rallies that hit the target are memorable; the slow losses are forgotten. Count every case, as our bench does, to see whether the setup can work for you.

Tools and trading strategies for spotting failed reversals

Technical analysis tools such as TradingView let you draw the neckline and set alerts on a close above the right shoulder or under it. Scanners differ on how they define the peaks, so two tools can flag different patterns on the same chart. Day trading and swing trading strategies both rely on the same check. Before entering with money, mark the last 30 head and shoulders tops on your market, write down which ones failed and what happened next, and compare with random entries. That small exercise is the same test we run at scale, and it is a useful investment of time before any investment of capital.

Frequently asked questions about failed head and shoulders patterns

Summary of key points

• A failed head and shoulders pattern is a complete top where price closes above the right shoulder before any break lower.
• It is less common than the top: 5,567 daily cases on our bench against 7,405 confirmed head and shoulders tops.
• On daily charts, 53.7% were higher 10 days later, against 54.3% for random dates on the same assets: weak, below chance.
• On those daily cases, the 2:1 setup reached its target 21.0% of the time against 22.3% at random, on 5,478 trades: no clear edge.
• On its own, the failed pattern was not reliable: it did not beat chance in any market or timeframe we tested.

If you trade it, treat the failure as context, not as a signal on its own: define the stop before entering, size the position on it, and keep the risk per trade small. This page is educational material, not investment advice: past results on historical prices do not guarantee future ones, and you can lose capital.

About the author and the data. Written by Amaury, founder of PatternsWizard, who is passionate about finance, quantitative trading and data analysis. He built the PatternsWizard test bench, which detects each pattern with fixed, published rules on full market history: crypto, forex, US stocks, ETFs, indices and commodities. Trades and win rates are compared with random bars on the same assets, excess returns with the rest of the market, with error margins and a correction for multiple testing, and the whole pipeline is checked on random-walk prices, where nothing can be predicted. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.