Butterfly Pattern Trading: Rules, Ratios and How to Trade It

The butterfly pattern is a harmonic reversal pattern with five points, X, A, B, C and D, in which point B is a 78.6% retracement of the first leg and point D completes beyond point X, at the 127.2% extension of XA. Traders use it to identify the end of a move that has just made a new extreme: a bullish butterfly forms after a fall to a new low, a bearish one after a rally to a new high. The trade is taken in a potential reversal zone calculated in advance with Fibonacci tools, based on the length of each leg.

This guide covers butterfly pattern trading from start to finish: how to identify the pattern, the exact Fibonacci ratios, a worked example with numbers, how the butterfly differs from the Gartley and the Bat, entry, stop-loss and target placement, three trading strategies, and what we can honestly say about its reliability. We tested it on our bench, and we give the results where they matter. This page is about the harmonic chart pattern, not the butterfly spread used in options trading.

What is the butterfly pattern in harmonic trading?

The butterfly is one of the harmonic patterns, a family of chart patterns in which each swing must respect specific Fibonacci ratios. Harmonic patterns in technical analysis are based on the idea that financial markets move in waves and that corrections tend to end near Fibonacci retracement and extension levels, on the view that traders watching the same levels make them matter. Scott Carney, author of The Harmonic Trader (1999), credits Bryce Gilmore with discovering the general structure of the butterfly, and defined the ratios used on this page.

What sets the butterfly apart from other harmonic patterns is where it ends. In the Gartley and the Bat, point D stays inside the XA leg. In the butterfly pattern, point D goes beyond point X: price action makes a new low (or a new high) and then may reverse. That is why it is called an extension pattern, and why, in Carney's reading, it marks the final swing of a strong trend, when the move is losing strength. Unlike a retracement pattern, the butterfly harmonic pattern trades the failure of a breakout.

The anatomy of a butterfly pattern: XABCD structure

The butterfly pattern uses the same five-point structure as every XABCD pattern. Four legs join the five points, and each leg is measured against the one before it. The anatomy chart in this section is the median shape of a random sample of real daily setups from all markets; every setup has the same five points, so its caption's "most common number of swing points" covers all of them.

Defining the XA impulse leg

XA is the first, strong leg. In a bullish butterfly pattern, price rises from point X to point A; in a bearish butterfly pattern, it falls. Every other measure is based on the XA leg, so it should be a clear, long swing that stands out on the chart. A weak or choppy XA leg makes every later ratio meaningless.

Measuring the AB retracement

Price then retraces 78.6% of XA to form point B, the end of the AB leg. This deep retracement is the signature of the butterfly pattern: a point B near 61.8% points to a Gartley, a B at 38.2% to 50% points to a Bat. A precise Fibonacci reading at point B is the first filter.

Identifying the BC retracement

From point B, the BC leg moves back in the direction of A and retraces 38.2% to 88.6% of the AB leg. Point C never goes beyond point A. A C in the middle of that range is the easiest to read; a C very close to A leaves the pattern harder to read.

Pinpointing the CD completion zone

The final leg, CD, is longer than AB and BC: it extends beyond X and ends at point D, where three measures should meet: the 1.272 extension of XA, a 1.618 to 2.24 extension of BC (some traders allow up to 2.618), and an AB=CD measure, with CD about 1.27 times AB in Carney's version. The area where these Fibonacci levels cluster is the potential reversal zone (PRZ), the completion zone of the pattern. Pattern completion happens when price trades into that zone, not before.

Key Fibonacci ratios for the butterfly pattern

The importance of the 1.272 extension

The 1.272 Fibonacci extension of the XA leg is the anchor of the butterfly pattern. It is what makes D a new extreme: price has gone 27.2% further than the whole first leg. Some traders accept D up to the 1.618 extension, which is where the Crab completes, with its own rules. Our bench is stricter than that range: it takes D within 5% of the 1.27 extension (1.207 to 1.334) and leaves deeper D values to the Crab. Both Fibonacci extensions are drawn with the same tool, from X to A, while the Fibonacci retracements at B and C are drawn on the leg before each point.

A worked example with numbers

Illustrative values, not a real chart. In a bullish butterfly pattern, X is at 100 and A at 150, so XA is 50.

• B at a 78.6% retracement of XA: 150 minus 39.3, so B is at 110.7.
• C rebounds to 135. BC is 24.3, which is 61.8% of AB (39.3): inside the 38.2% to 88.6% range.
• D at the 1.272 extension of XA: 150 minus 63.6, so D is at about 86.4, below X.
• Check the BC extension: CD runs from 135 to 86.4, which is 48.6, or 2.0 times BC: inside the 1.618 to 2.24 range.
• Check AB=CD: 1.27 times AB is 49.9, which would put D at 85.1.

The potential reversal zone is therefore about 85.1 to 86.4: a tight cluster of key levels, which is the point of the method. The 1.618 extension of XA, at 69.1, is the level beyond which the setup is no longer a butterfly pattern.

Comparing butterfly vs. Gartley vs. Bat patterns

These harmonic patterns share the XABCD structure. The position of point B and the depth of D tell them apart.

The Butterfly and the Crab are extension patterns; the Gartley and the Bat are retracement patterns. The practical difference is the stop-loss placement. In a Gartley or a Bat, X is a natural stop just beyond D. In a butterfly pattern, X has already been broken when the trade starts, so the stop must go beyond the next extension of XA. The trade buys a new low (or sells a new high), which feels wrong and is the main reason the butterfly needs strict rules.

How to identify and trade the butterfly pattern

Step-by-step identification process

To identify a butterfly pattern on a chart, work through the points in order. If one ratio fails, stop: it is another pattern, or no pattern at all.

Validating the pattern with confluence

Bullish butterfly pattern: entry and management

Identifying the bullish reversal zone

A bullish butterfly pattern forms after a rise from X to A, a pullback to B, a bounce to C, and a strong drop to D below X. The shape on the chart is an M. Traders who identify a bullish butterfly pattern forming, and then completing, look to buy in the potential reversal zone, prepared to exit if price keeps falling. The signals they watch for are a slowing decline, long lower wicks and a first higher low.

Setting stop-loss and take-profit levels

• Entry: a limit order inside the PRZ, or a buy order once price action shows a reaction (a bullish reversal candle, a close back above the zone).
• Stop-loss: placed below the 1.618 extension of XA, or just below the PRZ for a tighter stop that is hit more often.
• Targets: the 38.2% and 61.8% Fibonacci retracement levels of AD, then B, then A for an aggressive target.

With the numbers above, buying at 86.4 with a stop at 69.1 risks 17.3. The 38.2% retracement of AD (63.6) is at about 110.7, a gain of 24.3, about 1.4 times the risk. The 61.8% retracement is at about 125.7, a gain of 39.3, about 2.3 times the risk. These figures describe the plan, not the odds.

Our bench tests one standardized version of this trade, with the same entry and target rules for every harmonic pattern so that results can be compared, and this pattern's own stop: an entry at the close of the first bar that reaches the reversal zone, a stop at the 1.618 extension of XA and a target at twice the risk. The 61.8% retracement of AD is measured separately.

Bearish butterfly pattern: entry and management

Spotting the bearish reversal signal

The bearish setup is easy to spot once you know the bullish one: turn the chart upside down.

A bearish butterfly pattern is the mirror image: a fall from X to A, a rebound to B at a 78.6% retracement of XA, a dip to C, then a strong rally to D above X. The shape is a W. Traders look to sell at point D, after price has made a new high that may fail. A bearish reversal candle in the zone, or a lower high formed after the first drop, is the confirmation that the conservative entry below waits for.

Risk management strategies for short positions

The stop goes above the 1.618 extension of XA, or just above the PRZ: the point where the bearish pattern is proved wrong. Short positions in stocks can face gaps and borrowing costs, and in crypto a fast squeeze in a period of high volatility can run through a stop: size the trade so that a loss at the stop is a small, fixed share of the account. Targets mirror the bullish case: the 38.2% and 61.8% retracement levels of AD, then B and A.

Butterfly pattern trading strategies

The pattern gives a location in the market; a strategy adds the rules for acting on it. Here are three trading strategies for the butterfly harmonic pattern, from the most aggressive to the most conservative; the right one depends on how much time you can spend watching the market. Our bench tests a standardized form of the first one, entering at the close of the first bar that reaches the zone rather than with a limit order; the other two are not tested. Test any strategy on your own market before you risk money on it.

Strategy 1: the limit order in the PRZ

In a bullish setup, the trader places a limit order at the 1.272 extension of XA, with the stop beyond the 1.618 level, before the pattern completes. Some traders split the order across the zone, between the levels given by the two extensions and the AB=CD measure. This gives the best price and a clear risk, but it buys before any sign of a turn: in a strong trend, price can run through the whole zone.

Strategy 2: the confirmation entry

The trader waits for price action to confirm the reversal in the zone: a reversal candle followed by a close back inside the range, or a break of the short trend line drawn along the CD leg. Fewer signals, a later entry, and a wider distance to the stop, in exchange for a sign that buyers (or sellers) have stepped in. This confirmation approach is meant to filter out some failed reversals; we have not measured how many.

Strategy 3: trading with the higher timeframe trend

The trader only takes bullish butterfly patterns when the higher timeframe trend is up (for example, price above its 200-day moving average) and bearish ones when it is down. The pattern then marks the end of a pullback rather than a reversal against the main direction. On our own butterfly data, the split goes the way this strategy expects, without proof. On daily charts, 319 bullish setups formed above the 200-day average and 1,640 below it; the closed 2:1 trades entered above it reached the objective 42.0% of the time, against 29.8% below it. 237 bearish setups formed below the average and 2,526 above it; the closed 2:1 trades entered below it reached the objective 31.6% of the time, against 23.5% above it. Setups that formed too early in an asset's history to have a 200-day average are in neither group. That split is descriptive: we did not test it against random dates or correct it for multiple testing.

Advanced tools for harmonic pattern recognition

Using indicators for confirmation

Technical indicators can be used to confirm the reversal at point D, especially when the potential reversal zone is wide: a bullish divergence on the RSI (price makes a lower low while the RSI makes a higher low), an RSI leaving oversold territory, or a MACD crossover. Used as confirmation, each indicator delays the entry and removes some failed trades, but also moves the entry further from the stop. We have not tested an RSI or MACD filter on this pattern. In this use, a momentum indicator confirms that the final leg is losing strength; it does not say where the leg ends.

A candle such as the hammer did not beat random dates on its own: on 12,005 closed trades on daily hammers, the 2:1 trade reached its objective 35.4% of the time against 36.1% for the same trade on random dates, and its average result per trade (+0.06 R against +0.09 R) is rated No clear edge, too weak to count as an edge.

The role of volume and momentum

Harmonic traders look for exhaustion at point D: a spike in volume followed by a fast reversal, a jump in volatility, or fading momentum as price makes its final extreme. Our butterfly data gives no clear answer. On daily charts, bullish setups on high volume (above 1.5 times the average) reached the 2:1 objective in 30.6% of trades, against 33.3% on normal volume, while bearish ones did the reverse (24.6% on high volume, against 23.8%). That breakdown is descriptive, not tested against random dates: volume at D is not a filter our data supports.

Drawing and scanning tools

TradingView includes an XABCD drawing tool that shows the Fibonacci ratios as you place the five points on the market chart. Community scripts and paid scanners find harmonic patterns automatically; check which ratios and tolerances each one uses, and test it on past data before you rely on it.

Is the butterfly pattern reliable?

A fair test needs a fixed rule for each swing and for the tolerance around each Fibonacci level, and we set ours from chart reviews, never from results. Turning points are swings of 2, 4 or 8 times the average true range. X must be a true turning point: over as many bars as XA took, before X, price never went more than 10% of XA beyond it. A ratio published as a single value is accepted within 5% of it (0.786 becomes 0.747 to 0.825 for B), and a published range is used as is. Once C is confirmed, the reversal zone is where the 1.27 extension of XA (1.207 to 1.334) overlaps the 1.618 to 2.24 projection of BC. This is narrower than the textbook range up to 1.618 given above: deeper D values are left to the Crab. The signal is the first bar that trades into the zone without trading beyond the 1.618 extension of XA.

On 1,972 closed trades on bullish setups on daily charts, all markets together, the 2:1 trade reached its target 31.8% of the time, against 31.4% for the same trade on random dates. On 2,784 closed trades on bearish setups, it was 24.2% against 23.4%. The 61.8% retracement of AD was reached before the stop, within 100 bars, in 36.4% of bullish daily setups and 28.5% of bearish ones. Those rates are not in the tables: they describe the shape, not an edge.

Trying many versions of the detection rules until one fits would produce a number that means nothing. Be wary of win rates quoted elsewhere without the number of trades, the markets, and a comparison with chance.

What the bench measured on the butterfly

Every occurrence our rule finds on full price history, in crypto, forex, US stocks, ETFs, indices and commodities, compared with the same trade placed on random dates on the same assets.

Read each row in pairs: the setup against the same entry, stop and target placed on random dates. On daily charts, all markets together, neither the butterfly's average excess return nor its average result per trade differed measurably from chance, in either direction. Market by market, no daily row stands out either: bullish setups on daily crypto charts (208 cases) were followed by a higher price 10 days later in 50.2% of cases, against 44.2% after random bars, but their average excess return over the rest of the market is not statistically significant (the row reads No clear edge), and their 2:1 trade reached its target less often than on random dates (15.0% against 22.8%). On shorter charts, the bearish butterfly on 4-hour charts stands out, almost all of it on crypto: 10 bars after the signal, price was lower in 57.1% of cases, against 51.4% after random bars on the same assets, and the average excess return over the rest of the market (+0.86%) survives our correction for multiple testing and holds in both halves of the history. The 2:1 trade did not share it: its average result per trade (0.00 R against −0.06 R on random dates) does not pass our statistical test. On hourly charts, the bearish butterfly also fell more often than random bars: 55.5% of 4,440 setups, all markets together, were lower 10 bars later, against 50.4%, with an average excess return of +0.13% over the rest of the market, and +0.15% on the 3,846 crypto setups. Both rows are rated Promising: the gap passes a simple test but not our correction for multiple testing. The bullish 2:1 trade on hourly charts does not pass our statistical test, on crypto (+0.04 R against +0.01 R on random dates) or on all markets combined. With thousands of tests across markets, timeframes and patterns, one confirmed result and a few that pass a simple test but not our correction, among many null ones, is a lead to watch on the next run, not a strategy. The full methodology explains how we test.

Common mistakes and how to avoid them

With harmonic patterns, the ratios are only half the job. These specific mistakes are easy to make with the butterfly.

• Calling a Gartley a butterfly: if point B is a 61.8% retracement rather than 78.6%, D belongs at 78.6% of XA, inside X. Trading it as a butterfly puts the entry in the wrong place.
• Entering before the zone: buying as soon as price breaks X, before the 1.272 level, means buying a falling market with no structure yet.
• Ignoring the larger trend: a bullish butterfly pattern against a strong downtrend on the higher timeframe is a bet against the main move, and so is a bearish one in a strong uptrend. The move can continue straight through the 1.618 level.
• Moving the stop: once price passes the stop, the pattern has failed. Redrawing it to keep the trade alive turns a small loss into a large one.
• Remembering only the winners: on past charts, the butterfly patterns that worked are easy to spot. Keep a record of every one you would have traded.

Pros and cons of trading the butterfly pattern

Advantages

• Entry, stop and targets are defined before the trade.
• It aims at the extreme of a move, where reversals, when they come, leave room for a large potential gain relative to the risk.
• It gives specific, measurable points, so two traders can identify the same pattern and compare their analysis.
• The Fibonacci ratios are precise, so the pattern can be checked rather than guessed.
• The rules are the same in stock, forex and crypto markets, on every timeframe.

Disadvantages

• It buys a new low or sells a new high: against the short-term trend by design.
• The stop beyond the 1.618 level is wider than in the Gartley or the Bat.
• Perfect butterfly patterns are rare, which makes it tempting to stretch the ratios.
• Our own test found no measurable edge on daily charts, all markets together.

Frequently asked questions

Summary

• The butterfly is a five-point harmonic pattern: B at a 78.6% retracement of XA, C at 38.2% to 88.6% of AB, D at the 1.272 extension of XA, beyond X.
• Traders enter in the potential reversal zone at D, place the stop beyond the 1.618 extension of XA, and target the 38.2% and 61.8% retracements of AD.
• A B at 61.8% makes it a Gartley; a D at the 1.618 extension makes it closer to a Crab.
• Tested on our bench with a fixed rule: no measurable edge on daily charts, all markets together; bearish setups on 4-hour charts moved in the expected direction relative to their market after 10 bars, without improving the 2:1 trade.
• Following the rules strictly, and cutting the loss at the stop, matters more than finding more patterns.

This guide is general information about technical analysis, not financial or investment advice. Trading carries a risk of loss: never trade money you cannot afford to lose.

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 setup 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. The butterfly rule (swings of 2, 4 or 8 times the average true range, ratios within 5%) was set from chart reviews, never tuned on results. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.