Short answer: harmonic patterns are multi-leg reversal shapes, usually labelled X, A, B, C and D. Each leg must land within a fixed band of fibonacci ratios measured against the leg before it. Gartley, bat, butterfly, crab, shark and ABCD all belong to the same family. They differ mostly in how far point B pulls back and where point D completes. Traders use them to mark a reversal zone in advance, place a stop just beyond it, and take profit at ratio-based targets. The pattern is only as good as the ratios that define it, so measurement discipline matters more than memorising shapes.
This is the family overview. Each pattern gets a short paragraph and a table row rather than a full walkthrough. The useful part is the logic they share: the ratios, how to tell them apart, how to validate one, and how the method behaves under prop-firm limits.
What every harmonic pattern has in common
Strip the names away and every harmonic pattern is the same object. Price makes an initial impulse (XA), pulls back (AB), bounces (BC), then pushes past B into a final leg (CD) that finishes at a projected zone called D. The pattern is bullish when the shape ends with a low at D and bearish when it ends with a high. A trader is not buying the shape. They are buying the expectation that D marks exhaustion.
Three features tie the family together.
- Every leg is measured, never eyeballed. B is a retracement of XA, C is a retracement of AB, and D is a retracement or an extension of XA.
- D is defined before price gets there. You know your entry, stop and first target while the last leg is still forming.
- The completion zone is a band, not a line. Two or three ratios usually converge near D, and that overlap is the potential reversal zone.
An ordinary swing high and swing low will scarcely ever fit these rules by accident. That scarcity is the filter. Most five-point shapes on a chart are unremarkable, and a harmonic trader spends far more time rejecting candidates than trading them.
The ratio logic: why fibonacci ratios define each pattern
The method rests on a small set of fibonacci ratios. Retracements use 0.382, 0.5, 0.618, 0.786 and 0.886. Extensions use 1.13, 1.272, 1.618, 2.24, 2.618 and 3.618. Their origin is covered in our guide to fibonacci retracement, which explains how the levels are drawn on a single swing. Harmonic trading applies the same tool three or four times in a row, with each measurement constraining the next.
Here is how the constraints stack. Suppose XA is a 100 pip rally. A gartley requires B to retrace 0.618 of XA, so B sits roughly 62 pips below A. A bat requires B to stop at 0.382 to 0.50, so B sits 38 to 50 pips below A. A butterfly wants B near 0.786, 79 pips down, and then sends D violently beyond X, to 1.27 or 1.618 of XA. The shallow or deep B pullback is the first fork in the road, and D falls out of it.
Neither the 0.886 level nor the 1.13 extension describes anything physical about markets. The first is the square root of 0.786, the second is its reciprocal. Bluntly, they matter because enough traders measure the same way that the zones become crowded, and crowded zones produce reactions.
The main patterns, one paragraph each
Each deserves its own walkthrough, so these summaries are deliberately short and focus on what makes each pattern distinct.
Gartley
The original and most conservative pattern. B retraces 0.618 of XA and D completes at 0.786 of XA, inside the XA range. Because D never breaks X, the stop sits just beyond X and the risk is tidy and easy to size. Gartleys form as deep pullbacks, so the trade is mostly with the larger trend.
Bat
A gartley with a shallower B (0.382 to 0.50) and a deeper D (0.886 of XA). The difference sounds small but is consequential. The 0.886 completion sits so close to X that the stop can be very tight, and the reward-to-risk ratio is often the most generous in the family. Bats are also the least forgiving of sloppy measurement.
Butterfly
B retraces 0.786 of XA, and D extends past X to 1.27 or 1.618 of XA. Butterflies mark new highs or new lows, which makes them reversal calls against the recent trend. The stop goes beyond the 1.618 extension, so risk is wider than in a gartley, and the pattern often ends a lumpy, exhausted move.
Crab
The most extreme extension pattern. B is shallow (0.382 to 0.618) and D stretches to 1.618 of XA, often with the CD leg at 2.24 to 3.618 of BC. Crabs catch sharp overshoots that reverse violently, but they demand patience, because price may decelerate long before the distant zone arrives.
Shark
A newer variant that skips the classic B retracement. It measures C as an extension of AB (1.13 to 1.618) and D at 0.886 to 1.13 of the initial swing. Sharks catch fast reversals after a failed breakout. Their structure is more awkward to label, which is why beginners misread them most often.
ABCD
The humble building block, with no X point. AB and CD are equal in length or related by 1.27 or 1.618, and BC retraces 0.618 to 0.786 of AB. Every larger harmonic pattern contains an ABCD, so learning the simple version quietly trains the eye for the rest.
Comparison table: how to choose between the patterns
| Pattern | B as retracement of XA | D completion | D relative to X | Stop logic | Character |
|---|---|---|---|---|---|
| Gartley | 0.618 | 0.786 of XA | Inside range | Just beyond X | Pullback with the trend |
| Bat | 0.382 to 0.50 | 0.886 of XA | Inside range, near X | Just beyond X, very tight | Deep pullback, tight risk |
| Butterfly | 0.786 | 1.27 to 1.618 of XA | Beyond X | Beyond 1.618 | Counter-trend at fresh extremes |
| Crab | 0.382 to 0.618 | 1.618 of XA | Well beyond X | Beyond 1.618, wider | Overshoot reversal |
| Shark | Not used | 0.886 to 1.13 of the first swing | Around X | Beyond 1.13 | Failed-breakout reversal |
| ABCD | No X point | CD equals AB or 1.27 to 1.618 of AB | Not applicable | Beyond D | Simplest, sits inside the others |
Choosing is easier than it looks, because the chart chooses for you. Measure B first. A B near 0.382 to 0.50 means bat or crab, near 0.618 means gartley, near 0.786 means butterfly. Then watch whether the CD leg stops short of X (gartley or bat) or blows through it (butterfly or crab). If the ratios sit between two patterns, the honest answer is that it is neither. Ambiguity at this step is a reason to pass, and hesitancy here is a virtue.
A second criterion is what you can afford to wait for. A crab completion may sit 60 percent further from the previous swing than a gartley completion on the same chart. That distance is dead time and wobbly price action, and an impatient trader will squander the setup by entering early.
Validation before entry: a checklist
Harmonic trading fails most often at the validation step, not at the measurement step. A pattern can be perfectly proportioned and still lose, so the checklist adds evidence rather than confidence.
- Measure the ratios on closes or on wicks, and never switch between them mid-pattern.
- Require at least two ratios to overlap at D within a few pips. A single ratio is a coarse guess.
- Wait for price to reach the zone. Entering during the CD leg because D looks close is the fastest way to be stopped repeatedly.
- Ask the zone to prove itself. A candle that closes back inside it, or a lower-timeframe higher low, is the minimum.
- Check the wider context. A bullish pattern completing into a strong daily downtrend is precarious, however clean its ratios.
- Skip anything that forms across a scheduled news release, because the CD leg can turn disorderly and the zone gets run through.
The tedious part is that most candidates fail one of these tests. That is fine. Whatever edge the method has comes from the few patterns that pass every test, not from the volume of shapes you spot.
Trading strategies: a worked bullish bat on a 10,000 account
Numbers make the strategy concrete. Assume a forex pair, a 10,000 account, and a rule of risking 1 percent, or 100, per trade.
- X, the swing low, is 1.1000. A, the swing high, is 1.1300, so XA is 300 pips.
- B retraces 0.50 of XA to 1.1150. That fits a bat.
- C bounces to 0.618 of AB, roughly 1.1243.
- D projects to 0.886 of XA, which is 1.1300 minus 266 pips, so about 1.1034.
- The 2.24 projection of BC lands near 1.1035, so two ratios converge within a pip or two. That overlap is the reversal zone.
Entry is a buy at 1.1035 after a rejection candle prints in the zone. The stop goes 50 pips lower at 1.0985, below X with a small buffer. With 100 at risk and a 50 pip stop, the position is 2 per pip, or 0.2 standard lots. The first target is a 0.382 retracement of the AD move, about 100 pips higher at 1.1135, which returns 200. The second target is 0.618 of AD, roughly 165 pips higher at 1.1200, worth 330 on the full position. A common approach is to close half at the first target and move the stop to entry. That roughly halves the open risk while leaving the second target alive.
Each harmonic entry is a single trade with one stop. The spread and commission cost per idea therefore stays modest compared with strategies that scale in and out repeatedly.
Limitations and common mistakes
Harmonic patterns are brittle. Their precision is their attractiveness and also their weakness, because a leg that finishes a few pips outside the band invalidates the whole structure. These are the usual ways the method goes wrong.
- Forcing the ratios. Moving an anchor from a wick to a close so the numbers fit is the most common error, and it turns a rule-based method into a fickle guess.
- Trading the shape without the reaction. A completed zone is where you start looking for a reason to enter. It is not the reason.
- Confusing patterns. A gartley with a slightly deep B is often a butterfly in the making. Entering as if D is inside the range when it is about to extend beyond X is expensive.
- Trading crowded timeframes. On one-minute charts the ratios are drowned by spread, and completions get run through needlessly.
- Ignoring the trend. Counter-trend butterflies and crabs in a strong market are lopsided bets, however tidy the geometry.
- Stale patterns. A zone that price reached several sessions earlier and drifted through has no fresh information in it.
The disadvantage that traders grudgingly admit last is the time commitment. Waiting for D is thankless on a sluggish day, and a jittery trader will take an early entry that the method never sanctioned. Results vary, and in my experience the patchy results most people report come from these process errors rather than from the ratios.
Harmonic patterns compared with other reversal tools
Harmonics are one branch of a larger family of trend reversal patterns. A head and shoulders pattern is confirmed after the neckline breaks, so the entry comes late but with structural proof. A harmonic pattern is confirmed by ratio agreement, so the entry comes early, at the extreme, with tighter risk but less proof.
That trade-off explains where each tool is underused. Harmonics suit traders who prefer a defined zone and a small stop and can tolerate more failed attempts. Classical reversal patterns suit traders who want the market to declare itself first. A harmonic completion that lines up with a right shoulder or a double-bottom retest is a stronger signal than either alone.
Harmonic patterns inside a prop-firm evaluation
Because the stop sits at a known distance, a harmonic trade fits neatly into a daily loss budget. If the daily limit is 5 percent and each trade risks 1 percent, a trader knows in advance that five full losses end the day. The precise stop also makes it easy to track a static total drawdown in your own head, since every position has a worst case before it opens.
The harder part is consistency. Harmonic setups arrive in clusters and then vanish for days. A trader who forces trades through the quiet stretches will erode the account with marginal patterns. Pivex has no minimum trading days and no time limit on its one-step evaluation. That combination suits harmonic traders well, because the method rewards waiting and punishes activity for its own sake. The built-in assistant that tracks your daily loss and total drawdown adds a second check on the arithmetic during a losing cluster.
One adjustment helps in practice. Halve the position when a completion zone forms within an hour of a scheduled release, or skip it. The limits are about the total, and a zone that gets run through during a release is the fastest way to reach them.
A practical next step is to pick one pair and scroll back through three months of four-hour candles. Label every five-point swing that fits a gartley or a bat, and note which ones completed inside a zone with two overlapping ratios. Doing that exercise before a Pivex evaluation shows plainly how few valid setups a month produces, and that number should shape your risk plan.
Related guides: fibonacci retracement and trend reversal patterns.
Before participating, review Pivex's current program rules and decide whether a simulated proprietary trading evaluation fits your experience and risk limits. Participation does not guarantee a reward, income, or any particular trading result.