How we tested this pattern
- Sample
- 959 occurrences on 6 markets (crypto, forex, US stocks, ETFs, indices and commodities), 1971 to 2026.
- Detection
- Found by code, never picked by hand: every match of the rule on full price history, winners and losers alike.bullish crab pattern: Five swing points X (low), A, B, C, D. B retraces 0.382 to 0.618 of XA, C retraces 0.382 to 0.886 of AB. When C is confirmed, the reversal zone is where a 1.537 to 1.699 extension of XA (1.618 within 5%, beyond X) overlaps a 2.24 to 3.618 projection of BC. Signal at the close of the first bar that trades into the zone without trading beyond its far edge. Stop at the far edge of the zone. The swing points before the reversal zone span at least 3 ATR, with at least 3 bars between consecutive points. The first swing point is a true extreme: over the duration of the first leg before it, price never went more than 10% of that leg beyond it.bearish crab pattern: Five swing points X (high), A, B, C, D. B retraces 0.382 to 0.618 of XA, C retraces 0.382 to 0.886 of AB. When C is confirmed, the reversal zone is where a 1.537 to 1.699 extension of XA (1.618 within 5%, beyond X) overlaps a 2.24 to 3.618 projection of BC. Signal at the close of the first bar that trades into the zone without trading beyond its far edge. Stop at the far edge of the zone. The swing points before the reversal zone span at least 3 ATR, with at least 3 bars between consecutive points. The first swing point is a true extreme: over the duration of the first leg before it, price never went more than 10% of that leg beyond it.
- Benchmark
- Each occurrence is compared with the same assets on random bars: the move 10 bars later, and a 2:1 reward-to-risk trade.
- Checks
- Block bootstrap and a correction for the thousands of tests we run, plus a random-walk self-test that the bench passes: on random prices, no pattern should win.
- Updated
- Data updated November 5, 2026, code version c2a1683f. Read the methodology · bullish crab pattern data (JSON) · bearish crab pattern data (JSON)
At a glance: what textbooks say, what we measured
- Pattern
- Bullish crab
- Occurrences tested
- 315 occurrences, all markets, daily
- Textbook: price should rise after it
- Too close to call · +0.29% vs the market, 10 bars
- Textbook: target reached most of the time
- 20.7% vs 21.6% on random dates
- Its best market and timeframe, still negative, US stocks left out
- Crypto, 1-hour: −0.04 R 2:1 trade vs random dates, not proven
- Verdict
- 10-bar move: No clear edge2:1 trade: No clear edge
- Pattern
- Bearish crab
- Occurrences tested
- 644 occurrences, all markets, daily
- Textbook: price should fall after it
- Too close to call · +0.29% vs the market, 10 bars
- Textbook: target reached most of the time
- 12.5% vs 11.6% on random dates
- Where it did best
- US stocks, daily: +0.06 R 2:1 trade vs random dates, not proven
- Verdict
- 10-bar move: No clear edge2:1 trade: No clear edge
On this page
- What is the crab harmonic pattern?
- Key Fibonacci ratios and identification rules
- Bullish crab pattern: identification and setup
- Bearish crab pattern: identification and setup
- Real examples
- How to trade the crab pattern
- Deep crab pattern vs. standard crab pattern
- Is the crab pattern reliable?
- Common pitfalls and technical limitations
- Software for pattern recognition
- Frequently asked questions about harmonic trading
- Summary
The crab harmonic pattern is a five-point reversal pattern, labeled X, A, B, C and D, in which D completes far beyond X, at the 161.8% extension of the first leg. Defined by Scott Carney, the crab pattern aims at the end of a stretched move: the setup buys at D in the bullish form and sells at D in the bearish one. This guide gives the exact rules, a worked case, the deep variant, a trade plan, and what our bench measured.
What is the crab harmonic pattern?
It belongs to the harmonic patterns, a family of chart patterns in which each swing must respect Fibonacci proportions on the same XABCD frame: five pivots joined by four legs. Its signature is the depth of the last leg. In the Gartley and the Bat, D stays inside the first leg; in the Butterfly, it reaches the 1.27 extension; here it goes to 1.618, the most extended of the four classic harmonic patterns. Scott Carney, who coined the term harmonic trading in The Harmonic Trader (1999), defined it on the idea that a move stretching 61.8% beyond its own starting point is exhausted.
Key Fibonacci ratios and identification rules

- Swing low where the pattern starts
- End of the XA leg
- Retraces 0.382 to 0.618 of XA
- Retraces 0.382 to 0.886 of AB
- D zone: signal1.537 to 1.699 extension of XA beyond X (1.618 within 5%), overlapping a 2.24 to 3.618 projection of BC; entry at the close of the first bar that reaches it

- Swing high where the pattern starts
- End of the XA leg
- Retraces 0.382 to 0.618 of XA
- Retraces 0.382 to 0.886 of AB
- D zone: signal1.537 to 1.699 extension of XA beyond X (1.618 within 5%), overlapping a 2.24 to 3.618 projection of BC; entry at the close of the first bar that reaches it
Defining the structural components
To identify it, find the five pivots on the price chart first, then check each leg. XA is the first, strong leg, and every other measure is based on it. B is formed by a pullback that gives back part of it, BC bounces without passing A, and CD, the longest leg, runs beyond X to D.
Precise Fibonacci measurement requirements
| Point | Rule | Measured on |
|---|---|---|
| B | 0.382 to 0.618 retracement | XA |
| C | 0.382 to 0.886 retracement | AB |
| D | 1.618 extension | XA |
| D | 2.24 to 3.618 extension | BC |
| D | alternate AB=CD, CD about 1.27 to 1.618 times AB (optional) | AB |
The most important measure is the 1.618 extension. The measures for D do not have to meet at one price: together they form a band of prices, the potential reversal zone (PRZ). A tighter band gives a narrower zone to act in; our bench does not test whether tighter bands do better. A B deeper than 0.618 points to another pattern, such as the deep crab.
A worked example with numbers
Illustrative values, not a real chart. For a long setup, X is at 100 and A at 150: the first leg is 50. B at a 61.8% retracement is 150 minus 30.9, so 119.1. C at an 88.6% retracement of AB is 119.1 plus 27.4, so 146.5. D at the 1.618 extension is 150 minus 80.9, so 69.1, well below X. Check BC: CD from 146.5 to 69.1 is 77.4, or 2.82 times BC (27.4), inside the 2.24 to 3.618 range; 2.618 times BC would put D at 74.8. The PRZ is about 69.1 to 74.8. CD is 2.5 times AB here, far from an AB=CD, which is why that measure stays optional.
Bullish crab pattern: identification and setup
A bullish crab pattern begins with a rise from X to A, followed by a pullback to B, a bounce to C below A, then price drops fast to D, well below X. On the chart it looks like an M with a long last leg. The buy is placed in the PRZ, and a confirmation entry watches the price action there: a slowing fall, long lower wicks, a first higher low.

Bearish crab pattern: identification and setup
A bearish crab pattern is the mirror image: a fall from X to A, a rebound to B, a dip to C above A, then a rally to D, far above X, like a W. The sell is placed at D, after price makes a new high that may fail; a confirmation entry waits for a bearish candle or a first lower high in the band.

The Real examples section shows crab setups from our bench, bullish and bearish, with the result of the 2:1 trade in R, the initial risk of the trade.
Real examples
Real occurrences from our test bench, each on its market's main timeframe. Crypto examples come from liquid assets only (the largest coins). Best and Worst are the best and worst 2:1 trade results, Textbook the cleanest shapes (chosen on geometry alone, never on the outcome), Most recent the latest detections among the assets the gallery draws from, Typical a random draw. Shaded area: the pattern. Dashed lines: entry, stop loss (SL) and 2:1 target (TP) of the tested trade. Dotted line: 10 bars after the signal. A result above +2 R or below −1 R means a bar opened past the target or the stop, and the trade was closed at that open. From the bullish crab gallery, we left out 1 card with a single bar so large that the pattern or the trade levels could not be read. It still counts in the statistics, and is not replaced: our bench picks a fixed number of cards per category.



















How to trade the crab pattern

- Entry
- Close of the first bar that reaches the near edge of the reversal zone
- Stop loss (SL)
- At the far edge of the reversal zone
- Target (TP)
- Twice the risk (2:1)

- Entry
- Close of the first bar that reaches the near edge of the reversal zone
- Stop loss (SL)
- At the far edge of the reversal zone
- Target (TP)
- Twice the risk (2:1)
A clean case from our bench: US stocksChosen on its shape alone, never on the outcome. Stock prices are adjusted for splits and dividends, so they can differ from the prices quoted at the time.

A clean case from our bench: US stocksChosen on its shape alone, never on the outcome. Stock prices are adjusted for splits and dividends, so they can differ from the prices quoted at the time.

Entry strategies at the potential reversal zone
A limit order at the 1.618 extension gets the best fill but acts on a falling move with no sign of a turn. A confirmation entry waits for a reaction in the band, such as a hammer or an engulfing candle: fewer signals, a later fill. Confirmation adds information, not certainty: on 12,005 closed trades on daily hammers, our bench found the 2:1 trade reached its objective 35.4% of the time against 36.1% for the same trade on random dates, a gap rated No clear edge: its average result per trade (+0.06 R against +0.09 R) does not differ measurably from chance, so it does not count as an edge.
Managing risk and setting stop-loss orders
Target levels and profit taking
The objectives of the classic plan are the 0.382 and 0.618 retracements of AD, then B, then A. With the numbers above, buying at 69.1 with a stop at 60 puts 9.1 at stake. AD is 80.9, so the 0.382 retracement sits at 100.0 (a gain of 30.9, about 3.4 times the amount at stake) and the 0.618 retracement at 119.1 (a gain of 50, about 5.5 times). These figures describe the plan, not the odds: a large reward against the loss comes with a smaller chance of getting there, because a target farther from the entry is, by construction, harder to reach than a closer one.
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 far edge of the zone and a target at twice the risk. The 0.618 retracement of AD is measured separately.
Deep crab pattern vs. standard crab pattern
The deep crab pattern differs in one point: B is a deep retracement, at 88.6% of XA. D still completes at the 1.618 extension, with a BC extension of 2.24 to 3.618. Our bench does not test it separately: its B at 0.886 falls outside the crab rule, so the figures on this page cover the standard crab only.
- ButterflyBullish or bearishBullish, 10-bar move, daily No clear edgeBearish, 10-bar move, daily No clear edge
- GartleyBullish or bearishBullish, 10-bar move, daily No clear edgeBearish, 10-bar move, daily No clear edge
- BatBullish or bearishBullish, 10-bar move, daily No clear edgeBearish, 10-bar move, daily No clear edge

| Pattern | Point B | Point D |
|---|---|---|
| Crab | 0.382 to 0.618 | 1.618, beyond X |
| Deep crab | 0.886 | 1.618, beyond X |
| Butterfly | 0.786 | 1.27, beyond X |
| Gartley | 0.618 | 0.786 |
| Bat | 0.382 to 0.50 | 0.886 |
The Butterfly and both crab forms complete beyond X; the Gartley and the Bat stay inside the first leg, with X as a natural stop. The Cypher and the Shark let C go beyond A.
Is the crab pattern reliable?
A fair test needs a fixed rule for each pivot and for the tolerance around each ratio, and we set ours from chart reviews, never from results. Turning points are swings of 2, 4 or 8 times the average true range, and 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, and a published range is used as is. Once C is confirmed, the reversal zone is where the 1.618 extension of XA (1.537 to 1.699) overlaps the 2.24 to 3.618 projection of BC. The signal is the first bar that trades into the zone without trading beyond its far edge.
Moved the expected way 10 bars later, compared with random bars.
Measured on 315 daily bullish crab occurrences across 281 assets. Random bars on the same assets are the honest baseline.
No clear edgeMoved the expected way 10 bars later, compared with random bars.
Measured on 644 daily bearish crab occurrences across 518 assets. Random bars on the same assets are the honest baseline.
No clear edgeBullish crab against random bars on the same assets, by market, on daily candles. Bars of almost the same length mean no edge.
- All markets
- US stocks
- All markets
- US stocks
Bearish crab against random bars on the same assets, by market, on daily candles. Bars of almost the same length mean no edge.
- All markets
- Crypto
- US stocks
- All markets
- US stocks
All statistics
Full results for the crab harmonic pattern, market by market
Every occurrence our rule finds on full price history, compared with what the same assets did on random bars. A pattern only earns an edge if it beats random entries after correcting for the thousands of tests we run.
Our verdict
Not a reliable signal on its own here. Use it as context, not as a trigger.
Bullish crab: 315 occurrences tested
Bullish pattern · All markets · Daily · 281 assets · 1982-07 to 2026-09

What this means. No measurable difference from the rest of its market over the next 10 bars.
Timeline
- Pattern length, first swing to signal
- 61 days · middle half 40 to 113 · on 315 occurrences
- Signal, after the last confirmed swing
- 19 days · middle half 10 to 43 · on 315 occurrences
- Time to the 2:1 target, when hit
- 4 days · middle half 2 to 15 · on 108 trades that hit it
- Time to the stop, when hit
- 2 days · middle half 1 to 5 · on 197 trades that hit it
- Time in the trade, all exits
- 3 days · middle half 1 to 8 · on 314 trades
Size and trade path
- Pattern height
- 19.9% · middle half 13.1% to 35.7% · on 315 occurrences
- Best open profit during the trade
- 0.98 R · middle half 0.00 R to 2.10 R · on 314 trades
- Stopped trades that were first up 1 R
- 21.3% · on 197 stopped trades
- Trades a stop moved to breakeven at +1 R would have closed flat instead of at a loss
Medians, each on the cases named on its line; "middle half" spans the 25th to 75th percentile; no value under 30 cases.
How we detect it. Five swing points X (low), A, B, C, D. B retraces 0.382 to 0.618 of XA, C retraces 0.382 to 0.886 of AB. When C is confirmed, the reversal zone is where a 1.537 to 1.699 extension of XA (1.618 within 5%, beyond X) overlaps a 2.24 to 3.618 projection of BC. Signal at the close of the first bar that trades into the zone without trading beyond its far edge. Stop at the far edge of the zone. The swing points before the reversal zone span at least 3 ATR, with at least 3 bars between consecutive points. The first swing point is a true extreme: over the duration of the first leg before it, price never went more than 10% of that leg beyond it.
All measurements (17 rows)
10-bar move: did price go the expected way?
| Market | Occurrences | Pattern | Random bars | Excess | 10-bar move verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 693 | 54.7% | 48.2% | +0.13% | No clear edge |
| Forex · 1-hour | 107 | 49.5% | 50.2% | +0.06% | No clear edge |
| All markets · 1-hour | 800 | 54.0% | 48.5% | +0.12% | No clear edge |
| Crypto · 4-hour | 158 | 53.8% | 47.5% | +0.04% | No clear edge |
| Forex · 4-hour | 26 | 50.0% | 49.8% | +0.04% | Not enough data |
| All markets · 4-hour | 184 | 53.3% | 47.8% | +0.04% | No clear edge |
| Crypto · Daily | 5 | 40.0% | 43.7% | +0.63% | Not enough data |
| US stocks · Daily | 283 | 56.2% | 53.5% | +0.29% | No clear edge |
| ETFs · Daily | 11 | 36.4% | 57.2% | −0.43% | Not enough data |
| Indices · Daily | 8 | 75.0% | 57.7% | +2.28% | Not enough data |
| Forex · Daily | 7 | 42.9% | 49.8% | −0.51% | Not enough data |
| Commodities · Daily | 1 | 0.0% | 53.9% | −3.74% | Not enough data |
| All markets · Daily | 315 | 55.2% | 53.5% | +0.29% | No clear edge |
| US stocks · Weekly | 30 | 51.7% | 55.5% | −0.46% | No clear edge |
| ETFs · Weekly | 1 | 100.0% | 54.3% | +7.22% | Not enough data |
| Commodities · Weekly | 1 | 0.0% | 59.9% | −8.20% | Not enough data |
| All markets · Weekly | 32 | 51.6% | 55.6% | −0.46% | No clear edge |
2:1 trade: target hit and average result
| Market | Trades | Target hit | Same trade, random dates | Per trade | 2:1 trade verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 693 | 28.7% | 30.1% | −0.08 R vs −0.04 R | No clear edge |
| Forex · 1-hour | 107 | 34.6% | 37.1% | −0.02 R vs +0.13 R | No clear edge |
| All markets · 1-hour | 800 | 29.5% | 31.1% | −0.07 R vs −0.01 R | No clear edge |
| Crypto · 4-hour | 157 | 26.8% | 26.9% | −0.11 R vs −0.10 R | No clear edge |
| Forex · 4-hour | 26 | 15.4% | 28.8% | −0.63 R vs −0.16 R | Not enough data |
| All markets · 4-hour | 183 | 25.1% | 27.2% | −0.18 R vs −0.11 R | No clear edge |
| Crypto · Daily | 5 | 20.0% | 5.0% | −0.09 R vs −0.41 R | Not enough data |
| US stocks · Daily | 282 | 34.0% | 35.9% | +0.10 R vs +0.06 R | No clear edge |
| ETFs · Daily | 11 | 45.5% | 43.2% | +0.34 R vs +0.26 R | Not enough data |
| Indices · Daily | 8 | 62.5% | 34.4% | +1.30 R vs +0.08 R | Not enough data |
| Forex · Daily | 7 | 14.3% | 21.4% | −0.83 R vs −0.36 R | Not enough data |
| Commodities · Daily | 1 | 0.0% | 25.0% | −1.00 R vs −0.25 R | Not enough data |
| All markets · Daily | 314 | 34.4% | 35.3% | +0.11 R vs +0.05 R | No clear edge |
| US stocks · Weekly | 29 | 44.8% | 37.1% | +0.38 R vs +0.17 R | Not enough data |
| ETFs · Weekly | 1 | 100.0% | 0.0% | +2.00 R vs −0.43 R | Not enough data |
| Commodities · Weekly | 1 | 0.0% | 0.0% | −1.00 R vs −1.00 R | Not enough data |
| All markets · Weekly | 31 | 45.2% | 34.7% | +0.39 R vs +0.11 R | No clear edge |
Our verdict
Not a reliable signal on its own here. Use it as context, not as a trigger.
Bearish crab: 644 occurrences tested
Bearish pattern · All markets · Daily · 518 assets · 1971-04 to 2026-10

What this means. No measurable difference from the rest of its market over the next 10 bars.
Timeline
- Pattern length, first swing to signal
- 105 days · middle half 53 to 217 · on 644 occurrences
- Signal, after the last confirmed swing
- 42 days · middle half 21 to 101 · on 644 occurrences
- Time to the 2:1 target, when hit
- 6 days · middle half 2 to 13 · on 193 trades that hit it
- Time to the stop, when hit
- 4 days · middle half 2 to 10 · on 438 trades that hit it
- Time in the trade, all exits
- 5 days · middle half 2 to 12 · on 639 trades
Size and trade path
- Pattern height
- 19.1% · middle half 12.8% to 28.2% · on 644 occurrences
- Best open profit during the trade
- 0.85 R · middle half 0.23 R to 2.07 R · on 639 trades
- Stopped trades that were first up 1 R
- 22.8% · on 438 stopped trades
- Trades a stop moved to breakeven at +1 R would have closed flat instead of at a loss
Medians, each on the cases named on its line; "middle half" spans the 25th to 75th percentile; no value under 30 cases.
How we detect it. Five swing points X (high), A, B, C, D. B retraces 0.382 to 0.618 of XA, C retraces 0.382 to 0.886 of AB. When C is confirmed, the reversal zone is where a 1.537 to 1.699 extension of XA (1.618 within 5%, beyond X) overlaps a 2.24 to 3.618 projection of BC. Signal at the close of the first bar that trades into the zone without trading beyond its far edge. Stop at the far edge of the zone. The swing points before the reversal zone span at least 3 ATR, with at least 3 bars between consecutive points. The first swing point is a true extreme: over the duration of the first leg before it, price never went more than 10% of that leg beyond it.
All measurements (19 rows)
10-bar move: did price go the expected way?
| Market | Occurrences | Pattern | Random bars | Excess | 10-bar move verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 653 | 55.1% | 50.7% | +0.13% | No clear edge |
| Forex · 1-hour | 133 | 57.1% | 48.9% | +0.05% | No clear edge |
| All markets · 1-hour | 786 | 55.4% | 50.4% | +0.11% | No clear edge |
| Crypto · 4-hour | 149 | 53.0% | 51.9% | +0.29% | No clear edge |
| Forex · 4-hour | 26 | 46.2% | 49.6% | +0.10% | Not enough data |
| All markets · 4-hour | 175 | 52.0% | 51.6% | +0.26% | No clear edge |
| Crypto · Daily | 30 | 50.0% | 53.8% | +2.70% | No clear edge |
| US stocks · Daily | 562 | 47.0% | 45.2% | +0.26% | No clear edge |
| ETFs · Daily | 25 | 32.0% | 43.1% | −0.45% | Not enough data |
| Indices · Daily | 12 | 58.3% | 41.8% | −0.44% | Not enough data |
| Forex · Daily | 7 | 28.6% | 45.9% | −0.93% | Not enough data |
| Commodities · Daily | 8 | 25.0% | 47.8% | −2.23% | Not enough data |
| All markets · Daily | 644 | 46.3% | 45.5% | +0.29% | No clear edge |
| US stocks · Weekly | 122 | 36.9% | 41.8% | −0.69% | No clear edge |
| ETFs · Weekly | 8 | 37.5% | 35.6% | −1.91% | Not enough data |
| Indices · Weekly | 5 | 80.0% | 39.7% | +3.20% | Not enough data |
| Forex · Weekly | 1 | 100.0% | 41.0% | +1.06% | Not enough data |
| Commodities · Weekly | 4 | 25.0% | 48.2% | −4.60% | Not enough data |
| All markets · Weekly | 140 | 38.6% | 41.6% | −0.72% | No clear edge |
2:1 trade: target hit and average result
| Market | Trades | Target hit | Same trade, random dates | Per trade | 2:1 trade verdict |
|---|---|---|---|---|---|
| Crypto · 1-hour | 652 | 30.4% | 30.3% | −0.02 R vs −0.03 R | No clear edge |
| Forex · 1-hour | 133 | 36.1% | 28.2% | +0.09 R vs −0.13 R | No clear edge |
| All markets · 1-hour | 785 | 31.3% | 29.9% | 0.00 R vs −0.05 R | No clear edge |
| Crypto · 4-hour | 149 | 29.5% | 28.1% | −0.05 R vs −0.11 R | No clear edge |
| Forex · 4-hour | 26 | 38.5% | 27.9% | +0.15 R vs −0.11 R | Not enough data |
| All markets · 4-hour | 175 | 30.9% | 28.0% | −0.02 R vs −0.11 R | No clear edge |
| Crypto · Daily | 28 | 32.1% | 31.5% | +0.04 R vs +0.02 R | Not enough data |
| US stocks · Daily | 559 | 31.1% | 28.7% | −0.08 R vs −0.15 R | No clear edge |
| ETFs · Daily | 25 | 4.0% | 28.0% | −0.98 R vs −0.12 R | Not enough data |
| Indices · Daily | 12 | 58.3% | 27.1% | +0.76 R vs −0.20 R | Not enough data |
| Forex · Daily | 7 | 28.6% | 28.6% | −0.14 R vs −0.14 R | Not enough data |
| Commodities · Daily | 8 | 0.0% | 21.9% | −1.00 R vs −0.57 R | Not enough data |
| All markets · Daily | 639 | 30.2% | 28.7% | −0.11 R vs −0.14 R | No clear edge |
| US stocks · Weekly | 122 | 25.4% | 31.1% | −0.24 R vs −0.07 R | No clear edge |
| ETFs · Weekly | 8 | 0.0% | 9.4% | −1.00 R vs −0.72 R | Not enough data |
| Indices · Weekly | 5 | 40.0% | 15.0% | +0.20 R vs −0.55 R | Not enough data |
| Forex · Weekly | 1 | 0.0% | 25.0% | −1.00 R vs +0.45 R | Not enough data |
| Commodities · Weekly | 4 | 0.0% | 43.8% | −1.00 R vs +0.31 R | Not enough data |
| All markets · Weekly | 140 | 23.6% | 29.6% | −0.29 R vs −0.11 R | No clear edge |
Data updated November 5, 2026 · code version c2a1683f · Read the methodology. We measure every pattern on crypto and forex (1-hour, 4-hour, daily and weekly charts) and US stocks, ETFs, indices and commodities (daily and weekly charts). A market or timeframe from that list missing from the tables above had no occurrence.
What you often read
“The bullish crab is a reliable buy signal.”
What our data shows
On 315 daily occurrences, it moved in the expected direction 55.2% of the time, against 53.5% for random bars: no clear edge, since the average excess return is not statistically significant.
What you often read
“The bearish crab is a reliable sell signal.”
What our data shows
On 644 daily occurrences, it moved in the expected direction 46.3% of the time, against 45.5% for random bars: no clear edge, since the average excess return is not statistically significant.
The crab is rare: on daily charts, all markets together, the bench found 315 bullish and 644 bearish setups. The 2:1 trade reached its target in 34.4% of the 314 bullish closed trades, against 35.3% for the same trade on random dates, and in 30.2% of the 639 bearish trades, against 28.7%. The 0.618 retracement of AD was reached before the stop, within 100 bars, in 18.5% of bullish daily setups and 11.5% of bearish ones. Trying versions until one looks good would produce a number that means nothing. Be wary of any claim that it is among the most profitable setups unless it says how the pattern was detected and how the outcome compares with chance. Here is what our bench measured on the crab itself, every occurrence our rule finds on full history against the same trade on random dates:
Table quoted in the text (verdicts and every market: see All measurements)
| Setup tested | Occurrences (2:1 trades) | 2:1 target hit | Same trade, random dates | Expectancy before costs |
|---|---|---|---|---|
| Bullish, all markets, daily | 315 (314) | 34.4% | 35.3% | +0.11 R vs +0.05 R |
| Bearish, all markets, daily | 644 (639) | 30.2% | 28.7% | −0.11 R vs −0.14 R |
| Bullish, all markets, hourly | 800 (800) | 29.5% | 31.1% | −0.07 R vs −0.01 R |
| Bearish, all markets, hourly | 786 (785) | 31.3% | 29.9% | 0.00 R vs −0.05 R |
| Bearish, all markets, weekly | 140 (140) | 23.6% | 29.6% | −0.29 R vs −0.11 R |
Read each row in pairs: none of these all-markets daily or weekly rows differs measurably from chance, in either direction, and bullish weekly setups, only 32, read no clear edge as well. In the full statistics table, no line stands out either: bearish setups on US stocks (559 trades) made −0.08 R against −0.15 R on random dates, read no clear edge: still a loss, and a gap that does not pass our statistical test. On hourly charts, the bearish 2:1 trade reads no clear edge too. That does not prove the crab pattern fails; it is a reason to test it on your own assets before trusting it. Our methodology explains the method.
Common pitfalls and technical limitations
- Mislabeling a Butterfly: with B at 78.6% and D near 1.27, the order and the stop belong elsewhere.
- Acting before the PRZ: buying as soon as price breaks X means buying a falling move with no structure yet.
- Moving the stop: once it is hit, the setup has failed; redrawing it turns a small loss into a large one.
- Trusting the drawing: pivots are chosen by eye; we know of no independent, published test showing that these proportions beat chance, and our own test found no established edge.
Software for pattern recognition
TradingView includes an XABCD drawing tool. Free community scripts and paid scanners flag harmonic patterns automatically on stocks, currencies or crypto; check which tolerances each one uses, and practice on a demo account before risking money.
Frequently asked questions about harmonic trading
Is the crab pattern bullish or bearish?
Both. One form finishes at a new low and points up; the other finishes at a new high and points down. The rules are the same.
What is the difference between the crab and the butterfly?
Both complete beyond X. The Butterfly has B at 78.6% of the first leg and D at the 1.27 extension; the crab pattern has B at 38.2% to 61.8% and D at 1.618.
Does it work in forex trading?
The rules are the same on currency pairs. On our bench the crab is rare there: 107 bullish and 133 bearish setups on hourly forex charts, with no measurable difference from chance (the bullish 10-bar move reads no clear edge), and fewer than 30 of each on 4-hour charts (26 and 26), daily and weekly charts.
Summary
- B at 38.2% to 61.8% of XA, C at 38.2% to 88.6% of AB, D at the 1.618 extension of XA, beyond X.
- The deep variant has B at 88.6% and the same D.
- Act in the PRZ, stop just beyond it, objectives at the 0.382 and 0.618 retracements of AD.
- Tested on our bench with a fixed rule: rare, and no difference from chance that survives our correction for multiple testing, on any timeframe.
This guide is general information about technical analysis, not 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 price 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 crab 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.





