XABCD Pattern

The XABCD framework descends from H.M. Gartley's 1935 work "Profits in the Stock Market," which first described a four-point retracement pattern (A\u2013B\u2013C\u2013D) based on the observation that price moves in symmetrical leg pairs. Gartley noticed that when the BC retracement fell at specific Fibonacci levels of AB, the resulting CD leg tended to complete at a precise projection of the initial AB move. His work predated the widespread adoption of Fibonacci retracement tools in retail trading platforms and relied on manual calculation of price ratios.

Modern harmonic trading was formalised by Scott Carney in the late 1990s and early 2000s, published across two volumes of "Harmonic Trading." Carney added the X origin point to Gartley's ABCD structure and introduced strict Fibonacci ratio requirements for each pattern variant. His key insight was that the XA leg \u2014 not AB \u2014 provides the most reliable Fibonacci projection for the D completion point. Carney identified the Bat (0.886 XA) and Crab (1.618 XA) patterns, while Larry Pesavento and Robert Bryant contributed to the Butterfly variant documentation. Carney's work shifted harmonic trading from a qualitative art to a quantitative discipline with measurable entry and exit rules.

Underlying all XABCD patterns is the concept of the Potential Reversal Zone (PRZ) \u2014 a cluster of Fibonacci levels at D where multiple ratios converge. The PRZ is the highest-probability entry zone because it represents a point of geometric confluence: the CD leg typically reaches 1.272\u20131.618 of BC, a specific retracement of XA, and often a prior support/resistance level. The tighter the PRZ cluster, the higher the probability of a reversal. A PRZ with three or more converging levels is considered a "strong" PRZ. Carney argued that the PRZ is the single most important concept in harmonic trading because it quantifies the point of maximum probability for a trend reversal.

The psychological basis of harmonic patterns lies in the self-fulfilling nature of widely watched Fibonacci levels. As more traders recognise the XABCD structure, they place orders at the PRZ, creating the very reversal they anticipate. This is similar to how round numbers and moving averages become support and resistance through collective focus. The most reliable patterns occur on higher timeframes (daily and weekly) where institutional traders operate and where Fibonacci levels carry more weight due to the larger capital flows involved.

Each XABCD variant is defined by three key Fibonacci ratios: the B retracement of XA, the C extension of AB, and the D retracement of XA. The D ratio is the most critical \u2014 it determines the PRZ entry zone. Below are the exact ratios for each variant as defined by Carney, Pesavento, and Bryant.

1. 1
   Identify X–A impulse

   Locate a clear, clean price swing with minimal overlap between bars. The X–A leg should be unambiguous — a strong trend move with expanding volume. Avoid X–A legs that form during low-volume or holiday periods.
2. 2
   Mark B retracement

   Place B where price pulls back from A. B must not exceed the X extreme, and must fall within the variant-specific Fib zone (e.g. 0.382–0.618 for Crab, 0.786 for Butterfly). Use the snap-to-Fib feature to ensure precision.
3. 3
   Mark C extension

   Place C where price extends beyond A, ideally at 1.272–1.618 of AB. The C point confirms the pattern direction and sets up the final CD leg. If C fails to reach 1.272 AB, the pattern may be invalid.
4. 4
   D projects automatically

   The tool projects D based on the selected variant. D is the Fib retracement of XA (e.g. 0.886 for Bat). Price reaching D completes the pattern; enter on reversal confirmation. The tool shows the PRZ as a highlighted band.

Not every five-swing sequence is a valid XABCD pattern. The following criteria must be met for the pattern to have a statistical edge. Applying these filters rigorously improves win rates from approximately 40% (unfiltered) to 65\u201370% (filtered).

• Clean X\u2013A leg: The X\u2013A swing must be a distinct, recognisable impulse with 3\u20135 bars minimum. Choppy, overlapping X\u2013A legs produce unreliable projections. The X\u2013A leg should have a clear trend direction with expanding volume.
• B within tolerance: B must retrace XA to the exact Fib zone for the chosen variant. A B point at 0.65 is acceptable for a Crab but not for a Bat (max 0.5). Use the settings panel to set tolerance levels.
• C exceeds A: C must extend beyond A by at least 1.272 AB. If C fails to exceed A, the pattern is not valid \u2014 it may be a different harmonic structure. In rare cases, a C point that barely exceeds A (1.05 AB) may still be valid with wider tolerance.
• D at PRZ: D must land at the precise Fib retracement of XA for the variant. The closer D is to the exact ratio, the higher the probability of reversal. A D point within 0.02 of the ideal ratio is excellent.
• AB=CD symmetry: In the Gartley, AB should approximately equal CD in price distance. Tools that enforce AB=CD increase reliability. For Bat and Crab, AB=CD is not required but adds confluence.
• Prior S/R at D: If the PRZ coincides with a prior swing high/low, a trendline, or a moving average, the confluence is significantly stronger. The more levels stacked at D, the higher the probability of reversal.

The XABCD pattern is traded as a reversal at the D completion point. Both bullish and bearish setups follow the same structure but mirrored vertically. The key to profitability is disciplined entry at the PRZ with confirmed price rejection.

Recognising when a harmonic pattern has failed is as important as recognising when it is valid. The following conditions signal that the XABCD pattern is no longer tradeable. Exiting quickly when these occur preserves capital for the next setup.

• B exceeds X: If B retraces past the X extreme, the pattern is invalid. The move may be a trend reversal rather than a harmonic retracement. Exit any pending entry and re-evaluate the trend direction.
• C fails to exceed A: If C is lower than A (in an uptrend) or higher than A (in a downtrend), the extension leg never materialised. The pattern degenerates into a flat or triangle, not a harmonic.
• D breaks beyond tolerance: If price slices through the D PRZ by more than 1 ATR without reversing, the pattern has failed. Exit any pending entry. A break beyond D signals that the harmonic structure has been overpowered by the larger trend.
• Variant mismatch: If B is at 0.65 and you are using the Bat variant (max B=0.5), the pattern is not a Bat. Re-classify as a Crab or Gartley, or discard the setup entirely.
• False break at D: Price tags D intraday but closes beyond it, then continues in the same direction. The PRZ failed as a reversal zone. Do not re-enter \u2014 the pattern has failed and should be removed from the chart.
• Trend resumption through PRZ: If the larger trend overwhelms the PRZ and price closes beyond D with momentum, the harmonic pattern has been invalidated by macro forces. Accept the loss and move on.

The XABCD pattern reflects a specific sequence of trader emotions. The X\u2013A leg represents initial conviction \u2014 traders are confident in the trend direction. The A\u2013B retracement creates doubt: is this a pullback or a reversal? The B\u2013C extension reignites confidence as price breaks beyond A, attracting latecomers. Finally, the C\u2013D leg into the PRZ represents exhaustion \u2014 the last gasp of the trend before the inevitable reversal. Understanding this psychological progression helps traders anticipate the reversal at D and avoid being caught on the wrong side of the trade.

The PRZ (Potential Reversal Zone) at D is where the greatest number of traders are trapped. Latecomers who entered during the B\u2013C extension are now underwater as price reverses at D. Their stop-loss orders, placed just beyond D, are triggered as the reversal unfolds, adding fuel to the new trend. The PRZ thus functions as a "trap zone" \u2014 a point where the maximum number of market participants are positioned incorrectly. Carney emphasised that the PRZ is not just a Fibonacci level but a "zone of maximum pain" for the wrong-positioned trader.

This trap dynamic is strongest when the PRZ coincides with a prior support/resistance level or a round number. The combination of Fib confluence and psychological price levels creates a powerful reversal magnet. On higher timeframes (daily and weekly), this effect is amplified by the participation of institutional traders who place large orders at these levels. The self-fulfilling nature of harmonic patterns means that the mere presence of a recognised PRZ increases the probability of a reversal, as multiple traders independently place orders at the same levels.

XABCD patterns should always be analysed in the context of higher and lower timeframes. A pattern on the 1-hour chart that is aligned with the daily trend direction has a significantly higher probability of success. Conversely, a pattern that forms against the higher timeframe trend requires additional confluence (more Fib levels, stronger divergence) before it can be traded. Carney recommended a "top-down" approach: identify the trend on the weekly chart, find harmonic patterns on the daily chart that align with that trend, and use the 4-hour chart for precise entry timing.

Fractal harmonics: the same XABCD pattern can appear on multiple timeframes simultaneously. For example, a daily chart may show a Bat pattern completing, while the 4-hour chart shows a smaller Gartley pattern within the XA leg. Trading in the direction of the larger harmonic pattern while using the smaller pattern for entry timing is a powerful technique.

The XABCD pattern\u2019s performance varies significantly by variant and market condition. The Gartley pattern (D at 0.786 XA) has the highest historical win rate of approximately 68\u201372% in backtests across major forex pairs and equity indices, but produces the smallest average risk/reward ratio (approximately 1:2.5). The Bat pattern (D at 0.886 XA) has a slightly lower win rate (63\u201367%) but larger risk/reward (1:3.5) due to the deeper retracement providing a tighter stop. The Crab pattern (D at 1.618 XA) has the lowest win rate (55\u201360%) but the highest risk/reward (1:5+), making it suitable for traders with a lower win rate tolerance. The Butterfly (D at 1.272 XA) sits in the middle with approximately 62\u201366% win rate and 1:3 risk/reward.

Market conditions matter: harmonic patterns perform best in range-bound markets where Fib levels act as reliable magnets. In strongly trending markets, the D projection is frequently overrun as the trend overpowers the harmonic structure. In low-volatility environments, the PRZ produces more false touches. The optimal conditions for XABCD trading are moderate volatility (ATR within the 30\u201370th percentile) and a clear alternating swing structure. The pattern\u2019s performance also degrades on lower timeframes (sub-1-hour) where price noise reduces the reliability of Fib projections. For best results, trade XABCD patterns on 1-hour to daily charts with a minimum ATR of 20 pips (forex) or 0.50 points (equities).

When implementing XABCD pattern detection algorithmically, the following parameters should be configurable: tolerance per ratio (B, C, and D tolerance should be independent), minimum XA leg length (in bars and price distance), maximum pattern duration (to avoid stale patterns), and confluence detection (distance to nearest support/resistance level). The algorithm should reject patterns where the B\u2013C leg takes significantly longer than the X\u2013A leg (more than 2x the time), as this suggests the pattern is degenerating. Backtesting across multiple market regimes is essential to validate that the tolerance settings are appropriate for the specific market being traded.

• 1.Harmonic Trading Volume 1 & 2 — Scott Carney
• 2.Profits in the Stock Market — H.M. Gartley (1935)
• 3.Harmonic Patterns guide — TradingView Help
• 4.The Bat Pattern — Investopedia
• 5.Fibonacci Ratios with Pattern Recognition — Larry Pesavento
• 6.Elliott Wave Principle — Frost & Prechter
• 7.Harmonic Trading on TradingView — TradingView Community