Understanding Harmonic Patterns: The Bat
The Bat harmonic pattern describes five alternating price points—X, A, B, C and D—whose leg lengths satisfy specified ratio relationships. Its defining feature is a deep retracement toward X, with D near the 88.6% retracement of XA. Traders use that area to investigate a possible reversal, not to establish that one must occur.
Scott Carney introduced the Bat framework in 2001. His original description distinguishes its ratios from other harmonic structures. Claims that it is exceptionally accurate are the author’s assessment; they do not substitute for a test of a complete trading rule.
Read the Bullish and Bearish Structures
In a bullish candidate, XA rises, AB falls, BC rises and CD falls toward the completion area. The possible trade thesis is a subsequent upward reversal. A bearish candidate mirrors those directions and investigates a downward reversal. A drawing of D ahead of price is a projection, not a completed swing.
Choose how swing points are identified before measuring the legs. Ratio rules reduce some ambiguity, but point selection, wick-versus-close inputs, tolerance and confirmation timing still involve choices. Do not redraw points after the outcome simply to make a successful reversal fit.
Measure the Ratios Consistently
Use absolute price distances for the leg ratios below. AB/XA means |B−A| divided by |A−X|, for example. A retracement measures a move back through an earlier leg; a projection greater than 100% of BC is an extension relative to BC, not a retracement within it.
| Relationship | Working definition | What to check |
|---|---|---|
| AB relative to XA | Common preferred B retracements are 0.382 to 0.50 | Carney requires B below 0.618; a B point exactly at 0.618 is not interchangeable with the preferred Bat alignment |
| BC relative to AB | A commonly used retracement range is 0.382 to 0.886 | Record the exact tolerance and confirmed C point |
| CD relative to BC | 1.618 to 2.618 projection | A 1.27 BC projection does not satisfy Carney’s Bat definition |
| AD relative to XA | Approximately 0.886 | Measure from A back toward X, rather than treating 88.6% as a distance from X |
| CD relative to AB | At least the equivalent AB=CD length in Carney’s description | An extended AB=CD relationship, often around 1.27, is part of his preferred structure |
All selected constraints must hold together for the same points. A match to the 0.886 level alone is not a complete Bat pattern. If using a looser or alternative definition, label it as such and evaluate it separately rather than transferring results from another rule set.
A Coherent Numerical Example
Consider a hypothetical bullish sequence with X = 100, A = 120, B = 110, C = 116 and a candidate D = 102.28. These values illustrate the geometry; they are not a recommended trade or a forecast.
| Check | Calculation | Result |
|---|---|---|
| AB/XA | 10 / 20 | 0.50 |
| BC/AB | 6 / 10 | 0.60 |
| CD/BC | 13.72 / 6 | About 2.287 |
| AD/XA | 17.72 / 20 | 0.886 |
| CD/AB | 13.72 / 10 | 1.372 |
The 0.886 retracement level is A − 0.886(A−X) = 102.28. The candidate D remains above X. The CD leg also exceeds AB and falls within the stated BC projection range. This demonstrates compatible measurements, not the likelihood or size of a subsequent reversal.
What Fibonacci Ratios Do—and Do Not—Establish
The Fibonacci sequence adds the two preceding terms to produce the next. Ratios of successive positive terms approach about 1.618 in one direction and 0.618 in the other; ratios across two positions approach about 2.618 and 0.382. Finite sequence ratios are approximations, not identical constants.
Harmonic methods also use derived and rounded levels. The 0.886 convention is related to taking a fourth root of approximately 0.618; it is not simply another adjacent-term Fibonacci ratio. A 50% retracement is also a trading convention rather than that adjacent-term ratio.
The presence of mathematical proportions in natural structures does not demonstrate that market prices must reverse at the same proportions. Treat each setup based on these proportions as a hypothesis. Predictive value requires evidence under specified data, timing, entry, exit and cost assumptions.
Define a Potential Reversal Zone Before Price Arrives
A potential reversal zone, or PRZ, is an area where the selected retracement and projection calculations come together. In this framework, the 0.886 XA retracement is central. Specify which additional projections are used and how close they must be to count as a cluster.
An arbitrarily broad band from 78.6% to 100% of XA is not equivalent to a precisely defined Bat completion rule. Widening a zone changes the number of apparent matches, entry prices and invalidations. Preserve that choice in the research record rather than expanding the zone after price misses the original level.
Distinguish three events: price approaches a projected zone, price reaches it, and price meets a separate reversal-confirmation condition. A rule may enter at the zone or wait for confirmation, but those approaches have different timing and execution assumptions. Waiting can change the entry price or miss a move; entering early can expose the position to a reversal that never develops.
Preserve Swing and Confirmation Timing
A confirmed local high or low may require subsequent bars. If C or D becomes known only after later observations, a test cannot use that knowledge at the earlier swing bar. TradingView’s repainting documentation explains why plotting a later-confirmed point into the past does not make it available earlier.
An oscillator divergence, candle condition or move away from the zone can be added to a rule, but it should be explicit. Such conditions do not guarantee independent confirmation or a profitable reversal. Test the complete rule with and without the added condition using comparable samples.
Historical Examples: Completed Drawing Versus Projected Setup
Partial exits at intermediate levels and a final reference at A are possible management choices. Define the share of the position exited at each level and the response if price reverses before the remaining target. A chart touching several lines does not prove the outcome of an unspecified trading plan.
Passing the price level of B may be part of a setup’s development, but it does not establish that D will be reached. Keep incomplete and invalidated candidates in the evaluation. Counting only finished patterns that later reverse creates a misleading picture of the process.
Separate Stop Placement from a Limit on Loss
A structural invalidation reference near X must be mirrored correctly: below X for a bullish long scenario and above X for a bearish short scenario. The actual distance, confirmation rule and execution method need to be specified. A tighter stop reduces the planned distance while potentially changing how often normal movement triggers an exit.
For a hypothetical long entry at 102.50 and planned exit trigger at 99.50, the price distance is 3 per unit. A planned price-risk amount of 150 corresponds to 50 units before costs. If a gap produces an exit at 98, the price loss is 225. The planned amount is not a guaranteed loss cap.
Investor.gov’s order guide explains that market orders do not guarantee a price and limit orders may not execute. Check order behavior for the actual instrument and broker, including short-sale access, borrowing or funding costs where relevant.
A target at A or a favorable-looking target-to-stop distance does not establish expectancy. Evaluate realized wins and losses, missed fills, spread, fees and slippage. A small reversal can still produce a losing trade under the selected entry and exit rules.
Do Not Treat Turning Rates as Win Rates
The percentage of selected patterns that show some price turn near D is a different measurement from the percentage of executable trades that make money. To interpret a study, identify how points were selected, what counted as a turn, the allowed time horizon, the sample and the treatment of failures.
A short sample or an author’s accuracy claim cannot establish that the Bat outperforms other patterns across markets. Compare a fixed definition with a suitable baseline and a later evaluation period. Avoid choosing ratio tolerances or exits after seeing which combination makes the historical examples look strongest.
Research the Rule in LuxAlgo’s Native Charts
Start with a specific instrument, interval, swing definition and ratio tolerance in LuxAlgo’s native charts. Keep the development period separate from later evaluation and retain failed or incomplete candidates alongside completed ones.
Ask Quant, our coding agent to help express a supported strategy hypothesis with explicit confirmation, exits and sizing. Inspect the generated code and run it manually. Review strategy settings, costs and individual trades to check that the implementation uses only information available at each decision.
Check native data coverage and history before comparing results. The documented US-equity source is Cboe EDGX rather than a consolidated all-venue feed. Changing the symbol, interval or assumptions changes the experiment and requires another run.
LuxAlgo’s TradingView toolkits are separate from native charts. The legacy Backtesting Assistant is also distinct from the current native strategy workflow. Follow the documentation for the tool and platform actually being tested.
Frequently Asked Questions
What defines the Bat harmonic pattern?
It uses five alternating points, X through D, with specified leg ratios. A completion near the 0.886 retracement of XA is central, but the other relationships must also fit the chosen definition.
Is the Bat pattern always a reversal signal?
No. A projected completion area is a hypothesis about a possible reversal. Price may never reach it, may pass through it or may turn too little to produce a profitable trade.
Is a 0.618 B-point retracement the preferred Bat alignment?
No. Carney specifies B below 0.618 and prefers 0.382 or 0.50. A different tolerance or pattern definition should be labeled and tested separately.
Does a high turning rate mean a high trading win rate?
No. Turning rates depend on how a turn is measured. Trading results also depend on entry timing, exits, sizing, costs and whether assumed fills were possible.
How can LuxAlgo help research Bat-pattern rules?
Use native charts and Quant to express a supported hypothesis with explicit swing confirmation, ratios, exits and sizing. Inspect generated code and run it manually, then review individual trades and a later evaluation period.
Read next