Gartley Pattern — The Harmonic Classic From 1935

If harmonic patterns had a family tree, the Gartley would sit at its root. Harold M. Gartley described it in "Profits in the Stock Market" (1935) — and because the setup appeared on page 222, it's still sometimes called "the 222 pattern" to this day. A detail few people mention: the original description had no Fibonacci ratios at all. Gartley described the shape and the logic of a correction inside a trend, and only decades later did Larry Pesavento add the ratios, with Scott Carney codifying the version you'll find in every scanner today.

The result is the most famous five-point XABCD structure: a stretched-out letter "M" in the bullish variant, an inverted "W" in the bearish one. The idea is simple — after a strong XA impulse, the market builds a complex ABCD correction that exhausts itself at 78.6% retracement of the whole impulse, and from there the trend is expected to resume. The Gartley is a pattern of CORRECTION inside a trend, not top-catching: D never goes beyond point X. You'll find the full map of the family and the shared rules in our harmonic patterns guide — here we take the classic itself apart piece by piece.

Pattern Structure

The skeleton is five points (X, A, B, C, D) and four waves. Bullish variant: an XA impulse up, then a three-wave correction down that ends at D, where you buy. Bearish variant — a mirror image.

Two things turn this table into a pattern rather than a list of numbers. First, point B: a 61.8% retracement of XA is the Gartley's calling card. If B stopped at 50%, you're more likely looking at a Bat pattern; if it reached 78.6%, it's a candidate for a Butterfly. The depth of B selects the pattern before you even start calculating D. Second, point D sits below (bullish) or above (bearish) point B, but still inside the range of XA — the Gartley is a retracement-type pattern; the correction has to stay a correction.

In practice a tolerance of roughly ±3 percentage points is applied, but that's a buffer for market noise, not permission for creativity. B at 70% of XA isn't "almost a Gartley" — it's no pattern at all.

How to Measure Point D

You calculate point D IN ADVANCE, once the first three waves (XA, AB, BC) are complete. In the Gartley, the PRZ zone is built from three independent measurements:

1. 78.6% retracement of XA — the primary measurement. Stretch the Fibonacci grid from X to A; the 78.6% level is the core of the zone.
2. 127.2–161.8% extension of BC — a supporting measurement. Wave CD usually ends inside this range measured from C.
3. AB=CD projection — measure the length of wave AB from point C. Inside every Gartley sits the simpler AB=CD pattern, and its completion is the third vote in the confluence.

The tighter these three levels overlap, the better the setup. A perfect Gartley has a PRZ only a fraction of the XA wave wide; if the measurements spread out widely, the zone stops being a zone and becomes a wish. One practical note to finish: measure by candle wicks, not bodies — unless a wick is an anomaly (one panic candle), in which case some practitioners use the close instead. What matters is doing it CONSISTENTLY, because switching your measuring method between setups is the easiest way to "find" a pattern where there isn't one.

How to Trade It

Entry. Price reaching the PRZ zone is an invitation to watch, not a signal. Wait for a reaction: a rejection candle (pin bar, engulfing), a structure shift on a lower timeframe, or at least a clear loss of momentum in wave CD. You're trading against the last wave — confirmation costs a few ticks of worse price, but filters out cases where "D" turns out to be just a rest stop in a strong decline.

Stop loss. Behind point X, with a buffer for noise. The logic is elegant: since a Gartley is a correction inside a trend, a break of X means the correction has retraced more than 100% of the impulse — meaning it has stopped being a correction. The setup is then dead by definition, and there's nothing left to defend. The distance from D (78.6% XA) to X (100% XA) gives a risk of roughly 21% of the length of wave XA plus a buffer — more than the Bat, less than the extension-type patterns.

Targets. First: 38.2–61.8% retracement of wave CD, which usually overlaps with the area around point B. Second: the area around point C. Third, on a full trend continuation: the area around A. A sensible execution template is to take part of the position at the first target and move the stop to breakeven — with entry at 78.6% XA and a stop behind X, even reaching C typically gives a risk-reward ratio around 1:1.5–1:2.5, depending on the geometry of the specific setup. Calculate this BEFORE entering, not after.

A numerical example, for the memory. A bullish Gartley: an XA impulse from 100 to 110 (wave = 10 units). B at 61.8% retracement, roughly 103.8. Bounce BC to around 106.5, then decline CD. D calculated at 78.6% XA, roughly 102.1 — that's where you wait for a reaction. Stop below X = 100 with a buffer (risk of about 2.3 units), first target near B ≈ 103.8 (gain of about 1.7), second near C ≈ 106.5 (gain of about 4.4 — an R:R near 1:2). You can see right away that only the second target makes this a sensible trade — which is why you calculate the geometry BEFORE entering.

The overriding filter. A bullish Gartley traded as a correction inside a confirmed higher-timeframe uptrend is a completely different trade from the same pattern drawn against a weekly downtrend. The pattern gives you the entry geometry; context decides whether it's worth taking.

Common Mistakes

Forcing point B. The most common sin: B at 55% or 70% of XA "rounded" to 61.8%. The ±3-pp tolerance exists to absorb noise, not to bend the market to your wish. The pattern draws itself — or it doesn't exist.

Confusing the Gartley with the Bat. The shape is nearly identical, the ratios different — and along with them the entry point (78.6% vs. 88.6% XA) and the real risk-reward ratio. Always start your classification with the depth of B.

Entering with a blind limit order at 78.6%. Without a price reaction you're buying a falling knife at a spot you calculated with a ruler. Sometimes it works out; systematically, confirmation inside the PRZ is cheaper than a string of stop-outs.

A stop "just under D" instead of behind X. Tighter means cheaper, but D is a zone, not a point — the market regularly pokes a few ticks into the PRZ before reversing. X is the natural invalidation boundary, and it's what defines whether the setup is alive.

Trusting success percentages from the internet. The standard for this whole series is brutal: for classic patterns we cite Bulkowski's measurements, and for harmonics such measurements DO NOT EXIST — the "Encyclopedia of Chart Patterns" doesn't catalog them at all. Every "Gartley hit rate" you find is an anecdote or marketing. There's a second risk few people mention out loud: subjectivity in drawing. Two traders looking at the same chart can mark different X and A swings — and one will see a Gartley, the other nothing. So before you trade it with real capital, test the pattern in your own journal, with your swing-selection rules fixed in writing. That's the only statistic that proves anything.

FAQ

What is the Gartley pattern? A five-point harmonic XABCD structure from 1935 (H.M. Gartley, "Profits in the Stock Market"), shaped like the letter M (bullish) or W (bearish). Key ratios: B = 61.8% XA, C = 38.2–88.6% AB, D = 78.6% XA. It's a correction-in-trend pattern — D doesn't go beyond point X.

How does the Gartley differ from the Bat? In proportions, with an identical shape. Gartley: B = 61.8% XA and D = 78.6% XA. Bat: B = 38.2–50% XA and D = 88.6% XA. The Bat's shallower B and deeper D mean a tighter stop behind X — which is why you check the depth of B first.

Where's the stop loss and what are the targets? Stop behind point X with a buffer — a break of X invalidates the setup by definition. Targets in order: 38.2–61.8% retracement of CD (near B), the area around point C, and on trend continuation, the area around A. After the first target, it's worth banking part of the position and moving the stop to breakeven.

FAQ

Rafał — Krypto Bez Ściemy

Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.

kryptobezsciemy.plInstagramYouTubeKlub Zero