Harmonic Patterns — The Complete XABCD and Fibonacci Guide

Harmonic patterns look like technical analysis for engineers: instead of a "more or less" double bottom, you get an XABCD structure in which every wave has to satisfy a specific Fibonacci ratio, and you calculate the turning point IN ADVANCE, before price ever gets there. The precision is impressive — and that's exactly why just as many legends have grown up around harmonics as around any other family of patterns.

This guide brings the whole family together in one place: XABCD geometry, the key ratios, a table of the six main patterns (Gartley, Bat, Butterfly, Crab, Shark, Cypher) and the rules for entering at point D. And in the section on the numbers — the honesty that's been missing most from write-ups on harmonics: what's actually measured, and what's just a repeated legend.

How a Harmonic Pattern Is Built (XABCD)

Lineage: the concept was started by H.M. Gartley in his book "Profits in the Stock Market" (1932–1935), Larry Pesavento added the Fibonacci ratios, and the modern canon — along with the Bat, Crab and Shark patterns — was codified by Scott Carney in his "Harmonic Trading" series.

The skeleton is five turning points and four waves:

• X — the start of the structure;
• XA — the impulse wave, the longest move in the structure;
• AB — the correction of wave XA; point B is the "eye" of the pattern and its first test: how deep it goes decides which pattern you're even drawing;
• BC — the correction of wave AB (usually 38.2–88.6% of AB);
• CD — the final wave, ending at the calculated point D — the potential reversal zone (PRZ), where you look for a trade AGAINST wave CD.

A bullish pattern draws the shape of a letter "M" stretched downward (D below or near the lows — you buy); a bearish one, an inverted "W" (D high up — you sell). Inside every five-point structure sits the simpler AB=CD pattern, which we cover separately — it's the primer worth starting with.

The ratios used are retracements: 38.2 / 50 / 61.8 / 78.6 / 88.6% and extensions: 113 / 127.2 / 141.4 / 161.8 / 224 / 261.8 / 361.8%. In practice a tolerance of roughly ±3 percentage points is accepted — the market doesn't stop dead on the tick at 78.6%. But careful: tolerance isn't permission to force a fit. If B retraced 70% instead of 61.8%, that's not "almost a Gartley" — it's not a Gartley.

The Six Main Patterns — Ratio Table

All six differ only in their wave proportions. Remember two things: the depth of point B (it selects the pattern) and the location of point D (retracement = D inside the XA range; extension = D beyond point X).

A quick word on personalities. Gartley — the classic and the benchmark: a shallow correction in a trend, D doesn't break X. Bat (Carney, 2001) — a shallower B, but a deeper D (88.6%); considered the most precise, because D sits right before X, so the stop is tight and invalidation is unambiguous. Butterfly — D goes BEYOND X (127.2%); instead of a correction it catches a fresh high or low, which is why it's sometimes used at exhaustion points. Crab — the deepest extension (161.8% XA), D furthest from the structure; commonly said to have "the best R:R," because reversals from such stretched levels can be violent. Shark — different notation (0XABC) and entry at point C, not D; a pattern from 2011, the forerunner of the 5-0 setup. Cypher (Darren Oglesbee, outside Carney's canon) — the only one where C goes ABOVE A, and D is measured from XC; measuring D from XA is the most common mistake with this pattern.

What the Numbers Say — and This Is Where It Hurts

In our articles on classic patterns we cite Thomas Bulkowski: tens of thousands of measured cases, failure rates, average moves, rankings. So where's the Bulkowski table for harmonic patterns? It doesn't exist. The "Encyclopedia of Chart Patterns" doesn't catalog harmonic patterns — not the Gartley, not the Bat, not the Crab. So there's no equivalent of the "13% failure rate, 46% average rise" numbers we give for the triple bottom.

So what is out there? First, educational material (StockCharts, Babypips, brokers) — it describes the rules but doesn't measure performance; phrases like "highly probable setups" in these pieces are marketing, not measurement. Second, private backtests by individual authors — sometimes interesting, but usually a small sample, one market, and an unverifiable methodology for recognizing the pattern. Third — and this is the crux of it — automatic detection of harmonics is hard (even their proponents admit it), and without automation there's no large, repeatable sample. The circle closes.

The conclusion, without the hype: the effectiveness of harmonic patterns is unmeasured to the standard we have for classic patterns. That doesn't mean they don't work — it means anyone quoting you a "success rate for the Gartley" is citing an anecdote, or an advertisement. Treat harmonics as a framework for precise entry and invalidation levels, not as a system with a known edge. One thing must be credited to them, though: of every pattern family, they have the most honestly defined FAILURE — the level beyond which the setup is dead, known before you ever enter.

How to Trade Harmonic Patterns

Step 1 — identify it while it's forming. The signal to watch for appears once the first three waves (XA, AB, BC) are complete: that's when you calculate where D will fall. Don't force the fit — if you have to "bend" the swings to make the ratios line up, it isn't a pattern.

Step 2 — a PRZ zone, not a point. D isn't a single price, it's a zone built from overlapping levels: the XA retracement + the BC extension + the AB=CD projection. The tighter the confluence, the better the setup. Example for a bullish Gartley: 78.6% XA, 127.2–161.8% BC and AB=CD equality all landing in the same spot.

Step 3 — enter on a reaction, not a blind limit order. Price reaching the PRZ is an invitation to watch, not a signal. The professional version (per Carney and practitioners): wait for confirmation — a rejection candle, a reversal formation on a lower timeframe — and remember that trading the PRZ is by definition a bet against the momentum of the last wave.

Step 4 — stop and invalidation. Stop behind point D with a buffer. In retracement-type patterns (Gartley, Bat), the natural boundary is X: if price breaks X, the correction has stopped being a correction. In extension-type patterns (Butterfly, Crab) — behind the outer ratio (e.g., behind 161.8% XA for the Butterfly).

Step 5 — targets. First zone: 38.2–61.8% retracement of the CD wave, or the area around point C. Second: the area around A, or 61.8–78.6% of XA measured from D. Sensible practice: bank part of the position at the first target and move the stop to breakeven, because by the nature of the setup you're catching a falling knife — risk control matters more than maximizing your hit rate.

Myth vs. Measurement — What This Fibonacci Thing Is Really About

The myth goes: "markets move in natural harmonic proportions because Fibonacci governs the universe." There's no measurement to back that metaphysics — and research into Fibonacci retracements alone produces, at best, mixed results. So why bother with harmonics at all?

For three hard, non-metaphysical reasons. First: selection discipline. Strict ratios filter out the 90% of "patterns" you'd draw everywhere with loose criteria — fewer setups means fewer random trades. Second: pre-defined risk. You know the entry level, the stop and the invalidation before the trade; that's rare in discretionary trading and a real, measurable advantage independent of whether 61.8% has magic properties. Third: confluence zones. Point D often overlaps with levels other traders are watching anyway — old support, a supply zone, a liquidity sweep — and then the harmonic is simply a precise record of a broader confluence. Notice, by the way, that a bullish D below the lows of a structure is functionally the same idea as a Wyckoff spring or an ICT sweep: buying where the market shook out the earlier buyers.

The standard disclaimer applies: the lack of large-scale statistics means the burden of proof is on you. Before you trade harmonics with real capital, test ONE pattern on your market and timeframe, tracking hits and misses in your own journal. That's the only statistic that proves anything.

FAQ

What are harmonic patterns? Five-point XABCD structures in which the waves must satisfy specific Fibonacci proportions (with a tolerance of roughly ±3 percentage points). The pre-calculated point D marks a potential reversal zone (PRZ), where you look for an entry against the last wave — with a stop just behind the structure.

Which pattern is the best? There's no measurement that settles this — rankings of harmonic patterns are opinions. The Bat is popularly considered the most precise (D = 88.6% XA, stop behind X), the Crab the most aggressive (D = 161.8% XA). It's wiser to pick one, test it, and measure, than to trust the legends.

How do you enter at point D? After a reaction from price inside the PRZ zone (a rejection candle, a structure shift on a lower timeframe), not with a blind limit order. Stop behind D or behind X, first target at 38.2–61.8% of the CD wave, partial profit and breakeven — because you're trading against momentum and risk control is the whole game.

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