AB=CD compares two price-leg lengths separated by a retracement. An equal-length drawing does not supply a complete strategy: define when pivots become available, the entry and exit rules, position size and costs. The procedure below distinguishes geometry, executable trades and a test of net outcomes.

What the AB=CD pattern is

AB=CD uses ordered A, B, C and D. AB and CD move in the same direction, with BC retracing between them; the equal-leg version projects the AB price distance from C. Equal price distance does not force equal duration or parallel slopes on linear time/price axes. A separate time-symmetry filter needs its own definition before testing. D is a projected completion area, not proof of an order filled there.

The pattern was popularised by Larry Pesavento and Scott Carney, who combined the old observation about equal legs with precise Fibonacci measurement. If you are starting out, work through the basics of trading harmonic patterns first — AB=CD is their common denominator.

Structure and Fibonacci levels

Equal A-B and C-D legs require compatible retracement and projection ratios. A completion zone is a candidate reversal area, not a forecast.

Measure A-B first, then the B-C retracement and C-D projection. A B-C retracement near 0.618 pairs with a C-D projection near 1.618 of B-C; a retracement near 0.786 pairs with a projection near 1.272. A deeper B-C retracement alone does not identify a Bat, whose classification requires the additional X-A leg and its ratios.

Point D is a candidate completion zone where the projected A-B distance from C overlaps a compatible B-C projection. Pair approximately 0.618 with 1.618, or 0.786 with 1.272, rather than choosing either extension indiscriminately. Use Fibonacci extensions to compare those levels. A bullish pattern has A high and B low; a bearish pattern reverses that orientation. Completion identifies a possible reversal, not a guaranteed entry outcome.

Hypothetical example — bullish AB=CD on EUR/USD (illustrative values)
Point Aswing high at 1.1000, start of the A-B leg down
Point Blow at 1.0800; the A-B leg measures 200 pips
Point Cillustrative rounded C = 1.0924 gives BC 124 pips and BC/AB = 0.6200; an exact 0.618 retracement would give 1.09236
Point D — candidate zoneEqual legs give 1.0724; the compatible 1.618 B-C projection gives about 1.07234

How to recognise the formation step by step

Step 1 — define the A-B leg

Record A, B and C prices, timestamps and the rule that confirms each pivot. For a separate hypothetical model, AB is 100 pips and C at bar 20 is 1.09618. Require two closed bars to its right with lower highs: bar 21 high 1.09500 and bar 22 high 1.09400. Equal-leg D is 1.08618, already passed by bar 21’s low 1.08600. C becomes confirmed only at bar 22 close. A rule requiring confirmed C cannot retrospectively enter at the earlier first touch. Trading unconfirmed C or a later retest would be a different model.

Step 2 — measure the B-C correction

In Carney’s general definition, BC can retrace from 0.382 to 0.886 of AB; reciprocal BC projections accompany equal-leg completion. If a tested variant permits only approximately 0.618 or 0.786, state that restriction as its own rule rather than the definition of every AB=CD. Compute the compatible projection and tolerance before looking at later prices.

Step 3 — locate point D from two measurements

Project the A-B distance from C. Then compare it with the compatible B-C projection: approximately 1.618 for a 0.618 retracement, or 1.272 for a 0.786 retracement. Their overlap defines the candidate D zone.

Entry, stop and targets — a hypothetical example

In this hypothetical example, A is 1.1000, B is 1.0800 and C is 1.0924. Equal 200-pip legs put D at 1.0724; the 1.618 projection of the 124-pip B-C correction reaches about 1.07234. Suppose price tests that completion area, produces a reversal signal and then an entry fills at 1.0745. An illustrative stop at 1.0705 gives a planned 40-pip distance. An ordinary stop can fill worse during a gap or rapid move.

With D at 1.0724, the A-to-D distance is 276 pips. Its 38.2 and 61.8 percent retracements give illustrative targets near 1.08294 and 1.08946. From an entry at 1.0745, these offer about 84.4 and 149.6 pips, or 2.11 R and 3.74 R against a planned 40-pip stop, before costs and slippage. These are hypothetical calculations; they do not establish the probability of reaching either target.

What if two signals arrive together?

Both candidates may have valid geometry. A test also needs a rule for conflicting signals.

Scroll sideways to see the full diagram

What if two signals arrive together?At 10:00, candidates 12 and 27 are available. The predefined limit is one position, with the lower number winning a tie. Candidate 12 is accepted, free capacity falls from one to zero, and 27 is recorded as skipped. Later prices do not affect this choice.Two setups, one slotRule defined before the test:max. 1 position; tie → lower numberCandidate 12signal: 10:00rules: satisfiedCandidate 27signal: 10:00rules: satisfied12 < 27select: candidate 12Free slots: 1 → 0Candidate 27: recorded as skippedLater prices do not selectthe signal accepted earlier.

Explicit model: two hypothetical candidates are available at 10:00 and satisfy the recorded criteria. There is one free position slot. Our rule selects the smaller number, assigned beforehand, when signals arrive together. This is a test policy, not a platform setting or guarantee.

Accepting candidate 12 leaves 1 − 1 = 0 free slots. Candidate 27 is logged with the reason for skipping it, even if it later looks better. Changing the selection after seeing future prices would create a different test. In the TradingView documentation strategy conditions and the execution model are separate components; do not equate this custom rule with the pyramiding parameter alone.

The selection rule belongs to the strategy, not to the AB = CD ratios.

The most common mistakes when trading AB=CD

  1. Accepting ratios outside the chosen AB=CD specification without defining a different variant. A B-C retracement alone cannot establish that a formation is a Bat.
  2. Treating the 1.272 extension alone as sufficient — without matching the C-D length to A-B you lose the formation's most important filter.
  3. Treating equal price distances as equal durations or parallel slopes. A time-symmetry condition is an additional chosen filter, not implied by AB = CD.
  4. Entering before point D completes, while the C-D leg is still unfolding — that means jumping into a move that may keep extending the correction.
  5. Assuming a requested stop beyond D guarantees a fixed loss. Specify exits and test adverse fills separately; a candle wick or completed shape does not determine the actual execution price.

How AB=CD differs from five-point patterns

The most important difference is structural: AB=CD has four points, while patterns such as the Gartley, Bat or Crab have five (X-A-B-C-D). The extra X-A leg gives them a different reference — point D is measured against the XA leg, not the leg symmetry. It then lands on a 0.786 retracement of XA in the Gartley pattern, a deeper 0.886 in the Bat pattern and a 1.618 extension in the Crab pattern (the guide to Gartley and Bat patterns breaks those down).

The corresponding equal or alternate AB=CD calculation is one part of Carney’s larger structures. Alternate projections multiply AB by a stated factor; do not silently rename an unsuccessful equal-leg setup after price has passed D. The retracement explanation helps keep reference legs and units explicit.

Who this formation is for

Define and test the method before classifying it as easy or effective. In an independent hypothetical USD-account calculation, assume an already available signal and a EUR 10,000 purchase filled at Ask 1.08640, ordinary stop requested at 1.08440 and target 1.09040. Exact alternative Bid exits at those levels imply −USD 20 or +USD 40. With assumed total additional costs USD 3, net becomes −USD 23 or +USD 37. Spread is already in the fills. If the stop instead fills at 1.08390, net becomes −USD 28. These levels are not optimized rules or actual trades. Record unfilled limits and no-entry cases as well.

What to do tomorrow to learn the AB=CD pattern

  1. Open TradingView on EUR/USD on the hourly timeframe and review recent clear, two-leg moves, marking points A-B-C in turn — this trains you to see the formation's context before a tradeable D-zone entry appears on the chart.
  2. For each candidate, compare equal A-B and C-D distances with a compatible pair of ratios: approximately 0.618/1.618 or 0.786/1.272.
  3. Set up a simple spreadsheet journal with columns for the leg proportions, the boundaries of the D zone, the stop-loss level and the achieved risk-to-reward ratio, then fill it in after every demo trade to see your results in numbers.
  4. Place a price alert at the C-D symmetry completion level of the pair you are watching, instead of staring at the chart for hours — when price reaches the D zone you can calmly judge whether a reversal candle confirms the entry.
  5. Retain all signals, rejected candidates, unfilled orders and net losses, then check unchanged rules on a separate later period. MT5’s forward split supports that separation. Its pip-profit mode excludes commission, swap and margin checks, so it cannot replace a complete final cash evaluation. Twenty demo trades is not a universal proof of readiness or an edge.