LiquiditySweeps.com
Mentorship
Home
›
Articles
›

Expansion vs Retracement in ICT Price Delivery

Expansion and retracement in ICT explained — the five real-time candle characteristics, the three retracement depth categories, the expansion-retracement ratio as an AMD quality metric, the AMD phase mapping, and what straight-line delivery signals.
When the algorithm executes a Phase 3 delivery, it does not move in a single continuous thrust from the FVG entry to the PDH target. It delivers in alternating legs: expansion — directional displacement advancing toward the ERL — followed by retracement — the return to the FVG zone to fill the unfilled institutional orders — followed by expansion again. Reading which leg is forming in real time is what separates traders who enter on the correct leg from those who enter on the wrong one.
Entering on an expansion candle means buying at the session high of the current delivery leg — the worst structural price — then watching the retracement stop the position out before Phase 3 resumes. Entering on the retracement means buying at the FVG CE — the densest institutional order concentration — with the structural stop at the far boundary and the full delivery ahead. This guide defines both legs precisely, gives the five real-time candle characteristics, maps the three retracement depth categories to their AMD quality meanings, introduces the expansion-retracement ratio as a live quality metric, and explains what straight-line delivery signals.

Key takeaways

  • The expansion is the signal; the retracement is the entry. The expansion creates the FVG — the retracement fills its CE.
  • Five real-time tells separate the legs: body size, close position, FVG creation vs fill, BOS vs no counter-BOS, volume character.
  • Retracement depth vs the prior leg: shallow <38% (max conviction), standard 38–62% (CE fills), deep 62–79% (OTE fills), >79% (reassess).
  • The E/R ratio (expansion pips ÷ retracement pips) scores AMD quality live: 3:1+ strong, 2:1 healthy, below 1.5:1 reassess.
  • The framework applies to Phase 3 only — applied to the Judas it produces the most expensive error in session trading.

ICT definitions — expansion and retracement

Definition

  • In ICT, the expansion is the signal and the retracement is the entry. The expansion candle creates the FVG zone; the retracement delivers price to the FVG CE where the institutional limit orders — and the ICT trader’s limit order — are filled. A trader who enters on the expansion candle is entering at the worst structural price in the delivery leg. A trader who enters on the retracement is entering at the best structural price: the FVG CE, at the densest institutional order concentration, with the structural stop at the far boundary of the zone.

Expansion — Phase 3 delivery in motion

An expansion candle or leg is a directional displacement sequence that: advances price toward the ERL in the AMD delivery direction; creates at least one C1-C2-C3 FVG (C1 high below C3 low for a bullish expansion); closes in the upper 25% of its range (bullish) or lower 25% (bearish), confirming the delivery state; and produces a new BOS above a prior swing high — the structural confirmation that Phase 3 is advancing. Expansion is what Phase 3 looks like while it is happening: the algorithm executing institutional order delivery toward the ERL at scale.

Retracement — the fill obligation in action

A retracement leg is the counter-directional move that follows an expansion and returns price toward the FVG or OB zone the expansion created. A valid Phase 3 retracement: moves counter to the delivery direction; does not close beyond the FVG far boundary (that would be structural mitigation); does not produce a counter-directional BOS (that would be a CHoCH, signalling potential direction change); and is smaller in absolute magnitude than the preceding expansion in a healthy delivery. The retracement is the algorithm completing its order fill at the CE before the next leg continues.

The central reframe

  • The retracement is not a threat to the position. It is the planned, expected institutional order fill event that ICT traders specifically position for with a limit at the FVG CE. The expansion creates the opportunity; the retracement delivers the entry. Without the retracement, the FVG CE limit order never fills.
EURUSD 5M
The expansion-retracement rhythm of one Phase 3 delivery Candlestick chart of a bullish Phase 3: a 40-pip expansion leg creates the first fair value gap, an 18-pip retracement fills its consequent encroachment for the first entry at an expansion-retracement ratio of 2.2 to 1, a 28-pip second expansion creates a second gap, a 12-pip retracement fills its midpoint at 2.3 to 1, and a final expansion delivers to the prior day high. 1.08401.0820 1.08001.0780 PDH 1.0842 — ERL TARGET E1 +40 R1 −18 · E/R 2.2:1 E2 +28 R2 −12 · 2.3:1 E3 → PDH ① CE fill 1.0801 ② CE fill 1.0816 PHASE 3 — EXPANSION · RETRACEMENT · EXPANSION · RETRACEMENT · EXPANSION
The rhythm, with the ratios live. E1 expands 40 pips and creates FVG1; R1 retraces 18 (E/R 2.2:1 — healthy) and fills the CE at 1.0801 (entry ①). E2 expands 28 and creates FVG2; R2 retraces 12 (2.3:1) and fills its CE (②). E3 completes the delivery at the PDH. Each expansion-retracement pair is one completed sub-delivery cycle — and each retracement, not each expansion, is where the entries happen.

Five candle characteristics — identifying each leg in real time

The distinction is observable within the first 30–60 seconds of a candle opening — before it closes — from five signals:
Expansion vs retracement — the real-time tells
CharacteristicExpansion candleRetracement candleReal-time identification
Body size vs rangeLarge body — 60%+ of the total candle range, minimal wicksSmall body — under 40% of the range, wicks proportionally largerMeasure the body against the range as the candle forms: a large body closing in one direction = expansion in progress.
Close position within rangeBullish: close in the upper 25%; bearish: lower 25% — the delivery state is strongly directionalClose near the midpoint (40–60%) or in the opposing 25% — neutral or counter-directional stateAsk: is this candle closing near its high (expansion) or its midpoint (retracement)? The close position identifies the type instantly.
FVG creation vs FVG fillCreates at least one C1-C2-C3 FVG — the displacement leaves a gap in the order bookReturns price into a prior FVG zone, filling it partially or fully; closes within the zone, not beyond the far boundaryThe single most definitive signal: is this candle (as C2) leaving a gap — or entering one?
BOS vs no counter-BOSProduces a BOS above the prior swing high — each expansion in a healthy Phase 3 makes a new higher highDoes NOT close below the prior swing low — that would be a CHoCH, not a retracementDid this candle close below the most recent swing low? YES = CHoCH warning. NO = valid retracement candidate.
Volume character (index futures)Higher volume — institutional conviction driving the displacement; volume expands with each legDeclining volume — reduced counter-pressure, demand absorbing the pullbackOn NQ/ES: rising volume on a counter-move = reversal warning; falling volume = healthy retracement.
The five characteristics work as a scoring system: 4–5 expansion signals = expansion with high confidence; 4–5 retracement signals = retracement with high confidence. Candles scoring 2–3 each way are transitional — accumulation candles belonging to neither, typical of early Phase 1 or the pause at the end of a retracement before the next leg. When in doubt, row three decides: a candle creating a C1-C2-C3 gap is expansion; a candle entering a prior gap is retracement.

The three retracement depth categories

Retracement depth — measured as a percentage of the prior expansion leg’s pip size — is the single most informative real-time AMD quality metric available from raw price action:
RETRACEMENT DEPTH GAUGE ICT
The retracement depth gauge Horizontal gauge of retracement depth as a percentage of the prior expansion leg: shallow under 38 percent means partial rebalancing and no fill, standard 38 to 62 percent is where the fair value gap consequent encroachment fills for the Model 1 and 2 entry, deep 62 to 79 percent fills the optimal trade entry levels, and beyond 79 percent is the reassessment threshold rather than a stop signal. RETRACEMENT DEPTH — % OF THE PRIOR EXPANSION LEG CE = 50% SHALLOW <38% STANDARD 38–62% DEEP 62–79% ⚠ >79% REASSESS 0% 38% 62% 79% 100% max conviction — partial rebalancing · no fill FVG CE fills — Model 1/2 entry 62/79% OTE levels fill · verify AMD reassessment trigger — NOT a stop signal MEASURE AGAINST THE MOST RECENT EXPANSION LEG — NEVER THE CUMULATIVE PHASE 3 DISTANCE
Depth is the diagnosis. Shallow (<38%): price reverses before the CE — extreme conviction, likely no fill. Standard (38–62%): the CE at 50% fills — the primary Model 1/2 scenario. Deep (62–79%): the OTE levels within the zone fill — valid while no close breaches the far boundary. Beyond 79%: a reassessment trigger — check the daily AMD, the CHoCH risk, and the mitigation status, but never widen a structurally placed stop.
Shallow — under 38% of the prior leg
Extreme institutional conviction

Example: 40-pip expansion → 14-pip retracement = 35%. The CE at 50% of the leg is never reached — sufficient institutional buy orders filled at the zone boundary alone. Action: the CE limit did not fill (partial rebalancing). Do not chase above the CE. Journal: “Shallow retracement — strong AMD quality signal. No fill — valid no-trade session.” Multiple shallow retracements in a week = exceptionally strong weekly AMD direction.

Standard — 38–62% of the prior leg
Healthy delivery, the CE fills

Example: 40-pip expansion → 16–24-pip retracement. The FVG CE at the leg midpoint fills — the most common depth in a normal Phase 3 and the scenario the Model 1/2 limit is designed for. Action: the limit fills; stop at the far boundary minus 2–3 pips; 50% partial at the PDH (IRL), trail the remainder.

Deep — 62–79% of the prior leg
Moderate conviction, the OTE levels fill

Example: 40-pip expansion → 26–32-pip retracement. Demand concentrates at the deeper OTE levels rather than the CE; the far boundary sits near the 79% level of the leg. Action: the 62%/79% levels within the zone become the entry references; valid while no candle closes beyond the far boundary; stop unchanged at the far boundary minus 2–3 pips.

Warning — beyond 79% of the prior leg
Reassessment trigger — not a stop signal

Example: 40-pip expansion → 33+ pips = 82%+. Either the “expansion” was inducement, or the AMD direction is changing (an HTF CHoCH may be forming). Action: pause and check three things — daily AMD intact (no daily CHoCH)? FVG fully mitigated (close beyond the far boundary)? HTF structure supportive? All intact and the stop untouched: hold. Any check fails: close and reassess. And if the stop is structurally placed at the far boundary: let it do its job — the 79%+ depth is the warning, never a reason to widen the stop.

The depth table carries a counterintuitive truth: a partial rebalancing — price reversing before the CE is reached — is a stronger AMD quality signal than a standard CE fill. The algorithm reversed early because demand absorbed all counter-pressure at the zone boundary. The missed fill is frustrating — $0 P&L for the session — but the quality signal is the strongest available, and it belongs in the journal as a high-quality AMD day even though no trade was taken.

The expansion-retracement ratio — a real-time AMD quality metric

Definition

  • The expansion-retracement ratio (E/R ratio) = expansion leg pip size ÷ retracement leg pip size. Above 3:1 signals strong institutional delivery — expansion legs three or more times larger than retracements. 2:1 is healthy. Below 1.5:1 indicates weak delivery or potential inducement. Below 1:1 — the retracement larger than the expansion — indicates the prior “expansion” was not genuine Phase 3 delivery at all: likely Phase 2 inducement or Phase 1 accumulation masquerading as Phase 3.
The E/R ratio in practice — six worked readings
Expansion (pips)Retracement (pips)E/R ratioAMD qualityICT implication
40104.0:1ExcellentHigh-conviction Phase 3. Shallow retracement — partial rebalancing likely; expect fast PDH delivery.
40162.5:1StrongHealthy Phase 3, standard CE fill — the Model 1/2 limit fills; good probability of PDH delivery within the session.
40221.8:1HealthyNormal delivery, deep-standard retracement; CE fill at ~55% of the leg. Viable entry.
40281.4:1ModerateDeep retracement into the OTE zone (70%); price near the far boundary. Reduce confidence; verify the daily AMD.
40351.1:1WeakWarning threshold exceeded (87.5%). Possible inducement — daily AMD reassessment required.
20181.1:1Very weakNear-equal legs: not genuine Phase 3 — likely accumulation or inducement misread as delivery.
Calculate the ratio after each expansion-retracement pair completes. A 3-hour London kill zone typically produces 2–3 pairs within Phase 3; their average is the session AMD quality score. Sessions averaging above 2.5:1 should correlate with higher journal win rates; below 1.5:1, with more frequent stop-outs — because the AMD was not delivering with conviction. Practically: record it in the journal’s AMD-quality field — “1st pair: 40/18 = 2.2:1. 2nd pair: 28/12 = 2.3:1. Session average 2.25:1 — healthy delivery.” Over 30+ sessions, the E/R-to-win-rate correlation becomes one of the most useful backtest insights the journal produces.

How expansion and retracement map to the AMD phases

The framework applies to Phase 3 exclusively. Applied to Phase 1 or Phase 2 it produces incorrect conclusions and dangerous entries:
When the framework applies — and when it must not
AMD phasePrice characterExpansion/retracement roleICT action
Phase 1 — AccumulationOverlapping small-body candles; range-building, minimal displacementNo reference exists yet — no directional leg to measure againstMark the range boundaries (future Judas targets). Do not apply the framework. Observe only.
Phase 2 — Judas sweepLarge directional candles in the WRONG direction; may create bearish FVGs; visually identical to Phase 3 expansion invertedThe framework does NOT apply. The Judas is manipulation, not delivery — applying it produces short entries from bearish Judas FVGs on a bullish dayObserve only. The bearish “expansion” candles are Phase 2 liquidity collection — the setup for the bullish Phase 3 that follows. No entries.
Phase 3 — DistributionAlternating legs: large displacement (expansion) → smaller counter-move to the FVG CE (retracement) → next expansionApplies exclusively here. Each expansion-retracement pair is one completed sub-delivery cycleLimit at the FVG CE; stop at the far boundary; log the E/R ratio per pair; exit at the ERL or the kill zone close.
The most dangerous misapplication is explicit in the middle row: the Judas produces large bearish candles on a bullish day that look identical to bearish Phase 3 expansion. The difference is direction relative to the AMD bias. Bullish day + bearish expansion candles = Phase 2 (observe only); bullish day + bullish expansion candles = Phase 3 (enter on the retracement). The framework activates only after the Judas is confirmed and the MSS has signalled Phase 3 — applying it earlier produces short entries during the sweep, the most common and most expensive session trading error.

Expansion without retracement — straight-line delivery

What it looks like

Occasionally Phase 3 delivers from the MSS displacement directly to the PDH without a meaningful retracement. The MSS creates the FVG, Phase 3 begins, and consecutive expansion candles run to the target without any counter-directional leg. The FVG zone is never revisited; the limit at the CE never fills; the session produces no position and $0 P&L.

What it means — the strongest AMD quality signal

Straight-line delivery is the strongest institutional conviction signal a session can produce. The algorithm did not pause to rebalance because demand at every tick was sufficient to absorb all retail counter-pressure without returning to the zone. The resting orders technically remain — but the delivery pace is so aggressive that price never retraces far enough to fill them before advancing.

The straight-line response

  • Do NOT enter at market while straight-line delivery is in progress — a market entry above the FVG CE is the worst structural price without the institutional fill reference. The limit did not fill, which is the correct outcome. Record: “Straight-line delivery — strong AMD quality signal. FVG CE limit not filled. No trade taken. Session successful.”

Straight-line delivery as a weekly signal

Track the count per week. One or more straight-line sessions signals an exceptionally strong weekly AMD direction — the IPDA delivering at maximum velocity toward the PWH or PWL. A week with two or more is a week where the swing trade in the same direction should be held with maximum confidence toward the weekly ERL.

Four common expansion/retracement mistakes

Entering on the expansion candle

It feels like the strongest signal — large body, momentum, a new high forming. It is the worst entry in the leg: the maximum price before the retracement, no FVG reference below (the gap only exists once C3 closes), no precise structural stop, and maximum risk of being stopped out before Phase 3 resumes. The expansion is the SIGNAL. The retracement is the ENTRY.

Applying the framework to the Judas sweep

Phase 2 candles look identical to inverted Phase 3 expansion — and treating them as expansion produces bearish FVG “entries” in the sweep direction on a bullish day. The diagnostic is one question: is this expansion advancing in the AMD direction or against it? With it = Phase 3 (enter on the retracement). Against it = Judas (observe only, never enter).

Measuring depth against the wrong leg

The reference is always the most recent expansion leg — never the cumulative Phase 3 distance. A delivery 40 pips off the MSS whose most recent leg is 10 pips, followed by an 8-pip retracement, is an 80% retracement of the recent leg (warning breach) — not a comforting 20% of the full move. The cumulative misread produces false confidence exactly when caution is required.

Holding through a counter-directional BOS as if it were a retracement

A close below the prior swing low on a bullish day is a CHoCH — a structural break, not a deep pullback. A genuine retracement reverses inside the FVG zone without breaking the prior swing low. The difference is a single candle close, and that close is the invalidation signal: reassess the AMD immediately rather than holding in hope that Phase 3 resumes.

FAQ — expansion and retracement in ICT

What is the difference between expansion and retracement in ICT? +
Expansion: large-body directional candles advancing toward the ERL, creating FVGs, producing BOS events, closing in the top or bottom 25% of their range. Retracement: small-body counter-directional candles returning to the FVG/OB zone, producing no counter-BOS, closing near their midpoint. The defining relationship: expansion creates the FVG; retracement delivers price to its CE where the limit fills. The expansion is the signal; the retracement is the entry.
How deep should an ICT retracement go? +
Measured against the prior expansion leg: shallow (<38%) = strong conviction, partial rebalancing, the limit may not fill; standard (38–62%) = healthy delivery, the CE fills — the primary Model 1/2 scenario; deep (62–79%) = moderate strength, the OTE levels fill. Beyond 79% is the reassessment threshold: check the daily AMD, the CHoCH risk, and whether the FVG has been fully mitigated.
What does it mean when Phase 3 delivers with no retracement? +
Straight-line delivery — the strongest conviction signal available. The algorithm never needed to rebalance; the limit does not fill ($0 P&L, no trade), which is the correct outcome. Record it as a strong AMD quality signal and never chase at market above the CE. Multiple straight-line sessions in a week = a swing continuation signal toward the weekly ERL.
What is the expansion-retracement ratio and how do I use it? +
E/R ratio = expansion pips ÷ retracement pips. Above 3:1 strong; ~2:1 healthy; below 1.5:1 reassess; below 1:1 the prior “expansion” was likely not genuine Phase 3. Log the per-pair ratios and the session average in the journal as the AMD quality metric — over 30+ sessions the E/R-to-win-rate correlation becomes one of the most useful insights the journal produces.

Conclusion — every expansion is followed by its retracement, and every retracement is an entry

The expansion-retracement alternation is the fundamental rhythm of Phase 3 delivery. Every expansion creates the FVG; every retracement fills the CE. The expansion is the evidence that Phase 3 is advancing; the retracement is the specific entry window the limit order is designed to capture. A trader who reads the rhythm enters at the CE on every standard retracement — the best structural price in the leg — with the stop at the far boundary and the delivery ahead.
The E/R ratio converts the observation into a live metric: the larger the expansion relative to the retracement, the stronger the conviction. Shallow retracements signal maximum conviction; deep ones signal reassessment; straight-line delivery signals the strongest possible day — a missed entry that is simultaneously the best quality signal available. And the 79% breach is the single most important mid-session trigger: not a stop level, but a structural warning that the session’s balance has shifted and the AMD needs re-evaluation before the next entry decision.
The context pieces: the FVG guide covers the C1-C2-C3 mechanics the expansion creates and the retracement fills; the AMD cycle covers the three-phase structure that determines when the framework applies; the algorithm guide covers the institutional mechanics behind each leg; the CHoCH guide covers the structural signal separating retracement from reversal; and the 1st presented FVG guide implements the retracement-to-CE limit entry in full. Or join the mentorship for direct feedback on your leg identification and E/R journaling.
Lio
Founder & ICT trading educator, LiquiditySweeps.com

Lio has traded ICT and Smart Money Concepts on forex majors and US indices since 2021 and built LiquiditySweeps.com to teach the framework the way it should be learned: in sequence, on real charts, with free live tools instead of paid indicators. Every article on this site follows the same rule — nothing gets published that wouldn’t survive a trade review.

About All articles Mentorship
Ready to go further?
The mentorship programme covers the full framework — weekly live sessions, daily bias review, and personal trade review on every entry you submit.
View mentorship — $1,500/year

On This Page

Reading progress
0%Just started
Quick check
A large bullish candle just closed — strong body, new high, delivery confirmed. What now?
✓ Correct. The expansion candle is the signal — it creates the FVG. The entry is the retracement to that FVG’s CE, with the stop at the far boundary. Entering on the expansion means buying the leg’s maximum price right before the pullback.
Expansion-retracement routine
Related tools
🕐Killzone clockLive⚡Silver Bullet timerLive🌍Session clockLive