ICT Fractal Theory — How Each Timeframe Mirrors the Next

ICT fractal theory explained — the five fractal levels from quarterly to micro, the nesting of AMD cycles within AMD cycles, fractal alignment as the maximum-probability entry condition, the 1M-precision fractal entry, and the three fractal failure modes.
Take any ICT chart and hide the time scale. You cannot tell if it is a monthly chart or a 5M chart. The accumulation zone — overlapping small candles building a range — looks the same on the monthly as on the 5M. The Judas sweep — a sharp wick through the boundary, a close back inside, an immediate reversal — looks the same on the weekly as on the 15M. The MSS displacement creating the FVG looks the same on the daily as on the 5M.
This is the ICT fractal: the same institutional delivery sequence — accumulation, manipulation, distribution — repeating at every timeframe scale, because the same institutional mechanics that produce a monthly AMD also produce a 5M micro-AMD. The scale changes; the structure does not. This guide explains the five fractal levels, the nesting relationship between them, fractal alignment as the maximum-probability entry condition, the 1M-precision fractal entry, and the three fractal failure modes behind the most common ICT trading errors.

Key takeaways

  • One structure, five scales: quarterly, weekly, daily, session, micro — each AMD contains the next smaller one inside each phase.
  • The Tuesday London Judas is three Judas sweeps in one event: session, daily, and weekly Phase 2 simultaneously.
  • Wednesday–Thursday are the best days because three fractal levels sit in Phase 3 together — alignment, quantified.
  • The fractal entry pairs the tightest stop (1M CHoCH, 3–6 pips) with the largest target (IPDA 40–60-day, 200–500+ pips).
  • Top-down analysis isn’t a discipline — it’s the fractal read in the only logical order: outermost level inward.

What ICT fractal theory is

Definition

  • The ICT fractal is the observation that the AMD cycle — Accumulation, Manipulation/Judas sweep, Distribution — repeats structurally identically at every timeframe scale. A monthly AMD contains weekly AMDs within each phase. Each weekly AMD contains daily AMDs. Each daily AMD contains session AMDs. Each session AMD contains micro-AMDs. At every scale the structure is the same: Phase 1 builds a range, Phase 2 sweeps one boundary to collect liquidity, and Phase 3 delivers price to the ERL in the true direction. The time duration, pip magnitude, and institutional scale differ. The structural sequence does not.

Not a mathematical fractal — a structural fractal

The ICT fractal is not a fractal in the Mandelbrot sense — it does not have infinite self-similarity at every level of magnification. It is a structural fractal: the same three-phase sequence repeating at the specific scales that correspond to institutional AMD planning and execution horizons. Those scales are not arbitrary. The quarterly AMD exists because central banks and sovereign wealth funds plan quarterly. The weekly AMD exists because institutional desks operate weekly delivery cycles. The daily AMD follows the 24-hour FX session structure. The session AMD exists because the London and NY kill zones are the execution windows within each day. And the micro-AMD exists because even intraday scalp positions are filled with the same three-phase mechanic at the smallest executable scale.

Why the fractal exists — the universal institutional mechanic

The fractal exists because accumulate → sweep liquidity → deliver to the ERL is the only viable execution strategy for any order large enough to impact the market. A central bank filling a $10 billion EUR/USD position and a proprietary trader filling a $1 million intraday position both use the same three-phase mechanic, because it is the only way to fill size without adverse impact. The central bank’s quarterly AMD and the prop trader’s micro-AMD are the same mechanical logic at different scales — the fractal is the observable consequence of a universal execution constraint.

The zoom test

  • Any ICT chart with the time scale hidden is unidentifiable as monthly, weekly, daily, or 5M — because the OBs, FVGs, liquidity sweeps, and AMD phases are structurally identical at every scale. This is the fractal’s most compelling proof: the tools work at every timeframe because the structure is the same at every timeframe.
THE ZOOM TEST ? ? ?
The zoom test — the same AMD structure at three hidden scales Three chart panels with the time scales hidden, each showing the identical AMD sequence: an accumulation range, a Judas wick sweeping below the range low, a reclaim candle, a displacement candle breaking above the range, and a delivery leg. One panel is a monthly chart, one a daily, one a five-minute chart — and they cannot be told apart, which is the fractal argument. WHICH IS THE MONTHLY? WHICH IS THE DAILY? WHICH IS THE 5M? CHART ①CHART ②CHART ③ TF: ▓▓▓ · RANGE: ▓▓▓ TF: ▓▓▓ · RANGE: ▓▓▓ TF: ▓▓▓ · RANGE: ▓▓▓ ACCUMULATION → JUDAS → MSS → DELIVERY — IN ALL THREE answer: it doesn’t matter which is which — the structure is identical. that is the fractal.
The zoom test, rendered. Three charts, time scales hidden. Each shows the same sequence: an accumulation range, a Judas wick through the range low, a reclaim, an MSS displacement above the range, and the delivery. One is a monthly chart, one a daily, one a 5M — and there is no way to tell which is which. The tools transfer across every timeframe because the structure they describe exists at every timeframe.

The five ICT fractal levels — from quarterly to micro

The five levels correspond to the five institutional AMD planning and execution horizons. Each contains all lower levels within it — a quarterly fractal contains many weekly fractals, which contain daily fractals, which contain session fractals, which contain micro-fractals:
Level 1 · Monthly chart · 3-month cycle
The quarterly fractal

Phase 1: 3–6 months of monthly candles building the quarterly range; monthly OBs forming; the quarterly CE establishing the macro premium/discount reference. Phase 2: the quarterly Judas — a 1–3 month sweep of the quarterly boundary, targeting the IPDA 20/40/60-day high or low. Phase 3: monthly delivery toward the IPDA 60-day extreme, with multiple weekly AMDs nested inside it. Entry: daily OB after the weekly Judas confirms the quarterly Phase 3 · stop at the daily OB low (40–80 pips) · target the IPDA 60-day high (200–600+ pips) · the position/swing trader’s level. Tools: monthly OBs, IPDA 20/40/60-day ranges, monthly CE, weekly MSS.

Level 2 · Weekly chart · 5-day cycle
The weekly fractal

Phase 1: Monday’s narrow range establishing the weekly CE and the week’s Judas targets. Phase 2: the Tuesday–Wednesday Judas sweep of the Monday boundary — the weekly liquidity collection that identifies the true weekly direction. Phase 3: Wednesday–Friday delivery toward the PWH or PWL, with daily AMDs nested inside. Entry: daily OB or 4H FVG after the Tuesday Judas and weekly MSS · stop at the daily OB low (20–60 pips) · target PWH/PWL · the swing trader’s level. Tools: PWH/PWL, weekly CE, daily OBs in weekly discount, 4H FVGs, weekly MSS.

Level 3 · Daily chart · 24-hour cycle
The daily fractal

Phase 1: the Asian session — the daily range builds, the Asian range boundaries become the Judas targets. Phase 2: the London-open Judas (02:00–02:40 EST) sweeping the Asian boundary — the daily liquidity collection. Phase 3: the London kill zone delivery (02:33–05:00 EST) toward the PDH or PDL, with session AMDs nested inside. Entry: 5M FVG CE after the 02:33 MSS · stop at the Judas extreme minus 2–3 pips (8–15 pips) · target PDH/PDL · the session model trader’s level. Tools: PDH/PDL, Asian range, 5M FVG CE, 02:33 MSS, London kill zone.

Level 4 · 15M/5M charts · 3–4 hour cycle
The session fractal

Phase 1: the first 15–30 minutes of the kill zone — the session range forming on overlapping 5M candles. Phase 2: the 5M-level Judas within the kill zone, collecting the session-scale stops at the range boundary. Phase 3: the 5M delivery from the FVG CE to the PDH, with micro-AMDs nested inside. Entry: 5M FVG CE limit or 1M CHoCH after the 5M MSS · stop at the FVG far boundary (6–12 pips) · target the PDH (session ERL) · the Model 1/2 trader’s level. Tools: 5M FVG CE, 1M CHoCH, 5M BOS trailing, kill zone timing.

Level 5 · 1M/15-second charts · 15–30 minute cycle
The micro fractal

Phase 1: 1M accumulation inside the 5M FVG zone — small candles building the micro-range within the gap. Phase 2: the 1M micro-Judas — a 1M wick sweeping the micro-range’s SSL within the FVG. Phase 3: the 1M CHoCH above the prior 1M swing high — the scalp entry trigger — delivering to the 5M FVG near-boundary as the first micro-ERL. Entry: 1M CHoCH close (market order) · stop at the micro-Judas extreme (3–6 pips) · target the 5M FVG near-boundary · the scalp model trader’s level. Tools: 1M OBs, 1M CHoCH, 15-second AMD observation, micro-accumulation within the 5M FVG.

The scale relationships are quantitative: a quarterly AMD contains roughly 13 weekly AMDs; each weekly AMD contains 5 daily AMDs; each daily AMD contains 2 session AMDs (London + NY AM); each session AMD contains 3–5 micro-AMDs within its kill zone window. A single 1M scalp entry at the micro level is therefore simultaneously a sub-delivery event within all five levels above it — and when all five are in Phase 3 in the same direction, the entry has maximum fractal backing.

Fractal nesting — how one week contains every level

The clearest way to see the nesting is to trace a single bullish weekly AMD from Monday to Friday and identify the fractal event at each scale on each day:
Monday · weekly Phase 1
Accumulation — the week’s targets form

Weekly fractal: Phase 1 — Monday’s range establishes the weekly Judas targets: the Monday high (BSL) and Monday low (SSL). Daily fractal: Monday still contains its own complete daily AMD — Asian accumulation, London Judas, delivery toward the Monday PDH — a sub-fractal running inside the weekly Phase 1. Session fractal: each London and NY session on Monday runs its own session AMD inside that daily AMD.

Tuesday · weekly Phase 2
The triple Judas — three sweeps in one event

Weekly fractal: the weekly Judas — price drives to the Monday range boundary, collects the weekly liquidity, and begins the weekly MSS. Daily fractal: Tuesday’s London kill zone IS the daily Judas AND the weekly Judas simultaneously — the same candle sweeping the Asian range low is sweeping the daily-level SSL and the weekly-level SSL (the Monday boundary) at once. Session fractal: the 02:15–02:40 EST Judas is a session Phase 2 + daily Phase 2 + weekly Phase 2 in one event — maximum fractal alignment at Phase 2, which means the Phase 3 that follows carries maximum alignment too.

Wednesday–Thursday · weekly Phase 3
Primary distribution — the alignment peak

Weekly fractal: Phase 3 delivery toward the PWH — the highest-delivery days of the weekly cycle. Daily fractal: each day runs its own full AMD, and each daily Phase 3 is a sub-delivery within the weekly Phase 3. Session fractal: the Wednesday 02:33 EST entry is simultaneously a session Phase 3 entry, a daily Phase 3 entry, AND a weekly Phase 3 entry — full three-level alignment on the primary delivery day.

Friday · weekly Phase 3 completion
Transition — the alignment trough

Weekly fractal: the delivery completes at the PWH; the BSL above it is collected and becomes the next weekly AMD’s Phase 1 reference. Daily fractal: a completion day — the PDH target may coincide with the PWH; position squaring ahead of the weekend lowers conviction. Session fractal: Friday entries carry the lowest alignment of the week — the weekly Phase 3 is ending, not beginning, and weekend gap risk reduces institutional positioning. Apply the conservative 75% partial protocol.

The nesting reveals the single most important insight in this article: Wednesday and Thursday have the highest fractal alignment of the week — session Phase 3, daily Phase 3, and weekly Phase 3 simultaneously. Every ICT convention that emphasises Wednesday–Thursday as primary delivery days is grounded in exactly this. The Friday flip side matters equally: the same 5M FVG CE entry that carries three aligned levels on Wednesday carries at best two on Friday, with the weekly Phase 3 winding down toward the next week’s accumulation.

Fractal alignment — the maximum-probability entry condition

Definition

  • Fractal alignment is the condition in which multiple consecutive fractal levels are simultaneously in Phase 3 delivery in the same direction. The more levels aligned, the higher the probability of the lower-timeframe entry — because more institutional scales are delivering in the same direction at once, and the opposing order flow at each scale is being absorbed by institutional buying at that scale. The pre-session checklist is a fractal alignment audit: each item confirms one level’s Phase 3 status.
ALIGNMENT LADDER ICT
The fractal alignment ladder Five bars of increasing width representing how many fractal levels are simultaneously in Phase 3: one level, session only, is below threshold and not a valid entry; two levels adds the daily and is acceptable; three levels adds the weekly and is good, the Wednesday to Thursday standard; four levels adds the quarterly and is strong with full risk; five levels adds the one-minute micro fractal and is the theoretical maximum probability entry. HOW MANY FRACTAL LEVELS ARE IN PHASE 3 TOGETHER? 1 · SESSION ONLY 2 · + DAILY 3 · + WEEKLY 4 · + QUARTERLY 5 · + MICRO (1M CHoCH) below threshold — ✗ not a valid entry acceptable — minimum Model 1/2 good — the Wed–Thu top-down standard strong — full 1% risk MAXIMUM EACH PRE-SESSION CHECKLIST STEP AUDITS ONE LEVEL — THE FULL ROUTINE IS A FULL ALIGNMENT AUDIT
Probability grows with the stack. One level (session only) is below threshold — a naked 5M structure without higher-TF backing. Two levels is the Model 1/2 minimum; three is the Wednesday–Thursday top-down standard; four adds the quarterly IPDA direction and carries full risk; five — the 1M CHoCH firing inside the aligned stack — is the theoretical maximum-probability ICT entry.
The five alignment tiers in detail
AlignmentLevels in Phase 3TierChecklist requirementEntry quality
Minimum (1)Session onlyBelow thresholdSession Judas swept + MSS formed; no daily/weekly/quarterly checkNot a valid ICT entry — single-fractal sessions lack higher-TF backing.
Basic (2)Session + dailyAcceptableDaily bias confirmed + session Judas + MSSMinimum valid Model 1/2 entry — two levels aligned.
Good (3)+ weeklyGoodWeekly Phase 3 (Wed–Thu primary delivery day) + daily + sessionThe standard full top-down entry — the pre-session minimum for experienced traders.
Strong (4)+ quarterlyStrongIPDA delivering toward the 20-day high + weekly + daily + sessionMaximum standard quality — all four analysis levels aligned; carry full 1% risk.
Maximum (5)+ micro (1M CHoCH)MaximumFull 4-level alignment + the 1M CHoCH trigger inside the 5M FVG zoneAll five levels in Phase 3 — the theoretical maximum-probability entry.

The pre-session routine as an alignment audit

Each step of the pre-session routine checks one fractal level: the monthly/quarterly AMD check confirms Level 1; the weekly AMD and PWH/PWL confirm Level 2; the daily bias and PDH/PDL confirm Level 3; the Asian range marking identifies the daily fractal’s accumulation boundaries; and the session Judas + MSS confirmation captures the Level 4 Phase 2 → Phase 3 transition. A session that passes all steps has confirmed alignment at Levels 1–4 before the 5M limit is placed — and adding the 1M CHoCH at execution brings the count to five.

The fractal entry — 1M precision for the quarterly AMD

The concept

The fractal entry applies the lowest-TF precision — the 1M CHoCH scalp trigger — to the highest-TF direction — the quarterly AMD Phase 3. The result: the tightest possible stop (typically 3–6 pips, at the 1M micro-Judas extreme) on a trade whose target is the IPDA 40–60-day high, 200–500+ pips away. The fractal makes this structurally valid because the 1M micro-AMD is a genuine sub-delivery event nested at the bottom of the quarterly stack — not an unrelated squiggle.

Why the 1M stop is structurally valid

The 1M stop invalidates the 1M micro-AMD — not the quarterly AMD. If it triggers, the quarterly Phase 3 rebalancing simply went deeper than the 1M micro-structure could contain; the quarterly direction is untouched, and the FVG CE limit (if not already filled) catches the next rebalancing at a deeper level. The 1M stop is the tightest valid stop for the current micro-fractal position: the highest theoretical RR available for the quarterly AMD while maintaining structural precision at the smallest scale.

The fractal entry requirement

  • The fractal entry requires full alignment at every level above the 1M — quarterly + weekly + daily + session all in Phase 3 simultaneously. Without full alignment, the 1M CHoCH is simply a scalp entry. Full alignment converts the scalp into a position entry with scalp precision: the tightest stop for the longest-duration trade.

The fractal and top-down analysis — structurally non-negotiable

Top-down is the fractal made operational

Top-down analysis is not a discipline or a preference — it is the fractal structure of the market made operational. A 5M FVG cannot be evaluated without the daily, weekly, and quarterly fractal phases, because its significance depends entirely on which phase those levels are in. A 5M bullish FVG within a daily Phase 3 bullish AMD = a valid long entry. The identical 5M bullish FVG within a daily Phase 2 bearish Judas = a counter-trend retracement, not a valid long. Identical chart pattern; opposite fractal context; opposite correct response.

The bottom-up fractal error

Starting the analysis from the 5M and working upward reads the most nested fractal first and then tries to infer its context from the outer levels — like reading a sentence backward. The meaning of every lower-TF element is defined by the higher-TF context: a 5M MSS within the daily Phase 2 is not a Phase 3 commencement but a temporary sub-delivery within the Judas. Worse, bottom-up reading introduces the most damaging cognitive bias in the process: the entry intention forms from the 5M MSS first, and the higher-TF review degrades into confirmation-seeking rather than genuine context assessment.

The three fractal failure modes

Treating the 5M AMD as primary context

A clean 5M AMD — accumulation, Judas, MSS, Phase 3 — gets traded from the 5M FVG CE without checking the daily or weekly phase. If that 5M bullish Phase 3 is unfolding inside a daily Phase 2 bearish Judas, the long is a counter-trend retracement within the larger bearish structure — the highest-failure-rate ICT entry type. Correction: before any 5M entry, confirm the daily phase. Daily Phase 3 bullish → the 5M Phase 3 is a valid sub-delivery. Daily Phase 2 bearish → the 5M “Phase 3” is a retracement inside the Judas — invalid.

Ignoring the quarterly fractal (IPDA context)

Weekly and daily both bullish — but the quarterly AMD is delivering toward the IPDA 60-day LOW. The weekly and daily bullish AMDs are then sub-deliveries within the quarterly bearish structure: temporary rallies inside a larger AMD that eventually continues lower. Correction: before each week, check the IPDA delivery direction. Quarterly bearish Phase 3 → weekly bullish rallies are contra-quarterly entries: maximum caution, targets reduced from the PWH to 50% of the PDH distance.

Misreading a same-candle micro-fractal as “noise”

On the 15M: a single large candle wicks through the Asian range low and closes back above it. The retail read: “one volatile candle.” The fractal read: a complete micro-AMD inside one candle — the wick is the micro-Phase 2 Judas (SSL collection), the body is the micro-Phase 3 MSS (bullish close above the swept level). Correction: a wick through a structural level that closes back beyond it is a micro-fractal Phase 2→3 transition. The following candles are the micro-Phase 3 delivery — from which the 5M FVG entry should be taken.

The three failure modes share one structure: applying the framework at the wrong scale, without the higher-fractal context that defines the lower-fractal signal’s validity. The fix is the same for all three: top-down fractal reading, always from the highest available level to the lowest, before any entry decision at the 5M or 1M level.

FAQ — ICT fractal theory

What is ICT fractal theory? +
The observation that the AMD cycle (Accumulation → Manipulation/Judas → Distribution) repeats structurally identically at every timeframe scale. The same OBs, FVGs, sweeps, and three-phase sequences appear on the monthly, weekly, daily, 15M, 5M, and 1M charts because the same institutional delivery mechanics operate at every scale. The timeframe stack is not a collection of independent charts — it is a nested system, each scale containing the next smaller one.
How does fractal theory explain top-down analysis? +
Top-down analysis is the fractal made operational. A 5M FVG’s validity depends entirely on the daily, weekly, and quarterly phases: the same bullish FVG is a valid long inside a daily Phase 3 but a counter-trend retracement inside a daily Phase 2 Judas. Checking quarterly → weekly → daily → session before the 5M entry confirms the phase at every containing level. It is not optional, because the nesting exists whether the trader checks it or not.
What is fractal alignment? +
The condition in which multiple consecutive fractal levels are simultaneously in Phase 3 in the same direction. More aligned levels = more institutional scales delivering together = higher entry probability. Full five-level alignment (quarterly, weekly, daily, session, micro) is the maximum-probability entry condition — and the pre-session routine is the audit: each step checks one level’s Phase 3 status.
Why are Wednesday and Thursday the best trading days? +
Highest fractal alignment of the week: the weekly AMD is in Phase 3 (its primary delivery days), each day’s daily AMD runs its own Phase 3 inside it, and the session AMD adds a third level. Three levels in Phase 3 on Wednesday–Thursday versus one or two on Monday (weekly Phase 1) or Friday (weekly Phase 3 completing). More levels in Phase 3 = higher institutional backing.

Conclusion — the fractal is why ICT is a unified system

The fractal is not an abstract layer on top of the framework — it is the structural explanation for why every element of the framework exists. The OB appears at every timeframe because the fractal places an accumulation event at every scale. The FVG appears at every timeframe because it places a displacement event at every scale. The kill zones exist because the fractal operates on institutional time windows, and London and NY are where the session-fractal Phase 2→3 transition consistently occurs.
Understanding the fractal converts the toolkit into a system: the 5M OB and FVG are the same accumulation and displacement mechanics as their monthly counterparts, at the session scale instead of the quarterly. Top-down analysis is the fractal read in the only logical order — outermost containing level inward to the entry level. Alignment scoring quantifies how many levels are in Phase 3 together. And the fractal entry — 1M precision for the quarterly AMD — is the most elegant expression of what the nesting makes possible: the tightest stop for the longest-duration trade, because the smallest fractal lives inside the largest.
The fractal at each scale, in the existing guides: the AMD cycle covers the three-phase structure the fractal repeats; the expansion & retracement guide covers the sub-deliveries within each Phase 3; the algorithm guide explains why the mechanics are scale-independent; the time-price theory and weekly bias guides cover the timing stack the levels run on; and the 1st presented FVG guide implements the session-fractal entry with the nested 1M micro-AMD. Or join the mentorship for direct feedback on your fractal alignment audits.
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.