What Are ICT Methods? An Honest Assessment

If you have spent any time in retail forex communities since around 2018, you have run into ICT: Inner Circle Trader, the methodology developed and taught by Michael J. Huddleston. It is, by a wide margin, the most-discussed trading framework in retail FX of the past decade. The terminology (order blocks, fair value gaps, liquidity sweeps, killzones, the power of three, optimal trade entry) saturates trading Twitter, YouTube tutorials, Discord servers, and the inevitable paid mentorships.

This article is the honest assessment. It is not a takedown and it is not an endorsement. The methodology contains real insights about how institutional order flow works, recycles several older ideas under new names, makes some claims that are well-supported and others that are not, and exists inside a marketing ecosystem with all the usual distortions. A reader who wants to evaluate ICT on the merits, or who wants to talk usefully to friends who follow it, needs the same careful breakdown we apply to everything else on this site.

What ICT actually is

Inner Circle Trader is a body of teaching content developed by Michael Huddleston starting in the early 2010s. Originally distributed through paid mentorships at significant cost (some packages were priced in the four-figure range), the bulk of the core curriculum has been re-released as a free YouTube series over time. The methodology sits within a broader umbrella now usually called Smart Money Concepts (SMC), a term applied more loosely to ICT-derived material and to overlapping work by other educators.

The underlying thesis is that retail trading patterns and institutional trading patterns are different, and that retail traders consistently lose because they read the market the way the textbooks teach (support, resistance, indicators) rather than the way institutional desks actually behave (liquidity-seeking, inventory-management, order-flow). ICT positions itself as teaching the institutional perspective.

That core premise has a real basis. Market microstructure research does show that dealer behaviour systematically differs from naive-retail behaviour, that liquidity provision creates predictable flow patterns, and that retail-style stops cluster at observable levels (see, for instance, the academic work on broker-trader strategic interaction, including arXiv:2412.20847 and 2412.17712). The question is how much of the ICT-specific apparatus on top of that premise is also well-founded, and how much is methodology that sounds institutional without being institutional.

The core ICT concepts, in plain English

A short tour of the most-used terms.

Market structure

The basic skeleton: identifying higher highs and higher lows (uptrend), lower highs and lower lows (downtrend), and the break of structure (BOS) or change of character (CHoCH) that signals a regime shift between them. This is essentially classical Dow Theory restated with newer vocabulary. The underlying observation is sound: trends end when they stop making new highs (or lows), and the first lower high in a confirmed uptrend is meaningful.

Order blocks

An order block in ICT terminology is the last opposite-coloured candle before a strong move in the other direction. The premise is that institutional orders were filled at that candle’s price area and that price will return to “mitigate” those orders before continuing in the original direction. Traders use the order block as a re-entry zone.

The classical analogue is the supply/demand zone, a concept predating ICT by decades. The order-block formulation adds specific criteria for identification, but the underlying claim (that price revisits areas of significant prior activity) is older than the ICT vocabulary.

Fair value gaps and imbalances

A fair value gap (FVG, also called an imbalance) is a three-candle pattern in which the middle candle’s body creates a visible gap between the first and third candles’ wicks. The premise is that price will return to “fill” the gap before continuing.

The observation underneath is genuine: visible gaps in price action do tend to be revisited, and the academic literature on mean reversion at short horizons supports a weak version of the claim. The ICT-specific framing as a directional trade trigger is a much stronger claim than the literature supports.

Liquidity pools and sweeps

A liquidity pool is a cluster of stop orders, typically just beyond an obvious support or resistance level (under a swing low, above a swing high, beyond a session high/low). A liquidity sweep or stop hunt is a move that reaches into the pool, triggers the stops, and reverses.

This is the most empirically defensible ICT concept. Stops do cluster at observable levels. Stops are sometimes deliberately triggered by larger participants, especially in less-liquid windows (see arXiv:2412.17712 for a formal model). The phenomenon is real, and reading the market with awareness of where retail stops cluster is genuinely useful. The ICT teaching here overlaps substantially with what professional traders call liquidity-seeking behaviour and what microstructure researchers call stop-running.

Killzones

Killzones are specific intraday time windows where, ICT claims, the highest-probability moves of the session occur. The most-cited ones are the London Killzone (roughly 02:00-05:00 New York time, the London open), the New York Killzone (07:00-10:00 New York time, US open), and the Asian Range (the overnight low-volatility window).

The session-volatility-pattern observation is real (FX volatility does cluster around session opens; see What Is the Forex Market). The ICT-specific framing of these as windows of predictable directional bias is a stronger claim that does not survive straightforward backtesting.

Optimal trade entry (OTE)

A specific Fibonacci retracement zone (between roughly 62% and 79%) where ICT teaches that trades have the highest probability of working. The claim is that retracements to this zone before continuation are the institutional re-entry pattern.

This is one of the weaker ICT claims. The academic record on specific Fibonacci ratios as predictors is unsupportive: the 62-79% zone is not statistically different from arbitrary retracement ranges of similar width. The cleanest published tests of Fibonacci predictive power consistently find no edge.

Power of three (PO3)

A daily-cycle model in which ICT divides the trading day into accumulation (range-building), manipulation (the false move in one direction, often during the London session), and distribution (the real move in the other direction during New York). The trader’s job is to wait for accumulation and manipulation and to position for distribution.

This is an interesting frame but, again, it is harder to validate than to articulate. Days that fit the PO3 model are visible in hindsight; days that do not fit it are easy to re-label. The pattern is suggestive, not testable in the form ICT typically teaches it.

What’s substantive

A fair summary of the ICT material that holds up on its merits:

• Awareness of where stops cluster is genuinely useful. Most retail traders place stops at obvious levels (under prior swing lows, above swing highs), and large participants do sometimes trade against those levels. Knowing this is part of mature market reading.
• The break-of-structure framework is a clean restatement of Dow Theory and is a defensible way to identify trend changes. Whether you call it BOS, CHoCH, or “the first lower high in an uptrend”, the observation has weak but real predictive content.
• Session-volatility awareness is real. Trading the London or New York open is structurally different from trading the Asian-overnight session. Understanding which window you are in matters.
• Order-flow-style thinking is healthier than pure pattern-spotting. Even where the specific ICT mechanics are questionable, the meta-habit of asking “who is buying, who is selling, where are the stops, who needs to hedge” is closer to how professional traders think than indicator-stacking.
• The framework forces traders to articulate a thesis. Even an imperfect framework that requires the trader to identify the setup, the entry, the stop, and the target before clicking is better than ad-hoc discretionary trading.

What’s marketed

The honest part of the assessment requires noting what the methodology, and the surrounding ecosystem, oversell:

• Specific predictive claims rarely survive testing. “The FVG always gets filled before continuation” is a stronger claim than the data supports. “Price always returns to the order block” is stronger still. The careful version of these statements (“price often returns to gaps and significant zones, at horizons that vary widely”) is much harder to trade than the marketed version.
• The terminology rebrands older ideas. Order blocks are supply/demand zones. BOS is Dow Theory’s trend reversal. OTE is Fibonacci retracement. The relabeling can give the impression of a unique edge where there is mostly classical content.
• The success-story selection is structural. ICT-affiliated social-media accounts post the trades that worked; the trades that did not work are absent from the marketing. This is the same pattern that affects signal-selling (see What Are Forex Signals).
• The mentorship economics matter. Paid ICT mentorships, paid Discord rooms, and “funded account” courses sit downstream of the methodology and generate the revenue that funds the marketing. The economic incentive of any educator is to grow the audience, not to publish negative results. The base-rate caution applies to ICT as to anything else.
• The “institutional perspective” framing is partly aspirational. Most ICT material is taught and traded by retail participants on retail platforms. Calling a methodology “institutional” because it uses institutional-sounding vocabulary is a marketing move, not a description of who is actually doing the trades.

The 74-89% rule still applies

The most important fact to anchor any discussion of ICT (or any other retail methodology) is the regulator-disclosed loss-rate data: 74-89% of retail CFD accounts lose money according to the mandatory ESMA disclosures. That population includes ICT followers in significant numbers. If ICT were a reliable solution to the retail-loss problem at the scale its marketing implies, the disclosure figures would have moved meaningfully over the years the methodology has been mainstream. They have not.

This does not prove ICT is worthless. It does mean that the “institutional secret” framing is incompatible with the population data. Whatever the methodology does, it does not transform the retail loss-rate.

What to take from ICT (and what to leave)

A pragmatic synthesis a reader can apply tomorrow:

Take:

• The habit of asking where stops cluster before placing your own.
• The discipline of identifying market structure (trend / range / transition) before looking for setups.
• The session-awareness habit: knowing whether you are trading the London open, the NY overlap, or the Asian quiet, and adjusting expectations accordingly.
• The meta-habit of articulating the trade in writing before clicking: where is the entry, where is the stop, where is the target, and what would invalidate the thesis.

Leave:

• The specific predictive claims about FVGs, order blocks, and OTE as standalone signals. They are weakly supported at best and the marketing routinely overstates them.
• The Fibonacci-ratio mysticism. The 62-79% zone is not meaningfully different from adjacent zones.
• The mentorship-pipeline framing. The economics of paid trading education work the same way they work for signal-selling.
• The “I now understand how the market really works” certainty. That phrase has been the leading indicator of an account blowup in every era of retail trading.

How to evaluate any methodology

The deeper lesson is methodological. The same questions apply to ICT, to classical technical analysis, to fundamental analysis, to any system you might encounter:

1. What is the predictive claim, precisely? Strong, falsifiable statements are easier to evaluate than vague ones.
2. What does the out-of-sample evidence say? Hand-picked chart examples are not evidence; systematic backtests on never-before-seen data are.
3. Who is making money from teaching the method? A method primarily monetised through education subscriptions, mentorships, and broker affiliate revenue has a different incentive structure than one primarily monetised through proprietary trading of the method.
4. Does the method survive the 74-89% test? If a method genuinely solved retail trading, the regulator-disclosed loss rates would be falling. They are not.
5. What does the methodology say about its own failure modes? Honest methods are explicit about when and where they fail. Marketing-driven methods describe themselves as if they always work.

ICT passes some of these tests and fails others. So does almost every methodology in the retail-trading literature. The right frame is not “is ICT right” but “which parts hold up and which parts are marketing”, and the answer is “some of both, in measurable proportions”.

The takeaway

ICT is a real, internally-coherent trading methodology with identifiable origins (Michael Huddleston, early 2010s), a large follower base, distinctive vocabulary that rebrands several older ideas, some well-grounded observations about market structure and liquidity behaviour, several specific predictive claims that do not survive rigorous testing, and a paid-mentorship economic layer that shapes how it is marketed. The 74-89% retail loss-rate disclosed by regulators applies to ICT-followers as much as to anyone else.

The substantive ICT content (structure-awareness, liquidity thinking, session timing, explicit trade theses) is worth internalising and overlaps with what professional traders actually do. The strong predictive claims, the specific Fibonacci-zone trading, and the “I now know what the institutions are doing” framing are weaker than the marketing implies and should be held loosely.

For the broader analysis context, read Fundamental vs Technical Analysis. For the structural reasons most retail systems struggle, read the risk-management and trading-costs pieces. The same discipline applies to any methodology, including this one.