What Is ICT Trading? The Inner Circle Trader Method, Tested on 578 Days of NQ Futures
Michael Whitman
·
Trading since 2017, full-time since 2020
·
YouTube creator, 137K subscribers
·
Published 2026-09-05
·
Updated
·
Reviewed
ICT is the Inner Circle Trader methodology, a named approach built on smart money concepts. Most of its vocabulary renames older ideas; its own additions are clock windows called killzones, a false move at the session open, and a fixed retracement band for entries.
Covered on this page: TradingView and LuxAlgo.
ICT trading is the Inner Circle Trader method taught by Michael J. Huddleston: find where stop orders rest, expect price to take them at set times of day, then trade the move that follows. Its real additions to smart money concepts are a clock and a narrow retracement range for entries.
Everything else on an ICT chart, from order blocks to fair value gaps, has an older and plainer name. This page translates the vocabulary, works through one real Nasdaq futures day by the method’s own labels, and then tests the clock on 578 days of the same market.
How it works
ICT sits on top of smart money concepts, and its core loop has five steps. The trader marks liquidity, meaning the clusters of stop orders above swing highs and below swing lows. Price is expected to run one of those clusters first: a liquidity sweep, which ICT calls a judas swing when it happens near the start of a session.
After the sweep, the trader waits for a market structure shift, a break of the last swing in the opposite direction (see market structure shift). The entry is a pullback into a gap or the last opposite candle, ideally inside the deeper part of the move. The exit aims at the next obvious pool of stops, the draw on liquidity.
Power of 3 frames the whole day as three phases: accumulation, manipulation and distribution. The power of three page covers it in full. On a bullish day the manipulation is a drop below the day’s opening price before the rally; on a bearish day it is a push above it.
The clock: killzones and the midnight open
The clock is the part of ICT that is genuinely its own. Two kinds of time anchor matter.
Killzones are windows in New York time when the sweep is expected. LiteFinance’s guide, read on 1 November 2026, gives the London killzone as 2:00 to 5:00 and the New York killzone as 7:00 to 10:00, the hours tested below. The silver bullet cuts the idea down to a single hour, and the trading sessions page explains why those opens matter at all.
The midnight open is the price at 00:00 New York time. LuxAlgo’s ICT library, read on 1 November 2026, describes it as the start of the “true day”. Above it is the expensive half of the day for a buyer, below it the cheap half, which is premium and discount applied to a clock instead of a swing.
This is the testable part. “Price seeks liquidity” can describe any chart after the fact, but a claim that the day’s false move lands in particular hours can be counted. The original data below counts it.
The vocabulary, translated
| ICT term | Plain meaning |
|---|---|
| Judas swing | a liquidity sweep near a session open |
| Power of 3 | range, false move, real move, as one day’s script |
| Midnight open | the price at 00:00 New York time, used as the day’s reference |
| Draw on liquidity | the next obvious high or low price is expected to reach |
| Optimal trade entry | a 62% to 79% pullback, a deeper slice of the discount half |
| Order block | an order block, the last opposite candle before a strong move |
| Fair value gap | a fair value gap, a three-candle gap |
| Breaker | a failed level that now works from the other side, the flip on the support and resistance page |
| Mitigation block | a level price came back to before continuing |
| Killzone | a window on the clock |
Three rows are new ideas: the two clock rows and the entry range. The rest are relabels, so a beginner who learns the plain versions first already knows most of ICT.
A worked example
Thursday 29 October 2026, Nasdaq-100 E-mini futures (NQ), 5-minute bars from Yahoo Finance. This day was picked by rule, not by eye: it is the most recent up day in the data whose low printed inside a killzone hour and below the midnight open.
The midnight open was 29,549.50, the first price Yahoo’s feed prints after midnight (its 00:00 to 00:05 bar is missing, so this is the 00:05 open). The overnight high came early, at 29,764.25 in the 1:15 AM bar, 214.75 points above the open.
Price drifted lower through both killzones and was at 29,611.00 at the 9:30 AM stock market open. In the 9:35 bar it printed the day’s low, 29,405.00: 144.50 points, or 0.49%, under the midnight open, inside the New York killzone.
By 4:00 PM it closed at 29,566.25, 161.25 points off the low and 16.75 points above the midnight open. Read with ICT labels, this is a textbook bullish day: a drop under the midnight open in the killzone, then a rally into the close.
Now read it again without hindsight. At 2 PM the hour’s low was 29,423.25, only 18.25 points above the morning low, and the day still had to finish above 29,549.50 to count as bullish at all. Had it closed 17 points lower, the same morning would be relabeled as part of a bearish day, and the 1:15 AM high would become the “manipulation”. The label is only certain at 4:00 PM, which is the core difficulty with trading it live.
The original data
The test. Yahoo Finance hourly bars for NQ futures, 11 June 2024 to 29 October 2026. Each weekday runs from the midnight open to the 4:00 PM close, 16 hourly bars. Days without all 16 are left out (22, mostly holidays and early closes), as are three days where an hour inside the window opened more than 0.4% away from the hour before it, a bad print or a contract-roll mix-up in the continuous series (16 and 17 December 2025 and 30 October 2026).
That leaves 578 days, 309 up and 269 down. On an up day the wrong-way extreme is the day’s low, on a down day its high, placed in the first hour to print it. The killzone hours are the six that start at 2, 3, 4, 7, 8 and 9 AM: 37.5% of the window, and 36.47% of its summed hourly range.
What came out. The wrong-way extreme printed in a killzone hour on 283 of 578 days, 48.96%. The single busiest hour was 9 AM, with 87 days (15.05%), the hour of the 9:30 stock open.
Two benchmarks, and why the second is the fair one. Shuffling each day’s own 16 hours into random order 1,000 times put the extreme in a killzone hour a median 35.47% of the time, and no shuffle reached 48.96%. That comparison flatters ICT: shuffling spreads the busy 9 AM hour evenly across the day. A fairer control builds each fake day from the same hours of randomly chosen real days, so 9 AM stays as busy as 9 AM really is but the order inside a day is lost. That control gave a median of 47.23% (5th to 95th percentile 43.77% to 50.69%), and 218 of 1,000 control runs matched or beat the real 48.96%. The difference is within chance.
The same check on S&P 500 futures (ES) gave 280 of 577 non-flat days in killzone hours, 48.53%, against a control median of 45.85% (42.39% to 49.31%), with 93 of 1,000 control runs at or above it. Closer to the edge of chance than NQ, still inside it. The midnight open itself: price traded on both sides of it on 575 of 578 NQ days, and the shuffled days gave the same 99.48%.
So on these two markets the killzones mark when the market is busy, not a sequence beyond that. Every day is listed in the NQ day file and the ES day file. The limits: hourly bars cannot tell 9:05 from 9:55, the midnight price is the first one after midnight in Yahoo’s feed, and two years is one market regime.
Two earlier tests fit the same reading. On the euro, ICT’s home market, 575 days of hourly bars from January 2024 to October 2026 put 31.04% of daily highs and lows in the six killzone hours, which carried 34.40% of the day’s movement. The busiest single hour was 17:00, the first bar of the forex day, with 10.09% of extremes on 3.46% of movement. The euro hour table and pound hour table list every hour, and Yahoo’s noon price sat a median 2.69 pips from the Fed’s noon buying rate for the euro.
On SPY, 60 regular sessions of 5-minute bars from July to October 2026 put the move against the day inside the first 30 minutes on 30 sessions by closing prices, against about 35% when each session’s moves were shuffled; the session file lists all 60. That bunching sits at the open, the widest part of the day, as the best time to day trade page measures.
What gets taught. Across our study of 24,971 trading videos, 685 carry ICT or Inner Circle Trader in the title, and half get fewer than 13,316 views. Only 12 name a killzone, the silver bullet, the judas swing or a London or New York open, so the clock, the testable half, is the least taught.
When it fails
The day changes its label at the close
The worked example finished 16.75 points above its midnight open. A down day the next morning reads the same chart in reverse. Waiting for a sweep inside a window means guessing which kind of day it is, and that is only settled at the close.
The window is the busy part, not a trap
Most of what killzones capture is activity: 9 and 10 AM are the widest NQ hours, and the busy-hours control matched the real data. A plan that only looks inside the windows still misses the 295 days whose extreme came at other hours.
The sequence runs and then fails
A session can sweep, shift and retrace exactly as drawn and still reverse. A sequence that often comes before a move is not one that causes it. Test your own version on your own market with a backtest before trusting it.
Thirty names describe any chart
The vocabulary can label every bar after the fact. A description is not a prediction, and the ICT concepts page shows how much of a chart the renamed half marks by rule.
Questions people ask
What is ICT in trading?
ICT stands for Inner Circle Trader, the online name of Michael J. Huddleston, and for the method he teaches: liquidity, sweeps, structure shifts and fixed times of day. The ICT 2022 model is its best-known single setup.
Is ICT a good trading strategy?
It is a vocabulary and a framework more than one fixed strategy, so it has no single track record. On 578 days of NQ futures, its clock windows held no more of the day’s false moves than busy-hours chance.
Which is better, ICT or SMC?
They overlap almost completely. Smart money concepts is the shorter core; ICT adds the clock, the midnight open and the 62% to 79% entry range on top. Learning the core first makes ICT easier.
Is ICT trading profitable?
This page found no audited public record of results, for its teacher or its students, to point to. Any result depends on the exact rules a trader writes down, their costs and their discipline, which is why a written, tested rule matters.
Related
Smart money concepts is the layer underneath, five ideas worth learning before any of the names here.
Liquidity sweep is what a judas swing actually is, with the rule spelled out; the liquidity sweep vs support and resistance test found a deeper poke above IWM swing highs did not change the outcome (77.63% vs 85.53%, p = 0.21).
And premium and discount is the simpler version of the entry range, the more useful place to start.
Take pieces of this without taking the whole framework. Watch for a false move before the real one, but do not let a clock decide whether it is allowed to happen. The vocabulary is much bigger than the ideas underneath it, so learn the four or five underlying ideas properly and then treat the names as names.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.