The fifth and last module of the ICT method puts the four before it together into one reading: a bias, a target, a zone and an hour, then an entry with its stop. Applied to real sessions, losing ones included, and set against what the site has measured.
Step 53
The bias: which way?
Modules 7 to 10 provided the pieces; this one puts them together, in the order you use them before a session. First question: buy or sell?
One rule, on the 1-hour chart
The bias is read on a timeframe above the one you enter on: here the 1-hour, for 5-minute entries. You look for the last break of structure (steps 35 and 36):
a close above the last swing high: bullish bias;
a close below the last swing low: bearish bias;
no break for more than 8 hours: no bias, so no trade.
Set beforehand, valid while it holds
The bias is decided before the New York killzone, at 8:30. If a 1-hour close breaks the other way before the entry, it falls, and the day stops there.
Having no bias is an answer: the model does not trade every day.
What it gives
A direction, not an entry. It ties structure to the draw on liquidity of step 43: bullish, the target is looked for above; bearish, below.
The chart shows the gold session used as the example up to step 56.
Bullish candleBearish candle
Gold (MGC), 1-hour candles, 18 and 19 February 2025. On the 18th, price climbs from 2,920.6 (5:00) to a swing high at 2,959 (20:00), then drops back to a swing low at 2,942.3 (22:00). On the 19th, the 3:00 candle goes above the previous day high but closes at 2,955.9, below 2,959; the 4:00 one closes at 2,962.6, above it: that is the last break of structure before 8:30, 3 hours 30 old. Bullish bias. At 7:00, price drops back to 2,948.8 without closing below 2,942.3: the bias holds.
Key pointThe bias is the direction of the last 1-hour break of structure, set before the session. No break for 8 hours means no bias and no trade; if it turns before the entry, you stop.
The bias gives a direction; the target says how far. It is the draw on liquidity of step 43, chosen by a rule.
The nearest intact liquidity
Before the session, list the levels of module 8 and check they are still intact:
the previous day’s high and low, and the previous week’s;
the Asian range (20:00-midnight) and the London open (2:00-5:00): their highs and lows;
the equal highs and equal lows of the last two days.
The target is the nearest intact level in the direction of the bias. A level already exceeded no longer draws.
Far enough, not too far
At the entry (step 56), it must sit at least 1.5 times the entry-to-stop distance away (1.5 R): closer, aim for the next level. Beyond 10 R, the level is too far for one session, and the trade is not taken.
The other side counts too
The opposite liquidity is often taken first: that is the Judas swing of step 52. With a bullish bias, the nearby lows are where to watch for the entry.
Bullish candleBearish candle
Gold (MGC), 15-minute candles, from 18 February at 20:00 to 19 February at 8:30. Above price, the previous day high (2,956.5) and the Asian high (2,959) were exceeded during the night, at 3:15 and 4:15: they no longer draw. The London high, 2,964.1 (4:15), is intact: it is the target of the bullish bias. Below remain the London low (2,946.9, at 2:00) and the Asian low (2,942.3, at 22:00): the liquidity on the other side.
Key pointThe target is the nearest intact liquidity in the direction of the bias, between 1.5 and 10 R. The liquidity on the other side says where to watch for the entry.
The direction and the target are set. What remains is where and when to wait for the entry: the model does not look for it just anywhere, nor at just any hour.
The zone: in discount to buy
The dealing range is the 1-hour leg that gave the bias, from its start to its extreme (step 44). Bullish bias: buy only below its middle, in discount. Bearish bias: sell only above it, in premium.
In that half, the trigger will come from a liquidity sweep or from a zone of module 9: a FVG, an order block, the OTE.
The hour: the killzone
Wait only during the New York killzone, from 8:30 to 11:00 (step 50). An entry that shows up at 13:00 does not count.
The session often opens with a move against the bias that takes the liquidity on the other side: the Judas swing of step 52. That return into discount is what you wait for.
And if nothing comes?
At 11:00, with no entry in discount, the day is over. Many days give none: that is expected.
Bullish candleBearish candleKey zone
Gold (MGC), 15-minute candles, from 18 February at 22:00 to 19 February at 9:15. The leg that gave the bias starts at 2,942.3 (22:00) and climbs to 2,964.1 (4:15): that is the dealing range, and its middle is at 2,953.2. Above, premium; below, discount, where you may buy. The New York killzone opens at 8:30. At 8:45, price drops back to 2,949.3: it is in discount, on time, and that is where you wait for the trigger.
Key pointBuy in discount of the dealing range (sell in premium), and only during the killzone. Without both, no trade.
The direction, the target, the zone and the hour are ready. Only the trigger is missing, on the 5-minute chart: one of the battle engine’s three setups, in the direction of the bias.
A sweep: a candle goes beyond a low (or a high) and closes back inside (step 41).
A 5-minute break of structure (BOS), by a displacement candle (steps 35 and 37).
A first return into a 5-minute FVG, and a close back out of it in its direction (step 45).
The entry and the stop
You enter at the close of the setup candle. The stop goes where the idea becomes wrong: one tick beyond the wick of the sweep, beyond the swing the BOS starts from, behind the FVG.
Never closer than half of a usual 5-minute candle: any closer, mere market noise hits it.
The target and the size
The target is the one from step 54. The stop distance gives the position size (step 11): the possible loss is known before you enter.
Then you touch nothing. The trade ends at the stop, at the target or at 16:00, and it is the only one of the day.
Bullish candleBearish candle
Gold (MGC), 5-minute candles, 19 February 2025. At 9:25, a candle trades down to 2,946, below the London low (2,946.9), and closes at 2,950.3: a sweep, in discount, inside the killzone, in the direction of the bias. Entry at 2,950.3, stop one tick under the wick at 2,945.9 (4.4 points), target at the London high, 2,964.1, at 3.1 R. Price climbs to 2,957.5 at 9:50, then goes round in circles. At 10:55, a candle trades down to 2,941.4: the stop is hit, −1 R.
Key pointEnter at the close of the setup, stop where the idea becomes wrong, target on the liquidity aimed at. The stop sets the size: the possible loss is known before you enter.
The model applied with nothing picked: on each market, the first session of February 2026 that meets all its conditions, whatever the outcome.
Step by step
Nasdaq (NQ), 10 February: bullish bias; at 9:35, sweep of the 8:30 swing low, in discount; target at the previous day high, at 3.6 R.
6E, 5 February: bearish bias; at 8:30, sweep of the London high, in premium; target on equal lows, at 1.7 R.
Gold (MGC), 2 February: bearish bias; at 9:00, sweep of the 8:50 swing high, in premium; target at the London low, at 9.2 R.
Three stops
All three hit their stop within twenty minutes. On the Nasdaq and the 6E, price then reaches the target, after the stop. That is no reason to widen it (step 10): it sat where the idea became wrong.
The table adds the next session on each market: a stop, an exit at 16:00 on a small gain, and the target reached on gold, at 7.1 R.
What six sessions say
Nothing about what the model is worth: over so few trades, chance decides, and the further the target, the less often it is reached (step 12). They show how it is applied, losers included.
Session
Bias
Trigger
Entry · stop · target
Outcome
NQ, 10 Feb
Bullish
Sweep, 9:35
25,327.25 · 25,299 · 25,428.75 (3.6 R)
Stop at 9:40: −1 R
6E, 5 Feb
Bearish
Sweep, 8:30
1.18245 · 1.18355 · 1.1806 (1.7 R)
Stop at 8:50: −1 R
Gold, 2 Feb
Bearish
Sweep, 9:00
4,728.3 · 4,752.8 · 4,502 (9.2 R)
Stop at 9:05: −1 R
NQ, 12 Feb
Bullish
Sweep, 9:45
25,251.5 · 25,226.25 · 25,465.25 (8.5 R)
Stop at 9:55: −1 R
6E, 10 Feb
Bearish
BOS, 9:45
1.1929 · 1.19485 · 1.1829 (5.1 R)
Exit at 16:00, at 1.1914: +0.8 R
Gold, 4 Feb
Bearish
Sweep, 8:55
5,077.4 · 5,087.4 · 5,006.7 (7.1 R)
Target at 9:35: +7.1 R
The first three rows are the charts below; the next three, the following session on each market. Prices read on the 5-minute candles; 1 R = the distance between entry and stop.
Nasdaq-100 (NQ), 10 February 20266E (euro FX future), 5 February 2026Gold (MGC), 2 February 2026
Key pointApplied with nothing picked, the model gives three stops, then, the next session, a target reached on gold. Six sessions do not say what it is worth: they show how it is applied.
What was measured: why no setup is “high probability”
The model in this module is made of pieces that have all been measured. On this site, fourteen quality criteria of a setup were tested over thousands of trades of this method; the bias, the swept level, the zone and the hour are among them.
What the measurement says
None of them separated good setups from bad ones. With or without them, everything lands between 48 and 50% winning trades at 1 R, that is with a target as far as the stop.
At 1 R, winning one time in two is break-even (BE), before costs (step 12). A “high-probability” setup would do better: the measurement found none, and a page that promises one contradicts these figures.
What the method brings all the same
A reading frame: a reason to look at one place rather than another, and to do nothing the rest of the time.
A stop that means something: the price where the idea becomes wrong, so a risk known before you enter.
A discipline: the same rules every day, one trade, then you stop.
Where the outcome is decided
In what the setup does not decide: the size, the exit, the repetition. That is the subject of the last module, « Durer » (Lasting), which opens with a written trading plan (step 59).
Piece of the model
Criterion measured
Separates good setups from bad?
The bias (step 53)
Higher-timeframe bias
No
The swept level (step 56)
Previous day high and low, level strength
No
The zone (step 55)
Premium and discount, FVG grading
No
The hour (step 55)
Session, killzones
No
All of them
Fourteen criteria, thousands of trades
Between 48 and 50% winners at 1 R
Seven of the fourteen criteria measured: those that match the pieces of this model. With or without each one, the result does not move.
Key pointNo criterion measured makes an ICT setup “high probability”: everything lands between 48 and 50% at 1 R. The method gives a frame, a stop that means something and a discipline; the outcome is decided by the size, the exit and the repetition.
