ICT Concepts for Futures Trading: A Practical Framework

ICT concepts give names to things many chart readers already see: a spike through an obvious high, a three-candle gap, the last candle before a sharp move. The hard part on ES and NQ isn't learning the vocabulary. It's deciding which pieces you can define precisely enough to test. ICT material says large participants engineer price to collect resting orders. That's an explanation, not something a chart proves, so I treat every concept below as a rule to test, not a fact.

The thesis and what to do with it

The core idea is that stop orders cluster around obvious swing points, and that price tends to travel to those clusters. One part is mechanical: a stop order becomes a market order when its price trades. Whether anyone deliberately hunts those orders, and whether trading around them pays after costs, is a question for your own records.

Liquidity pools: equal highs and lows

Two or more highs at nearly the same price, or two or more lows, are treated as pools. Mark them on the 5-minute and 15-minute charts each morning. The practical rule is to avoid initiating a long into equal highs and wait for the sweep: a brief trade through the level followed by a rejection. The sweep is the trigger, not the approach.

You have to define equal. Pick a tolerance, for example within 2 ES points, and write it down. Without a number, every pair of similar highs becomes a pool in hindsight.

Fair value gaps

A bullish gap is a three-candle pattern where the first candle's high is below the third candle's low. The space between is the gap. A bearish gap is the mirror image. ICT's explanation is that the range was traded through quickly, leaving unfilled orders, so a return to the zone may produce a reaction.

Hypothetical example: candle one high 5,300.00, candle three low 5,304.50. The gap is 4.50 points, or $225 of price range per ES contract. If you enter on a pullback at the midpoint, 5,302.25, with a stop at 5,299.75, you risk 2.50 points: $125 on one ES contract or $12.50 on one MES, before costs.

Common filters are gaps formed by a strong directional move, gaps in line with the higher-timeframe trend, and gaps not already filled. Those are filters to test against your log, not proven distinctions. Gaps in choppy, overlapping price and gaps inside earlier gaps are the ones people usually skip.

Order blocks

An order block is the last opposing candle before an impulsive move. A bullish block is the last down candle before a strong up move. A typical rule is entry at the candle body with the stop beyond the wick. The weak point is subjectivity: which candle counts, body or wick, how strong the move must be. Write those choices as rules before you backtest or paper trade, or you'll keep redrawing zones until they fit.

Kill zones

Kill zones are time windows teachers use to filter setups. Definitions vary. The ones in common use here, all in Eastern Time:

• London: 2:00 am to 5:00 am ET
• New York morning: 9:30 am to 11:00 am ET, starting at the regular US cash open
• New York lunch: 12:00 pm to 1:30 pm ET, usually avoided
• New York afternoon: 2:00 pm to 4:00 pm ET

These aren't exchange rules, and holidays and early closes change the schedule. The testable question is whether your setups perform differently by window. Tag each logged trade with its time and compare.

Power of three and market structure shifts

Power of three describes a day as accumulation, manipulation, then distribution. On ES and NQ the idea is that price sweeps one side of the overnight range early, reverses, and moves toward the other side. Plenty of days don't follow that script, so define in advance what counts as a valid sweep and a valid reversal.

A market structure shift is a close through a prior swing high or low with displacement, meaning a strong move rather than a drift. Entry is on the first pullback, often into the gap or block that formed the move, with the stop beyond the swing that created the shift. Put a number on displacement, such as a candle range at least some multiple of recent average range.

A pre-market routine

1. Mark the overnight high and low and the midnight open (ICT uses midnight New York time) on the 15-minute chart.
2. Mark unfilled daily gaps from the prior session.
3. Mark prior day and prior week highs and lows.
4. Mark equal highs and lows that fit your tolerance.
5. Note whether price is above or below the daily open and the prior day's high.

My suggested filter is to act only when a setup lines up with at least two of those items inside a window you've defined. That's a process choice to test, not a proven rule. YMI's KPL levels are a separate framework: fixed reference prices anchored to the regular-session opening print, which isn't known until the cash open. You can read how they're defined on the KPL trading page.

How to test it

Log every setup, including those you skip, with entry, stop, target, time and result. Count real costs: an ES tick is $12.50, and commissions and slippage come off every trade. Thirty trades is not proof of an edge, and 50 simulated trades are not live proof, because simulated fills don't validate queue position or slippage. Put your expectancy through the edge calculator and treat a thin sample as an estimate.

What ICT doesn't fix

It doesn't set position size, cap your daily loss, or tell you when the market isn't giving pullbacks. Those need their own rules, written before the session.