Here is an uncomfortable observation about most trading journals: if you sort your results by the hour you entered, the spread is often wider than the spread between your strategies.
That isn't a coincidence, and it isn't about you. Markets don't distribute their movement evenly across the day. Liquidity arrives in waves tied to when the world's financial centres open and close, and the setups that work beautifully inside those waves behave very differently outside them. A Fair Value Gap during the London Open sits in a market with real two-sided flow. The same gap at 1:00 PM Eastern sits in a market that has largely gone quiet.
This is what killzones formalise, and it's why ICT KillZones is often the first thing a trader adds and the last thing they'd remove.
The four windows
All times Eastern, as the ICT framework defines them:
- Asian Session — 8:00 PM to 12:00 AM. Typically the quietest and most range-bound. Its value is less about trading it and more about the range it builds, which frequently becomes the liquidity that London raids.
- London Open — 2:00 AM to 5:00 AM. European liquidity arrives. Often the session that establishes the day's directional tone and sweeps whatever the Asian range left sitting.
- New York Open — 7:00 AM to 9:00 AM. US participation enters while London is still active. The overlap makes this the highest-liquidity window of the day for index futures.
- London Close — 10:00 AM to 12:00 PM. European desks square up. Frequently produces sharp counter-moves against the morning's direction as positions are unwound rather than initiated.
Notice that these aren't equal in character. Two of them are about liquidity entering the market; one is about liquidity leaving it; one is about a range being built. Treating all four as generically "good hours to trade" misses most of the value.
Why the windows matter mechanically
The concept can sound like folklore, so it's worth being concrete about the mechanism. Three things change inside a killzone:
Participation. More active participants means tighter spreads, deeper books, and moves that can actually be filled at reasonable prices. Outside these windows the book thins and slippage on the same order size gets worse.
Intent. The flow arriving at a session open is largely positioning — desks establishing or adjusting exposure. That produces directional movement with follow-through. Midday flow is disproportionately noise and small-order churn.
Liquidity targets. Sessions build ranges, and ranges accumulate stops at their extremes. The Asian range's high and low are known, visible levels by the time London opens — which is exactly why London so often takes one of them out first. The session structure creates the liquidity that the next session hunts.
That last point is the one that connects killzones to the rest of the toolkit. A liquidity sweep at 3:00 AM taking out the Asian high is a recognisable, repeatable sequence. The same sweep at 1:00 PM has no session narrative behind it.
Timezone handling, which sounds boring and isn't
Every killzone definition is anchored to Eastern time, because that's where the institutional activity is centred. If you trade from London, Dubai, Karachi or Sydney, that means constant mental arithmetic — and arithmetic that breaks twice a year when daylight saving shifts on different dates in different regions.
The indicator takes all session times in EST and converts to your local time automatically. You enter 2:00 AM for the London Open regardless of where you are, and the zone lands correctly on your chart. It's a small thing that removes an entire category of quiet, hard-to-notice error — the kind where your killzone has been an hour off since the last DST change and you never realised.
Key levels and the true futures day
Alongside the sessions, the indicator plots previous day, previous week and current day highs and lows — with a configurable trading-day anchor, defaulting to 17:00 EST.
That default is doing real work. Futures don't open at midnight; the CME session begins at 6:00 PM Eastern, and a trading day anchored to midnight splits the actual session in the wrong place. Anchoring at 17:00 gives you a previous-day high and low that match the real futures day rather than a calendar convenience. If you've ever wondered why your PDH didn't match someone else's on the same instrument, this is usually why.
These levels matter because they're the ones ICT Bias uses to build its directional read. Having them anchored consistently across both tools keeps the two in agreement.
Integrated Fibonacci from the session range
Once a session completes, the indicator builds Fibonacci retracements from its high to its low — up to seven levels: 0, 0.236, 0.382, 0.5, 0.618, 0.782, and 1.
The reasoning is that a session's range is not an arbitrary span. It's a period during which a known cohort of participants transacted, bounded by two levels the market itself established. Retracements measured from that range are anchored to something structural, unlike a Fib drawn by hand between two swings you picked.
In practice the 0.5 and 0.618 of the London Open range are the levels most traders watch during the New York session — a pullback into the prior session's equilibrium before continuation. Having them drawn automatically, from the correct range, is the difference between using them and meaning to.
Custom sessions
Four additional user-defined sessions sit on top of the presets, each with its own name, colour and times. Common uses:
- The CME futures open at 6:00 PM EST, which isn't one of the four killzones but matters for gap analysis
- A regional session relevant to your instrument — Tokyo for JPY pairs, Frankfurt for European indices
- Economic release windows you want visually marked
- A personal "do not trade" block, shaded so you notice when you're about to break your own rule
The session flag: the most useful plot in the toolkit
The indicator exposes a session-active flag as a simple 0 or 1, along with the key levels and Fibonacci values.
That binary flag deserves attention out of proportion to its simplicity. Drop it into Strategy Builder as a condition and your automated strategy only trades during killzone hours. No new logic, no additional signal — you've just removed every trade the strategy would have taken during the dead hours.
Because the sessions are time-based and fixed, this also behaves perfectly in backtesting. There is no pivot to confirm, no bar to wait for, no possibility of the session moving after the fact. Time is time. That makes killzones one of the few filters where the backtest and the live behaviour are guaranteed to match.
Common mistakes
Treating all four killzones as interchangeable. London Close is a position-unwinding window and behaves differently from London Open, which is a position-establishing one. A continuation strategy that works at 2:00 AM may be systematically wrong at 10:30 AM.
Trading the Asian session as if it were London. The Asian range is usually the thing to observe, not the thing to trade. Its high and low become tomorrow's liquidity. Breakout strategies inside it tend to fight the range rather than benefit from it.
Assuming a killzone guarantees a move. It's a window where movement is more likely, not a schedule. Plenty of London Opens go nowhere. The killzone tells you when to be watching, not that something will happen.
Entering times in local time. The indicator expects EST and converts for you. Enter your local time and every zone will be shifted, usually by enough to look plausible and be wrong.
Using killzones as the whole strategy. Timing is a filter, not a direction. Killzones tell you when; ICT Bias tells you which way, and the execution tools tell you where. All three, or you're just trading at better hours in a random direction.
The cheapest improvement available
Most attempts to improve results involve adding something — a new indicator, a new pattern, a new confirmation. Session filtering is subtractive. You keep the strategy you have and stop taking it during the hours where it doesn't work.
Next week: the macros — the twenty-minute windows inside these sessions where the actual algorithmic activity concentrates.
Disclaimer: Trading futures and other leveraged instruments involves substantial risk of loss and is not suitable for all investors. Past performance and indicator signals are not indicative of future results.
