A three-hour window contains the day's high 12.5% of the time for no reason at all. Any killzone worth its name has to beat that — and one of the most famous ones does not.
A killzone is a specific window of the trading day said to contain a disproportionate share of the important price action, which makes it one of the few trading claims that can be checked against a simple baseline.
A killzone is a specific window of the trading day that is supposed to contain a disproportionate amount of the important price action. The ones taught most often are the Asian killzone, the London killzone, the New York morning killzone and the London close.
The claim is testable in a very direct way. If a three-hour window really is special, the day's high or low should form inside it more often than three hours out of twenty-four would suggest — which is 12.5% of the time. That baseline is the whole test, and it is the number almost never printed alongside the claim.
A shop claims its busiest three hours are between two and five. To check, you do not count the customers between two and five — you count them in every three-hour block and see whether that one stands out.
If a quarter of the day's customers arrive in an eighth of the day, the claim is real. If it is an eighth of the customers in an eighth of the day, the window was never special; you were just looking at it.
Every UTC trading day with a full set of hourly candles, with the hour containing the day's high and the hour containing its low recorded. The figure below is how often each window held the daily high, against the 12.5% a three-hour window gets by chance.
| Instrument | Asia 00–03 | London 07–10 | NY 12–15 | By chance | Days |
|---|---|---|---|---|---|
| XAUUSD | 21% | 8.9% | 23.2% | 12.5% | 1,445 |
| GBPUSD | 16.1% | 14.8% | 20.2% | 12.5% | 1,411 |
| EURUSD | 17.4% | 15.3% | 21.8% | 12.5% | 1,420 |
| US30 | 11% | 7.7% | 23.3% | 12.5% | 1,040 |
| NAS100 | 11.1% | 8.2% | 23.9% | 12.5% | 1,059 |
What was counted: Windows are fixed in UTC. For each trading day with at least eighteen hourly candles, the hour containing the day's high and the hour containing its low are recorded, and each window's own high-to-low range is measured. "Expected by chance" is simply the window's length as a share of the day.
The lows behave almost identically to the highs on every instrument, so only the highs are shown here to keep the table readable.
Two findings, and the second one is the one nobody publishes.
The New York morning killzone is real. It holds the daily high around 23.2% of the time on gold against a 12.5% baseline — close to twice chance — and it accounts for a median of 59.1% of the entire day's range in three hours. Every instrument agrees. This window earns its reputation.
The London killzone does not. On gold it holds the daily high 8.9% of the time — below the 12.5% baseline. The Dow is the same. Only the currency pairs put it slightly above chance, and even there the margin is small.
Extremes are not spread evenly through the day even in a market with no sessions at all, because the first hours of any measurement window sit right where price starts. Splitting the gold day into eight three-hour blocks shows exactly that: the 00:00–03:00 block holds 23.2% of all daily extremes, well above chance.
So the Asian killzone's strong-looking number should be discounted — a good part of it is the day boundary rather than the session. Note though that the final block, 21:00–24:00, sits at only 6.3%, so this is not a simple both-ends artefact.
The comparison that survives all of this is London against New York. Both are mid-day, both three hours, both measured identically — so whatever distorts one distorts the other. New York is at roughly twice chance and London is at or below it. That gap is real.
Fatima trades gold and has been getting up early for the London killzone because that is what the videos told her to do. She trades 07:00 to 10:00 UTC and finds it frustrating.
What the data says about her schedule. On gold, the London window holds the daily high 8.9% of the time, against 12.5% by chance. She has been getting up early for a window that is, on her instrument, slightly worse than picking three hours at random.
Where the action is. The New York morning window holds the daily high 23.2% of the time and carries 59.1% of the whole day's range. If she moved her three hours to 12:00–15:00 UTC she would be at the screen for the part of the day that actually moves.
What has not changed. The window does not tell her which way to trade. It tells her when the opportunity exists. Her entries still come from structure, and her stop still has to clear the noise — noise that is considerably larger in the New York window than in London, which she now has to size for.
The honest caveat. If she traded GBPUSD instead, London would look better. The rule is not universal, and that is the point of measuring your own instrument.
Gold, the Dow and GBPUSD gave three different answers about London. Replay your own instrument, note the hour the day's high and low form, and tally it over a few months. That table is worth more than any general rule about killzones, including this one.
Test the windows in the backtester →Run the drill in the free backtester → Free, no sign-in to begin.
What are ICT killzones?
Specific windows of the trading day said to contain most of the meaningful price action. The commonly taught ones are the Asian killzone, the London killzone, the New York morning killzone and the London close. This page tests them against the share a window of that length would get by chance.
Which killzone actually works?
The New York morning window, 12:00–15:00 UTC. On gold it held the daily high 23.2% of the time against a 12.5% baseline, and it carries a median 59.1% of the entire day's range in three hours. Every instrument tested agrees.
Is the London killzone real?
Not on every instrument. On gold it held the daily high 8.9% of the time, below the 12.5% a random three hours would get, and the Dow is similar. GBPUSD and EURUSD put it slightly above chance. If you trade metals or US indices, this window is not doing what you were told it does.
Why does the Asian killzone look so strong?
Partly because it sits at the start of the UTC day, and extremes cluster near the boundary of any measurement window. The 00:00–03:00 block holds 23.2% of gold's daily extremes, and some of that is the boundary rather than the session. The clean comparison is London against New York, since both are mid-day and both are distorted equally.
Do killzones tell me which direction to trade?
No. A window measures where the range and the extremes concentrate, not which way price goes. Use it to decide when to be at the screen; the direction has to come from structure.
Should I use New York time or UTC?
Measure in UTC so the buckets stay comparable, then convert for your own schedule. A window taught in New York time drifts by an hour against UTC twice a year, which is enough to move a one-hour window like the silver bullet completely out of position.
Gold, the Dow and GBPUSD disagreed about London. Replay the market you trade, note where the daily high and low form, and bucket them into three-hour blocks. Then trade the block that wins.
Open the free backtester →Beginner exploration
Open each answer for a plain-language way to read Killzones, Tested Against The Clock, test it carefully and decide what to explore next.
This page focuses on “Killzones, Tested Against The Clock”.Killzones measured against the 12.5% a three-hour window gets by chance. On gold the New York window held the daily high 23.2% of the time; the London killzone held it 8.9% — below chance.For “Killzones, Tested Against The Clock”, a beginner should identify what the learning guide measures, assumes or teaches before acting on its conclusion.Treat this page's account of “Killzones, Tested Against The Clock” as a learning reference rather than a prediction, signal or promise of future performance.
For “Killzones, Tested Against The Clock”, translate the idea into a definition you could apply the same way on two different charts.While exploring “Killzones, Tested Against The Clock”, work through one example slowly and record which inputs or observations determined the result.Keep your “Killzones, Tested Against The Clock” record honest: list the limitation or counterexample before using the concept in a trading plan.Before leaving “Killzones, Tested Against The Clock”, practise the definition on unseen history and review consistency before judging performance.
Turn one idea from “Killzones, Tested Against The Clock” into a rule with explicit inputs, dates, costs and pass-or-fail conditions.Ask AI to expose missing assumptions in that “Killzones, Tested Against The Clock” test, not to guess the next market move.Use the FXAbsolute AI Backtesting Lab to inspect calculations connected to “Killzones, Tested Against The Clock” and the assumptions behind them.Reproduce any important “Killzones, Tested Against The Clock” result and reserve unseen data before deciding that an apparent pattern is useful.
Continue your exploration of Killzones, Tested Against The Clock with the beginner AI prompt guide, or inspect public calculations in the AI Backtesting Lab.