The 'kill zone' idea, tested
Posted: Wed Oct 28, 2026 5:01 am
Some educators use the term kill zones for time windows when large moves are supposed to occur: the London open, the New York open and the London close. Trading outside them is discouraged.
There's something to the idea. Volatility and volume do concentrate around session opens, and many strategies perform better during these windows. The specific hours chosen vary between teachers, and the claim that these windows hold special power is harder to confirm.
A test is simple. Measure average range per five-minute block across the day in UTC, over a year. Mark the supposed kill zones and compare. Likely, the London open and the New York open show clear spikes, since those are real events. The boundaries drawn by individual teachers, down to the minute, are less clear.
The practical lesson is using data to set personal windows, not borrowed ones. A pair's own volatility profile shows when it moves, and spread data shows when it's cheap.
Daylight saving changes shift the real market times relative to fixed UTC windows, a detail that borrowed kill zone lists often ignore.
Many online lists of kill zones are based on US or UK time without adjusting for daylight saving, so the windows drift by an hour for part of the year. A rule written as "7 to 10 AM New York" is different from "12 to 3 PM London" for several weeks each spring and autumn.
How do the supposed kill zones compare with the measured volatility profile for the pairs traded?
There's something to the idea. Volatility and volume do concentrate around session opens, and many strategies perform better during these windows. The specific hours chosen vary between teachers, and the claim that these windows hold special power is harder to confirm.
A test is simple. Measure average range per five-minute block across the day in UTC, over a year. Mark the supposed kill zones and compare. Likely, the London open and the New York open show clear spikes, since those are real events. The boundaries drawn by individual teachers, down to the minute, are less clear.
The practical lesson is using data to set personal windows, not borrowed ones. A pair's own volatility profile shows when it moves, and spread data shows when it's cheap.
Daylight saving changes shift the real market times relative to fixed UTC windows, a detail that borrowed kill zone lists often ignore.
Many online lists of kill zones are based on US or UK time without adjusting for daylight saving, so the windows drift by an hour for part of the year. A rule written as "7 to 10 AM New York" is different from "12 to 3 PM London" for several weeks each spring and autumn.
How do the supposed kill zones compare with the measured volatility profile for the pairs traded?