Mastering ICT Killzones: Timing Institutional Liquidity for Better Entries

ICT Killzones explained: key trading sessions and times

Throughout the trading day, global order flow compresses into narrow windows where institutional participation spikes. According to a recent breakdown by equiti.com, these windows — what SMC practitioners label "ICT Killzones" — are when liquidity sweeps, displacement, and market structure shifts tend to cluster. They are not entry signals. They are session filters. That distinction is the difference between a probabilistic framework and a retail superstition, and it is where most published Killzone content quietly fails the reader.

London: The 2:00–5:00 AM NY Window

London opens while the Asian session is still cleaning up its range. Equiti's note puts the London Killzone between 2:00 and 5:00 AM New York time, with daylight saving shifts occasionally nudging the boundaries. The Asian session typically prints a tight range with obvious highs and lows where resting orders accumulate. London frequently breaks one side of that range before choosing direction. The break itself is noise unless it confirms with a market structure shift or a change of character.

This is the leg most amateurs skip. They see the sweep, mark the box, and chase the wick. The framework is sound. The execution requires waiting for displacement confirmation and then trading the retracement into a fair value gap or order block — not the initial impulsive leg.

New York: Where Conviction Peaks

The New York Killzone begins as the US session opens while London remains active — the only overlap of the two highest-liquidity venues in the FX calendar. Displacement during this window tends to carry further because both pools of institutional orders are stacked on the same side of the book. According to equiti.com's read of the methodology, this is where most SMC traders wait for a clean setup rather than force London.

If you run a single-session algo, this is the window your sample size lives or dies on. The London–New York overlap delivers the cleanest displacement-to-retracement sequences, which is why nearly every published ICT case study screenshots this hour.

The Filter Problem

A Killzone is a prior on time, not a guarantee of direction. The edge comes from confluence — sweep plus structure shift plus a fair value gap or order block in proximity. Strip any one of those legs and the trade degrades toward noise. That is the mechanical reality of session-based trading and the thing no influencer chart ever mentions.

The same discipline applies to any external input you let into your stack. Spotting red flags in a supplier follows the same logic as filtering a session signal — verify the structural tell, ignore the surface.