Killzones
Execution Precision
8 min read
timeInMarket
Identify high-probability time windows (killzones) where institutional activity concentrates and setups have the best odds.
Loading
8 min read
Identify high-probability time windows (killzones) where institutional activity concentrates and setups have the best odds.
Crypto trades 24/7, but pretending all hours matter equally is how amateurs blow accounts on 22:00 UTC wicks. The session structure is the single highest-EV environmental filter you have — and most retail ignore it. Institutional desks, algorithmic systems, and concentrated liquidity pools activate during specific session windows -- and knowing when these high-activity windows (popularised as "killzones" by ICT/SMC literature, though the framework is debated) align with empirical volume peaks determines whether your stop placement is robust or vulnerable.
All times in this lesson are UTC. London (BST/GMT) and NY (EDT/EST) shift by one hour during their respective DST windows; verify boundaries each spring/fall.
Although cryptocurrency markets never close, they inherit the rhythm of traditional financial centers. Each major session brings its own participants, volume characteristics, and behavioral patterns.
The causal chain is concrete, not mystical: CME BTC futures trade EU/US hours and settle 4pm ET, basis-arbitrage desks rebalance into spot/perp around those flows, Deribit options expire 08:00 UTC Fridays, and US macro prints (NFP, CPI) land 13:30 UTC. Each mechanism imports volatility from TradFi into crypto. "Killzone" is ICT vocabulary — the empirical phenomenon (volume clustering during TradFi hours) is real; the mystical framing isn't required.
| Session | UTC Time | Local Reference | Primary Characteristics |
|---|---|---|---|
| Asia (Tokyo) | 00:00 - 08:00 | Tokyo 09:00 - 17:00 | Low volatility, range-building, accumulation |
| London | 07:00 - 16:00 | London 07:00 - 16:00 | Trend initiation, breakouts, stop hunts (large orders pushed through visible stop clusters to harvest liquidity) at Asian extremes |
| New York | 12:00 - 21:00 | New York 08:00 - 17:00 | Highest volume, continuation or reversal of London moves |
| Late New York / Pre-Asia | 21:00 - 00:00 | New York 17:00 - 20:00 | Low liquidity, wide spreads, position squaring |
The London/New York overlap from 12:00 to 16:00 UTC consistently produces the highest volume, widest ranges, and most decisive price movements in BTC/USDT. On rolling 90-day BTC/USDT samples (cf. Kaiko / Coinglass session-volume reports), this four-hour window has historically accounted for ~30-40% of the daily true range — confirm against your own data before sizing on it.
Overlaps between sessions create windows of amplified activity. When two major participant pools are active simultaneously, order flow intensity spikes and price moves become more directional.
| Overlap | UTC Time | Duration | Significance |
|---|---|---|---|
| Asia / London | 07:00 - 08:00 | 1 hour | London opens, sweeps Asian session highs/lows |
| London / New York | 12:00 - 16:00 | 4 hours | Peak liquidity, strongest trends, largest candles (note: DST shifts this overlap by one hour twice yearly — verify session boundaries in UTC after each clock change) |
| New York / Asia | 00:00 - 00:00 | Minimal | Brief handoff, generally low impact |
Volume distribution across sessions is not uniform. Understanding where volume concentrates informs both trade timing and stop placement.
| Session | Share of 24h Volume | Average Hourly Range | Wick Frequency |
|---|---|---|---|
| Asia | 15-20% | $150-250 | Low |
| London (pre-overlap) | 20-25% | $250-400 | Medium |
| London/NY Overlap | 30-40% | $400-700 | High |
| Late New York | 10-15% | $150-300 | Medium |
| Pre-Asia Dead Zone | 5-10% | $80-150 | Low |
BTC/USDT Volume Share by Session
Midpoints of cited ranges. London/NY overlap consistently dominates.
Average BTC/USDT Hourly Range by Session (USD)
Midpoints of cited ranges. Overlap windows produce ~5x the dead-zone hourly range.
The pre-Asia dead zone (21:00-00:00 UTC) is notorious for erratic price spikes on thin order books. A $300 wick that would be absorbed instantly during the London/NY overlap can trigger stops and cause genuine displacement during this window. Avoid entering new positions in dead zones unless your strategy specifically targets low-liquidity conditions.
The "London open sweep of the Asian range" is a frequently cited ICT pattern. Empirically, BTC does often probe the prior 8h range early in the London session, but the published hit-rate is sensitive to definition (sweep depth, reversal threshold) and survivorship bias is real — backtest before trading. Defined narrowly as "price prints a new Asian-range extreme in the first 60 minutes of London then closes back inside within 30 minutes," the setup has historically reversed in roughly 55-65% of BTC sessions in 2023-24 — meaning ~35-45% of attempts continue through your stop. Size accordingly. As London traders come online, they frequently drive price through the Asian session high or low to trigger resting stop orders and build liquidity for the real move.
Strict definition: new Asian-range extreme in first 60m of London, closes back inside within 30m. ~35-45% continue through your stop. Backtest before sizing.
Asian session range was $66,700-$67,400. London opened and swept below the Asian low to $66,650 before reversing sharply. Entry on the reversal confirmation with stop below the sweep low.
The stop was placed below the London sweep low rather than below the Asian range low. This distinction matters because the sweep itself creates a new structural level that is far more significant than the original range boundary.
The New York open at 12:00 UTC either continues the move that London initiated or reverses it. This decision point is critical for stop management on trades entered during the London session.
If New York continues the London move:
If New York reverses the London move:
Entered short during London session after a breakdown. New York opened and reversed the move aggressively, sweeping the London high. Stop hit at $68,700.
This trade failed because the New York session chose to reverse the London move. Recognizing this risk and tightening the stop before the New York open would have reduced the loss.
Stop Buffer = Base Buffer x Session Multiplier
Where Base Buffer is defined per the ATR-Based vs Structural Stops lesson (e.g., 1.0x ATR-15m or 1.0x recent swing).
Session Multipliers: Asian session: 1.5x (wider, compensates for thin liquidity) London session: 1.0x (baseline) London/NY overlap: 0.8x (tighter, high conviction moves) Late NY / Pre-Asia: 1.8x (widest, erratic conditions)
Session Stop Multipliers
Tighter during conviction windows, wider during thin-liquidity windows.
Large participants -- market makers, proprietary trading firms, and fund managers -- operate on predictable schedules tied to traditional market hours. Their activity creates observable patterns in crypto markets.
| Behavior | Typical Timing (UTC) | Impact on Stops |
|---|---|---|
| Stop hunts / liquidity sweeps | 07:00-08:00 (London open) | Place stops beyond likely sweep targets |
| Trend initiation | 08:00-10:00 (early London) | Enter with trend, tight stops behind initiation structure |
| Momentum acceleration | 12:00-14:00 (overlap start) | Trail stops, expect extended moves |
| Profit taking / reversal | 15:00-17:00 (late overlap) | Tighten stops, consider partial exits |
| Position squaring | 20:00-21:00 (NY close) | Expect reduced volatility, widen stops if holding overnight |
Killzones define windows of opportunity, not exact timestamps. London may "open" at 07:00 UTC, but the significant sweep or breakout might not occur until 07:30 or 08:15. Wait for the pattern to develop within the killzone window rather than forcing entries at the exact session boundary.
A practical approach to incorporating killzones into your stop placement:
The London/New York overlap runs from 12:00 to 16:00 UTC and consistently produces the highest volume, widest ranges, and most decisive price movements in BTC/USDT. DST shifts the overlap by one hour twice yearly, so verify boundaries each spring/fall.
Use a Base Buffer multiplied by a session factor: Asian session 1.5x (wider, thin liquidity), London session 1.0x (baseline), London/NY overlap 0.8x (tighter, high-conviction moves), Late NY / Pre-Asia 1.8x (widest, erratic conditions).
The empirical phenomenon — volume clustering during TradFi hours — is real and observable in BTC/USDT data. The ICT framing layered on top (precise timestamps, mystical "kill" language) is contested. Treat session windows as a probabilistic environmental filter, not a deterministic signal, and always backtest hit-rates against your own data before sizing on them. See also MAE/MFE statistics by session for the empirical foundation behind the multipliers.