ICT Kill Zones are 2–3 hour windows of peak institutional liquidity. The two primary ones for forex EAs: London Kill Zone (07:00–10:00 GMT) and New York Kill Zone (13:30–16:00 GMT). Restricting your EA to only enter during these windows eliminates dead-hour chop, reduces false signals by 30–50%, and improves profit factor without changing the entry strategy itself.
Table of Contents
Most retail EAs trade 24 hours a day, five days a week. That sounds like an advantage — more time in the market, more opportunities. In practice it is one of the primary reasons most EAs underperform on live accounts after looking good in backtests. Roughly 60% of H1 bar closes happen during low-liquidity windows where price movement is random noise, not institutional order flow.
Kill zones, a concept from trader Michael J. Huddleston (known as ICT), identify the specific hours when banks and institutions place orders that create real, directional moves. Filtering an EA to only trade during these windows is one of the highest-leverage improvements available to any automated strategy.
What Are ICT Kill Zones?
Kill zones are time windows when institutional order flow is concentrated. Banks and large funds operate on fixed schedules — London traders start at 07:00 GMT, New York opens at 13:30 GMT. At these moments, large stop-hunt sweeps occur, liquidity pools are raided, and real price displacement begins. Outside these windows, price drifts without directional conviction.
For a trend-following EA whose logic is built around momentum and continuation patterns, dead-hour entries produce the worst backtest results. A pin bar that forms during European lunch has a far lower completion rate than the same pattern during the London kill zone — the setup looks identical on the chart but the institutional backing is absent.
The Four Kill Zones — Exact GMT Times
🇬🇧 London Kill Zone — PRIMARY
The most important kill zone for CHF pairs, GBPUSD, and EURUSD. London accounts for ~34% of global daily forex volume. The first 90 minutes after London open (07:00–08:30) produce the largest daily moves on CHF crosses. This is where TITAN’s highest win-rate trades occur.
🇺🇸 New York Kill Zone — PRIMARY
The NY open overlaps with the tail of London. This overlap is the highest-volume window of the day. Large stop sweeps occur immediately after US economic data (typically 13:30 GMT). For CHF pairs, NY entries show slightly lower win rates than London but produce larger average winners due to US volatility.
🇦🇺 Asian Kill Zone — SECONDARY
Targets Tokyo open. Relevant for JPY, AUD, NZD pairs. For CHF crosses specifically, the Asian session adds minimal value — CHF liquidity is not concentrated in Tokyo hours. Most CHF strategies perform significantly worse with Asian entries enabled.
🚫 Dead Hours — Avoid Entries
European lunch (10:00–13:00) and post-NY drift (16:00–22:00) produce the most false signals in trend-following EAs. Price action is characterised by thin order books, tight ranges, and frequent fakeouts. Any H1 pattern in these hours has a significantly lower completion rate.
Why Kill Zones Improve EA Performance
| Metric | Before Kill Zone Filter | After Kill Zone Filter | Change |
|---|---|---|---|
| Total trades (annual) | ~180 | ~80 | −56% |
| Win rate | 62% | 72% | +10pp |
| Profit factor | 1.38 | 1.86 | +35% |
| Max drawdown | 8.4% | 3.2% | −62% |
Trade count drops significantly, but every other metric improves. The removed trades were specifically the low-conviction dead-hour entries dragging down the average. Fewer high-quality trades always beats more mixed-quality trades on a prop firm challenge account.
Kill Zones for CHF Pairs Specifically
CHF pairs (CADCHF, AUDCHF, USDCHF) have overwhelmingly London-concentrated liquidity — the Swiss National Bank operates on European hours. The following priority applies to CHF crosses:
| Session (GMT) | CHF Priority | Spread | False Signal Rate |
|---|---|---|---|
| 07:00–09:30 (London open) | ✅ Highest | Low | Low |
| 09:30–10:30 (London mid) | ✅ High | Low | Medium |
| 13:30–15:30 (NY open) | ⚠ Medium | Medium | Medium |
| 00:00–06:00 (Asian) | ❌ Low | Wide | High |
| 10:30–13:00 (EU lunch) | ❌ Avoid | Elevated | High |
TITAN’s 78.6% win rate on CADCHF is built on London + NY kill zone filtering (07:00–10:00 and 13:00–15:00 GMT). Removing the filter and allowing 24-hour trading drops win rate to approximately 64% — a 14 percentage point degradation from a single parameter change.
Kill Zone Filter in MQL5
Add this function to any EA and call it at the top of OnTick():
TimeCurrent() returns broker server time, not GMT. Most ECN brokers (IC Markets, Pepperstone) run on GMT+2 (GMT+3 in summer). Set LondonStart=9 and NYStart=15 if your broker is GMT+2. Verify in Terminal → server time display.
3 Mistakes When Using Kill Zone Filters
- 1. Not accounting for broker server time offset. Raw hour values assuming GMT will be wrong by 2–3 hours on most brokers. Always verify
TimeCurrent()timezone before going live. - 2. Not verifying improvement in backtest. Add the filter, run a fresh backtest, and confirm the metrics actually improve. Some strategies perform better in specific dead hours on certain pairs — never assume, verify.
- 3. Making windows too narrow. A 90-minute window only can cut trade count so aggressively that results become statistically unreliable. Aim for 2–3 hour windows maintaining at least 25+ trades per year per pair.
How TITAN Uses Kill Zones
TITAN AutoTrader uses a dual kill zone filter as one of its four core entry gates. The EA only fires during London (07:00–10:00 GMT) and New York (13:00–15:00 GMT). In 17-month backtesting, the kill zone filter was the single highest-value gate in TITAN’s confluence stack — more impactful than the ADX filter. It directly explains why max drawdown stayed at 1.39% on CADCHF: the most dangerous entries simply never fired.
TITAN AutoTrader — Kill Zones Built In
London + NY kill zone filter is one of TITAN’s four entry gates. 78.6% win rate on CADCHF, 1.39% max drawdown, 17 months tested.
See TITAN Pricing → View Results