1. ICT-Style Killzone Sweep
Trade only inside chosen session windows: sweep, change of character, fair value gap, entry in discount.
2.What it is
This strategy turns popular ICT-style vocabulary into one testable rule: during a chosen high-volatility window (a killzone), wait for a sweep of the session's range, a change of character and a fair value gap, then enter inside the gap in the discount half of the move (for buys) with the stop beyond the sweep.
In simple words
Only look for trades in two or three busy hours. In those hours, wait for price to grab the stops of the quiet range, jump back and leave a gap. Buy when price returns into the gap's lower half, with the stop under the lowest point.
3.Why it may work
Volatility and participation concentrate at certain hours, which gives price room to move. Combining a sweep, a structure shift and an imbalance asks for several confirmations, which means fewer trades. It is one interpretation of the market and must be judged by its measured results.
4.Best market conditions
- The trade is inside your chosen killzone window
- A clear prior range or session extreme to sweep
- A change of character after the sweep
- A fair value gap from the displacement
- Entry in the discount half (buys) or premium half (sells)
5.When NOT to trade it
- Outside your killzone
- No clear sweep or structure shift
- The gap is tiny
- A major news release is imminent
- You cannot define the stop beyond the sweep with acceptable risk
6.Timeframes
Entry: 5M, 15M · Context: 1H
Mark the session range on the 1H chart; trade the sweep and the gap on the 5M or 15M chart.
7.Entry rules and step-by-step setup
Entry rules
- The window is open
- Sweep, change of character and gap are all present
- Price returns to the gap in the discount half
Step by step
- 1.Choose one or two killzone windows in your time zone
- 2.Mark the Asian (or prior session) high and low
- 3.Wait for a sweep inside the window
- 4.Wait for a change of character and a displacement that leaves a gap
- 5.Mark the 50% level of the move: buy below it (discount)
- 6.Enter inside the gap in the discount half; stop beyond the sweep
8.Stop loss
Beyond the sweep extreme.
9.Take profit
The opposite side of the session range or a fixed 2R; partial profit at 1R is common.
10.Risk / reward
At least 1:2. The stop sits beyond the sweep, so measure its distance first: if the target is closer than twice that distance, skip the trade.
11.Confirmation
- A reaction candle inside the gap
- The 50% level of the move holds as the limit of your entry zone
- Higher-timeframe bias agrees
12.Invalidation
- A close beyond the sweep extreme
- Price fills the entire gap and closes beyond it
13.Common mistakes
- Trading all day instead of in the window
- Using the 50% level or any retracement as a magic number
- Ignoring daylight-saving changes to session times
- Skipping the structure shift
14.Example chart
How it performed in our own backtests
TESTINGBacktested on 820 trades but not verified: 6 of 9 checks not met (win rate above 50%; profit factor at least).
- Win rate
- 33.5%
- Sample size
- 820 trades
- Average planned RR
- 1:2
- Profit factor
- 0.87
- Worst max drawdown
- 27.5%
- Average per trade
- -0.092R
- Tested timeframes
- 15m, 30m, 5m
- Tested markets
- BTCUSDT, ETHUSDT, SOLUSDT
Period 2026-03-09 to 2026-10-03 · OKX spot candles (real exchange history, closed candles only) · engine bahari-bt-1.0.0 · run on 2026-10-03 · 1% risk per trade, 0.05R extra cost per trade plus spread · one position at a time, ambiguous candles resolved stop-first · parameters fixed before running.
London / New York windows only: sweep, change of character and a fair value gap, stop beyond the sweep, 1:2 target.
VERIFIED checklist
- ✔Sample size of at least 300 trades(820 trades)
- ✘Win rate above 50%(33.5%)
- ✘Profit factor at least 1.2(0.87)
- ✘Average result per trade above 0.05R after costs(-0.092R)
- ✘Worst max drawdown at most 25%(27.5%)
- ✔Average planned reward:risk at least 1:1(1:2)
- ✔Tested in at least 4 market/timeframe conditions (2+ markets, 2+ timeframes)(9 conditions, 3 markets, 3 timeframes)
- ✘Profitable in at least 67% of the conditions(2 of 9)
- ✘Profitable in BOTH halves of the history (profit factor at least 1 in each)(first 0.8, second 0.95)
Results per market and timeframe (9 runs)
| Market | TF | Trades | Win rate | Total R | Max DD |
|---|---|---|---|---|---|
| BTCUSDT | 15m | 67 | 29.9% | -13.8 | 20.58% |
| BTCUSDT | 5m | 73 | 34.2% | -10.1 | 12% |
| BTCUSDT | 30m | 113 | 34.5% | -6.5 | 14.79% |
| ETHUSDT | 15m | 83 | 36.1% | -0.5 | 16.4% |
| ETHUSDT | 5m | 72 | 38.9% | 3.7 | 14.14% |
| ETHUSDT | 30m | 96 | 27.1% | -24.9 | 27.51% |
| SOLUSDT | 15m | 93 | 37.6% | 3.3 | 9.66% |
| SOLUSDT | 5m | 92 | 34.8% | -4.0 | 13.65% |
| SOLUSDT | 30m | 131 | 30.5% | -22.3 | 26% |
Reading this honestly: VERIFIED needs 300+ trades, a win rate above 50%, profit factor 1.2+, drawdown under 25%, profit in at least two thirds of the conditions and in both halves of the history. A strategy that fails here can still be worth learning, but do not risk money on it before testing it yourself in the Strategy Lab.
Past and backtested results never guarantee future results. Nothing on this page is financial advice.
15.Quiz
16.Practical challenge
Check the session window, then look for sweep, change of character and a gap in a challenge.