Sweep and Engulf
AUDJPY, US500, XAUUSD · 4H chart · strategy s01_sweep_engulf · 4 variations
Buys gold when a 4-hour candle first dips below the previous candle's low and then closes above its high, and lets winning trades run with a trailing stop.
What it is about
A sweep-and-engulf candle does two things in one candle. First it pushes below the previous candle's low, where many traders keep their stop-loss orders, and triggers them (the 'sweep'). Then it turns around and closes above the previous candle's high, covering its whole range (the 'engulf'). Within four hours, the sellers who pushed price down were overpowered.
The idea is that this shakeout followed by a strong close shows buyers taking control, and that the move often keeps going. The pattern comes from Victor's own TradingView indicator, which marks it in both directions. Testing found the useful part was buying gold in the direction of its longer-term trend. Sells lost money in 2024-26.
There is no fixed profit target. Gold often trends for days, so the strategy keeps winning trades open behind a trailing stop: a stop-loss that moves up as price rises. A few large winners pay for the many small losses.
How it trades
Gold, around the clock from Sunday evening to Friday. A signal can come at any of the six 4-hour closes. About 2-3 trades a month (2.7 for A, 2.2 for C). Trades last from a few hours to several weeks: a typical trade takes about 1-2 days and the average is about 4 days.
- A bit fewer than half the trades win (44-45%), but the average winner is close to 3R against an average loss of about 1R.
- About 2-3 trades a month, so there can be quiet weeks with nothing to do.
- Losing streaks of up to 10 trades in a row happened in the backtest (variation A).
- Largest drop from a peak (drawdown) about 13.6R-14.2R. Backtest 2018-today at zero cost: +195.7R (A) and +178.8R (C).
- Two of nine years lost money (2018 and 2022, under 6.5R each). Most of the profit came in 2019, 2024 and 2025.
- Buys only, on a market that rose strongly in 2024-26. The trend filter should keep it out of a long gold decline, but that has not been tested at length. Sells were tested and lost money in 2024-26 (26% wins).
- Several rules (Daily EMA 50, the straight-line skip, buys only, the rejection wick) were chosen while looking at the same history they are judged on. The past results probably look a little better than the future will.
- Trades can stay open for days or weeks. Overnight financing (swap) is not included in the results.
- The backtest uses Dukascopy price data. A broker's prices differ slightly. Checked against TradingView (FOREX.com data), both found 491 trades, and the totals differed by about 4% (+164.7R vs +158.0R).
- With only 2-3 trades a month, a flat or losing year is normal and does not by itself mean the strategy has stopped working.
Where it comes from
Victor's own TradingView indicator ('Sweep and Engulf'). It was rebuilt in Python and checked against TradingView, which found the same trades. It was then tested on 2018-2023 (the 'train' period, used to choose settings) and checked on 2024-2026 (the 'test' period). The trailing exit, Daily EMA 50, straight-line skip, buys-only and rejection-wick rules were added during that research.
Main variation: s01-D-AUDJPY ★ Main
The variations
Each variation is the same strategy with a different approach, trading its own paper account. Click one to see exactly what it changes.
D's rules on the AUDJPY 4-hour chart: buys only above the Daily EMA 50, and from +1.5R the stop trails 1.5R behind the best price. Found when every variation was run on 11 currency pairs; it beat random entries in every run and all 68 nearby settings were profitable in both periods.
s01-ATrailing buys with trend filtersThe base indicator trades both directions with a fixed 2R target. A keeps only buys, requires the rejection wick, the Daily EMA 50 uptrend and no straight-line run, and replaces the target with a 2R trailing stop that starts at +2R. In the research, the sell side and the fixed target left money on the table on gold. Letting trends run paid for the losers.
s01-CA + Daily market structureSame as A, plus a market-structure check on the Daily chart. Buys only while the Daily trend counts as 'up'. That starts when the Daily chart makes a higher low and then a 4-hour candle body closes above the highest point between the two lows, and it lasts until the mirror pattern appears. Swing points follow Victor's video rule: a swing high only counts once a later candle body closes below the low of the candle that made it. The aim is to trade only when the bigger picture is clearly rising.
s01-DUS500 4-hour buys, 1.5R trailing stopThe same sweep-and-engulf candle on the S&P 500 4-hour chart, buys only while the last Daily close is above its 50-day EMA. No fixed target: once a trade is 1.5R in profit the stop trails 1.5R behind the best price. Added after LuxAlgo backtested the same idea and our own test found indices work on the 4-hour chart.
Variations side by side
Live result of all 4 together: −2.0R (zero cost).
| Live paper account (R, zero cost) | Backtest 2018 – today (R, zero cost) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| s01-D-AUDJPY★ MainSweep and Engulf | AUDJPY | 4H | −1.0R | 1 | 0% | – | +142.8R | +0.48R | 8/9 | 2.8 | 51% | −9.8R |
| s01-ASweep and Engulf | XAUUSD | 4H | – | 0 | – | – | +195.7R | +0.70R | 7/9 | 2.6 | 44% | −14.2R |
| s01-CSweep and Engulf | XAUUSD | 4H | – | 0 | – | – | +178.8R | +0.78R | 7/9 | 2.2 | 45% | −13.6R |
| s01-DSweep and Engulf | US500 | 4H | −1.0R | 1 | 0% | – | +111.5R | +0.25R | 7/9 | 4.2 | 45% | −18.1R |
Main = the variation with the most backtest R per R of max drawdown (2018 – today, zero cost). It is picked on that history, so the live accounts are the real test.