FREDDoes the yesterday's level sweep (buy) work? Tested on 52,191 trades
Not on its own, but 2 filtered versions did. Across 52,191 yesterday's level sweep (buy) trades on 8 markets from 2020 to 2026, after real broker spreads, the setup averaged −8% of the risk per trade (−£0.78 per £10 risked) and reached a 1.5× take profit 34% of the time.
Older years vs recent years
2020 to 2023: −7% per trade over 30,146 trades. 2024 to 2026: −9% per trade over 22,045 trades.
The versions that made money
These made money in both 2020–23 and 2024–26, which is how Fred rates a setup green. Each uses its own best stop and take profit.
Got a chart open? Upload a screenshot and Fred checks it against six years of tests.
Read my chart free5 free reads. For learning and demo practice. Not financial advice.
By market
| Market | Trades | Per trade | Hit take profit |
|---|---|---|---|
| Gold (XAUUSD) | 7,141 | −4% | 35.2% |
| US30 (Dow) | 7,222 | −6% | 34.6% |
| USDJPY | 6,193 | −7% | 34.6% |
| EURUSD | 7,756 | −7% | 34.2% |
| GBPJPY | 6,637 | −8% | 33.3% |
| NAS100 (Nasdaq) | 2,200 | −9% | 32.5% |
| AUDUSD | 7,360 | −10% | 33.5% |
| GBPUSD | 7,682 | −11% | 32.9% |
By timeframe
| Chart | Trades | Per trade | Hit take profit |
|---|---|---|---|
| 5-minute | 27,575 | −10% | 33% |
| 15-minute | 16,321 | −7% | 33.5% |
| 1-hour | 8,295 | −1% | 38.1% |
How Fred tested it
Every time the setup appeared on MT5 price history from 2020 to 2026, Fred entered at the next candle's open, paid the real spread, placed the stop just past the pattern (plus half an average candle) and a take profit 1.5 times the risk. If neither was hit within 48 candles, the trade closed at the last price. "Per trade" is the average result as a share of the amount risked.
Back-tested results, not live trading. Past results don't guarantee future ones. Checked 28 October 2026. All tested patterns · AI chart tools compared