XAUUSD M15 S/R Liquidity Sweep and Reclaim, M5 Entry
A mechanical XAUUSD sweep-and-reclaim strategy (15-minute levels, M5 entries, trend and rejection filters, 2R target) lost 91% over five years with costs included. Other readings of the rules lost similarly. Tuning and a quiet-market filter looked better on past data but lost money on the unseen months, from very few trades.
From r/Forex. The idea, in our words: On gold, mark support and resistance zones on the 15-minute chart, then on the 5-minute chart trade in the trend direction when price briefly pierces a zone and closes back inside it with a strong rejection candle.
didn't hold upAs posted · the rules as described (see what we assumed below)
Optimized: best settings Quant Core found · last months held back as unseen data
Improved: skip the quietest markets (volatility in its bottom 30%)
We then tried 8 changes a quant would test first (640 settings in all) and kept the best on past data only. The held-back months played no part in choosing it. It did not hold up on the unseen months.
What we had to assume
- M15 S/R zones taken as M15 fractal swing highs/lows (5-bar pivots) from the last 100 M15 bars, each widened into a zone of ±0.1 x M15 ATR(14).
- Trend defined as M15 close above/below the M15 EMA(50).
- Liquidity sweep = M5 bar whose low (high) trades beyond the zone's outer edge; reclaim = the same or the next M5 bar closes back inside/above (below) the zone.
- Strong rejection = reclaim candle's wick beyond the zone is at least 50% of its range and it closes in the top (bottom) third of its range.
- Stop placed just beyond the sweep extreme (sweep low minus 0.1 x M5 ATR(14) for longs; sweep high plus the same for shorts).
- Take profit at 2R, since the post gives no target.
- Fixed 1R read as a constant fraction of equity (e.g. 1%) risked per trade.
A different reading of the rules could give a different result.
Backtested with the StratOS optimizer on historical data, with spread and costs charged, and the most recent months held back as unseen data. A backtest is not a forecast, and nothing here is investment advice.