Multi Strategy Breakout EA
- エキスパート
-
Andrew Mcculloch
I build real tools that can help you with your trading. If you want AI characters selling you a dream, I cannot help you.
Message me if there is indicators/features you want added to current indicators or have an idea for an indicator to be developed. - バージョン: 1.61
XAUUSD H1 breakout system — honest, tick-validated, free.
What this is
A gold (XAUUSD) breakout expert advisor that trades multiple support/resistance breakout strategies on the H1 timeframe, concentrated in the New York session, with a trailing stop and a false-break margin filter. It places pending stop orders at breakout levels and manages exits with a breakeven-then-trail stop. No grid, no martingale, no averaging-down. Every position carries a hard stop.
It is offered free. I build trading utilities and prefer to put my work out openly. The description below tells you exactly what it does and, more importantly, what its limitations are, because most of what determines whether a system helps you is the honest part that listings usually leave out.
How it was tested
This system was developed and validated on genuine tick data with real embedded bid/ask spreads, not candle approximations. The development deliberately tried to break it at each stage:
- Kill test across the full parameter space before any optimisation.
- Walk-forward validation: strategy parameters selected on past data, tested on unseen future data, rolling across years.
- Monte Carlo resampling (block bootstrap) to confirm the edge is not a lucky ordering of trades.
- Execution stress test with added slippage on entries and stops.
- Full commission and overnight swap modelling, calibrated to real broker costs.
Several variants and filters were tested and discarded because they failed out-of-sample. What remains is what survived honest testing.
Honest performance — net of all costs
Tested 2019–2026 on real ticks, net of commission and swap, the system is positive in trending years and roughly breakeven to slightly negative in choppy years. Profit factor net of costs is modest in the mid-1.1s. This is a real, modest edge, not a miracle. Here is the per-year character (relative, illustrative):
| Year | Net character |
|---|---|
| 2019 | ≈ breakeven |
| 2020 | negative (choppy) |
| 2021 | ≈ breakeven / slightly + |
| 2022 | negative (choppy) |
| 2023 | positive |
| 2024 | positive |
| 2025 | positive (strong trend) |
| 2026 | positive (strong trend) |
The system makes money when gold trends and loses or treads water when gold ranges. It is regime-survivable, not regime-neutral. If you require a system that profits every year, this is not it.
Critical limitations — please read
- XAUUSD ONLY. The edge was tested on seven FX majors and was negative on six of seven. Do not run it on other symbols.
- Trend-dependent. It relies on gold producing directional moves. In prolonged ranging conditions it will bleed slowly. It has had losing years.
- Modest expectancy. Net of costs the profit factor is in the mid-1.1s. Returns are real but not large relative to the drawdown, so position sizing matters more than usual.
- Backtest is not the future. Validation reduces the chance of self-deception; it does not guarantee forward results. Run it on demo first and form your own view.
Settings
All parameters are configurable; the defaults are the validated values.
| Input | Default | Meaning |
|---|---|---|
| Lookbacks | 80 / 100 / 120 | three parallel breakout strategies (H1 bars) |
| Margin filter | 0.50 ATR | break must exceed the level by this margin |
| Initial stop | 2.0 ATR | hard stop on entry |
| Trail activate | +1.0 ATR | move stop to breakeven after this profit |
| Trail distance | 1.5 ATR | trail behind best price once activated |
| Entry window | 13:00–20:00 UTC | new entries only in this window |
| Risk | 0.25% per strategy | conservative by design |
| NFP filter | on | skip Non-Farm Payroll spike |
| Weekend close | on | flatten before weekend to avoid swap |
Sizing guidance
The EA defaults to conservative risk. Because three strategies can each hold a position, your concurrent risk is up to three times the per-strategy figure. At the modest expectancy of this system, I recommend 0.15–0.25% per strategy. Higher sizing produces drawdowns (25%+) that are uncomfortable for the return on offer. Start small, on demo, and scale only if your own forward testing satisfies you.
Time / GMT setup
The EA auto-detects the broker GMT offset for live trading. For backtesting in the Strategy Tester, set the time mode to Manual and the winter/summer offsets to match your broker (the EA logs the detected offset on startup). Run the tester in "Every tick based on real ticks" for accurate trailing-stop behaviour. Note the entry window is defined in UTC; the effective broker-time window shifts by one hour across daylight-saving changes.
This EA is free. If you find the approach useful, my other trading utilities are on my profile. Feedback and honest reviews are welcome.
