Revanche Breche Gold

A one-year test · XAUUSD · 207 trades, both directions. The markers that separate a real edge from a curve-fit:

A 75% win rate with a profit factor near 1.8 at every setting and low-single-digit drawdown at the calm tier — balanced long and short. Fewer, cleaner trades that let winners run past the capped losses. The full tier breakdown is below.

Levels hold until they don't. When gold finally breaks structure — closing decisively through a level it had respected — real money has moved through it. That is a decision, not noise.

The clean breaks get retested: price returns to the broken level, finds it now holds from the other side, and pushes on. Old resistance becomes support; old support becomes resistance. Where most robots chase the first move or grid into it, Breche waits for that retest to confirm, then enters with the break — a single position with a hard stop, trading the continuation rather than guessing a top.

One position at a time. No grid, no martingale, no averaging down.
A hard stop on every trade. Sizing is deterministic — the exposure is always known.
Stealth stop-loss. Your real level stays off the server; a protective stop still guarantees an exit.
Every loss is capped. A hard stop sits on each trade from the open — the rare break that fails and runs the other way can't turn into a blow-up.
Long and short. Netting and hedging both handled; orders are magic-stamped, so your manual trades are untouched.

Trade a Fixed lot, or let Breche size from account balance. One decision — the strategy underneath is the same at every setting.

XAUUSD · 1 Oct 2025 to 1 Oct 2026 · $3,000 start · 207 trades · 75% win rate. The same trades, sized by each setting:

Figures are from a historical test and describe the past. They are not a forecast or a promise of future performance.

No VPS required. Breche acts on bar closes, not the tick, so it is not latency-bound.
No special broker. A built-in spread ceiling and stealth stops keep it broker-agnostic, standard or raw accounts alike.
Any gold feed. Arbitrary symbol names (XAUUSD, GOLD, XAUUSD.e), 2- or 3-digit quotes, any account currency, cent accounts.
Netting or hedging. No DLLs. A single EX5.
A stable feed and reasonable spread give the best fills — favourable conditions help, they are not required.

Because the setup lives in specific hours, Breche needs to know your server's time zone. It auto-detects the broker's GMT offset on attach and re-checks it for daylight saving, then anchors its trading hours and day filters to one internal reference clock. If auto-detect is ever off, a single manual GMT offset pins it exactly. That is the only broker-specific setting Breche ever asks about.

There are no strategy knobs to second-guess. The engine's logic is sealed in the compiled build; you control risk and housekeeping only.

Trading leveraged instruments carries risk of loss. Test on a demo account first, choose a risk setting appropriate to your capital, and never commit funds you cannot afford to lose. Test results reflect historical market conditions and do not guarantee future outcomes.