Gold EMA Cross Scalper

How to read these numbers

• The risk model engaged as configured. On a $10,000 deposit a 2% risk resolves to real position sizes — lots ranged 0.02 to 0.21, none floored at the broker minimum or capped at the 2.00 lot ceiling — and realised risk came out at a median 1.89% of balance per trade. The +58% headline is a fair reflection of the risk that was actually taken.
• The edge is concentrated in a handful of trades. The three largest winners account for 85% of the $5,811 net profit, and stripping out the top five leaves the remaining 633 trades net negative. That is the expected shape of a let-winners-run system, but it means the result rests on catching a small number of large moves — a window that missed them would look very different. The 1.16 profit factor says the same thing: $42,953 of gross profit against $37,142 of gross loss.
• Only 27% of trades win, and the test ran through 14 consecutive losses at one point on the way to a 24.7% peak-to-trough drawdown in closed balance (28.1% on equity). A system with this distribution only works if every signal is taken; skipping trades after a losing run is how the winners get missed.
• The worst trade lost 3.4R, not the 1R the stop was sized for. A position held over the weekend of 16–18 January 2026 gapped through its stop and closed $630 down — 6.6% of the account at the time. Stops bound risk under normal conditions; they do not guarantee it across a weekend gap.