Red flags in expectancy numbers when scalping gold

[FXS London Scalper]

Gold scalping can look great in a spreadsheet. Before trusting it, check how the numbers were built. Three common errors:

• R from the planned stop. On XAUUSD, real stop fills are often worse than the stop price in fast moves. Compute R from the actual fill, not the plan.
• Concentration. If a handful of trades in one news week made most of the profit, take that week out and see what's left. If expectancy drops to near zero, the strategy is a news-week strategy, not a daily one.
• Wrong commission. Brokers often charge per lot on a 100-ounce contract. A fixed commission figure copied from EURUSD can understate the cost per trade.

A worked example of the first point: if stop fills average 30 cents worse on a $3.00 stop, every losing trade costs 1.1R instead of 1R. With 50% losers, that takes 0.05R off the average trade, which can be a third of a thin edge.

After corrections, a strategy that looked like 0.2R per trade may be 0.05R. That's still positive, but it changes how much attention gold deserves.

What do you check first when a strategy looks too good on paper?