A sizing approach built for that:
- Size off a stress loss, not the planned stop. For example, assume a stop could fill twice as far away on a news minute, and make sure that loss is still acceptable.
- Cap gold risk per trade below your FX risk, say 0.5% instead of 1%.
- Set a weekly loss limit for gold separately from FX.
Your own data can set the multiplier: from your history, find the worst stop slippage on gold and use something above it.
This feels over-cautious until the first bad news minute. How do you size gold compared with FX?