How long should I test a gold strategy before going live?

Long enough to cover at least 100 trades and one full macro cycle — a CPI print, an NFP and an FOMC meeting. For an intraday XAUUSD strategy that is usually eight to twelve weeks; for a swing strategy it can be six months.

Is backtesting gold enough on its own?

No. A backtest cannot reproduce the spread widening, slippage and rollover behaviour that decide whether a gold scalp is profitable. Backtest to find the pattern, forward test on demo to find the cost.

What is the single most useful thing to record?

The intended stop next to the executed stop. On gold the gap between the two is where most accounts quietly die.

TESTING A GOLD STRATEGY

Testing a gold trading strategy requires more evidence than testing an FX or index strategy because XAUUSD carries wider spreads, larger average true range and heavier reaction to US macro data.

The four-stage method is backtest, demo forward test, small live allocation and continuous journal review, with a minimum sample of 100 trades covering CPI, NFP and FOMC releases.