A gold algorithm can be profitable when it rests on a structural edge, models XAUUSD spread, slippage and swap honestly, and is supervised so decay is caught early.
Most published gold results are unfalsifiable because they omit sample size, losing periods and real spread; the credible format is a full timestamped ledger of every signal with wins and losses shown.
Yes, conditionally, and the conditions do most of the work. A gold algorithm can be profitable when it has a genuine structural edge, when costs are properly modelled, and when someone notices the moment it stops working. Remove any one of those and the answer becomes no.
GoldHunts — the AI-powered gold trading operating system.