Really cool idea. A fully self-sufficient EA could definitely outperform humans if risk management and adaptive logic are solid. Interested to see how it handles different instruments and timeframes.
I disagree with the “EAs can’t be profitable long-term” argument. The problem is most bots are overfit and collapse when market conditions change. A fully hands-off system is possible but the hard part is proving it survives different regimes, spreads, slippage, and real drawdowns. Higher timeframes should give you a cleaner test.
