As 100% of COMEX withdrawals in April 2026 were pulled from [E], I will start estimating a run rate for the [E] category.
As the custodian for the iShares SLV ETF, JPM publishes a daily vault stock report which includes the total vault stock held in their NYC/COMEX [E] vault. We can safely discount this amount (currently 85,168,871.700 ozt) from the [E] total vault stock when calculating an upper bound estimate.
We know that more of the COMEX [E] vault stock is not free float - it is also held in custody for private vaulting services. There are no public records indicating how much of the remaining COMEX [E] vault stock is held under these constraints though. We do have a statement from the COMEX some years ago where they claimed ~50% of the COMEX [E] vault stock was illiquid. 50% at the time of the statement may not be exactly true today as the [E] vault stock has fluctuated since then, but I will use it as to calculate a possible lower bound estimate.
I think averaging the upper and lower bound estimates should yield a reasonable "realistic" estimate for an [E] liquid vault stock. Good enough for ball park run rate estimating anyways!
As the custodian for the iShares SLV ETF, JPM publishes a daily vault stock report which includes the total vault stock held in their NYC/COMEX [E] vault. We can safely discount this amount (currently 85,168,871.700 ozt) from the [E] total vault stock when calculating an upper bound estimate.
We know that more of the COMEX [E] vault stock is not free float - it is also held in custody for private vaulting services. There are no public records indicating how much of the remaining COMEX [E] vault stock is held under these constraints though. We do have a statement from the COMEX some years ago where they claimed ~50% of the COMEX [E] vault stock was illiquid. 50% at the time of the statement may not be exactly true today as the [E] vault stock has fluctuated since then, but I will use it as to calculate a possible lower bound estimate.
I think averaging the upper and lower bound estimates should yield a reasonable "realistic" estimate for an [E] liquid vault stock. Good enough for ball park run rate estimating anyways!