currency is not reserves. currency earns nothing. it’s astonishing that $2.4T of zero interest currency is outstanding while interest rates are 4%ish. kind of wow. (according the FRED, 2.8T of FRNs are now outstanding. that’s a lot of seignorage!) fred.stlouisfed.org/series/RESPP... 1/

current reserve balances are about $3T. securities are about $6T, so I was wrong to say the Fed would run short of FFR collateral. (my intuition was formed in the theory that zero-interest currency would be small, not 1/3 to 1/2 of the balance sheet!) 2/

a lot of factors can affect the relationship between reserves outstanding and the FFR that prevails absent IOR. again, we did live in this world for a long time. it’s not fanciful. 3/

it’s not based on the relationship between the Fed’s balance sheet and currency or reserves outstanding, however. it depends upon how banks manage reserves, a game of balancing minimizing the cost of zero-interest reserves vs the cost and risks of having to borrow if short. 4/

the cost and risks of having to borrow if short are a function of the scale and volatility of net-settlement requirements. 5/

from the central bank’s perspective it’s an empirical art, not a game of first principles. 6/

(but it’s an empirical art whose rules central banks can change if necessary. beyond settlement requirements, in theory there are reserve requirements. in a no-IOR world we might revisit the sweeps loophole, if CBs find they want both control of rates + contemporary girth in balance sheets.) /fin

sure. if IOR goes away but RRP, the facility maintained to allow nonbanks to earn effectively IOR, was maintained, then IOR would just be replaced by banks using that facility. 1/

as you suggest, there are potential glitches: RRP is limited by the Fed’s collateral. “the Fed’s balance sheet” is too vague a descriptor for constraints: reserve liabilities are a greater share of that balance sheet than suitable RRP collateral. 2/

but setting that aside, this is all a way of saying if the Fed did away with IOR but didn’t REALLY do away with IOR by maintaining RRP, little might change if technicalities are addressed. 3/

if the Fed is forced to actually stop paying IOR—whether directly or via interest paid to repo lenders—on the theory that banks don’t deserve the interest, presumably the RRP loophole would be eliminated as well. 4/

in which case, short-term risk-free rates are pinned at zero until the quantity of reserves is made scarce relative to banks — still extant! — need for reserves to clear and settle overnight. 5/

(yes, other facilities, other technicalities, but none are designed to replace actual settlement at scale. absent IOR, unless facilities are retooled to literally replace reserve-based net settlement, banks will come to demand and pay to borrow reserves when quantity grows sufficiently scarce.) 6/

(new systems and tools and facilities are layered on top of old systems and practices which, like a reptile brain, remain, sit beneath, backstop our monetary system. no existing facility is capable of replacing net settlement at scale, with FFR and the discount window and all of that.) /fin

Treasury would be paying very little interest to banks, because banks would bid short-term bills to yields of zero+ε, while the banking system holds such a large pool of suddenly zero-interest, hot potato, reserves. 1/

Effective FFR and T-bill rates would get pinned at near zero, unless/until the Fed drains reserves to render them scarce relative to demand for clearing purposes. /fin