Bank at the Fed

"Segregated Balance Accounts" is a nice new paper by Rodney Garratt, Antoine Martin, James McAndrews, and Ed Nosal.

Currently, large depositors, especially companies, have a problem. If they put money in banks, deposit insurance is limited. So, they use money market funds, overnight repo, and other very short-term overnight debt instead to park cash. If you've got $10 million in cash, these are safer than banks. But they're prone to runs, which cause little financial hiccups like fall 2008.

But there is a way to have completely run-free interest-paying money, not needing any taxpayer guarantee: Let people and companies invest in interest-paying reserves at the Fed. Or, allow narrow deposit-taking: deposits channeled 100% to reserves at the Fed.

(I'm being persnickety about language. I don't like the words "narrow banking." I like "narrow deposit-taking" and "equity-financed banking," to be clear that banking can stay as big as it wants.)

That's essentially what Segregated Balance Accounts are. A big depositor gives money to a bank, the bank invests it in reserves. If the bank goes under, the depositor immediately gets the reserves, which just need to be transferred to another bank. This gets around the pesky limitation that the Fed is not supposed to take deposits from people and institutions that aren't legally banks.

...the funds deposited in an SBA would be fully segregated from the other assets of the bank and, in particular, from the bank's Master Account. In addition, only the lender of the funds could initiate a transfer out of an SBA; consequently, the borrowing bank could not use the reserves that fund an SBA for any purpose other than paying back the lender. ...The bank receives the IOER rate for all balances held in an SBA. The interest rate that the bank pays the lender of the funds deposited in an SBA would be negotiated between the bank and the lender

The reverse repo program achieves the same thing, but many at the Fed seem to regard it with suspicion.

Why is this such a good idea? First, from my perspective, it opens the door to narrow banking; to government provided run-proof electronic money.

Second, emphasized in the paper, SBAs could help "pass through" interest rate rises. Suppose the Fed wants interest rates to be 5%  It starts paying banks 5% on reserves.  Banks will probably start demanding 5% or more on loans, since they can get 5% from the Fed. But banks may not compete on deposits, merrily taking our money at 0% and investing at 5%.  Large institutional investors, who can invest in money market funds, aren't going to sit still for that however, so they SBA accounts should very quickly reflect interest on reserves. In turn, that will put upward pressure on short-term commercial paper, Treasury, and other markets, and provide competition for deposits.

I learned an interesting legality. Are the SBA accounts really run free, exempt from bankruptcy proceedings? Not totally

Under the FDI Act, and subject to certain exceptions that are not applicable here, creditors of a DI [Depository Institution] that is in FDIC receivership are prohibited from exercising their right or power to terminate, accelerate, or declare a default under any contract with the DI, or to obtain possession or control of any property of the DI, without the consent of the receiver during the 90-day period beginning on the date of the appointment of the receiver. For purposes of this paper, it is assumed that the FDIC would act quickly to permit lenders to gain access to SBAs that collateralize their loans. However, this treatment has not been approved by the FDIC, and the decision by the FDIC on treatment of an SBA account in resolution could affect the willingness of firms to participate in these accounts.

That's all putting it mildly. It could also affect the willingness of firms not to run at the first hint of trouble, which is the whole point. Evidently, the FDIC needs to carve exemption from bankruptcy in stone.

A few quibbles

The near elimination of credit risk, which is the hallmark of SBAs, would level the playing field so that all banks could borrow in the overnight money market on equal footing..

Well, not really. Sure, they can borrow on equal footing so long as they put the results right in to the Fed. They cannot borrow for other purposes, like to lend it out to you and me, on equal footing.

The paper also echoes the worry that firms might run to these programs in a crisis

One concern is that SBA take-up could be too large. .. in times of intense stress, which may be characterized by a flight to quality, flows into SBAs could produce a scarcity of reserves that banks use to meet reserve requirements and could also cause (temporary) dislocations in funding markets for nonbank entities.  

I beat up on this view in discussing the overnight RRP program here, so I won't make the same points again. It still makes no sense to me. Flows into SBAs have to come from somewhere; and we're $3 trillion dollars away from required reserves anyway. And will be even further away once this program goes in.

Update: In fact, when you dig in to the paper, it pretty much concludes that these "financial stability" arguments are not important. From p 18

Recently, market observers and policy makers have expressed concerns that uncapped ON RRPs could exacerbate flight-to-quality flows, by providing a risk-free alternative to bank deposits, thereby causing a removal of much needed liquidity from the financial system.  For these reasons, an aggregate cap on the amount that can be invested at the ON RRP facility has been imposed and an auction pricing mechanism has been introduced to ration ON RRPs in the event that bids exceed
the aggregate cap. 

A similar concern could arise with SBAs. During a crisis, SBAs might be seen by lenders as an attractive near risk-free investment. However, a "surge" into SBAs i.e., an increased supply of funds by lenders for SBA collateral arrangements, would be accommodated by counterbalancing price movements.... an increase in the federal funds rate, as usable reserve become scarce. Further, because SBAs are supplied competitively, their rate would not adjust, since the rate is "competitively tied" to the IOER rate. The result would be an increase in the spread between the federal funds rate and the rate paid on SBA balances, which would help to arrest the surge and mitigate potential dislocations in funding markets. 

Additional factors could limit the ability of investors to suddenly place large sums of money into loans secured by SBAs. ...

I think there are deeper conceptual problems with the whole argument that offering SBAs, ON RRPs, or floating-rate Treasuries contributes to a run by offering a safe alternative, but in the end we are agreeing just for slightly different reasons.

Reserves for all! Via money funds and overnight RRP, or via narrow deposits at banks. Or, via fixed-value floating-rate Treasuries. Let the run-proof financial system begin to emerge.

Now, if the Fed would only say "and, by the way, any bank that puts all of its deposits in SBAs, and finances all of its lending with equity capital, will be exempt from all the Dodd-Frank regulation and stress tests, because it is obviously completely un-systemic."