IOER, negative rates, and Ben | FT Alphaville

IOER, negative rates, and Ben

The ECB’s recent decision to lower its deposit rate to zero raised speculation in the market that the FOMC might be considering the reduction or elimination of the 0.25 per cent interest the Fed pays on excess reserves.

Bernanke himself hadn’t mentioned the idea lately prior to his testimony before Congress this week, and it hasn’t come up in FOMC minutes since last September’s meeting. Yet the market continues to price in the possibility of a cut, as Barclays analysts noted this morning:

Term OIS has fallen between 3 and 4bp during the past week – even though last week these rates were practically the same as the overnight rate. Interestingly, despite expectations of a cut in IOER priced into OIS, there has not been any pricing into the repo market – 3m and longer-term repo remains stuck in the 20-23bp range.

During his testimony on Tuesday, Bernanke did include the possibility of lowering IOER in a list of measures the Fed could potentially take in future, but he did so only casually, as if taking inventory rather than suggesting he wanted to use it.

Or as a UBS strategist writes: “It’s on the menu, but not on the table.”

We agree, and very much doubt the idea is indeed under serious consideration. But for reasons that will become clear, we also think it’s worth explaining why.

The theoretical argument in favour is straightforward: lowering the rate means less incentive for banks to keep excess reserves on deposit with the Fed and more incentive to lend them out. Sounds good.

But the argument against is often summarised only briefly in vague terms of market disruptions. We wanted to be a little more specific, as the answer is related to broader (and very important) money market stability, negative rates, and much else besides.

First is a step-by-step explanation of what the Fed is likely worried could happen, followed by a few thoughts of our own.

1) We’ll begin with this helpful summary, via Barclays, of what it means when IOER is above the rates at which banks can fund themselves in short-term money markets vs when IOER is below financing rates:

When IOER exceeds prevailing financing rates, banks have an incentive to raise money to leave on deposit at the Fed, earning a positive spread. But cutting the rate, say, to 0%, eliminates the arbitrage and makes the reserves more of a burden. Provided market rates are higher, banks reverse the arbitrage: buying repo, bills, and other short-term instruments with above 0% yields.

2) But of course, if market rates are higher, then banks will just be “buying repo, bills, and other short-term instruments” from each other, pushing down rates on all of these instruments until they get to zero or even turn negative.

3) In a way, IOER represents a source of “supply” for money market funds because of the incentive for banks to play this arbitrage (between their funding rates they pay to MMFs and the IOER).

4) Lowering or eliminating IOER therefore creates a particular problem for money market funds, as many MMFs depend on lending to financial institutions in repo markets, through reverse repos, and in the market for commercial paper, though the latter continues to decline.

Money market funds’ business models and margins remain pressured by a shortage of supply and by an amended Rule 2a-7, which limits investments in lower-tier commercial paper and requires shorter-maturity and more-liquid holdings.

As for government-only MMFs, well, yields on the relevant government paper would plunge alongside everything else. Right now the 2-year, for instance, is yielding 0.21 per cent, slightly below IOER. But demand for these bills would spike, with both their yield and repo rates rates collapsing, if IOER were lowered.

A recent surge in dealer hoarding of Treasuries has pushed up repo rates and alleviated some of the pressure for now, but as we just said above, repo rates would likely fall along with everything else as both money markets and banks begin competing to squeeze yield from the same collateral.

5) Eliminating IOER, and thereby sending rates on short-term instruments to zero, will force money market funds to put their money on deposit with the banks, and earning essentially nothing, because the credit risk is no longer worth the paltry yield (if there even were any) they would get in short-term lending to the banks. Investors in money market funds would surely demand the switch to deposits. Consider that the unlimited deposit insurance on non-interest bearing accounts at banks has taken quite a bit of investor money from money market funds.

6) But banks will be less than keen on receiving all this cash given their own lack of options. Back to Barclays:

As deposit balances swell, bank deposit insurance assessments (which are based on total assets less Tier 1 capital) also rise. Although unrelated to a decline in market rates, a hint of the “buck passing” between money fund depositors andbanks occurred last summer. One large bank began charging a fee (effectively a negative interest rate) for abnormally high cash balances in response to the inflow of money pouring out of government-only money funds ahead of a politically messy debt ceiling negotiation.

Banks, quite simply, would lose money. And banks don’t like losing money, which means they would consider charging fees on some of these deposits. In other words, negative rates.

7) This is similar to the reason given by EUR-denominated money market funds when they stopped accepting new money after the ECB lowered the deposit rate to zero. According to Bloomberg Businessweek: “More than half of Europe’s money market funds by assets have closed because securities they invest in pay negative returns after the ECB cut interest rates”. Indeed it didn’t take too long after the ECB meeting for European bill and repo rates to turn negative.

8) Then what? To be honest, it’s hard to say. The risks, in sum, are that rates will plunge to zero or negative, money market funds and their investors would panic as their sources of yield disappeared, and that banks will follow Bank of New York Mellon’s lead last year and consider the (politically uber-controversial, btw) possibility of charging fees on deposits.

Money market funds would likely be subsidised for a time by their sponsors, but that can’t be counted on to the extent that it was before the crisis. Were this to pass, we couldn’t with any certainty predict the consequences — but given the panic that ensued when Reserve Primary broke the buck, it’s worth taking none of this lightly.

But in addition to general chaos in money markets, here are three more possible worries resulting from the above:

— The ensuing fall in rates would (further) unlink the explicit policy rate from other short-term rates,

— The policy would signal would be unclear given that Operation Twist involves the Fed selling short-end Treasuries, keeping collateral in the market to bolster repo rates, and

— There are problems such as delivery fails that could arise because of the zero per cent floor on repo transactions, which Izzy did a marvelous job of explaining last year. (Proving once again that it’s her world and we’re all just living in it.)

Or, you know, perhaps none of this would come to pass. But the reason we think the Fed isn’t seriously considering the possibility is that it seems as focused as ever on the potential effects of monetary policy on market functioning. And whether or not the above consequences are likely, it should be obvious why the Fed would at least have to think strongly about whether the stimulative benefits of reducing IOER (also unclear!) are worth risking the possibility.


Now, we doubt things would ever get to this point:

There are formidable logistical obstacles involved in the banks collectively converting a significant chunk of their excess reserves into cash. In the first instance the Eurosystem has to print currency on a gargantuan scale. Second, the commercial banks have to pay non-trivial sunk costs of setting up secure facilities in which to store that cash.

The problems do not end here. Cash on this scale starts to lose its function as a medium of exchange – only the other banks which have set up the secure facilities to hold cash on an industrial scale are likely to willingly accept million euro notes.

But even this extreme scenario of physical cash hoarding can give a sense of the larger issues involved.

Would negative short-term rates, under current conditions, spur people to spend more of their income and institutions to invest more of their retained earnings — or to seek credit from banks and capital markets against their future income and earnings? Would banks and bond investors give them the money?

Maybe. But it is just as possible to imagine how, regrettably and almost perversely, negative rates would actually make things worse. How they would lead to higher demand for money itself — to deflationary expectations.

If you knew that your savings vehicle would give you less money than you put in and didn’t see any superior alternatives, then it wouldn’t be unreasonable for you to choose to save even more money. Not everyone would make such a choice if rates were to go negative, but these are unusual times and the point is that the perceived existence of better alternatives cannot be taken for granted.

Andy Haldane has described the current state of the world as one in which “return on capital is no longer investors’ priority. Return of capital is… Investors are taking literally the notion of a security.”

In a sense, interest on excess reserves is like a safe asset for banks — you can consider it an imperfect near-substitute for the Treasuries and agency MBS that the Fed has removed from the market. And along with the unlimited deposit insurance on non-interest bearing accounts, it is (from a certain point of view) a kind of ongoing bailout for money market fund investors.

To build on our earlier posts, much of the case for paying interest on reserves is of a part with the case for considering fiscal policy (in terms of safe debt issuance) inseparable from monetary policy in a world where safe assets have money-like properties (and where credit creation is money creation).

In other words, if you want zero IOER and more lending, what you should be pushing for isn’t just a straightforward decision by Ben Bernanke, but rather a push for more expansionary fiscal policy.

Scrolling back through what we’ve just written, we notice the repeated use of the words “might” and “could”. Because truthfully there is so much about what we’ve just written that we remain uncertain about. So much that feels… philosophical, theoretical, metaphysical, whatever.

We’d be especially keen to hear from market monetarists on this, who have traditionally called for the elimination of IOER. (We’re not instinctively opposed to NGDP level targeting and agree that a clear statement from the Fed that it would temporarily tolerate higher inflation is needed; we just don’t agree that a target would eliminate the role for fiscal policy.)

As difficult as we find it to think through these issues, it must be infinitely more confounding and frustrating for policymakers.

Related link:
Why cutting IOER would be suicidal – FT Alphaville
The base money confusion – FT Alphaville
The high powered money problem – FT Alphaville