Here’s another reminder, this time from New York Fed staffer Lorie Logan: While the Federal Reserve plans to cut the size of its balance sheet, it probably won’t end up shrinking that much.

Even before the Fed addresses the tougher questions — how it wants to control interest rates, and who it wants to trade with — the balance sheet should probably be around $2.1tn, Logan says. (For context, the Fed currently owns about $2.5tn of Treasuries and $1.8tn of mortgage-backed securities.)

With our emphasis, and explanation in brackets:

In projecting the size of the normalized balance sheet, one can make assumptions about the long-run trends that will determine future levels for each of these non-reserve liabilities, subject to various kinds of uncertainty. Even taken at today’s levels, the TGA [the Treasury’s account with the Fed], other deposits, and the foreign repo pool alone account for about $600 billion of the Federal Reserve’s current balance sheet size. Adding on currency outstanding brings this set of non-reserve liabilities to $2.1 trillion today, and this set is likely to grow as the economy continues to expand. This sum does not yet consider reserves or ON RRPs.

It’s notable that Logan assumes the foreign repo pool — a decades-old Fed program for foreign central banks — will remain near its current size. The pool is much larger than it was before the financial crisis. To simplify a bit, regulations have prompted private banks to charge more for repo transactions, so foreign central banks have been going to the Fed instead. (The trend is analysed quite nicely over at Concentrated Ambiguity.)

That $2.1tn also includes deposits from Fannie, Freddie and major clearinghouses, Logan says:

…designated financial market utilities (DFMUs) can now establish and hold cash deposits in Reserve Bank accounts. The buildup of DFMU balances has prompted “other deposits” on the Federal Reserve’s balance sheet to grow steadily over recent years to a current level of about $85 billion, compared to a pre-crisis average of less than $1 billion. The other deposits category also includes balances held by GSEs and international and multilateral organizations.

As we get ready to tackle the tougher questions about policy control, here’s some background:

Before the financial crisis, bank reserves were scarce, so banks sometimes had to borrow reserves from peers to meet requirements. The Fed carried out policy by buying or selling reserves, in order to control the rate at which banks lent them to one another overnight. That process was complicated and sometimes imprecise, Logan says.

Then, the Fed’s bond-buying programs created trillions of dollars of excess bank reserves. So instead of trading reserves, the central bank controls rates by paying interest to banks and money-market funds.

This system seems to be working very well! It even managed to weather a massive shock from money-market fund reform.

But the presence of money-market funds raises two questions:

(1) Should the Fed trade with anyone besides bank-owned dealers? Money-market funds have not historically been counterparties for the Fed. That’s usually the realm of bank-affiliated trading firms known as primary dealers.

(2) Should it use only reserves to control policy? The funds also can’t earn interest on reserves. To achieve a similar outcome, the Fed does a reverse repurchase transaction, meaning it sells them a Treasury and agrees to buy it back for the price, plus interest, at a future date. Of course, that makes the Fed (kind of) active in the Treasury market.

Before it started raising rates, the Fed’s answers were (1) No and (2) Yes.

But recent events provide some argument that the Fed should consider other possibilities, Logan says. With our changes in brackets:

The FOMC has said that it plans to phase out the [overnight reverse repurchase] facility when it is no longer needed to help control the federal funds rate… it may be that some such facility would be helpful in maintaining effective interest rate control… Of course, a decision to maintain an ON RRP facility would require longer-term choices about counterparty relationships and consideration of possible financial stability risks.

Before the Fed expanded the RRP facility in 2015, staffers found “evidence… that the payment of interest on reserves may not provide the degree of control” the FOMC would want over interest rates.

And Logan says that funds don’t even need to use the facility for it to be useful:

We’ve observed that a high volume of actual ON RRP usage has not been necessary to achieve interest rate control. In the first half of 2016, for example, average daily usage of the facility was $63 billion. In principle, even with zero usage, the ON RRP facility can support market rates by ensuring that counterparties demand rates on other investments at least as attractive as the rate offered on the Federal Reserve’s ON RRPs.

But let’s say the Fed wants to control interest rates without money-market funds’ involvement.

One option, Logan says, is keeping a $145bn “buffer” of bank reserves that could help the Fed control rates in the case of an unexpected shock to demand for physical currency or to the Treasury’s deposits. And if it wants to keep using RRP, it might want another $215bn buffer on top of it. Both would be backed by securities holdings.

Add those figures to the $2.1tn proposed above, and the balance sheet ends up around $2.46tn. Hey! That is almost exactly the amount of Treasury securities currently held by the Fed. How… convenient.

Anyway, there could still be a chance that the Fed will want to send the New York Fed’s markets desk back to the hustle — trying to tweak reserve levels just so, in order to reach a certain short-term rate.

However, if the FOMC wanted to maintain a somewhat smaller balance sheet…the Desk would have to use regular reserve-adding open market operations to offset transitory shocks to reserve-absorbing factors in addition to secular trends. The smaller the buffer, the more the Desk’s operational schedule and supporting analysis would need to focus on short-run changes in demand for the Federal Reserve’s liabilities.

But that sounds a bit complicated, right?

In any event, this view matches up pretty well with the New York Fed’s projections for the balance sheet found in its annual report on market operations. The NY Fed expects us to reach “normal” balance sheet size at the end of 2021:

At that time, the domestic securities portfolio is estimated to be about $2.8 trillion, with a slightly higher concentration in Treasury securities than in agency MBS. Thereafter, Treasury-driven growth of securities holdings supports trend balance sheet growth, and agency debt and agency MBS holdings continue to run off.

Not that small at all!

Related links:
A conversation about the Fed’s balance sheet — FT Alphaville

Copyright The Financial Times Limited 2024. All rights reserved.
Reuse this content (opens in new window) CommentsJump to comments section

Follow the topics in this article

Comments