Normalizing the FOMC’s monetary policy tools
Thank you, Ken [Bentsen]. It’s great to be back at SIFMA.
The last time I spoke at a SIFMA event, in April 2021, I was serving as manager of the Federal Reserve’s System Open Market Account. I discussed how the expansion of our balance sheet to support the economy amid the pandemic was influencing money market conditions and policy implementation.
Today, I will return to the topics of money markets and the Fed’s balance sheet, but from a new perspective as a policymaker and in a very different economic environment.
The economy
Let me start by reviewing the economic situation. As always, these views are mine and not necessarily those of my colleagues on the Federal Open Market Committee (FOMC).
The FOMC held the fed funds target range at the effective lower bound from March 2020 through March 2022 to support the economy during the pandemic, then raised the target range rapidly in 2022 and 2023 in response to surging inflation. Last month, with inflation and employment both in striking distance of the FOMC’s goals rather than seriously overheated, we decided to lower the target range.
Less-restrictive policy will help avoid cooling the labor market by more than is necessary to bring inflation back to target in a sustainable and timely way. In thinking about the policy choices ahead, I continue to apply the same foundational considerations I’ve applied since I became president of the Dallas Fed two years ago. First and foremost, I’m committed to sustainably achieving both of the FOMC’s dual-mandate goals: maximum employment and stable prices. I’m also carefully considering how the risk environment affects our strategy. The appropriate policy strategy depends on whether the path ahead is clear or cloudy. And I continue to pay close attention to financial conditions and to the information I receive from the Dallas Fed’s surveys, from business and community contacts, and from market participants like you.
Two takeaways stand out to me from the current economic and financial picture. First, the economy is strong and stable. But second, meaningful uncertainties remain in the outlook. Downside risks to the labor market have increased, balanced against diminished but still real upside risks to inflation. And many of these risks are complex to assess and measure.
If the economy evolves as I currently expect, a strategy of gradually lowering the policy rate toward a more normal or neutral level can help manage the risks and achieve our goals. However, any number of shocks could influence what that path to normal will look like, how fast policy should move and where rates should settle. In my view, the FOMC will need to remain nimble and willing to adjust if appropriate.
Balance sheet normalization
As we begin to normalize the fed funds target range, we are also continuing the process we began in 2022 of normalizing our balance sheet by allowing longer-term asset holdings to mature.
Changes in the fed funds target range are, as the FOMC has stated, our primary means of adjusting the stance of monetary policy. With the target range well away from the effective lower bound, at this time we can accomplish our inflation and employment goals by adjusting the target range, while allowing balance sheet runoff to continue smoothly in the background as we move toward an efficient balance sheet size.
It's sometimes suggested that the reduction in asset holdings works at cross purposes to the strategy of reducing policy restriction by lowering the fed funds target range. I don’t see it that way, for two reasons. First, we are moving the target range toward neutral from above. The policy rate remains restrictive, which is consistent with also creating restriction by reducing asset holdings. Second, we are now normalizing both of our policy tools. Normalizing the fed funds rate means bringing it down from the elevated levels that were needed to restore price stability and returning to a level that will be consistent with sustaining maximum employment and price stability over time. Normalizing our balance sheet means bringing our asset holdings down from the elevated quantity that was necessary to support the economy during the pandemic and returning to a balance sheet size that will be consistent with implementing monetary policy efficiently and effectively. Those two normalization processes work in tandem and are consistent in my view. A number of other central banks are similarly reducing their asset holdings while lowering their policy rates in response to the changing economic outlook.
The normalization process will affect both sides of the Federal Reserve’s balance sheet. In the rest of my remarks, I will describe key policy considerations that I expect to affect the path of the Federal Reserve’s liabilities and assets in the long term, as well as some relatively near-term considerations that may arise along the way.
Liabilities
Let me start with liabilities. In 2019, the FOMC decided to implement monetary policy in the long run in a regime of ample reserves. In this regime, the Fed supplies sufficient liquidity to meet banks’ demand with money market rates close to the interest rate on reserve balances (IORB). The ample reserves regime has three main benefits. It provides robust interest rate control in an environment where both banks’ reserve demand and the autonomous factors that drain reserves are less predictable than in the past. The ample reserves regime also continues to effectively control interest rates when the central bank expands its balance sheet to provide macroeconomic stimulus or support financial stability. And, by remunerating reserves in line with other money-like assets, the ample reserves regime ensures banks don’t pay a penalty for appropriately managing their liquidity risk by holding reserves, the most liquid asset there is.
At present, liquidity appears to be more than ample. Reserve balances are around $3.2 trillion, compared with around $1.7 trillion in early 2020. The economy and financial system have grown, and the dash for cash at the start of the pandemic as well as the banking stresses in March 2023 may have led banks to increase their demand for liquidity. Still, I think it’s unlikely banks’ liquidity demand has nearly doubled in half a decade.
One sign liquidity remains in abundant supply, and not merely ample, is that money market rates continue to generally run well below IORB. The tri-party general collateral rate (TGCR) on repos secured by Treasury securities has been averaging 8 basis points below IORB. Because reserves and Treasury repos are both essentially risk-free overnight assets – and reserves are, if anything, more liquid – the spread of IORB over TGCR indicates reserves remain in relatively excess supply compared with other liquid assets.
The spread between IORB and the secured overnight financing rate (SOFR), which includes a broader set of Treasury repo transactions than TGCR, has recently been slightly tighter than the IORB-TGCR spread. However, some SOFR transactions include compensation for intermediating funds from the triparty segment of the repo market to cash borrowers who lack direct access to that segment. For this reason, I am watching TGCR for a cleaner read on liquidity conditions in secured markets.
Unsecured funding conditions also continue to reflect abundant liquidity. The effective federal funds rate has been running 7 basis points below IORB and remains insensitive to short-term fluctuations in reserve levels.
And the continuing substantial balances in the Fed’s overnight reverse repo (ON RRP) facility provide another sign that liquidity remains more than ample. The ON RRP facility accepts overnight investments from money market funds and certain other market participants at a rate currently 10 basis points below IORB. Thus, the facility reinforces the floor on money market rates created by IORB. Institutional and market frictions can influence investors’ choices to place funds in the ON RRP facility versus other instruments. However, should liquidity shortfalls emerge that create meaningful upward pressure on money market rates, I would expect market participants to move cash out of the facility in search of higher returns. For now, the remaining ON RRP balances provide a buffer of additional excess liquidity.
We have started to see some temporary money market fluctuations on statement dates such as the recent quarter end. As the supply of liquidity declines, the financial system needs to redistribute liquidity to the firms that need it most. That process isn’t always frictionless. The emergence of month-end and quarter-end pressures appears to reflect temporary intermediation frictions rather than shortfalls in aggregate reserve supply. For example, on September 30 and October 1, the spread between SOFR and TGCR widened by 7 to 12 basis points. This widening reportedly resulted from limited balance sheet availability at dealers that intermediate between the triparty and centrally cleared market segments.
As in the previous episode of balance sheet normalization in 2018 and early 2019, we are likely to continue to see occasional, modest rate pressures as our balance sheet shrinks. Such temporary rate pressures can be price signals that help market participants redistribute liquidity to the places where it’s needed most. And from a policy perspective, I think it’s important to tolerate normal, modest, temporary pressures of this type so we can get to an efficient balance sheet size.
In the long run, I believe it will be appropriate for the Federal Reserve to operate with money market rates close to, but perhaps slightly below, IORB. As I have discussed elsewhere, operating with money market rates close to IORB is consistent with the Friedman rule. In an environment where the central bank pays interest on reserves, the Friedman rule calls for making the cost of liquidity similar across markets and institutions by having roughly equal interest rates on reserves and comparable instruments such as repos. I say rates should be roughly equal because technical differences and minor competitive frictions between money markets can justify small spreads. The last time we normalized our balance sheet, in 2016 through 2019, money market rates became notably more volatile as they moved above the interest rate on reserves. A configuration of rates close to, but perhaps slightly below, IORB may provide the best combination of efficiency and effective rate control.[1]
I also believe it will be appropriate in the long run to operate with only negligible balances in the ON RRP facility. The facility is a backstop tool to reinforce the floor on money market rates created by paying interest on reserves. The lower interest rate on ON RRP balances than reserves reflects the facility’s backstop nature. As such, meaningful ON RRP balances should not form a permanent fixture on the Fed’s balance sheet, in my view.
ON RRP balances came down rapidly in 2022 and 2023, but the pace of decline has slowed more recently. It is not particularly surprising that the most elastic balances would leave the facility first. The remaining balances may be stickier as some participants particularly value overnight assets or face counterparty risk limits for repos in the private market. I anticipate the remaining balances will move out of the facility as repo rates rise closer to IORB, but if they do not, reducing the ON RRP interest rate could incentivize participants to return funds to private markets.
Assets
Turning to the other side of the ledger, balance sheet normalization also has implications for our policy choices about the composition of the Federal Reserve’s assets and the design of our lending facilities.
The FOMC has said repeatedly, most recently in 2022, that we intend to hold primarily Treasury securities in the long run. At present, however, we are rather far from that benchmark and not moving appreciably closer. Mortgage-backed securities (MBS) are running off slowly because homeowners have little incentive to refinance with mortgage rates higher than they were when we acquired most of the securities. Agency debt and MBS currently represent 34.4 percent of the par value of our securities holdings. That is actually up half a percentage point from the start of 2024.
In its most recent annual report, the New York Fed’s Open Market Trading Desk projected that the share of MBS in the portfolio would not decline materially until balance sheet normalization is complete and the Desk begins purchasing Treasuries to keep pace with growth in reserve demand. MBS would remain more than 15 percent of the portfolio through the end of the decade in those projections. As indicated in the minutes of the May 2022 FOMC meeting, a number of FOMC participants have suggested it could be appropriate at some point to sell MBS to move the mix of assets closer to our goal. But that’s not a near-term issue in my view.
Even with a portfolio consisting primarily of Treasuries, the FOMC will face choices about which Treasuries to hold. I believe the two most plausible options in the long run are for the FOMC to hold a roughly neutral Treasury portfolio, meaning one with a maturity composition similar to that of the Treasury universe, or to tilt toward shorter maturities. A neutral portfolio would leave long-run decisions about the maturity composition of government debt to the Treasury Department, as is appropriate under our institutional separation of monetary and fiscal policy. A tilt toward shorter maturities would allow more flexibility. Either way, the System Open Market Account portfolio is significantly underweight Treasury bills, and its weighted average maturity remains significantly longer than that of marketable debt outstanding. The tradeoffs around how to move toward a more neutral portfolio are complex and will require thoughtful policy deliberations.
Lending and funding transactions
The Federal Reserve’s assets also include loans we make to depository institutions at the discount window and repo operations that deliver liquidity to our various counterparties. Balance sheet runoff will influence the policy choices on this component of our balance sheet as well.
When reserve supply is closer to ample, individual banks will be more likely to face payment outflows that leave them temporarily below their desired liquidity levels. Money markets generally work well to redistribute liquidity in those circumstances, but frictions can arise. All banks should have strong contingency liquidity plans for addressing liquidity shortfalls, including those that private markets cannot meet.
In particular, as I have said before, I believe every bank in the United States should be operationally ready to access the discount window. That means completing the legal documents, making collateral arrangements and testing the plumbing. As a longtime market practitioner who has worked extensively on readiness for potential stress, I am convinced banks should practice using the window before a need arises. Take out and repay small-dollar test loans. And practice moving collateral between the window and other collateralized funding sources, in case a scenario arises where you can’t get the funding you want from those sources. The window is an important tool for healthy banks to meet their liquidity needs, but it works only when banks are ready to use it. Banks should also consider the potential benefits of establishing access to the Standing Repo Facility (SRF).
The Federal Reserve’s discount window policies have been largely unchanged for two decades, yet the structure of money markets and the banking and financial systems has changed dramatically over that time. The FOMC has moved from a scarce-reserves policy implementation regime, in which liquidity was costly and banks routinely experienced small liquidity shocks, to an ample-reserves regime, with much less costly liquidity and less frequent shocks. Regulations, especially for large banks, changed significantly after the Global Financial Crisis, calling for banks to manage liquidity risk much more carefully. Reforms adopted by the Securities and Exchange Commission (SEC) reshaped the money market fund landscape, effectively requiring much more use of secured transactions and shrinking the market for unsecured short-term lending to banks. And in parallel with these policy changes, markets are moving faster than ever before, culminating in the unprecedented speed of bank runs that we saw in 2023.
In light of these environmental changes, it is important to consider whether our discount window operations and policies should adjust to ensure the window remains an effective tool for meeting depository institutions’ liquidity needs and mitigating liquidity shocks that could lead to broader financial and economic stress. Last month, the Federal Reserve Board issued a formal Request for Information seeking public views on how best to strengthen discount window operations. I hope all market participants will consider commenting. Your views are important and help us do the best job of serving the public.
In addition to the discount window, the SRF, the Foreign and International Monetary Authorities Repo Facility and international dollar liquidity swap lines are important tools for providing liquidity to the financial system when temporary needs arise. As with the discount window, it is important to ensure the continued effectiveness of these facilities as we move toward ample reserves. I was pleased to see the SRF drawn on over the quarter-end turn as market participants worked through frictions in the redistribution of liquidity. To further enhance the effectiveness of our tools, the FOMC could consider the benefits of centrally clearing SRF and other open market operations. Central clearing would reduce our counterparties’ cost of intermediating funding to the broader market. The SEC also adopted a central clearing mandate earlier this year for many private-sector participants in the Treasury market. While that mandate does not apply to the Fed, centrally clearing our operations on a voluntary basis could reinforce this important market transition.
The primary benefit of the clearing mandate, in my view, will be to strengthen risk management by applying the strong and uniform requirements of central counterparties, instead of leaving risk management up to individual firms. Broader central clearing in the private sector will also enhance transparency. And it may reduce balance sheet costs for some intermediaries, which could help mitigate the statement-date repo market frictions I discussed earlier. As the industry works to implement the clearing mandate, it will be important to design access models and operations to allow efficient central clearing by a wide range of market participants so that the mandate’s benefits can be fully realized. SIFMA’s work developing master clearing agreements is an important step to support the clearing transition, alongside the work under way at many other organizations. With the first clearing deadline at the end of 2025, it is crucial for all market participants to prepare.
Thank you. I would be happy to answer your questions.
Note
I am grateful to Sam Schulhofer-Wohl for assistance in preparing these remarks.
- Rates that include returns to intermediation could be slightly higher.
The views expressed are my own and do not necessarily reflect official positions of the Federal Reserve System.