Thank you for the opportunity to speak to you today.1
There has been a lot of discussion of late about reducing the size of the balance sheet of the Federal Reserve to reduce our “footprint” in the financial system. I think it is important to frame the discussion, first by being clear about the nature of the problem to be solved and then weighing the tradeoffs of any remedies. I think shrinking the balance sheet is the wrong objective, and many of the proposals to meet this objective would undermine bank resilience, impede money market functioning, and, ultimately, threaten financial stability. Some would actually increase the Fed’s footprint in financial markets.
That’s because the Fed’s footprint in the financial system consists not only of the duration, composition, and size of our balance sheet (which are distinct issues), but also our roles in promoting the safety and soundness of banks, running the backbone of the payment system, and supporting financial stability. It doesn’t make sense to talk about “the Fed’s footprint” without taking into account these key functions and the way they interact. Some of the prominent proposals to reduce the Fed’s balance sheet would have perverse effects that would actually increase the Fed’s footprint in the financial system. For example, some proposals would increase the frequency of Fed lending and transactions in markets, both to implement monetary policy on an ongoing basis and, in extremis, to engage in interventions to preserve financial stability.
Today I am going talk about what I see as efficient and effective central banking—central banking that holistically implements monetary policy, provides meaningful oversight of financial institutions, supports payment system functioning, and, in the Fed’s case, also includes serving as the fiscal agent to the Treasury Department. We can’t ignore any of these responsibilities, so when considering changes to our policies and practices, we need to consider the spillovers to our other duties and evaluate the tradeoffs. This integrated approach to central banking is essential to achieving the Fed’s multiple objectives.
I will make two broad points today. First, the size of the Fed’s balance sheet is the wrong measure of the Fed’s footprint in financial markets, and, second, many of the proposals being floated to address this purported problem would make our monetary policy operations less efficient and effective and raise financial stability risks.
The Fed’s Balance Sheet
Let’s start with an understanding of the Fed’s balance sheet. The Fed’s liabilities largely consist of reserve balances (the deposits that commercial banks hold with the Fed), currency in circulation, and the balance of the Treasury General Account (TGA), which is the Treasury’s account with the Federal Reserve.2 The Fed’s assets are mostly Treasury securities, and we also have a sizable portfolio of agency mortgage-backed securities (MBS), most recently acquired in response to the COVID-19 crisis to address the extraordinary conditions at that time.3
The Federal Open Market Committee (FOMC) conducts monetary policy with an ample-reserves regime. This means that control of our policy rate, the federal funds rate, is accomplished primarily by setting our administered rates, particularly the interest on reserve balances (IORB) rate. You can think of this regime graphically, as the Fed supplying reserves on the flat portion of the reserve demand curve, so that normal changes in the supply and demand for reserves lead to only limited variations in price.4 In an ample-reserves regime, when we are getting it right, there should be limited money market volatility under normal conditions.
The Role of Reserves in the Banking System
Currently, reserves, which as I said are deposits of depository institutions at the Fed, represent $3 trillion of the Fed’s $6.5 trillion in liabilities. Some recent proposals have focused on shrinking reserve demand as a means of facilitating a shrinking of our balance sheet. Let me take a step back and explain why supplying reserves is so important.
As Chair Jerome Powell explained in a speech last year: “Reserves are the safest and most liquid asset in the financial system, and only the Fed can create them. The adequate provision of reserves is essential to the safety and soundness of our banking system, the resilience and efficiency of our payments system, and ultimately the stability of our economy.”5
Our provision of reserves has broader and important benefits beyond monetary policy implementation. If banks don’t have enough reserves, the payment system suffers, because it gives them an incentive to economize on their liquidity by slowing down their outgoing payments, leading to bottlenecks and stresses in funding markets. And, as we know, during stress if banks do not have enough reserves when depositors ask for withdrawals, panic can ensue. These aren’t theoretical problems, which is why we have regulations that support sound liquidity risk management, and reserves play an important role in banks’ portfolios of high-quality liquid assets (HQLA). The indispensable benefits of the Fed’s reserve provision need to be at the forefront of any discussion about our balance sheet.
Furthermore, creating reserves is costless to the Fed. The Fed pays interest on reserves but also receives interest on the other side of the balance sheet, which is largely in the form of Treasury securities.6 All the Fed’s excess earnings go back to the taxpayer.7
Implementing Monetary Policy Efficiently and Effectively
Let me now turn to efficient and effective monetary policy implementation. In keeping with our commitment to effective implementation, after substantially shrinking the balance sheet for a couple of years, the Fed is now slowly growing our balance sheet to keep up with demand for our liabilities. Over the many decades since our creation, we typically have grown our balance sheet through gradual purchases of Treasury securities, to ensure we continue to meet demand for our liabilities as the economy grows, consistent with our monetary policy implementation regime. Currently, the additions to the Fed’s balance sheet are incremental, and we are buying Treasury bills. Others have explained our reserve management purchases, so I will not repeat that here.8
Instead, I will get straight to a discussion on how I see monetary policy implementation today as efficient and effective. Effective monetary policy implementation means that the federal funds rate, as well as other short-term interest rates such as repo rates, should trade near the rate the Fed pays on reserve balances. The current regime has achieved that for many years. Effective policy implementation also supports smooth market functioning.
On the question of efficiency, Dallas Fed President Lorie Logan has discussed how monetary policy arrangements can minimize an economic cost associated with holding money, leaning on a principle put forth by economist Milton Friedman.9 In judging whether monetary policy is efficient, this approach uses the criterion of whether the opportunity cost to banks of holding reserves is close to the central bank’s cost of supplying reserves. This is, in fact, the case right now. When this convergence of rates occurs, banks do not have a return-based incentive to economize on reserves.10 And because banks receive an interest rate on reserves that reflects safe asset status, banks are also not gaining excessive returns on holding reserves, so they have no incentive to load up on reserve holdings at the expense of lending to businesses and households. The fact that current monetary policy implementation is abiding by the Friedman principle is a sign that it is efficient. Under normal circumstances, this is a good place to be, but, of course, there are times when money market rates may trade more substantially below the IORB rate. Monetary policy implementation can remain effective in such conditions because we have a floor tool: overnight reverse repo operations, supporting rate control at the bottom of the target range.
The footprint of reserve scarcity
Some people believe that we should return to a scarce reserves regime, rather than an ample one. That policy would mean a smaller balance sheet, but it would not reduce the Fed’s footprint in the market, given the degree of regulation and intervention that would be needed to operate this regime.11 Moreover, as I noted, there is no net cost to providing reserves, and making a free good scarce makes little economic sense.12 I will return to this topic in a bit, when I discuss attempting to implement monetary policy with minimally ample reserves.
The footprint of balance sheet duration
Some people argue that the Fed’s footprint is too big in the sense that the duration of our portfolio is longer than that of Treasury securities held by the public and, thus, that we are putting downward pressure on long-term rates.13 As I mentioned earlier, much of our current longer-term asset portfolio was accumulated in dire circumstances when short-term rates were near zero and more accommodation was needed.
Concerns about the duration or the composition of our asset portfolio need to be differentiated from concerns about the size of the balance sheet. In my judgment, the Treasury should determine the maturity composition of debt held by the public, and in normal times, our balance sheet should mirror the maturity distribution of outstanding issuance to the extent possible, thus having a neutral effect on the composition of securities available to the public over time. We are moving in this direction now.14
A different point about the duration of the Fed’s portfolio is that it exposes us to interest rate risk, given that the value of our assets fluctuates, as do the rates we pay on some of our liabilities. This is not about the size of the balance sheet, but about its duration. Again, the whole point of what people refer to as “quantitative easing” was for the Fed to take on duration risk to reduce interest rates at the effective lower bound. When economic circumstances demand it, the Fed is well positioned to take on interest rate risk beyond what is implied by mirroring Treasury issuance, because the Fed’s balance sheet can temporarily absorb losses without affecting its resilience or the ability to conduct monetary policy. But concerns about duration or composition are not about the size of our balance sheet.
Proposals to Reduce Reserve Demand
I’ve already discussed the duration and composition of the Fed’s balance sheet and why returning to a scarce reserves approach doesn’t make sense. There are also arguments that we should reduce the size of the balance sheet within the ample-reserves framework.
Many commentators understand that to reduce the Fed’s balance sheet in a significant way, one would have to reduce reserve demand. Reserve demand is driven by banks’ payment needs, their own assessment of liquidity needs, and liquidity regulations, including the Liquidity Coverage Ratio (LCR), internal liquidity stress tests, and resolution requirements. Some commentators have argued that we should reduce liquidity requirements to reduce demand for reserves. As I have explained, that’s the wrong goal. But if we’re pursuing it, this is also the wrong means.
Post-GFC Liquidity Regulation, and Why Liquidity Self-Insurance Matters
After the Global Financial Crisis (GFC), bank regulators designed and adopted rules that require the largest banks to hold a portfolio of HQLA in amounts calibrated to what their needs would be in stress scenarios. This approach, along with capital requirements, was designed to help the largest banks survive stress events. Rather than allowing these firms to threaten the viability of the financial system, necessitating their bailouts, we instead said they have to basically self-insure by holding an appropriate amount of HQLA. Conditions change, and it is reasonable to consider if these regulations need to be recalibrated. But if anything, the bank stresses of 2023 suggest that liquidity requirements should go up and not down, as I have discussed in a number of previous speeches.15 Moreover, after the events, we saw banks increase their collateral pledged at the discount window by over $1 trillion. This improved level of operational readiness will be important to retain as memories of this episode fade, which history shows tends to happen.16
Lowering Liquidity Coverage Ratio requirements
The LCR treats reserves and Treasury securities the same, and it also requires banks to demonstrate that they can access private markets, which helps to ensure banks have the operational experience needed to raise liquidity when they need it. Fed liquidity provision is also factored into these rules. For example, right now banks can use term primary credit to improve their LCR, if needed, but this requires actual borrowing rather than just borrowing potential. Also, HQLA securities pre-positioned at the discount window fully count toward meeting liquidity requirements.17 The upshot is that Fed liquidity provision is already well integrated with liquidity requirements.
To date, regulators have drawn the line at allowing banks to decrease the amount of HQLA they hold by using our facilities, but some proposals are seeking to cross this line. One prominent proposal is to let banks count some percentage of non-HQLA assets they pledge to the discount window (such as corporate loans) to meet their liquidity requirements. This is really just a way to lower liquidity requirements, and it only reduces the amount of self-insurance and resilience the Fed requires of the largest banks. Moreover, I am also skeptical that adjusting the LCR will have a meaningful effect on reserve demand.18
Reducing other liquidity requirements
Other proposals to adjust regulations could have more of an effect on reserve demand. Some proposals would adjust other liquidity regulations, such as the regulation that requires big banks to perform internal liquidity stress tests to ensure that they can meet stressed outflows over a variety of time horizons, and the expectation that banks have liquid assets that are sufficient to support the process of orderly resolution.19 But here, again, we have to look at the tradeoffs when reducing self-insurance. The largest banks need to take responsibility for their systemic role in the markets, and that means they need to take out some amount of self-insurance to manage shocks to their balance sheets. The alternative is that these banks can’t handle shocks, and, in the past, these dynamics have required exceptional interventions to preserve financial stability. This is an outcome no one considers desirable and one that obviously would not constitute a reduction in the Fed’s footprint in the financial system.
Proposals about the Treasury General Account
Another prominent proposal from those who want to reduce the Fed’s balance sheet is for the Treasury to reduce the size of the TGA, a liability on the Fed’s balance sheet. You can think of the TGA as the Treasury’s checking account, and it is intended to hold a cash balance sufficient to pay its bills, including maturing principal and interest on securities, even if there is a temporary disruption to its ability to auction securities. The size of the TGA is up to the Treasury, not the Federal Reserve. If the Treasury were to reduce the size of the TGA, it would not in any meaningful sense reduce the official sector footprint in the market. And my view is that we all benefit from a system in which the Treasury can meet its obligations under a variety of challenging circumstances.
Another idea is for the Fed to “sterilize” fluctuations in the Treasury’s cash balance so that they would not affect the supply of reserves. That would mean buying assets when the TGA goes up and selling assets when the TGA goes down.20 In this way, the average size of the Fed balance sheet over time would be somewhat smaller compared to the current approach. But implementing such a program would require frequent and at times very large transactions to actively manage our balance sheet, particularly around debt ceiling episodes. That’s another form of a larger footprint. Our current approach of gradually growing our balance sheet in response to trend growth in demand for our liabilities, so that our transactions are as boring as possible, is much preferable. A similar effect on reserves could be achieved if the Treasury were to invest some portion of its funds in repo transactions or bank deposits, so that there was less volatility in TGA. Of course, this again would be a decision of the Treasury. The Treasury Borrowing Advisory Committee recently provided analysis of options along these lines. The volume and variability of market transactions that would be required were highlighted.21
Operating with “Minimally Ample” Reserves
Another set of proposals would call for the Fed to operate at somewhat lower levels of ample reserves—that is, without a buffer to ensure that shocks to reserve supply and demand do not bring reserves below ample. The risk of operating at lower levels of reserves (for a given reserve demand) is that we would see significant volatility and increased risk of losing control over rates. That’s what we saw in 2019 when repo market rates spiked. The Fed avoided this outcome in late 2025 by acting in advance to incrementally grow our balance sheet.
Of course, the Federal Reserve’s tools to put a ceiling on the policy rate could help. If banks overcame their reluctance to use standing repo operations and the discount window, rate control could be maintained with a lower level of reserves. But it would mean that the Fed would be lending into the market on a regular basis, again, a form of a bigger footprint. I’m not opposed to Fed lending, but it would not reduce the Fed’s footprint in the market.
Although it is not my preferred approach, I recognize that there are some positive aspects to a system where central bank operations have to be used more frequently. Other central banks rely on ceiling tools as a key means of reserve provision.22 A benefit of this approach is that if ceiling tools are more frequently “in the money” and firms fully understand that they are expected to tap them, these tools would not atrophy from lack of use. Functional ceiling tools put the Federal Reserve in a better place to manage through bouts of market stress.23
In sum, operating with minimally ample reserves poses significant tradeoffs. We might have a slightly smaller balance sheet, but we would need to transact more frequently. There would be higher market volatility and less predictable rate control. If rates are consistently above the IORB rate, banks have an incentive to economize on reserves, and for some, they might economize beyond what’s prudent, raising risks for them and the broader financial system.
Other Ideas: Buffer Usability, Intraday Lending, and Payment Sequencing
Some commentators have proposed more technical steps to reduce reserve demand, and the costs and benefits of these approaches are worth exploring.24
We should take steps to make sure banks understand that required liquidity is usable. My experience during the stress of 2023 was that banks in fact did use their liquidity, and there may be technical adjustments that would ease banks’ fears about using their liquidity buffers. It is also worth taking a look at the Fed’s daylight overdraft policies to see if they can be improved, encouraging usage. In addition, investments in our payment infrastructure—such as through a liquidity savings mechanism that some have suggested—might reduce reserve demand but might have other downside costs and should be weighed against other priorities for system improvement, including movement to a near continuous availability for Fedwire® Funds and expanded use of FedNow, both of which I’d prioritize.25
Conclusion
The Federal Reserve has an important role in the financial system, but its footprint in financial markets is not well defined by the size of its balance sheet. Some proposals to shrink the balance sheet increase the Fed’s footprint in other ways, and all of them would entail significant tradeoffs. Some of the most prominent ideas, such as reducing liquidity requirements, would undermine financial stability. Letting banks treat some amount of assets because these assets are pledged at the discount window—as though they are the same as Treasury securities or reserves—as some have proposed, would, as I said, just reduce self-insurance and decrease the resilience of the banking system. In any consideration of a change in the Federal Reserve’s approach to managing its balance sheet, we ought to go back to basics and ask what problem we are trying to solve. In sum, shrinking the Fed’s balance sheet is the wrong goal, and reducing the resilience of the banking system is the wrong means.
1. The views expressed here are my own and are not necessarily those of my colleagues on the Federal Reserve Board or the Federal Open Market Committee. Return to text
2. A detailed balance sheet is published every Thursday; see Board of Governors of the Federal Reserve System, Statistical Release H.4.1, “Factors Affecting Reserve Balances of Depository Institutions and Condition Statement of Federal Reserve Banks.” Return to text
3. The System Open Market Account annual report provides a helpful discussion and depictions of Fed balance sheet developments and projections; see Markets Group of the Federal Reserve Bank of New York (2026), Open Market Operations during 2025 (PDF) (New York: Federal Reserve Bank of New York). Return to text
4. See Board of Governors of the Federal Reserve System (2025), “Policy Normalization,” webpage. Return to text
5. See Jerome H. Powell (2025), “Understanding the Fed’s Balance Sheet,” speech delivered at the 67th Annual Meeting of the National Association for Business Economics, Philadelphia, Pa., October 14, paragraph 4. Return to text
6. See Jane E. Ihrig and Scott A. Wolla (2026), “The Fed’s Balance Sheet and Ample Reserves,” Federal Reserve Bank of St. Louis, Page One Economics, February 18. Return to text
7. After payment of expenses and transfers to surplus, all the net earnings of the Reserve Banks are transferred to the U.S. Treasury; see Board of Governors of the Federal Reserve System (2021), The Fed Explained: What the Central Bank Does (PDF), 11th ed. (Washington: Board of Governors, August). Historically, asset income has exceeded liability costs, and the resulting net earnings are remitted to the U.S. Treasury; see Miguel Faria-e-Castro and Samuel Jordan-Wood (2023), “The Fed’s Remittances to the Treasury: Explaining the ‘Deferred Asset,’” Federal Reserve Bank of St. Louis, On the Economy Blog, November 21. and see also Markets Group of the Federal Reserve Bank of New York, Open Market Operations (note 3). Return to text
8. How the FOMC conducts monetary policy is explained in Board of Governors of the Federal Reserve System (2025), “Interest on Reserve Balances (IORB) Frequently Asked Questions,” webpage. Several years ago, the FOMC chose to implement monetary policy with an ample-reserves regime (see Board of Governors of the Federal Reserve System (2019), “Statement Regarding Monetary Policy Implementation and Balance Sheet Normalization,” press release, January 30), which had implications for the size of our balance sheet. This means that reserves should not be too abundant or too scarce and that the interest rate we pay on reserve balances will be the primary tool to guide short-term interest rates to the level we target. Among the reasons the FOMC chose an ample-reserves regime was that it minimizes the need for frequent interventions in the market, particularly in comparison to the implementation of the “scarce” reserves regime; (see Jane Ihrig, Zeynep Senyuz, and Gretchen C. Weinbach (2020), “Implementing Monetary Policy in an ‘Ample-Reserves’ Regime: The Basics (Note 1 of 3),” FEDS Notes (Washington: Board of Governors of the Federal Reserve System, July 1)). Return to text
9. See Lorie K. Logan (2023), “Ample Reserves and the Friedman Rule,” speech delivered at the European Central Bank Conference on Money Markets 2023, Frankfurt, Germany, November 10. Return to text
10. Trying to minimize reserve holdings is an undesirable practice because the holding of reserves is, as I noted, essential to healthy liquidity management, supports payment system function, and protects financial stability. Return to text
11. The monetary policy implementation regime before 2008 involved complex regulations and data collections to impose and assess those requirements and resulted in regulatory avoidance by banks. Implementing a scarce reserves regime again would likely necessitate creating a stable demand for reserves through similarly complex regulations. Return to text
12. See Christopher J. Waller (2025), “Demystifying the Federal Reserve’s Balance Sheet,” speech delivered at the Federal Reserve Bank of Dallas, Dallas, Texas, July 10. Return to text
13. A similar point is made about our MBS holdings, with some pointing out that our large portfolio puts downward pressure on mortgage rates, a form of credit allocation that some argue goes beyond monetary policy. Importantly, we are already reducing our MBS holdings carefully and gradually to avoid causing market disruptions. Return to text
14. For information on how the Fed’s portfolio has been normalizing, see Markets Group of the Federal Reserve Bank of New York, Open Market Operations (note 3). Return to text
15. See Michael S. Barr (2023), “The 2023 U.S. Treasury Market Conference,” speech delivered at the Federal Reserve Bank of New York, New York, New York, November 16; Michael S. Barr (2024), “Supporting Market Resilience and Financial Stability,” speech delivered at the 2024 U.S. Treasury Market Conference, Federal Reserve Bank of New York, New York, New York, September 26; and Michael S. Barr (2023), “Monetary Policy and Financial Stability,” speech delivered at the Forecasters Club of New York, New York, New York, October 2. Return to text
16. See Michael S. Barr (2025), “Booms and Busts and the Regulatory Cycle,” speech delivered at the Brookings Institution, Washington, July 16. Return to text
17. For information on how banks can use the discount window to improve the LCR, see Board of Governors of the Federal Reserve System (2024), “Discount Window,” webpage. Return to text
18. In the LCR, both reserves and Treasury securities are treated as the highest form of HQLA. Allowing banks to count lower-quality assets in this portfolio if these assets are pledged to the discount window would likely lead banks to substitute similar-duration assets for these securities, to the extent that banks have already optimized the duration of their balance sheets. This would mean that this proposal would lead to reduced demand for Treasury securities rather than reduced demand for reserves. Return to text
19. Moreover, banks can already use access to Fed liquidity provision as part of the mix for monetization sources in these tests. See Board of Governors of the Federal Reserve System (2024), “Subparts D and O—Enhanced Prudential Standards” in “Frequently Asked Questions about Regulation YY,” webpage. Return to text
20. See Annette Vissing-Jorgensen (2025), “Fluctuations in the Treasury General Account and Their Effect on the Fed’s Balance Sheet,” FEDS Notes (Washington: Board of Governors of the Federal Reserve System, August 6). Return to text
21. See Treasury Borrowing Advisory Committee (2026), “Treasury Investment in Repo (PDF),” May 5. Return to text
22. See Lorie K. Logan (2025), “Opening Remarks for Panel Titled ‘Post-Pandemic Challenges for Monetary Policy Implementation,’” speech delivered at the Conference on the Future of Central Banking, Bank of Mexico, Mexico City, August 25. Return to text
23. See, for example, Barr, “Supporting Market Resilience” (note 15). Return to text
24. See Darrell Duffie (2026), “The Payment System Puts a Floor on the Fed’s Balance Sheet (PDF),” paper presented at the Brookings Papers on Economic Activity Conference, held at the Brookings Institution, Washington, March 26–27. Return to text
25. For more information about liquidity savings mechanisms, see Darrell Duffie (2026), “Liquidity Savings Mechanisms,” in “The Payment System Puts a Floor on the Fed’s Balance Sheet (PDF),” paper presented at the Brookings Papers on Economic Activity Conference, held at the Brookings Institution, Washington, March 26–27, pp. 40–46. For details about the Fedwire® Funds Service, see Board of Governors of the Federal Reserve System (2025), “Federal Reserve Action to Expand Fedwirer® Funds Service and National Settlement Service Operating Hours (PDF),” Federal Register, vol. vol. 90 (November 17), pp. 51356–68. Return to text
