By Jeff Huther
ABA DataBank
The payment of interest on reserves by the Federal Reserve is a key component of the Fed’s monetary policy implementation framework. Proposals to discontinue interest on reserves would penalize the U.S. banking system by imposing an implicit tax on reserves. This implicit tax in turn would negatively impact U.S. banks’ ability to supply credit to their customers — Main Street households and businesses — without any short-term budgetary gains. These effects would be disproportionally felt by small banks since large banks would use the Fed’s overnight reverse repo program to retain interest payments.
Reserves are loans from commercial banks to the Fed. Reserves are created when the Fed buys Treasury securities or mortgage-backed securities — banks’ liabilities rise because the Fed credits both the asset sellers’ deposit accounts and the banks’ reserve accounts. That is, reserves are funded by deposits from banks’ customers that, instead of being lent to businesses and consumers, are placed in accounts at the Fed.
The relationship between reserves and the Fed’s quantitative monetary policies
The Fed finances quantitative easing by creating reserves. When the Fed buys assets (such as Treasury securities), it credits the seller’s bank with reserves (the bank’s asset) who then credits the seller’s bank account with a deposit (the bank’s liability). The recipient of the deposits (seller) can then take those funds and make an investment elsewhere, thereby moving the reserves to another bank.
So long as the Fed has the Treasury security on its books, it will have an offsetting liability. So, while the sellers of securities to the Fed can take their money and invest it anywhere, the movement of that money will simply move the corresponding reserves from one bank to another.
Prior to 2013, banks in aggregate had very few options for reducing the amount of reserves in the system. Recognizing that Fed asset purchases risked creating more reserves than banks wanted, the Fed introduced a programmatic way to convert reserves into another Fed liability – an overnight reverse repo. So, while the Fed still controls the aggregate amount of its liabilities, the composition of its largest liabilities is now determined by the private sector.
Before 2008, the Fed was not authorized to pay interest on reserves but the cost to banks was relatively small since reserves were low (generally under $20 billion). At the end of February 2025, they were $3.4 trillion (see figure 1). The $3.4 trillion in reserves are the result of the Fed’s various quantitative easings since early 2009. The Fed’s unwinding of the most recent QE — quantitative tightening — began in 2022 but, thus far, has only affected its non-reserve liabilities.
The Fed long sought the ability to pay interest on bank reserves before legislation was passed in 2006 with a 2011 effective date. The 2008-09 financial crisis prompted Congress to move up the effective date to 2008. Without interest on reserves, the Fed’s various QEs would not have been a feasible policy choice, since large scale unremunerated assets would undermine bank health. Interest payments on reserves vary with the Fed’s target interest rate so interest payments have been low in most years since 2008. The Fed’s 2022-23 interest rate hikes, however, led to a large increase in overnight borrowing costs. In 2024, for example, the Federal Reserve paid banks $176 billion in interest (see figure 2). Without this interest income, bank earnings would have been significantly reduced, and in some cases completely wiped out.
A change in how the Fed conducts monetary policy led it to reduce reserve requirements to zero in 2020. The elimination of reserve requirements was cited by the Fed as a way to support bank lending to households and businesses. While the Fed no longer has reserve requirements, banks still value reserves both because they are high quality assets paying a market rate of interest and because they are an important part of the payments system in the U.S. Interbank payments are made by debiting or crediting each bank’s reserve account.
These interbank transactions are a critical part of the U.S. economy — in January 2025, an average of $4.7 trillion was transferred each day. Since most banks end up close to “flat” at the end of each day, the reserves needed to facilitate these transactions is small relative to the current supply of reserves. How small is unknown, but the Fed has said that it will continue to reduce reserves through quantitative tightening as long as the supply remains “ample.”
Interest on reserves is compensation to banks for lending to the Fed. Removing interest would be the same as a 100 percent tax on banks’ interest income from holding reserves. The implicit tax is equal to what banks could have earned through loans or other income-generating assets instead of holding reserves earning no return. Increased lending can reduce a single bank’s reserves, but doing so simply shifts the reserves to another bank — it is the Fed that can control the stock of reserves.
Would removing interest on reserves help the federal budget?
Prior to 2008, the Fed remitted less than $35 billion annually to the Treasury. It received interest on its assets (mainly Treasury securities) and paid nothing on its liabilities (mainly currency). The Fed’s QEs led to much larger remittances to the Treasury (averaging $83 billion a year from 2010 to 2022, as shown in figure 3). The increase in interest rates in 2022, however, reduced the Fed’s net interest margin, eventually leading to negative remittances.
The Fed treats negative remittances as debt owed to the Treasury and calls this debt a “deferred asset” — a debt that is not in calculations of government accounts or the debt ceiling. That debt was $216 billion at the end of 2024 (see figure 4). When the Fed’s NIM turns positive, it will “pay off” its accumulated debt before restarting remittances. In other words, a positive NIM for the Fed does not immediately create remittances, so there would be no immediate budgetary gain from ceasing to pay interest on reserves.
The number of years needed for the Fed to repay its deferred asset is highly uncertain, since it depends on the pace of the Fed’s balance sheet reduction, the actions the banks would take to shift deposits into overnight reverse repurchase agreements and the evolution of the economy. As the Fed’s balance sheet shrinks, its earnings decline, so reductions in its deferred asset will be slower than its accumulation.
How removal of interest would disrupt banking operations
Not only would the cessation of interest on reserves fail to shore up the federal budget, it would also disrupt the banking system and hence the economy. Banks would attempt to maintain earnings by shifting to the Fed’s ONRRP facility. Participation there requires at least $30 billion in assets, so small banks would be forced to reduce their reserve holdings by pushing deposits to the larger banks that do have access. The largest banks would need to set up additional money market mutual funds to fully use the ONRRP, since participation is capped at $160 billion per counterparty.
A larger-scale increase in ONRRP would slow the reduction in the Fed’s deferred asset. A completely successful shift of reserves into ONRRP would mean that the cessation of paying interest on reserves would have almost no budgetary impact (that is, the effect would be the difference between the interest rate on reserves and the ONRRP interest rate, currently 15 basis points).
Banks of all sizes would likely hold Treasury bills instead of reserves, and bank equity prices would fall to reflect lower earnings. This in turn would increase the cost of capital for banks of all sizes, harming their ability to raise additional capital and consequently force them to reduce the credit they provide to U.S. businesses and households. This will be in addition to the credit contraction from reduced earnings imposed by the implicit tax. For the Fed, quantitative tightening efforts would likely be accelerated to reduce the tax on banks, and the fed funds market would contract sharply as borrowers seek to avoid holding reserves. Removing interest on excess reserves would disrupt the Fed’s careful unwinding of its balance sheet; at the end of February 2025, banks held $18 trillion in assets, of which almost 19 percent were loans to the Fed. These holdings, and thus the Fed’s management of monetary policy, would be seriously undermined by the elimination of interest on reserves.
Why would we needlessly risk financial market stability and economic growth?
For additional research and analysis from the ABA’s Office of the Chief Economist, please see the OCE website.
