U.S. Federal Reserve Bank Reserve Balances

Reserve balances are deposit liabilities of the Federal Reserve Banks owed to commercial banks and other depository institutions. They constitute the primary channel through which quantitative easing transmits to the banking system, and before the 2008 financial crisis aggregate reserves in the U.S. banking system stood at approximately USD 15 billion, barely sufficient to meet reserve requirements. Successive rounds of QE then expanded reserves to over USD 4 trillion by 2014, fundamentally transforming the Fed's operating framework from a scarce-reserves corridor system to an ample-reserves floor system.

Reserves earn the interest on reserve balances (IORB) rate, which replaced the earlier interest on excess reserves (IOER) rate in July 2021. Under the ample-reserves framework, IORB serves as the primary tool for implementing the federal funds rate target, since the abundance of reserves means banks have no need to borrow in the federal funds market at rates above IORB. The Fed estimates that the lower bound of 'ample' reserves lies near USD 3 trillion, and as long as reserve balances remain above this floor the floor system anchors the policy rate stably, whereas as balances approach it the demand curve for reserves steepens and overnight funding rates begin to exhibit upward pressure.

Sourced from the H.4.1 "Factors Affecting Reserve Balances" weekly statistical release. The reported figure is a Wednesday close-of-business observation of the total reserve balances maintained by depository institutions at the twelve Federal Reserve Banks, including both required and excess reserves. The raw data are reported in millions of USD and are converted to billions by dividing by 1,000.

Reserve balances are the operational fulcrum of the post-2008 monetary policy framework, and their aggregate level determines whether the Fed can implement its interest rate target through the floor system. Afonso, Kovner, and Schoar (2011) documented how the federal funds market was transformed by the dramatic expansion of reserves during QE1, showing that trading volume declined sharply as banks with excess reserves had diminished incentive to lend in the overnight market. This structural shift meant that the federal funds rate became pinned near the IOER rate rather than being determined by the interplay of reserve supply and demand, fundamentally altering the transmission mechanism of monetary policy.

The distribution of reserves across the banking system, not just their aggregate level, matters for financial stability and monetary policy transmission. Bianchi and Bigio (2022) developed a general equilibrium model in which heterogeneous banks manage liquidity through reserve holdings, demonstrating that the distribution of reserves across institutions affects both the pass-through of monetary policy to lending rates and the stability of the interbank market. Ennis and Wolman (2015) examined the demand for reserves in the post-crisis environment and found that large banks accumulated reserves well beyond regulatory requirements, which suggests a precautionary motive that makes the aggregate demand for reserves difficult to predict from regulatory parameters alone.

The relationship between reserves and repo-market functioning provides the most direct channel through which reserve adequacy affects broader financial conditions. Copeland, Martin, and Walker (2014) showed that the structure of the tri-party repo market concentrates reserve demand at clearing banks, implying that aggregate reserves can appear ample while specific institutions face localized scarcity. This distributional fragility was dramatically revealed in September 2019. The combination of quarterly corporate tax payments and Treasury settlement drained the reserves of key intermediaries below comfortable levels and sent repo rates sharply higher, which Afonso, Cipriani, and La Spada (2022) documented as evidence that the reserve demand curve features a steep, non-linear segment near the adequacy threshold.

For money-market participants and bank treasurers, the level of aggregate reserves is a critical input for assessing the risk of overnight funding-rate volatility. Pozsar (2014) mapped the complex plumbing of the shadow banking system and showed that reserves serve as the ultimate settlement asset in the hierarchy of money, with repo rates, commercial paper rates, and other money-market rates all anchored to the reserve supply through a chain of intermediation. When reserves decline toward the ample-reserves threshold, this anchoring weakens and money-market rates become more volatile and more sensitive to idiosyncratic demand shocks.

The pace of reserve drainage during quantitative tightening provides a forward-looking signal for the duration and intensity of balance-sheet normalization. Duffie and Krishnamurthy (2016) analyzed the pass-through of reserve changes to money-market rates and showed that the relationship is highly nonlinear, where large reserves have minimal marginal impact but reserve declines below a threshold produce disproportionate rate increases. Logan (2023), as President of the Federal Reserve Bank of Dallas, articulated the operational principles of the ample-reserves framework and emphasized that the Fed would need to slow and eventually stop balance-sheet runoff well before reserves approached scarcity, in order to avoid the abrupt market dislocations observed in September 2019.

The effective lower bound of reserve adequacy may lie substantially above the regulatory minimum, which makes the reserve balance series an essential monitoring tool for anticipating the terminal point of quantitative tightening. Acharya et al. (2024) documented a more subtle risk, namely that banks have become structurally dependent on ample reserves for intraday liquidity management, and that reducing reserves below banks' revealed-preference comfort levels could trigger precautionary hoarding behavior that amplifies funding stress. Smith and Sellon (2003) provided the pre-crisis baseline by documenting how reserve requirements shaped the demand for reserves under the scarce-reserves regime, offering a historical benchmark against which the post-2008 transformation can be measured.

Over 1225 observations from 2002-12-18 to 2026-06-03, the Federal Reserve reserve balances is integrated of order one on the log-level, with the Dickey and Fuller (1979) test in the Said and Dickey (1984) form not rejecting at p = 0.8545, the Phillips and Perron (1988) test concurring at p = 0.6497, and the Kwiatkowski et al. (1992) test rejecting stationarity. On the first difference of the log the Ljung and Box (1978) portmanteau rejects white noise at lags 13 and 26, Q = 1968.40 and Q = 3606.38 at p = 0.000 and p = 0.000. The Bai and Perron (1998, 2003) procedure finds no break in the mean, a reading consistent with the parameter-instability inference of Andrews (1993) on the differenced object (Perron 1989).

This Federal Reserve aggregate is released on a high-frequency reporting cadence with no low-integer seasonal period, so the seasonal machinery of Hylleberg et al. (1990) and Canova and Hansen (1995) carries no meaningful object and is not run (Beaulieu and Miron 1992; Ghysels and Osborn 2001).