Beyond the Repo Headlines: What the Liquidity Signals are Really Saying

In late October and early November 2025, usage of the Federal Reserve's Standing Repo Facility

(SRF) reached elevated levels exceeding $50 billion at month-end -- the highest utilization since

March 2020. Simultaneously, the Overnight Reverse Repo (ON RRP) facility has collapsed to

approximately $24 billion, down from peak levels exceeding $2 trillion in 2023. This

combination signals structural stress in U.S. money markets extending beyond seasonal factors.

These two facilities serve opposite functions in the Fed's monetary policy framework. The SRF is

an emergency lending facility where banks can borrow reserves overnight by pledging Treasury

or agency securities as collateral, paying the SRF rate (currently 4.50%). It acts as a ceiling on

overnight rates. The ON RRP works in reverse: money market funds and other institutions lend

cash to the Fed overnight, earning the ON RRP rate (currently 4.30%). It provides a floor on

rates.

The depletion of ON RRP removes a critical shock absorber. When the facility held trillions in

2021-2023, it functioned as a deployable liquidity reservoir. During stress events, as repo rates

in private markets rose above the ON RRP rate, money market funds would withdraw their cash

from the Fed and deploy it into higher-yielding private repo markets. This automatic flow of

liquidity would stabilize rates without Fed intervention. With ON RRP now depleted to $24

billion, this reservoir is empty. When liquidity shocks occur, there is no pool of cash to flow into

stressed markets. Instead, all pressure falls directly on bank reserves, currently at

approximately $2.8 trillion. The elevated SRF usage indicates that despite aggregate reserves

appearing adequate, banks are unable to efficiently reallocate liquidity across the system.

The core problem is that banks with surplus reserves face prohibitive costs to intermediating

due to post-2008 regulations, particularly the Supplementary Leverage Ratio (SLR) and G-SIB

capital surcharges. The SLR requires capital against all balance sheet assets, including reserves.

For a large bank to lend $1 billion overnight, it expands its balance sheet by that amount,

increasing SLR denominators and potentially triggering higher surcharge brackets. The capital

costs of holding additional assets on the balance sheet often exceed repo market spreads,

rendering arbitrage unviable. Banks with surplus reserves therefore park them at the Fed rather

than lending to institutions that need them.

Current conditions reveal that while dealer behavior around period-ends follows established

patterns, the magnitude of rate effects has grown substantially. Recent Federal Reserve

research documents that SOFR rose as much as 25 basis points above the ON RRP rate at recent quarter-ends, far exceeding the 5-10 basis point moves typical in 2017. The Fed's analysis

attributes this to "growing tightness in the repo market and a diminishing elasticity of supply

and demand" as reserves decline. Critically, the research shows that dealer quarter-end

behavior -- reducing triparty borrowing and shifting to central clearing -- has remained

"remarkably stable," yet rate impacts have intensified. This indicates the problem is not

changing behavior but deteriorating underlying conditions. The pattern mirrors 2018-2019,

when similar dynamics preceded the September 2019 crisis. Academic work from that episode

documented that foreign banks reached minimum reserve levels while domestic G-SIBs

maintained surpluses but declined to intermediate due to balance sheet constraints.¹

November 2025 differs critically from September 2019: the ON RRP buffer is now depleted. In

2021-2023, that buffer absorbed surpluses and prevented repo rate collapse. Its near-zero level

means the system lacks this stabilizer precisely when QT has reduced reserves and Treasury

issuance remains elevated. Additional liquidity pressure falls directly on reserves, leaving repo

markets vulnerable to quarter-end dynamics, tax payments, or Treasury settlement volatility.

Chairman Powell announced that QT will slow dramatically, with Treasury runoff ending while

mortgage-backed securities continue maturing. However, this addresses only aggregate levels,

not the structural issues driving period-end stress. The question remains whether current

reserve levels are sufficient given elevated post-pandemic deposits, outstanding credit line

commitments, tighter balance sheet constraints, and the expired Bank Term Funding Program.

What do these signals indicate? Three interpretations emerge. The most likely is that quarterend and month-end rate effects will continue intensifying as reserves decline further, with the

spread between SOFR and ON RRP at period-ends serving as a barometer of underlying

tightness. Federal Reserve research suggests that as Treasury issuance continues and reserves

decline, "the repo market is likely to tighten further and the effects of quarter- or month-ends

on repo rates may grow, providing another potential indicator that reserves are becoming less

abundant." This would manifest as larger SRF usage at period-ends and persistent elevated Fed

facility usage, though system functioning would remain generally stable between these events. A more adverse interpretation sees a triggering event during an already-stressed period-end

causing broader repo market seizure, forcing the Fed to resume asset purchases and confirming

that meaningful balance sheet normalization is impossible under current structures. An

optimistic interpretation requires regulatory reform -- SLR exemptions for reserves or changes

to quarter-end reporting requirements -- to reduce incentives for balance sheet window

dressing, though this appears politically unlikely.

For banks, the implication is that reserve buffers need to be higher than pre-2019 benchmarks,

and the ratio of demandable claims to liquid assets requires closer monitoring. For investors,

continued volatility in short-term interest rates should be expected, particularly around periodends. The Fed's weekly H.4.1 release tracking SRF and ON RRP levels provides leading

indicators. Money market fund flows have outsized impact as their allocation decisions directly

affect system liquidity buffers.

The transformation underway represents a fundamental shift from bank-intermediated to

partially Fed-intermediated money markets. Post-2008 regulations strengthened individual

bank resilience but broke private intermediation chains. The central bank now serves as both

lender and borrower of last resort, with private markets unable to efficiently connect flows.

September 2019, March 2020, March 2023, and November 2025 episodes demonstrate a

pattern: reserves appear adequate until buffers thin, after which modest events trigger

outsized disruptions.

1. Bostrom, E., Bowman, D., Rose, A., and Xia, A. (2025), "What Happens on Quarter-Ends

in the Repo Market," FEDS Notes, Board of Governors of the Federal Reserve System;

Copeland, A., Duffie, D., and Yang, Y. (2021), "Reserves Were Not So Ample After All,"

Federal Reserve Bank of New York.

2. Du, W. (2022), "Bank Balance Sheet Constraints at the Center of Liquidity Problems,"

Jackson Hole Economic Symposium.