Bank Reserves

What Are Bank Reserves?

Bank reserves are deposits that commercial banks hold at the Federal Reserve, the most fundamental form of money in the US financial system. They are the medium in which banks settle payments with each other, the asset against which regulatory capital requirements are calculated, and the variable the Fed monitors most closely to gauge whether its quantitative tightening program is causing stress.

Reserves are, quite literally, cash at the central bank. When JPMorgan sends $1 billion to Goldman Sachs to settle a Treasury trade, that transfer happens by moving $1 billion of reserves from JPMorgan's account at the Fed to Goldman's account at the Fed. Every wire transfer, every securities settlement, every interbank loan ultimately moves reserves. If reserves are insufficient, the payment system breaks down, as the world learned in September 2019 when a reserve shortage caused overnight lending rates to spike to 10%.

For traders, bank reserves are the pressure gauge of the financial system. They tell you how close the system is to stress, how much more QT the Fed can run before something breaks, and when the next liquidity pivot is approaching.

Reserves in the Modern Monetary System

The Pre-2008 World: Scarce Reserves

Before the 2008 financial crisis, the Fed operated in a "scarce reserves" framework. Total reserves in the banking system were approximately $15-50 billion, a tiny fraction of bank assets. The Fed controlled the federal funds rate by conducting daily open market operations: buying a few hundred million in Treasuries to add reserves (lower rates) or selling them to drain reserves (raise rates).

In this system, banks actively traded reserves in the federal funds market every night. Reserves were valuable and scarce, so the overnight rate was sensitive to small changes in supply. The Fed's daily operations could precisely control the rate.

The Post-2008 World: Ample Reserves

QE changed everything. Between 2008 and 2014, the Fed bought $3.6 trillion in bonds, creating $3.6 trillion in new reserves. Reserves went from $50 billion to $2.8 trillion. Suddenly, reserves were so abundant that banks didn't need to borrow them overnight, they already had more than enough.

In this "ample reserves" framework, the Fed controls rates differently:

• IORB (Interest on Reserve Balances): The rate the Fed pays banks on reserves parked at the Fed, this acts as the ceiling of the target range
• ON RRP rate: The rate the Fed pays non-banks (money funds), this acts as the floor

As long as reserves are "ample" (well above what banks need), these administered rates keep the market rate within the target range without daily operations.

The Critical Question: How Much Is Enough?

The transition from "ample" to "scarce" reserves is not gradual, it's a cliff:

The exact location of the cliff depends on the distribution of reserves across banks (concentration matters), the volume of daily payments, regulatory requirements, and bank risk appetite. The Fed estimates the transition zone is around $3.0-3.5 trillion as of 2025, but this is a rough estimate with a wide confidence interval.

The September 2019 Repo Crisis: The Defining Episode

What Happened

On September 16-17, 2019, the US overnight repo market, where banks and dealers borrow and lend cash against Treasury collateral, experienced its worst disruption since 2008:

Why It Happened

Three factors collided on the same day:

1. Reserve scarcity: QT had reduced reserves from $2.2T (2017) to $1.5T (September 2019). This was near or below the scarce threshold, but nobody knew exactly where the threshold was.

2. Tax payments: September 16 was a quarterly corporate tax payment date. Roughly $35 billion drained from corporate bank accounts into the TGA, reducing bank reserves.

3. Treasury settlement: $54 billion in newly auctioned Treasury coupons settled on September 16, requiring dealers to fund their purchases, further draining cash from the system.

The combination was lethal. Banks that had reserves were unwilling to lend them (preferring to hold reserves for their own next-day settlement needs). Banks and dealers that needed cash couldn't get it at any price below 5%, and some borrowers reportedly paid 10%.

The Aftermath

The Fed intervened within hours:

• September 17: Emergency repo operations, the Fed lent $75 billion overnight against Treasury collateral
• September 18: Increased to $100 billion
• October 2019: Began purchasing $60 billion/month in T-bills ("reserve management operations", effectively QE, though the Fed refused to use that term)
• March 2020: COVID triggered full QE restart, rendering the reserve debate temporarily moot

The lesson: The financial system cannot function when reserves are scarce. The transition from "ample" to "scarce" is abrupt, not gradual. And the Fed learned that it must err on the side of too many reserves, not too few, because the cost of a repo market seizure (systemic crisis risk) far exceeds the cost of maintaining excess reserves (slightly larger Fed balance sheet).

Reserves and the Current QT Cycle

The Drawdown Path

The current QT cycle has been the most watched reserve drainage in history:

*QT Treasury cap reduced from $60B to $25B in June 2024 **Further slowdown

The "Cushion" Phase vs. "Direct Hit" Phase

Cushion phase (2022-2024): QT reduced the Fed's balance sheet, but the drain was absorbed primarily by the RRP, not reserves. Money funds shifted from RRP to T-bills, so the cash left the Fed's balance sheet but reserves stayed roughly stable. This is why financial markets tolerated QT so well.

Direct hit phase (2025+): With RRP nearly depleted, further QT reduces reserves directly. Every dollar the Fed's balance sheet shrinks by is now a dollar fewer in bank reserves. This is the dangerous phase, the one that led to the September 2019 crisis.

When Will QT End?

The Fed has signalled that QT will end when reserves reach a level "somewhat above" the ample threshold. Market and Fed staff estimates:

• Optimistic estimate: QT continues through Q4 2025, ending at ~$3.0T reserves
• Base case: QT ends Q2-Q3 2025 around $3.2T as the Fed errs on the side of caution
• Hawkish scenario: QT continues longer, but the Fed creates a standing repo facility as a backstop

The end of QT is one of the most consequential liquidity events on the horizon. It means the balance sheet stops shrinking → the primary headwind for net liquidity disappears → risk assets receive a significant tailwind.

Reserves and Bank Behaviour

The Distribution Problem

Aggregate reserve levels can be misleading because reserves are unevenly distributed across banks:

• The Big Four (JPMorgan, Bank of America, Citigroup, Wells Fargo) hold a disproportionate share of total reserves, estimated at 30-40% of the total
• Regional banks hold far fewer reserves relative to their size and are more vulnerable to scarcity
• Foreign bank branches in the US hold reserves for dollar payment settlement and have different behaviour patterns

A system with $3.3T total reserves might be "ample" for JPMorgan but "scarce" for a mid-size regional bank. The September 2019 episode was partly caused by reserve distribution: the major banks had reserves but were unwilling to lend them because post-crisis regulations (LCR, internal stress tests) created incentives to hoard.

The Regulatory Incentive to Hoard

Post-2008 regulations require banks to hold "High-Quality Liquid Assets" (HQLA), and reserves are the highest-quality, most liquid asset available. Banks subject to the Liquidity Coverage Ratio (LCR) must hold enough HQLA to survive a 30-day stress scenario.

This creates a paradox: the regulations designed to make banks safer also make them reluctant to lend reserves, increasing the minimum level of reserves the system needs. The LCLOR is higher than it would be without post-crisis regulations, meaning the Fed must maintain a larger balance sheet permanently.

Trading Reserves: Practical Applications

Predicting QT Endpoints

Track reserves weekly and project the trajectory:

1. Note the current QT monthly cap (the maximum pace of balance sheet reduction)
2. Estimate how much of each month's QT hits reserves vs. RRP
3. Project when reserves reach the estimated LCLOR ($3.0-3.5T)
4. Compare to Fed communications about QT guidance

When QT ending is 3-6 months away, position for the liquidity pivot:

• Long duration (bonds rally as a tightening tool is removed)
• Long risk (equities and crypto benefit from the end of liquidity headwinds)
• Short volatility (the removal of QT uncertainty reduces tail risk)

Reserve Stress Indicators

Watch for early signs that reserves are becoming scarce:

When 2+ indicators enter the "warning" zone simultaneously, reduce leveraged positions and add liquidity hedges.

Cross-Asset Implications

What to Watch

1. Weekly H.4.1 reserve balances (Thursdays): The primary data point. Track the pace of decline.
2. SOFR rate and SOFR-IORB spread: Daily funding market health check. Widening spreads signal reserve scarcity approaching.
3. Fed meeting minutes: Look for discussion of "ample reserves," "reserve demand," or QT adjustments.
4. Standing Repo Facility (SRF) usage: The Fed created the SRF in 2021 as a backstop for repo market stress. Any usage is a warning signal.
5. Treasury issuance mix: Heavy coupon (vs. bill) issuance drains reserves faster. Monitor the Quarterly Refunding Announcement for shifts.