Bank reserves explained

What bank reserves are

When you deposit cash at a bank, the bank credits your account. Behind the scenes, the bank may hold vault cash or — far more commonly for large institutions — a balance in its master account at a Federal Reserve Bank. That Fed account balance is a reserve. Reserves are not physical currency sitting in a vault at the Fed; they are ledger entries, instantly transferable between banks for wire payments, ACH settlement, and interbank lending.

Only depository institutions (commercial banks, thrifts, credit unions with Fed access) hold reserves directly. The public holds deposits at banks; banks hold reserves at the Fed. When Treasury spends, reserves flow from the TGA into bank reserve accounts. When Treasury collects taxes or issues debt, reserves move the other way. The Fed creates or destroys reserves when it buys or sells securities in open-market operations or when loans through discount and emergency facilities settle.

• Reserve balances — the sum of individual bank Fed account balances; published weekly in H.4.1.
• Vault cash — physical currency banks hold on premises; counts toward some regulatory requirements but is not in the Fed ledger.
• Central bank money — reserves plus currency in circulation; the monetary base in a modern fiat system.

Required vs excess reserves

Historically, the Fed required banks to hold a fraction of certain deposits as non-interest-bearing reserves. Banks that fell short borrowed in the federal funds market from banks with excess. The effective fed funds rate was the price of reserve scarcity.

Two regime shifts changed this picture. First, after 2008 the Fed began paying interest on excess reserves (IOER), later renamed interest on reserve balances (IORB), letting the Fed raise short-term rates even with trillions of excess reserves outstanding. Second, in March 2020 the Fed reduced reserve requirement ratios to zero percent for all deposit types — so today most U.S. banks face no binding required reserve ratio, though foreign branches and reporting categories still track required and excess components for analytical purposes.

How to think about “required” today

• Regulatory required reserves — currently zero for most domestic banks; do not assume a 10% textbook fraction still applies.
• Internal required buffers — each bank maintains a comfort level for payment outflows, LCR stress scenarios, and intraday peaks; this is where scarcity bites in practice.
• Excess reserves — balances above regulatory requirements; in an abundant regime, the system holds large excess that earns IORB.
• Aggregate vs distributional scarcity — total reserves can look ample while individual banks are short; fed funds and repo specialness reflect the marginal borrower, not the average.

Even without statutory reserve ratios, aggregate reserve drain matters. When QT shrinks the Fed balance sheet or the TGA rebuilds, total reserve balances fall. Banks that relied on excess as a free liquidity pool must borrow more in overnight markets or shrink balance sheets. That is when SOFR and fed funds trade above IORB despite a stable policy target.

Abundant vs scarce reserve regimes

Economists describe post-2008 U.S. money markets as an abundant reserves regime: the Fed supplies enough reserves that the banking system as a whole does not need to borrow reserves to meet zero requirements. The Fed controls the price of overnight money primarily through administered rates — IORB and the overnight reverse repo (ON RRP) offering rate — rather than by rationing a scarce quantity.

Scarce-reserve intuition from textbooks still applies at the margin. When aggregate reserves fall toward the level where some banks’ internal buffers bind, the distribution of reserves tightens. Signs include:

• SOFR or fed funds above IORB on sustained basis (not just month-end turns).
• Repo specialness — specific collateral trades cheap to GC repo; cash borrowers pay up.
• Large FHLB advances — home loan banks lend to members when reserve redistribution through interbank markets is insufficient.
• Standing Fed facilities usage — banks tap the discount window or Standing Repo Facility when private markets are expensive.
• Bank bill issuance — institutions issue CDs and commercial paper to raise non-reserve funding.

The transition is not a single threshold published by the Fed. Practitioners watch the lowest reserve comfortable level (LCLOR) concept — the aggregate reserve level below which frictions emerge — estimated from past episodes (September 2019 repo spike, COVID injection, 2023 QT plus TGA rebuild). Your desk should treat LCLOR as a range, not a point estimate.

Reading reserves in H.4.1 and related data

The Fed’s weekly H.4.1 release is the canonical source. On the liability side, focus on:

1. Reserve balances with Federal Reserve Banks — total system reserves; the headline series for scarcity monitoring.
2. Treasury General Account — fiscal offset; TGA up often means reserves down.
3. Overnight reverse repurchase agreements — money-fund cash parked at the Fed; can absorb liquidity that would otherwise sit as reserves.
4. Currency in circulation — slow drain as the public holds more physical cash.

Assets matter too: Securities held outright (SOMA) shrink under QT, reducing reserves unless other liabilities expand. The Fed also publishes more granular tables in the Factors Affecting Reserve Balances (H.4.1 statistical release, Table 1) showing reserve supply and absorption items week by week.

Complement H.4.1 with the Financial Accounts of the United States (Z.1) for sectoral deposit stocks, and daily repo rates from the NY Fed for market stress. The effective federal funds rate is published daily; persistent prints above IORB signal distributional or aggregate tightness worth decomposing.

IORB, fed funds, and the rate corridor

In the abundant regime, the Fed sets a floor on overnight rates with IORB: banks should not lend reserves below what the Fed pays them. The ON RRP facility sets a subfloor for money funds and GSEs that cannot earn IORB directly. The effective federal funds rate and SOFR typically trade in a band between the RRP rate and IORB, often a few basis points wide.

When reserves become scarcer at the margin, the band widens upward. Borrowers pay more than IORB because the lender’s opportunity cost is no longer “leave it at the Fed” but “I might need this balance for payments tomorrow.” That is why monitoring reserves complements watching the FOMC target: the target is where the Fed wants overnight rates; reserve plumbing determines where they trade.

Harbor Credit Union liquidity desk refactor

Harbor’s 2023 miss came from conflating “policy rate on hold” with “liquidity unchanged.” The refactor added a reserve-scarcity sleeve with three inputs:

• 4-week change in H.4.1 reserve balances — smoothed weekly delta, same cadence as the TGA series.
• QT-implied reserve drain — expected SOMA runoff from published monthly caps.
• SOFR minus IORB spread — market confirmation that scarcity is binding, not just model-implied.

A calibrated rule: when the 4-week reserve change was worse than −$80 billion and SOFR–IORB exceeded 5 basis points for five consecutive days, Harbor added 8–12 basis points to its wholesale funding forecast. Combined with the existing TGA sleeve from the TGA guide, total forecast error on three-month funding costs dropped from 22 basis points to 6. The desk now labels weeks as “abundant,” “transition,” or “tight” in the ALM pack rather than reporting a single liquidity premium.

Technique decision table

Common pitfalls

• Assuming 10% required reserves still bind — ratios are zero for most U.S. banks since 2020; scarcity is about aggregates and internal buffers, not statutory fractions.
• Ignoring distribution — $3 trillion aggregate can still leave some banks short while others lend in GCF repo.
• Single-week H.4.1 noise — use 4-week changes aligned to your funding horizon.
• Conflating reserves with bank deposits — QE raises both, but they are different balance-sheet items with different velocity.
• Treating ON RRP as irrelevant — large RRP balances can mask reserve tightness until money funds redeploy cash.
• Month-end SOFR spikes as structural — distinguish balance-sheet reporting dates from sustained SOFR–IORB widening.
• Discount window stigma blind spot — published borrowing lags true stress; pair with private repo surveys.
• Static LCLOR from one episode — payment-system growth and Basel liquidity rules shift comfortable levels over time.

Production checklist

• Log H.4.1 reserve balances, TGA, and ON RRP every Thursday release.
• Maintain a 4-week rolling reserve change series alongside TGA delta.
• Track daily effective fed funds and SOFR vs IORB and RRP rates.
• Overlay QT monthly runoff caps on expected reserve drain.
• Monitor FHLB advance growth and large bank CD issuance as secondary stress signals.
• Stress-test ALM with −$200B / −$400B reserve scenarios.
• Read Fed minutes for discussion of reserve ample-ness and facility calibration.
• Document month-end and tax-date calendars separately from structural tightness.
• Re-estimate LCLOR range annually using the latest year of market frictions.
• Pair reserve monitoring with full balance-sheet decomposition when more than one liability line moves.

Key takeaways

• Bank reserves are Fed account balances held by depository institutions; they are the plumbing layer through which fiscal flows, QE/QT, and emergency lending move system liquidity.
• Statutory reserve requirements are zero for most U.S. banks today; scarcity shows up in aggregate drain and internal bank buffers, not textbook 10% ratios.
• In the abundant regime IORB floors overnight rates, but when reserves fall toward LCLOR, SOFR and fed funds can trade above IORB even with an unchanged policy target.
• Harbor Credit Union cut wholesale funding forecast error from 22 bps to 6 bps by adding reserve 4-week changes and SOFR–IORB confirmation to its liquidity sleeve.
• Read reserves alongside TGA, ON RRP, and QT — no single H.4.1 line explains front-end spreads alone.

Related reading

• Fed balance sheet explained — H.4.1 assets and liabilities, SOMA, and the full plumbing ledger
• Treasury General Account explained — how fiscal flows drain and refill reserves
• Quantitative tightening explained — SOMA runoff and the reserve drain from QT
• Repo markets explained — secured overnight funding where reserve scarcity first appears in prices