Money market funds (~$7.9T) — the constant-NAV run vehicle the bank lens misses

Built 2026-06-14. Structured data + edges: macro-money-market-funds.json. Companion to macro-uncovered-risk-pools and macro-stablecoin-failures-manipulation.

Money market funds are a ~$7.9T cash pool outside the FDIC bank lens that dominate repo and the Treasury-bill market and run by redemption, not credit — the institutional analogue of stablecoins (constant-value, redeemable-at-par, T-bill-backed, no FDIC).

1. Size & composition — fact (ICI)

~$7.87T total (week of June 10, 2026) — a record; up from ~$7.0T (Mar 2025) and ~$4.5T at the 2022 rate-hike start. The pool roughly doubled in four years.
Split: government ~$6.5T (~82%), prime ~$1.23T (~16%; institutional prime only ~$241B), tax-exempt ~$145B (~2%). The largest segment is the least reform-constrained.
Retail ~$3.10T; institutional ~$4.78T — the institutional tranche is the faster-moving, more run-prone cash.

2. The reforms shifted risk, didn't remove it

2014 (effective 2016): institutional prime floats its NAV → ~$1T+ migrated prime → government.
2023: removed gates; added a mandatory liquidity fee when daily net redemptions exceed 5% of NAV; raised liquidity minimums (25% daily / 50% weekly).
Residual: government MMFs (~82%) keep a stable $1.00 NAV with no fees/gates — and government-fund redemptions, not credit, were the March-2020 channel.

3. The run mechanism is proven — twice, each backstopped

2008: the $62.6B Reserve Primary Fund broke the buck (NAV $0.97) on $785M of Lehman paper; >$40B redeemed in two days; Treasury guaranteed MMFs via the ~$50B ESF.
March 2020: prime MMFs lost ~$125B (~11%) in <3 weeks — a pure liquidity run, no default. The Fed's MMLF drew ~$53B and stopped it.

4. Repo & Treasury plumbing — the shock-absorber is gone

The Fed's ON-RRP peaked >$2.5T (Dec 2022) and drained to ~$22B (early Sept 2025) — the excess-liquidity buffer is exhausted.
MMFs hold ~40% of outstanding T-bills (largest marginal buyer) into a deficit-driven supply surge (net bill issuance ~$475B Q3-2025, ~$416B Q1-2026).
2025-26 stress: Dec 31, 2025 repo prints ~4.0%; the Standing Repo Facility was tapped ~$74.6B (highest since COVID) before reversing to ~zero by Jan 2026. QT ended Dec 2025; the Fed began reserve-management T-bill purchases — the first real test of post-RRP-drain plumbing.

5. The stablecoin analogy — same collateral, weaker backstop

Major fiat stablecoins are T-bill/repo-backed, redeemable at par, run-prone, with no FDIC and no stable-NAV legal backstop — a constant-NAV money fund without Rule 2a-7 protections. USDC depegged to ~$0.87 (Mar 2023) on SVB fears.
The GENIUS Act (July 2025) requires ≥1:1 reserves in cash/T-bills/T-bill repo/government MMFs but exempts issuers from bank capital and gives holders no FDIC coverage. Issuers bought ~$109B of T-bills (2025) — putting stablecoins and MMFs into the same bill/repo collateral pool, so a run in either forces the same fire-sale.

Synthesis

At ~$7.9T, MMFs are the single biggest non-bank cash pool, the dominant repo lender, and the largest marginal T-bill holder — with the RRP cushion gone. The same long-rate move that re-opens the bank HTM hole (macro-bank-htm-marks) drives MMF and stablecoin redemptions into one shared bill/repo market.

What is NOT asserted

No MMF is currently breaking the buck — the point is structural run-proneness + a drained RRP cushion.
The ~40% T-bill share and ~$109B stablecoin figure are widely cited but approximate (contested); exact shares live in OFR/Fed data.
The stablecoin-MMF link is a structural analogy + shared-collateral channel, not a legal equivalence.
Overlay edges are excluded from the SCC / Z3 / TLA+ proofs.

Sources: ICI money-market-fund assets; SEC 2014 reform & 2023 reform; Reserve Primary Fund; Fed FEDS — prime-MMF behavior; Boston Fed — MMLF; FRED — ON RRP; Treasury refunding / bill supply; GENIUS Act text; Brookings — stablecoins; Wolf Street — year-end SRF.

← Research index · structured data: macro-money-market-funds.json · macro-money-market-funds.md