Charted: $7.85T Sits in Money Market Funds — Still 3.3 pp Above Bank Savings Yields

Bank deposit rates get the headlines when the Federal Reserve cuts. The quieter cash story is where households and institutions actually parked money during — and after — the hiking cycle. As of the week ended 29 July 2026, the Investment Company Institute puts US money market fund (MMF) assets at $7.85 trillion. That is more than double the $3.63 trillion year-end 2019 stock. Government funds alone are $6.47 trillion — 82.4% of the total.

Unlike our personal saving rate piece — which tracks the flow of saving out of disposable income — this post maps the stock of cash that left (or never entered) bank checking and savings books and instead sat in money market share classes. The Fed’s own November 2025 FEDS Note on deposit–MMF substitution is the bridge: when policy rates rose, deposits excl. large time deposits stalled near $15 trillion while MMF AUM kept climbing, and the MMF share of the combined cash pile jumped from the low-20s into the low-30s.

The headline pile: $7.85T and still growing

The composition is the first non-obvious fact. Retail brokerage UIs make prime and “cash” sweep products feel ubiquitous, but the dollar majority of US MMFs is government — funds that hold Treasuries, agency paper, and repo against government collateral. That matters for stress narratives: the 2026 cash pile is not a return of 2008-style prime commercial-paper risk at the same scale. It is a government-dominated parking lot that grew because yields cleared bank deposit pricing.

Institutional share classes still outweigh retail ($4.78T vs $3.08T). Both nearly doubled since 2019 YE (retail $1.37T → $3.08T; institutional $2.26T → $4.78T). The retail boom is real — app-based brokerage cash and high-yield savings alternatives pulled household dollars — but the institutional channel remains the larger ledger.

Two cash piles: MMFs rose while deposits stalled

The Fed FEDS Note (Im, Li, and Wang, November 2025) frames the substitution cleanly. Domestically chartered bank deposits excluding large time deposits sat near $15 trillion as of May 2025, while MMFs were about $7 trillion. In our dual-path series, MMF AUM rises from $3.63T (2019 YE) to $7.85T (Jul 2026) while the deposit pile — after the COVID stimulus peak — flattens in the mid-teens. The MMF share of the combined pile moves from roughly 23% pre-hike into the low-30s, hitting about 34% by mid-2026 on our estimated deposit path.

That share shift is the monetary-policy transmission story banks underplay. When the Fed hiked, wholesale funding costs rose immediately; many banks lagged on retail deposit betas. Money market funds, by design, pass through higher short-rate income quickly. Households and corporates that could move cash did. The FEDS Note estimates a substitution elasticity around −0.21 on weekly growth — not one-for-one, but large enough that deposit franchise value and MMF AUM became two sides of the same rate-cycle coin.

Pair this with our credit-card APR chart: cards stuck near 21% while policy rates fell. Banks can be slow to cut lending rates and slow to raise deposit rates. The MMF industry is the residual claimant on that stickiness — it harvests the yield gap households notice when they open a brokerage cash screen.

The yield gap that pulled the cash

Year-end 2025 ICI Fact Book commentary puts the average taxable MMF yield near 3.9% versus about 0.6% on money-market deposit accounts — a 3.3 percentage-point gap. Earlier in the hiking cycle the gap was wider still; our estimated path peaks near 4.6 pp around 2023 as policy rates crested and deposit betas lagged. Even after the first cuts, the gap did not collapse to zero. That is why AUM kept rising into 2026 instead of reversing the moment the Fed pivoted.

The gap is not a mystery of “financial literacy.” It is industrial organization. MMFs compete on yield and same-day liquidity within a regulated wrapper. Banks compete on branch networks, payment rails, FDIC insurance framing, and relationship pricing that often underpays the marginal cash saver. As long as the overnight complex clears above what banks choose to pay on MMDAs and savings, the cash will keep a toehold in fund share classes — even if some dollars eventually rotate into T-bills or short bond ETFs.

Government-fund dominance reinforces the point. Investors did not need to take prime credit risk to earn a competitive cash yield. They could buy government MMFs and still beat the bank quote by hundreds of basis points at the cycle peak.

Who wins and who is exposed

Winners: MMF complexes and the brokerages that distribute them; Treasury and agency markets that absorb government-fund demand; households and CFOs who moved cash and captured the gap. Exposed: regional and community banks whose deposit betas stayed low and whose funding mix leaned on non-maturity deposits that proved rate-sensitive once alternatives were one click away; any soft-landing narrative that treats “deposits stable” as proof households are flush when a large cash stock simply relocated.

The exposure is asymmetric. A $15T deposit pile that is flat is still enormous. But marginal funding — the next dollar of loan growth — got more expensive when the alternative yield was visible. That is how a 3.3 pp MMDA-vs-MMF gap shows up in loan officer surveys and in net interest margin compression even when headline deposit levels look calm.

Institutional cash managers matter for market plumbing. When institutional MMF assets grow by trillions, overnight repo and Treasury bill markets deepen as absorption valves. When they shrink, the reverse can tighten funding. The 2026 stock is large enough that any future rapid reverse — a sudden deposit-rate war or a flight into duration — would move short-rate markets, not just fund marketing decks.

Historical context: from COVID cash to hike-era parking

2019 YE MMFs at $3.63T were already a mature industry. COVID fiscal and monetary stimulus then flooded bank deposits; MMF totals rose more modestly into 2021 as yields collapsed toward zero and the opportunity cost of leaving cash in a checking account vanished. The regime change was 2022–23: policy rates ripped higher, MMF yields followed within weeks, and bank deposit pricing followed slowly. That is when AUM accelerated — $4.79T (2022) → $5.92T (2023) → $6.85T (2024) → $7.75T (2025 YE) → $7.85T (Jul 2026 weekly).

The composition shift toward government funds is itself a post-2010 regulatory legacy. Reform after the financial crisis made prime funds less convenient for many investors; government funds became the default institutional cash sleeve. The hike cycle poured gasoline on that already-preferred wrapper. So the 2026 chart is not “speculation returned to prime” — it is “safe cash got paid.”

What would change the story

A sustained bank deposit-rate war that closes the MMDA–MMF gap under 1 pp would slow inflows and could reverse some retail AUM. A sharp Fed easing cycle that drives taxable MMF yields toward 1% would shrink the opportunity cost of inertia — cash might stay put in funds out of habit, or rotate into short-duration bond products. A regulatory shock to MMF liquidity fees or gates could temporarily push cash back to insured deposits. A Treasury bill shortage that compresses government-fund yields relative to bank specials would flip relative pricing. None of those are in the Jul 2026 weekly snapshot — the snapshot still shows a wide gap and a rising stock.

Caveats and methodology

• Deposit path points outside May 2025 are estimated from the FEDS Note narrative and H.8 trends; only the May 2025 ~$15T deposits excl. large time deposits figure is treated as disclosed in that note. MMF totals from ICI weekly and Fact Book year-ends are disclosed.
• Yield gap history before YE 2025 uses estimated anchors; the 3.9% vs 0.6% YE 2025 comparison follows ICI Fact Book commentary and should not be read as a daily mark-to-market series.
• MMF share of the combined pile depends on the deposit definition (excl. large time). Including large time deposits would shrink the MMF share; the FEDS Note’s definition is the relevant one for substitution analysis.
• 2026 composition is a mid-year weekly snapshot, not a year-end Fact Book totallevels can move week to week with bill supply and rate expectations.
• Retail vs institutional is share-class reporting, not a perfect household-vs-corporate split; some household cash sits in institutional channels via advisers.

The shareable takeaway

US money market funds hold $7.85 trillion — with government funds at 82% — and YE 2025 taxable yields still cleared bank money-market deposit accounts by about 3.3 percentage points. Deposits excl. large time deposits stalled near $15 trillion while the MMF share of the combined cash pile climbed into the low-30s. The hiking cycle did not just raise loan rates; it relocated a multi-trillion-dollar cash stock into a yield-sensitive wrapper banks were slow to match.

Related reading: US personal saving rate and credit card APRs vs fed funds.