Spread on other people's cash — the higher-for-longer float-NII beneficiary basket
Spread on other people's cash — the higher-for-longer float-NII beneficiary basket
One-line summary: A hawkish, Warsh-tilted higher-for-longer Fed is a first-order tailwind for firms that earn a spread on other people's cash — brokerage sweeps, custody deposits, and exchange/clearinghouse margin collateral — because the asset side reprices up (or stays high) while what they pay clients stays sticky and low. The one variable that separates winners from also-rans is deposit beta.
The insight
The same forcing function (higher-for-longer) that boosts P&C insurer float income (p-and-c-insurer-float-income-rate-regime) boosts a broader basket of financials that all monetize the same thing: cash they hold for someone else, at a cost far below the prevailing rate. But the benefit is not uniform — it scales inversely with deposit beta (how much of a rate move gets passed through to the cash holder):
- Costless / mandatory collateral (lowest beta, cleanest): exchanges/clearinghouses (cme-group, intercontinental-exchange) earn interest on margin balances customers must post, at negligible incremental cost and with no deposit-pricing competition.
- Cheap, sticky sweep/retail funding (low beta): brokers (schwab pays ~0.19% on sweep cash; interactive-brokers on idle customer credits) and retail-funded regionals (regions-financial, mid-30s beta) retain most of the spread.
- Sophisticated institutional deposits (>100% beta, muted): custodians (state-street, bny-mellon, northern-trust) pass through more than the Fed's moves, so the NII benefit is real (BK NII +18% to $1.4B) but capped, and these names lean on fees.
The tradeable read: the thesis is strongest where funding is cheap and sticky (exchange collateral, broker sweeps, retail deposits) and weakest where clients are rate-sensitive institutions (custodian deposits).
Evidence
- From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: "A hawkish, Warsh-tilted 'higher-for-longer' Fed is a first-order tailwind for firms that earn a spread on other people's cash — brokerage sweeps, custody deposits, and exchange/clearinghouse margin collateral — because the asset side reprices up (or stays high) while what they pay clients stays sticky and low."
- From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: "custodians serve sophisticated institutions with deposit betas over 100% for STT and BK, meaning they pass through more than the Fed's moves — so the higher-for-longer NII benefit is real but more muted than for retail-funded banks."
- From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: CME earns interest on margin balances posted to its clearinghouse — "when rates were near zero those balances generated minimal income; that dynamic has changed as rates climbed."
- From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: Schwab's rate paid on sweep deposits "was just 0.19%," while IBKR's Q1 NII rose "17% YoY to $904M" on larger credit balances and margin loans.
Tradeable implications
- Basket, weighted by inverse deposit beta: cleanest = exchanges (CME, ICE) and the wide-spread broker (SCHW's sweep, IBKR's credits); muted = custodians (STT, BK, NTRS). The canonical custodian/exchange chain is client-float-interest-to-custodian-exchange-nii.
- Load-bearing macro assumption: "no cuts in 2026." A surprise easing cycle is the shared falsifier for the whole basket (IBKR alone loses ~$80–117M NII per 25bps cut).
- Sits alongside the two existing rate-regime chains: brokerage sweep NII (warsh-higher-for-longer-to-brokerage-nii-rerate) and P&C insurer float (p-and-c-insurer-float-income-rate-regime) — the same forcing function, three beneficiary expressions.
Contradictions / tensions
- Deposit beta cuts both ways. The custodian legs (STT/BK betas >100%) may see muted or negative incremental NII benefit even in a higher-for-longer hold — so "financials benefit from high rates" is not a blanket claim.
- Cash-sorting can shrink the cheap-funding balance even as the rate stays high (most acute for SCHW) — a balance risk, not a rate risk.
Open questions
- Do the SCHW / RF figures (currently snippet-sourced) hold against primary filings? SCHW and IBKR print 2026-07-21.
- Does higher-for-longer survive the July CPI print, or does oil-led disinflation force cuts back in (the warsh-higher-for-longer-to-brokerage-nii-rerate load-bearing fork)?