medium convictionactive · updated 2026-07-21T00:00:00.000Z
Client float interest → custodian & exchange NII re-rate
Under higher-for-longer, custodians and exchanges/clearinghouses convert client cash and mandatory margin collateral into near-costless interest income — but deposit beta decides how much survives: exchange collateral (near-zero beta) is the cleanest expression, custodian deposits (>100% beta) the most muted.
The chain
1
Rates stay elevated (higher-for-longer) → custodians (STT/BK/NTRS) reinvest client cash into higher-yielding short-term instruments, and exchanges/clearinghouses (CME/ICE) earn interest on customer margin and clearing collateral, while paying the cash holder near-zero.
From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: BNY (BK) "reinvests client cash into liquid, higher-yielding short-term instruments, profiting off the difference"; Q1 2026 NII "jumped 18% to $1.4B" on 13% YoY average deposit growth.
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: NTRS NII "rose 9% to $2.4B"; ICE earns interest on clearing-house collateral (more material at higher yields).
2
Because that collateral/deposit float is near-costless (collateral is *mandatory*; custody deposits are held for servicing), the interest income drops to the bottom line as high-margin NII.
From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: for CME, elevated rates convert "directly into bottom-line gains"; the incremental cost of the mandatory collateral is negligible.
From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: BK's NII acceleration drove net income up 36% (EPS +42% to $2.24) — the spread dropping through. (Custodian side; single-source snippet/blog — the near-costless claim is strongest for exchange collateral.)
3
→ The float-NII beneficiaries re-rate — cleanest at the exchanges (CME/ICE, near-zero-beta mandatory collateral), muted at the custodians (STT/BK/NTRS, deposit betas >100%).
From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: YTD 2026 STT +32%, BK +26%, NTRS +29%; ICE Q1 2026 revenue +20.4% YoY to $2.98B — the beneficiaries have already run.
From 2026-07-20-autoresearch-brokerage-bank-nii-rate-regime: "custodians serve sophisticated institutions with deposit betas over 100% for STT and BK ... so the higher-for-longer NII benefit is real but more muted than for retail-funded banks, and these names lean more on fees than NII." (Caps the custodian leg; the exchange leg is cleaner.)
Corroborating (brokerage leg, 2026-07-21). From 2026-07-21-schw-q2-2026-brokerage-custodian-nii-chain: Schwab's Q1 2026 NIM expanded to 2.88% (from 2.53% YoY) on *wholesale-funding paydown + lower funding cost* — a deleveraging NIM tailwind that persists without Fed cuts and is arguably *strengthened* by a higher-for-longer Warsh Fed; deposits/NII have stabilized since 2023. **Actuals note:** SCHW's *reported* Q2 print (consensus EPS $1.55 / NIR ~$3.28B) could not be fetched results-day morning — this is corroborating trajectory, not the confirming datapoint; the deposit-stabilization leg stays `partial` until the actual Q2 sweep-balance + cash-sorting commentary is folded in (earnings-ingest, next run).
What would falsify this
- Step 1: A surprise Fed easing cycle drops the rate level that powers the collateral/deposit interest income.
- Step 2: A custodian discloses that rising deposit costs (>100% beta) offset the reinvestment yield, so NII fails to expand despite higher rates.
- Step 3: Exchange margin/collateral balances fall (lower volatility / open interest), shrinking the near-costless float even with rates elevated.
Contradictions / tensions
- **Custodian deposit betas >100%** mean STT/BK may pass through *more* than the Fed's moves — the custodian leg's incremental NII benefit is muted and could be negative on the margin, unlike the exchange leg.
- **Already run:** YTD moves (STT +32%, BK +26%, NTRS +29%) suggest a good deal of the higher-for-longer benefit may be priced.
- Evidence per name is single-source (blog/snippet), not primary filings — conviction held at medium.
Implications
- **Tradeable basket, ranked by inverse deposit beta:** cleanest = exchanges cme-group / intercontinental-exchange (mandatory, near-costless collateral, no deposit-pricing competition); muted = custodians state-street / bny-mellon / northern-trust (betas >100%, fee-lean).
- Sibling to the brokerage-sweep chain (warsh-higher-for-longer-to-brokerage-nii-rerate) and P&C float (p-and-c-insurer-float-income-rate-regime); the unifying pattern is spread-on-other-peoples-cash.
- The whole basket rests on the "no cuts in 2026" assumption.
Companies
Concepts
Open questions