Cash float
Schwab pays 0.19% on sweep cash. Interactive Brokers lives on the spread. The Fed hiked 12–0; CME still clips the fight over the path.
On 17 September the Federal Open Market Committee hiked 12–0, twenty-five basis points, to a 3.75–4.00 percent funds range. Charles Schwab still pays 0.19% on the cash sitting in customer sweep accounts. The difference between that rate and what the money earns overnight is the business. The hike is the level. The argument about 2027 is the volume.
That pass-through has a name: deposit beta. It is the share of a rate change that reaches the customer. Net interest income, or NII, is what remains — yield on the assets minus what you pay on the deposits. The benefit scales inversely with beta.
Who keeps the spread
Interactive Brokers’ first-quarter NII rose 17% year-on-year, to $904 million. A quarter-point cut would cost the firm about $80 to $117 million of that income, on management’s own estimate. Schwab’s net interest margin had already widened to 2.88% from 2.53% a year earlier, helped by paying down wholesale funding. The second-quarter call then re-attributed the expansion: revenue $7.1 billion, adjusted earnings $1.62 a share, NII up 19%. CFO Mike Verdeschi put the year-on-year margin move “primarily [to] lending activity” — PAL originations up 59%, bank loans $67 billion, up 33% — not to the funds rate. The firm’s last disclosed scenario still assumed a December hike, with no 2026 P&L impact. “If rates resume a hiking pattern you’ll see even more expansion.”
The custodians print too, and keep less of it. BNY’s first-quarter NII jumped 18%, to $1.4 billion, on 13% average deposit growth; net income was up 36%. Northern Trust’s NII rose 9%, to $2.4 billion. State Street and BNY run deposit betas over 100% — they pass through more than the Fed’s move — so they lean on fees. ICE, on the exchange side, reported first-quarter revenue up 20.4%, to $2.98 billion, including interest on clearing collateral that paid almost nothing when rates were zero. CME’s second-quarter earnings from cash performance bonds and guaranty-fund contributions were $1,388.8 million. BNY’s second-quarter NII was up 20 percent year on year even as deposit margins compressed. ICE pledged collateral stood at $235.3 billion against $170.1 billion at year-end. The third-quarter prints are still ahead — State Street October 14, BNY the 15th, Northern Trust the 21st, ICE the 29th. The September hike is a 25-basis-point step-1 print. It is not, on the wiki, a post-hike NII model for Schwab.
Year-to-date the custodian stocks had already run when those prints were filed: State Street up 32%, BNY 26%, Northern Trust 29%. A good deal of the higher-for-longer bid may already be in the price. The 17 September print is the funds path the NII basket wanted. It is not, on the wiki, a brokerage re-rate. After the hike, Schwab sold off two sessions. A TIKR note attributed the drop to AI-disruption fear, not a NIM miss, and restated CFO Mike Verdeschi’s July guide: a fourth-quarter exit NIM of 3.25 to 3.30 percent and about $250 to $300 million of annualized net interest revenue per 25 basis points. A restatement of a July guide is not the post-hike sell-side NII model the hypothesis asked for. The Schwab question stays a hypothesis.
When the path is contested
CME Group has a second line that does not need rates to stay high. It needs them to stay argued. Short-term interest-rate futures — STIR, the SOFR, fed-funds, and related contracts — are how the market hedges a two-sided Fed. The June dots were split: nine of eighteen officials projected at least one hike, eight no change, one a cut. CME’s FedWatch tool had implied hike odds at 18% on July 2 and 36% on July 13. That was the setup. September coalesced the near vote.
Open interest in that complex went to records. SOFR futures above 15.8 million contracts. The three-month SOFR contract, SR3, at 14.2 million, with leveraged funds at record shorts. Fed-funds futures above 2.5 million. Total STIR open interest at a 2026 high of 18 million. CME earns a per-contract fee plus market-data revenue whether Kevin Warsh hikes, holds, or cuts.
The Summary of Economic Projections still holds funds at 4.1% at year-end 2026 and year-end 2027. Warsh, after the vote: he would be “hard-pressed to describe current financial conditions as restrictive.” The two-percent inflation objective is “a commitment, not a forecast. That standard has not been satisfied.” A unanimous print on this meeting is coalescence of the near vote. It is not, on the wiki, collapse of the path.
I think optionality, especially on the equity market side and more broadly, rate volume is out of step too low relative to the uncertainties. And I think that's, I have a very strong conviction there.
Dean Curnutt, Macro Risk Advisors, 14 September 2026
Curnutt was on Jack Farley’s show two days before the statement. Realized S&P correlation, in his telling, sat at 5 to 15 percent — “no comp.” People call the S&P the risk asset; he said the ten-year note is. Farley: they were recording with the ten-year at 5 percent and a hike mostly priced. The wiki’s own falsifier for the volume-toll is not this meeting. It is FedWatch into the next one — hike odds under 10 or over 90. Open interest is still not average daily volume. Transaction revenue tracks the trades. The recap that filed the records did not file the ADV. The directional NII basket needs the level. The volume basket needs the fight. Neither Schwab nor CME is a re-rate on the print.
Harley Bassman, recorded Wednesday afternoon after the vote, called the twenty-five-basis-point print — and a second hike before Christmas — “not a good result.” It should have been “0 or 50.” The load-bearing claim is trust, not the funds rate: “we have a lack of trust in the government in a lot of our institutions, and maybe about the Fed,” and “the Fed needs to go and regain the trust of the market.” Rates have risen 150 basis points while ten-year TIPS breakevens sit at 2.34–2.35 — “inflation. No one cares about inflation.” Hyperscaler bonds are “probably money good.” The equity he would not comment on. Non-bank private credit, he said, “reduces the ability of the Fed to transmit their policy.” Independent color on the same forcing function. It still does not supply a sell-side NII haircut on Schwab or Interactive Brokers. The Schwab question stays a hypothesis.