Maintaining Outperform: The Rate Shock Paused the Buyback, Not the Engine, as Record Revenue, a 2.84% Margin and a Raised Operating-Leverage Guide Leave the De-Rated Stock Cheaper Than the Business

Results vs. Consensus

Q3 2026 Scorecard

Year-Over-Year Comparisons

Quarter-Over-Quarter Comparisons

Quality of Beat

Revenue / NII: Net interest income on a taxable-equivalent basis grew 5.6% YoY and 2.3% sequentially to $4.49B, in the upper half of the +4–6% guide. The margin carried the quarter: NIM rose 5bps to 2.84%, the largest sequential gain since the 9bp jump in Q3 2025, as the earning-asset yield rose 6bps to 4.92% while the rate paid on interest-bearing liabilities rose only 3bps to 2.53%. Management attributed the sequential gain to fixed-asset repricing and the Amazon book, and noted that deposit costs rose 7bps (about 2bps from the September hike) while total funding costs rose just 1bp. Average earning assets grew only 0.3% QoQ as the bank shrank its securities book by $4.2B, so this was a mix-and-rate quarter rather than a volume quarter. The reported revenue line ($7,970M) sat a fraction below the ~$8.0B Street figure only because of a $65M securities loss.

Fees: Fee revenue of $3.55B grew 14.9% YoY, above the top of the guide. BTIG contributed $226M, so fee growth excluding BTIG was about 7.6%, a healthy organic pace. Capital markets revenue reached $640M (+69.3%), with legacy capital markets up roughly 10% on syndications and client derivatives. Corporate payment and treasury management revenue rose 12.3%, trust and investment management 9.9%, and investment products 19.6%. Two lines softened: merchant processing fell 0.4% (and 4.9% sequentially on seasonality), and mortgage banking fell 13.9% as rates rose. "Other" revenue of $166M included equity investment gains that management told investors to normalize to a $120–150M range, so roughly $15–45M of the fee beat is non-recurring.

EPS / expenses / credit: Expenses rose 7.1% YoY but only 2.0% excluding BTIG, for 240bps of reported operating leverage (440bps excluding BTIG). Below the operating line, three items held EPS flat sequentially despite 5.9% growth in pre-provision income: the $135M initial provision on the Amazon book (~$0.07), a higher effective tax rate (21.2% vs. 19.7% in Q2, roughly $0.03), and a larger deduction between net income and income to common shareholders, which covers preferred dividends ($107M vs. $79M, about $0.02). None of the three is a change in earnings power. Credit improved on nearly every line: net charge-offs fell to $523M (0.51%), card losses improved to 3.29% from 3.83% in Q2, nonperforming assets declined to $1,324M (0.32% of loans), and the allowance rose to 1.97% of loans (636% of nonperforming loans) mainly because of the acquired book. ROTCE was 18.5% and ROA 1.25%.

Segment Performance

Business Segments — Q3 2026

Fee Revenue Mix — Q3 2026

Wealth, Corporate, Commercial & Institutional Banking

This is now the earnings engine of the company: net income rose 29.8% YoY to $1.60B, about 73% of the consolidated total, on revenue up 19.9%. Net interest income grew 11.9% on average loans up 14.5% and deposits up 6.9%, while fee income rose 28.9% on the first full quarter of BTIG, syndications, client derivatives and trust growth. Expenses rose 17.3%, almost all BTIG compensation and technology, and provision fell 42.9% on favorable credit. The one soft spot was company-wide and sequential: total commercial loans were flat QoQ (−0.1%), even as the segment's average loans grew 1.2%, as nondepository financial institution (NDFI) balances ran off and some new deals did not clear USB's return hurdles.

"Capital markets growth of 69% year over year reflects a full quarter contribution from BTIG and 10% growth across legacy capital markets." — John Stern, Vice Chair & CFO

Assessment: The institutional segment is compounding faster than the company and is the reason group ROTCE holds near 18.5% through a reserve build and a rate shock. The flat sequential commercial book is a deliberate pricing choice rather than lost demand (management described pipelines as robust and expects C&I growth to resume in Q4), but company-wide commercial balances going flat is new for this cycle and makes Q4 a checkpoint.

Consumer & Business Banking

Revenue was essentially flat YoY (−0.5%) because mortgage banking revenue fell with higher rates, but net income rose 6.7% as expenses fell 4.7% and deposits grew 2.1%. The segment is where management is putting its next growth push: another record quarter of consumer deposits, savings balances company-wide up 26.2% YoY, the Business Essentials card suite launched, new small-business banker teams in expansion markets, and branch investment stepped up to ~$300M a year. Small business, which management sizes at more than two million clients, is named the priority for 2027.

"We have a truly differentiated set of products and an enviable client footprint with small business. And this segment will be an important priority going into 2027." — Gunjan Kedia, Chairman & CEO

Assessment: This segment is not a growth engine in reported revenue, but it is the funding engine that makes the margin story work: low-cost, relationship-based consumer and small-business deposits are why deposit costs rose only 7bps through a Fed hike. Mortgage will stay a drag while long rates stay around 5%. The 2027 small-business push (cards, deposits and treasury tools sold together) is the right extension of the Bank Smartly playbook.

Payment Services

Segment revenue grew 6.5% (management's "total payments revenue" figure), with net interest income up 6.8% on card loans including the Amazon book and fees up 6.3%. Corporate payment products and prepaid revenue grew 18.6% on corporate card volume up 15.4%, credit card fees grew 7.5% on card volume up 8.9%, and global merchant acquiring was flat (−0.4%) on flat merchant volume. Pre-provision income rose 10.5%, but net income fell 31.4% to $175M because the segment absorbed the Amazon Day-1 provision (segment provision $594M vs. $409M a year ago).

"The corporate payments landscape has the potential to be transformed by the growing convergence of treasury management rails like ACH and wires, card rails, and new stablecoins and tokenized deposits." — Gunjan Kedia, Chairman & CEO

Assessment: Read through the provision and the payments transformation is gaining traction: pre-provision income +10.5%, corporate payments growing high-teens on new business and lapped government-spend weakness, and card credit improving while balances grow. Merchant acquiring remains the laggard and management expects it to stay near zero growth for a couple more quarters. The segment's reported net income will look much better in Q4 without the Day-1 charge.

Treasury & Corporate Support

The treasury segment lost $204M (vs. a $78M loss a year ago) on lower NII from a smaller securities book and less cash at the Fed, plus the $65M of securities losses from repositioning. This is the visible cost of the balance-sheet actions management described: shrinking the fixed-rate securities portfolio, holding more cash, and repaying long-term Federal Home Loan Bank advances (long-term debt fell to $52.0B from $60.8B at year-end 2025).

Assessment: These are deliberate, small, near-term costs that make the balance sheet more asset-sensitive into a rising-rate cycle. We would not extrapolate the securities losses; we would watch whether the shrinking securities book keeps funding loan growth without more losses.

Payments KPIs

Average Loans by Category

Balance Sheet, Credit & Capital Snapshot

The capital walk explains the buyback decision. Earnings net of dividends added about 30bps to AOCI-inclusive CET1; roughly $12B of risk-weighted-asset growth (heavy loan demand plus the onboarded Amazon partnership) took some of it back; and the 87bp backup in the 10-year Treasury yield and wider credit spreads cost a further 40bps through AOCI. Net, the AOCI-inclusive ratio fell 30bps to 9.1%, and accumulated other comprehensive loss deepened to $9.46B from $6.99B at year-end 2025. The regulatory CET1 ratio (10.8%) and the tangible common equity ratio (6.6%) were unchanged, the latter helped by a $23B decline in period-end assets as the bank shrank its securities book and repaid long-term borrowings. USB repurchased no shares in the quarter and paid out about 40% of common earnings as dividends, against a 70–75% total payout framework.

Key Topics & Management Commentary

Overall Management Tone: Confident and growth-minded on operations, careful and conditional on capital. Management raised the full-year guide for the second straight quarter and talked readily about the next growth agenda (small business, corporate payments, capital markets), but answered every buyback question with a three-part framework of loan demand, rates and regulation instead of a date or a size. Compared with July's confident talk of a final lap of capital build, the posture moved from a planned glide to a prudent pause, while the operating tone stayed at the most assured level of our coverage.

1. The Buyback Pause: Temporary, but Unscheduled

"Despite a sharp move in rates this quarter, we maintained healthy capital levels and continue to have flexibility in how we manage the balance sheet. We expect our capital ratios to build from here. Positioning us to resume share repurchases over the next quarter or 2." — John Stern, Vice Chair & CFO

USB disclosed the pause at an industry conference in mid-September, on the day the Fed raised rates for the first time since 2023, and confirmed it on the call: no shares were repurchased in Q3. The CFO named three causes (strong loan growth, the capital demands of the Amazon book, and "the large rate move") and three conditions for restarting (where loan demand settles, the path of rates, and pending regulation). He stressed that the company generated 33bps of capital after dividends in the quarter and remains committed to the 70–75% payout.

The tension is arithmetic. In July, management tied the payout ramp to approaching a ~10% AOCI-inclusive CET1 ratio. That ratio is now 9.1%, further from the marker than it was in July, and the 10-year yield reached 5.31% in early October, with futures pricing another hike. Restarting buybacks within one or two quarters means either AOCI recovers, RWA growth slows, or the ~10% marker is softer than it sounded.

Assessment: This is the weakest outcome on our capital-return pillar since we initiated, and we do not dress it up. But the pause reflects mark-to-market losses on securities that pull to par as they mature, plus capital consumed by high-return loan growth, not a shortfall in earnings power. The test is simple and near: a restart by the Q1 2027 print keeps the pillar alive. A third quarter without repurchases would turn a temporary pause into a standing policy, and we would move to Hold.

2. The Margin Climbs 5bps, With Another 5bps Guided

"Net interest margin improved 5 basis points sequentially to 2.84% driven by improving loan yields and fixed asset repricing benefits. We currently expect net interest margin to increase by a similar amount in the fourth quarter." — John Stern, Vice Chair & CFO

The margin's step-up was larger than Q2's 2bp gain, and the Q4 guide implies an exit near 2.89–2.90%, close to the top of the range we had modeled. The drivers are the ones management has named for a year: roughly $10B a quarter of fixed-rate assets repricing at about 125bps higher (now slightly better because of the backup in rates), a loan mix tilting toward floating-rate C&I, CRE and card, and a deposit base growing in savings while time deposits run off (−19.3% YoY). The full-quarter Amazon book adds a further tailwind in Q4. The 3% target "at some point in 2027" was reaffirmed.

Assessment: The pillar that used to be our top risk is now the cleanest part of the story. Rising rates hurt USB less than they used to (floating-rate loans are 30% of earning assets, up from 22% four years ago), and total funding costs rose only 1bp in the quarter the hike landed, although two weeks of a hike is not yet a full beta test. With an exit rate near 2.90%, the 3% target needs only about 10bps more over 2027.

3. The Guide Is Raised Again: Revenue +8–9%, Operating Leverage ~300bps

"For the full year 2026, we now expect total net revenue growth of 8% to 9% compared to the prior year. Or 6% to 7% excluding BTIG. This is improved from our prior guidance range of 5% to 7%. We now expect to deliver approximately 300 basis points positive operating leverage for the full year or approximately 400 basis points excluding the impact from BTIG." — John Stern, Vice Chair & CFO

This is the second consecutive raise. In January the full-year revenue guide was +4–6%; in July it became +7–9% (+5–7% excluding BTIG); now it is +8–9% (+6–7% excluding BTIG). The larger change is operating leverage, raised by 100bps on both bases, which means the extra revenue is dropping through rather than being spent. Management also signaled, without giving 2027 guidance, that it expects positive operating leverage again next year on similar drivers.

Assessment: A raise in operating leverage matters more than a raise in revenue, because it shows expense discipline holding while two acquisitions are integrated. It also directly answers our Bear-2 concern that reinvestment would cap EPS growth. The raise rests on organic momentum: the ex-BTIG revenue range moved up too.

4. BTIG and Capital Markets: Ahead of Plan, and a 10%-of-Revenue Target

"We closed the BTIG acquisition in June, and it contributed $226 million of fee revenue in the third quarter. Our combined businesses are off to a wonderful start and already working very closely together to broaden how we serve our clients." — Gunjan Kedia, Chairman & CEO

BTIG's first full quarter beat the ~$200M per-quarter guide set in July, and management again guided ~$200M for Q4. Capital markets is now about 8% of company revenue, and the stated goal is 10% or more within three to five years through organic growth in both franchises, deeper wallet share where USB already lends heavily, and new product capabilities. Management said cultural fit, the main question at announcement, has proved good, and that cross-selling now works both ways, including trading flow from USB's sizable leveraged-ETF relationships. Management plans to discuss its capital-markets strategy at a conference later this fall and promised more disclosure as the business grows.

Assessment: The deal is tracking ahead of underwriting on revenue. The figure we cannot yet confirm is profitability: the expense bridge implies BTIG cost roughly $210–214M in the quarter against $226M of revenue, with integration costs not separated. The fall capital-markets presentation should show whether the ~15%-to-20% contribution margin management described in July is on track.

5. The Amazon Reserve: A Smaller Earnings Charge Than Flagged

"We reported earnings per common share of $1.34 with approximately $0.07 impact from reserve build this quarter. This equates to 10% year over year EPS growth or 16% excluding the reserve build." — John Stern, Vice Chair & CFO

The acquired small-business card portfolio closed in the quarter, adding about $0.9B to average loans and $1.7B to period-end loans. Accounting required a $135M initial provision through earnings plus a $55M allowance for acquired credit losses recorded through purchase accounting, so the total reserve set up was $190M. In July management had flagged a ~$160M reserve build, so the earnings charge came in lower and the total reserve higher. Excluding the $135M, Q3 provision would have been $519M against $523M of net charge-offs, a small underlying release consistent with the "improvement in the economic outlook" cited in the release.

Assessment: This is exactly the item we told readers in July to look through, and its cost to earnings came in smaller than flagged. The book's revenue contribution begins in earnest in Q4, while the Day-1 cost does not recur, so Q4 EPS should show the cleaner trajectory.

6. Loan Growth: Commercial Pauses by Choice, Card and CRE Lead

"No, I would add that the only place we have seen a little slowdown is commercial real estate and residential mortgage. Continues to be very muted. But otherwise, C and I loan demand is quite robust and broad based." — Gunjan Kedia, Chairman & CEO

Average loans grew 1.0% sequentially (about 0.8% excluding the Amazon book) and 8.0% YoY. Commercial loans were flat QoQ after several strong quarters, which the CFO attributed to runoff in NDFI balances and to deals that did not clear return hurdles, after management became "more prescriptive" with the business units about those hurdles. Credit card (+4.7% QoQ) and CRE (+3.4% QoQ) led, though management expects CRE to slow for a quarter or two given the backup in rates. Over time, residential mortgages are expected to fall from about 30% of loans toward the mid-twenties as relationship categories grow faster.

Assessment: We read the flat commercial line as discipline, not demand weakness: management chose return thresholds over volume in a quarter when capital was tight, which is the right trade-off. It does mean Q4 needs commercial growth to resume, as management said it will, or the "similar type growth" loan guide will look optimistic.

7. Deposits: Record Consumer Balances and a 1bp Funding-Cost Rise

"Deposit costs increased 7 basis points sequentially, with approximately 2 basis points of impact from the September Fed rate hike. Notably, our total funding costs are up just 1 basis point quarter over quarter as our funding mix continues to improve." — John Stern, Vice Chair & CFO

Average deposits grew 2.2% sequentially and 2.9% YoY to $526.6B, outpacing loan growth on a sequential basis, with another record quarter of consumer deposits and strong relationship deposits in investment services. Savings accounts grew 26.2% YoY while time deposits fell 19.3%, and noninterest-bearing deposits held at about $81B. Management described deposit and loan pricing after the hike as "pretty boring," with no unusual competition.

Assessment: A one-basis-point rise in total funding costs in the quarter the Fed resumed hiking is the strongest evidence yet that the consumer franchise is the low-cost funding asset management has argued it is. If that behavior holds through further hikes, the 50% starting beta management assumes could prove conservative.

8. Small Business and the Branch Strategy

"We have shared that our branch investment has increased from $200 million or thereabouts on average to $300 million, which is what we expect to do going forward. And those will be more new branches." — Gunjan Kedia, Chairman & CEO

Management spent part of the prepared remarks on small business: more than two million clients, the Amazon co-brand partnership, the Business Essentials card suite (fronted by a new celebrity ambassador), and new banker teams in expansion markets. On footprint, the CEO described a sequence: first densify branches inside the existing footprint toward 8–10% branch share in high-household-formation markets, then expand outward over time, using client centers as a fast interim way to reach national clients. On M&A, bolt-ons remain welcome, while on bank acquisitions management said it finds its organic strategies compelling and is focused on them.

Assessment: The strategy is coherent and capital-light compared with peers announcing hundreds of new branches. The stated preference for organic growth over bank acquisitions is helpful for the capital story, because it removes a common source of investor worry about how a buyback pause might end.

9. Corporate Payments and the USBDC Stablecoin Pilot

"Earlier this quarter, we launched our first USBDC stablecoin pilot. While these trends are emerging, we are uniquely positioned in corporate payments, and we are investing to create interoperability across these rails to deliver both the benefits of new technologies and the safety and soundness of the banking system." — Gunjan Kedia, Chairman & CEO

Corporate payments and treasury management together are about 6% of company revenue, and corporate payments grew about 19% YoY on new business and recovering corporate and government spend. In the quarter USB completed a live pilot of USBDC, its own dollar-backed stablecoin, for a cross-border transfer between its North American and European entities on the Stellar blockchain. Management framed the effort as building interoperability across card, ACH/wire and on-chain rails before demand arrives, citing 24/7 and cross-border settlement as the features corporate clients ask about.

Assessment: The candor is welcome: asked about demand, management said there is "not much client demand at all" yet, rather than overselling a pilot as a revenue line. The strategic logic is defensive and sound. If corporate treasury moves onto tokenized rails, the bank that already runs the card and treasury relationships should own the new rail too. We assign no value to it in our model.

10. AI and the Efficiency Ratio

"I would say the high end will be if capital markets growth turns out to be very robust, and it will be a good reason to be on a high end. that is a high efficiency ratio business. And the low end will be our AI program." — Gunjan Kedia, Chairman & CEO

Management credited productivity programs, including AI, for part of the efficiency gain (55.9%, down 130bps YoY) and said the low end of the mid-to-high-50s efficiency range depends on the AI program while the high end would come from a faster-growing capital-markets business, which runs at a higher efficiency ratio. It addressed the worry that AI agents will move cash out of bank deposits directly, saying it has seen no such activity and that the logistics are complex, while acknowledging that price transparency will rise over time.

Assessment: Management gave no dollar figure for AI savings, so we treat AI as one contributor to an expense discipline that is already visible in the numbers (expenses +2.0% excluding BTIG). The mix point matters more for modeling: as capital markets grows toward 10% of revenue, the efficiency ratio will drift up for a good reason, and investors should not read that as cost slippage.

11. Credit Under Higher Rates, and the Regulatory Calendar

"Right now, we are not seeing anything, on the consumer side. Spend levels are high. Employment is strong." — John Stern, Vice Chair & CFO

Credit metrics improved again: card net charge-offs fell to 3.29% from 3.80% a year ago while card balances grew 11.5%, and nonperforming loans fell to $1,284M. Management said commercial real estate is the area it watches as rates rise, and CRE charge-offs did tick up to 0.31% from 0.10% in Q2 on one quarter's data. On regulation, management expects rule changes in Q4 (it has estimated about a 5% risk-weighted-asset benefit from the Basel III endgame proposal) and is watching for tailoring changes, but stated that its buyback-restart plan does not count on any benefit from either, and that it is ready to move into Category II on any timetable.

Assessment: Credit remains a strength, not a risk, today. The rate path has hurt USB through capital (AOCI), not through credit losses. The regulatory calendar is a possible upside catalyst: a final rule with RWA relief or an AOCI phase-in would shorten the path back to buybacks, and management is right not to plan on it.

Guidance & Outlook

The Q4 guide asks for an acceleration in net interest income (+6–7% YoY, from +5.6% in Q3) on the full-quarter Amazon book, continued fixed-asset repricing and similar loan growth, alongside fee growth of 13–14% (roughly half from BTIG) and expense growth of ~8% (~4% excluding BTIG). Within fees, management guided mortgage banking to little or no YoY growth, merchant processing near zero for a couple of quarters, and "other" revenue back to its $120–150M normal range after Q3's equity gains. Corporate trust, capital markets and card are expected to stay strong.

Implied Q4: Applying the guided growth rates to Q4 2025 (NII $4,312M; fees $3,050M; expenses $4,227M) gives Q4 2026 NII of about $4.57–4.61B, fees of about $3.45–3.48B and expenses of about $4.57B. With provision near net charge-offs and no Day-1 charge, that supports Q4 EPS of roughly $1.40–1.45, and full-year revenue growth in the lower half of the 8–9% range.

Street at: Current Q4 EPS estimates sit in the high $1.30s. Our implied range is modestly above that, mainly because the Day-1 reserve does not recur and the margin is guided up another ~5bps.

Guidance style: Consistently conservative on operating lines in 2026. The Q3 guide was beaten on fees and expenses and met on NII, and the mid-quarter update in September flagged the upside early. The only line where the company's words have run ahead of its actions is capital return.

Analyst Q&A Highlights

What Has to Happen Before Buybacks Restart

A question asked management to list the main factors behind restarting repurchases and how it balances CET1 build against tangible-equity considerations. The answer named three variables (loan demand, rates, regulation) and left the timing between "this quarter or next."

Q: "Sounds like you will restart either this quarter or next quarter. Maybe just thoughts on kinda main factors you are thinking of as you anticipate returning to repurchase and maybe just sort of balance between common equity tier 1 build, TCE considerations, etcetera?"
— Scott Siefers, TD Cowen

A: "We had strong loan growth. We obviously had the Amazon book that added some capital demands. And then, the large rate move, which we all saw, of course. But, you know, for us, you know, given our unique position and just our prudent nature, we decided to pause. … I would just say what we are looking for this quarter is we are looking at loan demand, where does that actually, end up shaking out. … But interest rates, we are watching that very closely. And then, obviously, you have the added component of regulation."
— John Stern, Vice Chair & CFO

Assessment: The framework is honest but open-ended. Two of the three variables (rates and regulation) are outside management's control, which is why we set our own deadline rather than accept "a quarter or two" at face value. The "prudent nature" framing also signals a choice made with a margin of safety, not a forced stop.

Does a Final Basel Rule Change the AOCI Math?

A question noted that a five-year AOCI phase-in under a finalized Basel III endgame would differ greatly from the AOCI "cliff" that Category II status imposes, and asked whether finalization would change buyback timing or size. Management said the near-term decision depends more on RWA demand and rates than on the rule.

Q: "And I guess, like, do you have a sense on whether or not you know, if we do get finalization of Basel III endgame, if that would be sort of the superseding rule for you versus the quote, cliff vest on AOCI. In other words, you get to enjoy a 5-year phase-in versus the cliff vest."
— Erika Najarian, UBS

A: "But it is going to be probably more dependent on RWA demands as well as interest rates, particularly in the short run. We are very much interested in rule changes that we expect to see here in the fourth quarter. … But we are not planning on it. We are planning on growing our capital base."
— John Stern, Vice Chair & CFO

Assessment: This was the most important capital exchange on the call. Management explicitly decoupled the buyback restart from regulatory relief, which raises the bar on its own promise. If a final rule does arrive with an AOCI phase-in, it is pure upside to the capital path we model.

Why Commercial Loans Went Flat

A question asked whether flat sequential commercial balances and slightly lower utilization signaled weaker corporate demand or the bank's own capital constraints. Management blamed NDFI runoff and stricter return hurdles, and said pipelines are robust.

Q: "Wanted to ask on commercial loan balances. The average loan balances were essentially flat sequentially. And, I think utilization rates declined slightly."
— Manan Gosalia, Morgan Stanley

A: "You know, what I would say is that the underlying core C and I growth is very healthy and strong. We did, have some runoff in some balances related to more in that NDFI space. And some of this is just, new deals coming to market did not hurdle areas, and we have been more thoughtful on that of late. … We do anticipate that to increase. As we look into the fourth quarter."
— John Stern, Vice Chair & CFO

Assessment: The answer was candid but did not quantify the NDFI runoff. A bank that tightens hurdles in the quarter it pauses buybacks is rationing capital toward its best returns, which we view favorably, but Q4 commercial balances now carry a clear test.

How Rising Rates Feed the Margin

A question asked how rate hikes affect the NIM, and for updates on fixed-asset repricing and deposit-beta assumptions. Management described a deliberate shift toward asset sensitivity and gave specific repricing and beta numbers.

Q: "And just to follow-up on that, in terms of the net interest margin drivers, could you give us a little feel for how rates rising rates impacts it? Maybe an update on fixed asset reprice and your deposit beta assumptions."
— John McDonald, Truist Securities

A: "So, you know, we aim to be as neutral as we can be, from an interest rate perspective. However, given this environment, we have been shifting more into asset sensitivity, and, you are starting to see that, come through in terms of loans. … You know, we think that betas pick up right where they left off from the last rate hike, so we are assuming about a 50% or so, deposit beta to start out with."
— John Stern, Vice Chair & CFO

Assessment: The specific numbers (about $10B a quarter repricing at ~125bps higher, 50% starting beta) give investors a usable framework, and the 1bp rise in total funding cost this quarter is an encouraging first data point on betas. Repricing and asset remix help the margin as rates rise, while higher long rates hurt capital. That split is the core of the USB debate.

The Shape of Q4 Fees

A question asked why the guide implies lower fees in Q4, whether capital markets holds its Q3 level, and how much of Q3's "other" revenue was one-off equity gains. Management pointed to mortgage, normalizing other revenue and flat merchant processing.

Q: "Does capital markets hold up at the third quarter level? I know you are saying 200 million of BTIG versus $226 million And then you size the amount of equity gains that were in the other line and what that line should usually look like?"
— Ken Usdin, Autonomous Research

A: "You know, mortgages is going to be impacted by higher rates, and so we anticipate there to be more of a limited or low, year on year growth. … And so our other revenue tends to be in that a $120 million to $150 million range. … Merchant will probably be still in that 0% area for the next couple of quarters as we have talked about."
— John Stern, Vice Chair & CFO

Assessment: A clean, specific answer that helps modeling: Q3's other revenue was $166M, so $15–45M of the quarter's fee strength should not be annualized. The 13–14% Q4 fee guide is still strong after those adjustments.

Can AI Savings Actually Be Measured?

A question challenged management for citing AI as a reason for the higher operating-leverage guide, given that banks rarely quantify AI benefits. Management called AI one important lever among several and pointed to model-swapping to control token costs, but did not give a dollar figure.

Q: "So were you talking in a more general sense, or do you have numbers around it and how are you using AI and why do you think it is helping to improve your efficiency?"
— Mike Mayo, Wells Fargo

A: "So I think AI is 1, not the only reason to drive good productivity, but it is an important lever. And has growing momentum and growing use cases for productivity within our world. … We have invested in developing the capabilities to swap out models for cheaper models, and we have some confidence over time that we will continue to get positive returns from our AI program on productivity."
— Gunjan Kedia, Chairman & CEO

Assessment: Management answered in general terms, as the question anticipated. The credible evidence of productivity is in the reported numbers (expense growth of 2.0% excluding BTIG), not in the AI narrative, and we model it that way.

Is There Real Demand for a Bank Stablecoin?

A question asked whether the USBDC launch answered client demand or built capability ahead of it. Management said plainly that demand is not there yet.

Q: "Did you launch this due to strong client demand, or is this more of getting the capability up and running in anticipation of increased client demand down the road?"
— David Chiaverini, Jefferies

A: "it is certainly the latter. I very clearly have to say there is not much client demand at all. What there is now is client curiosity."
— Gunjan Kedia, Chairman & CEO

Assessment: An unusually direct answer on a topic where many banks overstate traction. It confirms that the stablecoin is optionality inside the corporate-payments franchise, not a 2027 earnings driver.

What They're NOT Saying

1. How a restart fits the ~10% AOCI-inclusive marker: In July, management tied the payout ramp to approaching a ~10% AOCI-inclusive CET1 ratio. The ratio is now 9.1%, 90bps short, yet the restart is promised within one or two quarters. Management did not reconcile the two, size the initial restart, or say whether the marker still applies.
2. An AOCI sensitivity to further rate moves: The quarter showed that higher Treasury rates and wider credit spreads (the 10-year rose 87bp) cost about 40bps of AOCI-inclusive capital. With futures pricing another hike and the 10-year around 5%, there was no guidance on how much a further 50–100bp backup would cost, or on the pull-to-par accretion that would rebuild capital if rates hold.
3. BTIG's profitability: Revenue ($226M) is disclosed; the margin is not. The expense bridge implies BTIG costs of roughly $210–214M in the quarter, and management did not say how much of that is one-time integration cost or update the ~15%-to-20% contribution margin it described in July.
4. The Amazon book's revenue run-rate: Management disclosed the loan balances ($0.9B average, $1.7B period-end) and the reserve, but not the revenue contribution in Q3 against the ~$75–85M per quarter full run-rate it described in July.
5. The size of the NDFI runoff: Management cited NDFI runoff and stricter hurdles as the reasons commercial loans went flat, but did not quantify either, which leaves investors unable to separate deliberate exits from softer demand.

Market Reaction

• Pre-print setup: USB closed at $56.26 on October 14, down 9.6% over the prior 30 days (from $62.22 on September 15) and 12.1% below its $64.01 close after the Q2 print. The stock was up only 5.4% YTD against the S&P 500's 12.8%, and up 21.1% over twelve months, within a 52-week closing range of $45.47–$65.42.
• The September break: The largest single drop came on September 16, when the Fed raised rates and USB disclosed the buyback pause at an industry conference: USB fell 4.0% that day, in line with large bank peers (PNC −3.9%, TFC −4.0%, WFC −3.0%). Over the full 30 days, PNC fell 10.4%, TFC 9.7% and WFC 9.8%, while the regional bank ETF fell 7.4%.
• Reaction day (October 15, BMO report): Shares opened at $56.14, traded a $55.97–$57.69 range and closed at $57.39, up 2.0% (+$1.13), on 12.1M shares versus a 7.7M 30-day average (1.6x). The S&P 500 rose 0.3% and the regional bank ETF rose 1.8%.
• Since the print: USB closed at $59.58 on October 23, up 5.9% from the pre-print close, versus +5.7% for the regional bank ETF and −1.2% for the S&P 500 over the same period.

Rate-Shock Beta, Not a USB Verdict: The most important fact about the stock's path is how closely it tracked its peers. USB lost almost exactly what its large bank peers lost in the month before the print, despite its own buyback-pause disclosure, and it has gained almost exactly what the regional group gained since. The market treated USB's 30-day decline as a sector rate trade (higher long yields mean AOCI losses for every bank with a securities book, and higher hurdle rates for every bank multiple), and it treated the pause as part of that trade rather than as company-specific bad news. The mid-September guide update had also moved fee and NII expectations to the high end, so the print confirmed rather than surprised.

What the reaction says about positioning: The 2.0% gain on 1.6x volume was a relief rally, not a re-rating. Investors saw the operating beat and the raised operating-leverage guide, but held back a larger move until the buyback actually restarts. That leaves the stock with a cleaner setup than in July: it now trades on a multiple that prices the pause, with a near-term catalyst (a restart) that is in management's own hands once rates stabilize.

Street Perspective

Debate: Is the Buyback Pause a Capital Problem or a Timing Blip?

Bull view: USB generates roughly 30bps of capital a quarter after dividends, regulatory CET1 is unchanged at 10.8%, and the AOCI loss pulls to par as bonds mature. The pause was a prudent response to the largest quarterly rise in the 10-year since 1994 and to high-return loan demand, and a restart within two quarters is credible. Any Basel finalization with an AOCI phase-in is upside.

Bear view: The buyback has been about to ramp for a year, and now it has stopped. The AOCI-inclusive ratio moved away from the ~10% marker management itself set, Category II will bring AOCI into regulatory capital, and the 10-year, already around 5%, could rise further. The pause could stretch well beyond two quarters.

Our take: Closer to the bull view, but with a deadline. The pause is explained by mark-to-market losses and loan demand rather than by earnings power, and management decoupled the restart from regulation, which makes its promise testable. We will hold management to the Q1 2027 print. If repurchases have not restarted by then, the bear case is right and we move to Hold.

Debate: Are Higher Rates a Tailwind or a Headwind for USB?

Bull view: The balance sheet has shifted toward asset sensitivity (floating-rate loans are 30% of earning assets), funding costs rose 1bp through a hike, fixed-rate assets reprice ~125bps higher, and the margin is guided up another 5bps toward a 3% target in 2027. Management says the remix and repricing are enough to offset deposit-beta pressure.

Bear view: Higher long rates cut tangible book, delay capital return, slow mortgage and CRE, and raise the risk of commercial-real-estate losses. Bank multiples compress when the 10-year rises, regardless of NII.

Our take: Both are true, and they act on different lines. Repricing and remix help earnings, while long rates hurt capital. For USB the earnings effect is durable while the capital effect is a timing cost that reverses as securities mature. On balance we see the current rate path as a net positive for 2027 EPS and a temporary negative for per-share capital return.

Debate: Can Fee Growth Stay in the Teens Once BTIG Is in the Base?

Bull view: Excluding BTIG, fees grew about 7.6%, with corporate payments near +19%, trust and investment management near +10% and legacy capital markets near +10%. The goal of taking capital markets to 10% of revenue implies years of above-average fee growth.

Bear view: BTIG enters the base in mid-2027, mortgage and merchant are flat-to-down, and Q3 included one-off equity gains. Fee growth will fall to mid-single digits once the acquisition annualizes.

Our take: Fee growth will slow from the teens when BTIG annualizes, and that is already in our model. The relevant number is the high-single-digit organic pace, which is above USB's own medium-term target and supports positive operating leverage in 2027.

Debate: After the De-Rating, Entry Point or Value Trap?

Bull view: The stock is 7% below its July reaction close while the earnings outlook has improved: two guide raises, a higher margin, better credit. At about 10x 2027 EPS and 2.0x tangible book for an 18.5% ROTCE franchise with a 3.6% yield, the risk/reward has improved since July.

Bear view: In a rising-rate regime, bank multiples have room to fall further, and without buybacks there is no per-share support. Street price targets have moved in both directions, a sign that the rate backdrop is outweighing the operating beat.

Our take: Entry point. The de-rating answered the Bear-1 concern we raised in July (a full valuation after the run) without any damage to the operating thesis. A mid-to-high-teens total return to our base value is available if the buyback restarts on time. That is the condition, and it is now our explicit trigger.

Model Update & Valuation Framework

Valuation: At $59.58 (October 23 close), USB trades at ~11.2x our ~$5.30 FY2026 EPS, ~10.0x our ~$5.95 FY2027 EPS and ~2.03x tangible book ($29.34), with a ~3.6% dividend yield. We set our 12-month fair-value range at $65–$71 (from $67–72), with a base of ~$68 (11.5x FY2027 EPS), a bull case of ~$76 (12.75x, on a buyback restart in Q4, a 3% margin in 2027 and stable long rates) and a bear case of ~$52 (about 9x a lower ~$5.80 FY2027 EPS, in a scenario where the 10-year keeps rising, CRE losses build and the buyback stays off through mid-2027). The base sits about 14% above the October 23 close, and adding the yield gives a mid-to-high-teens total return, comfortably above our Outperform bar.

Valuation impact: The base value barely moves. A lower FY2026 estimate and a lower tangible book are offset by rolling the multiple to 2027 earnings, which have improved with the margin and operating-leverage guide. What changed is the price: the stock fell while the earnings outlook rose, so the gap to fair value widened from high-single digits in July to about 14% now. The bear case is driven by rates and capital, not by operations.

Thesis Scorecard Post-Earnings

The Q2 Commitments, Graded

The Standing Thesis

Overall: Thesis unchanged in direction, with the risk moving from valuation to capital. Four of the five bull pillars confirmed or strengthened: fees beat, operating leverage was raised, the margin accelerated, and both 2026 deals are ahead of plan. Bear-2 (reinvestment) is effectively retired as a near-term risk. The capital-return pillar weakened again, from a flat buyback to a paused one, and the rate leg of Bear-3 moved from contained to emerging because it is the direct cause. At the same time, the July concern about a full valuation (Bear-1) corrected through a 12% de-rating that the operating results did not justify. The swing factor is now a single, dated event: the buyback restart.

Action: Maintain Outperform, conviction unchanged at moderate. The operating case is stronger than in July and the price is lower, which offsets the setback on capital return. We would add on further rate-driven weakness while credit holds. We would move to Hold if repurchases have not restarted by the Q1 2027 print, if AOCI-inclusive CET1 falls below 9% on a further rate backup, or if commercial-real-estate losses rise materially above the current 0.31%.