The Energy Scare Evaporates and ROTCE Touches 17.7%, But the Beat Is Lumpier Than It Looks: Maintaining Outperform Through a Fourth Straight Earnings Sell-Off

Rating Action: Maintaining Outperform

We upgraded Wells Fargo to Outperform at Q3 2025 when the post-asset-cap flywheel turned, and have maintained it through three prints since. Q2 2026 is the inverse of Q1's paradox. In April the reported quarter was soft while the structural case strengthened. This quarter the reported reads spectacular (EPS +25%, ROTCE 17.7%, revenue +9%, loans +12%) while the structural case advanced only modestly, because the flattering pieces of the print are precisely the pieces that do not repeat.

What decided the rating was not the headline. It was the resolution of the risk we ourselves put on the clock. In the Q1 note we wrote that a clear inflection in card and auto early-stage delinquencies under sustained energy prices “would move us toward Hold.” The opposite happened on every measurable axis: charge-offs down 10 bps year over year, consumer charge-offs improving sequentially, nonperformers down $824M, and management reporting delinquencies running ahead of model every month of the year across every cohort. Wells absorbed the scenario we were worried about and its credit book got better, not worse.

Against that, the bear case narrowed to a single durable point: core NII excluding markets is flat-to-stalling, and the earnings growth is being carried by fees, markets, venture gains, and buybacks. That is a real critique, and we do not dismiss it. But it is also, on this quarter's evidence, a mischaracterization of a deliberate trade. Wells is spending net interest margin to buy relationship depth, and the payoff showed up in the same print: markets revenue rose 24%, equities rose 64%, investment banking fees set a record at $939M, and the firm climbed from ninth to fourth in U.S. M&A league position. Marking Wells down for the margin cost while declining to credit the fee benefit is not conservatism; it is scoring one side of the ledger.

Meanwhile the stock has now fallen on four of its last five earnings days and trails the S&P by roughly 1,570 bps year-to-date, while ROTCE went from 15.2% to 17.7%, tangible book compounded 7%, and the share count shrank 6%. We are mindful that the multiple has not re-rated for four quarters despite the operating record, and we take the new late-cycle wholesale risk seriously enough to name it as a thesis pillar. Neither changes the conclusion. Maintaining Outperform.

Results vs. Consensus

On the headline, this was not a close call: a 16% EPS beat and a 3.3% revenue beat, with every operating segment growing both net interest income and noninterest income year over year, an outcome Wells has not produced in the modern era of this franchise. Pre-tax pre-provision profit, the cleanest read on operating momentum because it strips both the tax line and the reserve judgment, rose 20% to $8.96B as revenue grew 9% against 2% expense growth. The efficiency ratio improved four points to 60%.

Quality of Print

• Revenue: Genuinely broad-based beneath the venture noise. Every segment grew both NII and fees, which has not happened before in this recovery. Investment advisory fees rose 13% to $2,821M, deposit-related fees 9%, lending-related fees 9%, brokerage commissions 13%, and investment banking fees 35% to a record $939M. Strip venture gains entirely and revenue still grows roughly 6% year over year. The quality problem is concentration in one line, not breadth.
• NII: The weak spot, and the reason the stock fell. Total NII grew 5%, but markets NII contributed $501M against $104M a year ago. Excluding markets, NII grew 2% for a second straight quarter and remains below the Q4 2025 level. Interest expense grew 14% year over year while interest income grew 9%, the arithmetic of a deposit book shifting toward interest-bearing balances.
• Expenses: The cleanest line in the print. Total expense grew 2%, entirely from revenue-related and incentive compensation, which is self-funding. Non-revenue-related expenses declined outright. Headcount fell 7% year over year to 197,000, a 24th consecutive quarterly reduction and 79,000 below the level of six years ago. This is the least-discussed and most durable part of the story.
• Credit: Improving, not merely holding. The charge-off ratio fell 10 bps year over year to 34 bps, with commercial charge-offs at 10 bps (down from 24 bps sequentially) and consumer at 74 bps (from 78 bps). Nonperforming assets fell $824M to 0.77% of loans. The allowance held at 1.40% of loans, with a card/auto build offset by a lower office CRE allowance, which is itself a quiet positive on the CRE book.
• Capital and tax: The tax line deserves a note. Income tax expense of $1,402M more than doubled sequentially and rose 53% year over year, so the discrete benefit did not mask an unusually light tax quarter; the effective rate simply normalized against a much larger pre-tax base. CET1 held at 10.3% while the balance sheet grew 15%, which is capital generation doing real work.

Segment Performance

For the first time in this recovery, all four operating segments grew net interest income and noninterest income year over year. Corporate & Investment Banking carried the quarter on a 24% markets print and a 20% banking print, and its net income rose 34%, helped by a $181M provision release. Wealth & Investment Management delivered the best margin story (revenue +13%, net income +28%). The consumer franchise compounded quietly, with segment net income up 19% on a 6% revenue line.

Corporate & Investment Banking: the Markets Trade Starts Paying

CIB produced the quarter's decisive evidence. Revenue rose 16% to $5,425M with banking up 20% ($2,013M) and markets up 24% ($2,206M). Within markets, equities rose 64% to $635M and FICC rose 10% to $1,536M. Within banking, investment banking revenue rose 36% and the company-wide investment banking fee line hit a record $939M, up 35%. Average CIB loans grew 26% and deposits 16%. The segment released $181M of provision against a $103M charge a year ago.

This matters because it is the answer to the NIM critique. Wells has grown the markets balance sheet by roughly $198B since the end of 2024, about 60% of it financing, and financing is low-spread by construction. The bet was always that low-spread balance sheet buys flow, and the flow is now visible: management said financing revenue nearly doubled year over year and trading revenue grew north of 20% on the back of it, tracked client by client.

“When you start looking at the individual clients where we are deploying some of this incremental financing balance sheet to, every single one of them, all but a couple, have done significantly more business with us than they did just a year ago. And that is just getting started in terms of ramping up some of the volumes.” — Mike Santomassimo, CFO

Scharf added the strategic frame that we think is underappreciated: the margin compression is a choice, and choices are reversible.

“What we are seeing in NIM is not happening to us. What we are seeing in NIM is because of things that we are doing, and those are things that we do not have to continue to do or we can unwind at some point as well… we are either going to get paid for it or we are not going to do it.” — Charlie Scharf, CEO

The league-table movement corroborates that this is share capture rather than balance-sheet rental: year-to-date leveraged finance share of 7.2% (rank 3), equity capital markets share up 74 bps to 3.8%, and a climb from ninth to fourth among U.S. M&A advisers by announced deal volume. Wells is also first in non-agency CMBS bookrunning, first in real-estate loan syndications, and first in CRE CLOs.

Assessment: This is the most important segment in the thesis and it delivered the specific proof the bear case demanded. The financing-to-flow conversion is no longer a promise about future quarters; it is in the revenue line at +24%, with market share receipts attached. The reversibility point is the risk control: if the pull-through stops, Wells can shrink the book and the NIM recovers. That optionality is worth more than the 25 bps of margin it costs.

Consumer Banking & Lending: Compounding Under a Deliberate Profitability Drag

Segment revenue rose 6% to $10,288M and net income 19% to $2,290M, with provision flat year over year. The composition: Consumer, Small & Business Banking up 8% on higher balances and wider deposit spreads; auto up 33% ($320M) with originations up 41% and average balances up 31%; card up 2% ($1,614M); home lending down 7% ($762M) as the deliberate servicing shrink continues, though balances were roughly stable sequentially and the rate of decline keeps slowing. Primary checking accounts have now grown year over year for 13 consecutive quarters, mobile active users reached 33.7M, and Wells moved to second in the 2026 J.D. Power mobile app satisfaction study.

The card line is where a real economic decision is buried. New accounts grew 46%, but new vintages carry upfront marketing, promotional-rate, onboarding, and allowance costs that take two to three years to earn through. The 2022–2024 vintages are now accretive; the larger 2025–2026 vintages are offsetting that contribution.

“I do want to note that the rate of growth is a decision point for us. We could have higher profitability in the shorter term by reducing our growth, but we are prioritizing longer term results given the quality of the accounts we are generating.” — Charlie Scharf, CEO

Assessment: Card revenue growing 2% while accounts grow 46% looks like weakness and is actually deferred earnings. Wells is choosing to suppress near-term card profitability to build a book it will own for a decade, and it is acquiring those accounts through its own branches and website rather than third-party affiliates, which means lower acquisition cost and better credit. The honest disclosure of the tradeoff is the tell that this is managed, not stumbled into. This is a multi-year ROTCE tailwind that the current P/E does not pay for.

Wealth & Investment Management: the Quiet Margin Story

WIM revenue rose 13% to $3,892M, with NII up 17% on lower deposit pricing and higher balances, and fees up 12% on market valuations. Net income rose 28% to $537M, meaningfully faster than revenue, which is operating leverage arriving in the segment where it has been slowest. Client assets reached $2.4T (+15%), with four consecutive quarters of positive net flows. Securities-based lending average balances grew 31%. The segment launched Advisor Gateway, a GenAI-enabled desktop, against more than $1B of platform investment over recent years. Advisor attrition is at a record low and recruiting has been at or near record levels for three straight quarters. The Q3 setup is favorable, since advisory assets price at the start of the quarter and July 1 marks were above both April 1 and the prior year.

Assessment: WIM is the cleanest compounder in the firm and the one segment where the flywheel is unambiguous: better platform attracts better advisors, who bring investment assets plus banking needs, which lifts both fee and spread revenue. The 28% net income growth against 13% revenue growth is the number that matters. The July 1 pricing tailwind makes Q3 a likely repeat.

Commercial Banking: New Clients, Not Utilization

Revenue rose 6% to $3,118M with net income up 8%, on 3% NII growth and 13% fee growth (equity investments, renewable-energy tax credits, investment banking). Expenses fell 7%. Adjusting for last year's transfer of business customers to the consumer segment, average loans grew 9% and deposits 10%. Growth is coming from the roughly 20 high-density markets where Wells has been hiring coverage bankers into an underpenetrated position, and management confirmed again that revolver utilization is not the driver.

Assessment: The coiled spring from our Q1 note is still coiled. Wells is producing high-single-digit loan growth in this segment purely from new client acquisition, with the cyclical utilization lever untouched. That is the healthier way to grow, and it leaves a free option: if middle-market confidence turns, utilization adds to a base that did not depend on it. The 7% expense decline alongside 6% revenue growth is best-in-class operating leverage.

Key Topics & Management Commentary

Overall Management Tone: The most confident posture of the past five quarters, and notably self-aware about it. Management led with the venture-gain caveat rather than waiting to be asked, held guidance rather than chasing the beat, and volunteered a late-cycle risk warning that no analyst had raised. Pushback concentrated almost entirely on margin and the NII bridge, where responses were framework-level and strategic rather than numeric, leaving the ex-markets trajectory the one place the call was less than fully convincing.

1. The Energy Scare That Wasn't

Three months ago management devoted meaningful prepared-remarks time to warning that higher energy prices and geopolitical uncertainty would pressure the second-half consumer, with history suggesting households take months to absorb the cost. We tagged it as the key monitorable. This quarter the language inverted completely, and the data supported the inversion rather than merely asserting it.

“Consumers and businesses remain very strong. Consumer spending is higher, charge-offs and delinquencies are lower, and savings and investments are growing across consumer segments… The markets and US economy have absorbed macroeconomic and geopolitical uncertainty well.” — Charlie Scharf, CEO

The credit file corroborates it precisely where we said to look. Consumer charge-offs improved to 74 bps from 78 bps, auto and card charge-offs both fell sequentially, residential mortgage stayed in net recovery, and nonperforming assets dropped $824M.

“The delinquency trends are better than we model most months. Really, every month that we have seen now for all year, across each of the portfolios. We are not seeing any cohorts of clients, whether you break it by FICO or other ways to look at higher or lower income levels… change really at all.” — Mike Santomassimo, CFO

Assessment: This is the single most important continuity item in the quarter, and it resolves in the bull's favor about as cleanly as a risk can resolve. We put an explicit downgrade trigger on card and auto early-stage delinquencies; they improved. Bear #1 from the standing thesis moves from active to resolved. The credit conversation at Wells is now entirely about the wholesale book, which is a different discussion with a different answer.

2. The Venture Gain and the Honesty of the Disclosure

Wells booked $847M of net equity securities gains against $119M a year ago, or $640M after noncontrolling interest. Rather than let the ROTCE headline run, Scharf pre-empted it in his opening remarks.

“While outsized venture capital equity gains favorably affected our returns this quarter, we have said that they can be lumpy, but that we do expect strong returns from these investments over time. But more importantly, the growth and efficiency improvements that we have seen over the past several years are now broader based, and it is these trends that give us confidence in reaching our goal of a sustainable ROTCE of 17% to 18%.” — Charlie Scharf, CEO

Assessment: Management flagging its own flattering item before an analyst could is the behavior of a team that expects to be judged over years rather than quarters, and it is consistent with four quarters of refusing to date the ROTCE target rather than promise something rates could break. It also removes any argument that the quarter was dressed up. The gain is real cash economics, just not a run-rate, and Scharf drew that line himself. We take the 16.1% H1 ROTCE as the honest marker.

3. Core NII ex-Markets: the Number That Moved the Stock

The disclosure that mattered most was arithmetic. NII excluding markets was $11,816M, up 2% year over year and up 2% sequentially, and it sits below the $11,973M printed in Q4 2025. Management put a full-year ex-markets figure of roughly $48B on the record for the first time, against roughly $2B of markets NII, summing to the unchanged $50B total.

That framing creates a testable hurdle. First-half NII ex-markets was $23,431M, so the ~$48B guide implies ~$24,569M in the second half, about $12.28B per quarter. Against the Q2 run-rate of $11,816M, that is a required step-up of roughly 3.9% per quarter, and it means clearing the Q4 2025 peak by about 2.6%. For context, the equivalent second-half ramp last year was 3.5%. So the ex-markets guide asks for a slightly steeper climb than Wells delivered a year ago, in a quarter where NIM is still compressing.

The total-NII bridge is more forgiving. First-half total NII was $24,413M, so $50B implies $25,587M in the second half, a 4.8% increase versus the first half. Wells delivered a 4.6% first-half-to-second-half increase in 2025. On that basis the $50B is close to a seasonal repeat rather than a stretch, which is exactly how management framed it.

“When you look at what is going to progress for the second half of the year, it is very similar to what we saw last year in terms of the step-up as we went through each of the quarters. You do benefit from an extra day as you go into the third quarter… we still feel very good about getting to that $50 billion in total.” — Mike Santomassimo, CFO

Assessment: The $50B total is credible because it needs the same second-half shape Wells produced last year, with earning assets compounding and an extra day in Q3. The ~$48B ex-markets piece is the tighter of the two and is where the “plus or minus” will earn its keep. This is a legitimate bear point that we carry forward, but it is a growth-rate question, not an earnings-power question: NII dollars grew year over year and sequentially even with 25 bps of margin compression.

4. The NIM Trade: Deliberate, Reversible, and Nearly Done Compressing

NIM fell 4 bps sequentially to 2.43%, a marked deceleration from Q1's 13 bps decline, and is down 25 bps year over year. Management attributed it to two self-inflicted causes: growth in interest-bearing deposits, and growth in the low-spread markets financing book. They guided to modest further compression in Q3, roughly matching the 4 bps of Q2, then stabilization in Q4 as the markets balance sheet stops growing at its post-cap catch-up pace.

“We do not expect to see the markets balance sheet grow at the same pace, and so the impact that we have seen over the last few quarters moderates… we expect just a small decline potentially in the third quarter, hopefully it ends up maybe even being better than that, and then we stabilize from there.” — Mike Santomassimo, CFO

Average deposit cost rose to 1.51% from 1.43% sequentially, but sits essentially flat against 1.52% a year ago. Management was explicit that this is mix, not price competition: standard consumer rates have not moved, and there is no observed yield-seeking behavior.

Assessment: The deceleration from 13 bps to 4 bps is the tell that this compression is mechanical and self-limiting rather than competitive and open-ended. A margin falling because you are winning interest-bearing commercial deposits and financing prime clients is a different animal from a margin falling because you are paying up to hold retail money. The Q4 stabilization claim is the falsifiable part, and it is the single cleanest thing to grade next quarter.

5. Noninterest-Bearing Deposits: the Guidance Cut Nobody Headlined

The one genuine negative revision in the outlook received almost no attention. At the start of the year Wells assumed some growth in noninterest-bearing deposits; it now expects them to be relatively stable. Noninterest-bearing balances are the raw material of net interest margin, so this is a structural nick rather than a rounding change.

Management was careful to separate cause from symptom: this is not a failure of the checking-account engine, which is growing, but a rate-environment effect on where balances sit.

“It is actually not related to the checking account growth. That is actually progressing quite well… I think when you look at just the broader backdrop in terms of the rate environment, we expected a little bit more growth than we are seeing… But we are not seeing pricing pressure or client behavior drive any of the results.” — Mike Santomassimo, CFO

Assessment: Small in isolation and worth watching in aggregate, because it is the second consecutive quarter in which the deposit mix has disappointed relative to plan while the deposit franchise has performed. The offsetting argument, which we find credible, is that interest-bearing commercial and treasury-management relationships tend to drag noninterest-bearing balances along behind them with a lag. If that lag does not close by 2027, the ex-markets NII bridge gets harder.

6. ROTCE at 17.7%, and Why the Timeline Still Will Not Come

For a fourth consecutive quarter management declined to date the 17–18% target, even having just printed 17.7%. The reasoning has been consistent, and Scharf sharpened it by adding an ambition beyond the target.

“We are often asked about the timing of achieving this goal, and I know you all understand that interest rates, markets, and credit impact us, and are hard to predict, making it difficult to give a definitive answer. But assuming favorable conditions continue to exist, we remain confident that our favorable trends will allow us to achieve this goal in a reasonable time frame and then reset the bar higher for the future.” — Charlie Scharf, CEO

Pressed again in Q&A, he made the confidence trajectory itself the message: “our confidence is higher, not lower, as each quarter goes by.”

Assessment: Refusing to date a target you just hit is either excessive caution or intellectual honesty about which parts of the number you control. Given that management simultaneously disclosed the venture gain that got them there, we read it as the latter. The word doing the work in the target is “sustainable,” and on a clean basis Wells is at roughly 16%, not 17.7%. The undated timeline remains a sentiment drag and a legitimate Street frustration, but it is not evidence of weak visibility.

7. Late-Cycle Risk-Taking: an Unprompted Warning

The most interesting thirty seconds of the call came unbidden. In the middle of a quarter that beat on every line, Scharf inserted a warning about the environment producing the beat.

“Strong environments like this do not last forever, and we see large amounts of capital being deployed by both banks and nonbanks across a broad range of risk assets. Often, when times like this continue, leverage and risks develop that are sometimes hard to see… We will watch carefully for signs of outsized risks and stress and continue to deploy our resources carefully and deliberately.” — Charlie Scharf, CEO

Asked later whether competitors are underwriting loosely, he drew a sharp line between the two halves of the market.

“On the consumer side, we would say, not really… On the wholesale side, it is a very, very different story… there are more risk assets being created on the wholesale side, and there is a lot of capital out there that is there to support that. And we are doing the pieces of the transactions that we are comfortable with… and there are others that are willing to take more risks than we are.” — Charlie Scharf, CEO

Assessment: This replaces the energy-price cloud as the thesis's live risk, and it is a better-founded one because it comes from inside the underwriting room rather than from a macro forecast. Two readings coexist. The reassuring one is that a bank publicly narrating the late-cycle setup while its own charge-offs fall 10 bps is unlikely to be the one caught out. The cautionary one is that Wells is growing wholesale loans 26% in CIB into the exact environment its CEO just described as opaque. We resolve it in Wells' favor for now, on the evidence that growth is coming from investment-grade financing and new client acquisition rather than yield reach, but this is the item to grade every quarter from here.

8. The AI and Data-Center Second Derivative

The call's final exchange pushed on whether Wells has hidden exposure to an AI capex slowdown through suppliers and adjacent borrowers. Scharf's answer was the most granular underwriting commentary of the call, and it disaggregated a category the market treats as monolithic.

“There is core and shell, there is power, there are chips, and there are a whole series of things that go into the data center, and we underwrite those different pieces of those financings very differently… it is very, very different lending to a chipmaker that has 80% margins where we get paid back in a year and a half versus lending to someone else in the supply chain where it is going to take 15 years to get paid back, in the hope that the LLM provider who is renting that space is going to be there.” — Charlie Scharf, CEO

Assessment: The distinction between an eighteen-month payback against an 80% gross-margin chipmaker and a fifteen-year payback contingent on an LLM tenant still existing is the correct way to think about data-center credit, and hearing it articulated unprompted is a genuine risk-management signal. It does not size Wells' exposure, which management pointedly did not do, and that omission belongs in the unsaid column. But it establishes that the framework exists and that the duration-and-dependency risk is understood rather than aggregated into a single “data center” bucket.

9. Capital: RWA Relief Still Parked, Buyback Pace Eased

CET1 held at 10.3% inside the 10–10.5% target while average assets grew 15%, and the stress capital buffer stayed at the 2.5% floor. Wells repurchased $3.0B (37.4M shares) in Q2, down from $4B in Q1, bringing first-half buybacks to roughly $7B and total capital returned to $9.8B. The Q3 dividend rises 11% to $0.50. The share count is down 6% year over year. The estimated ~7% RWA reduction under the proposed capital rules remains intact but explicitly parked until the rules finalize.

“The capital rules might not necessarily change that CET1 minimum plus buffers, but it could certainly change what goes into the calculation relative to freeing up capital through the RWA calculation for us… we still expect our RWA to go down as a result of at least what was proposed by about 7%.” — Charlie Scharf, CEO and Mike Santomassimo, CFO

Assessment: The slower buyback is not a negative; capital went into 12% loan growth instead, which is a better use at 1.85x tangible book than shrinking the count. The RWA relief remains the single largest unpriced lever in the story, worth roughly 80 bps of CET1 capacity, and management confirmed it will not act until finalization. That is a timing question rather than a substance question, and it means the capital story has a catalyst still ahead of it rather than behind it.

10. Efficiency: 24 Straight Quarters, and a New Lever Opening

Headcount fell to 197,000, down 3,500 sequentially, 15,000 year over year, and 79,000 over six years, a 24th consecutive quarterly decline. The efficiency ratio improved four points to 60%. Critically, non-revenue-related expenses declined outright while Wells simultaneously funded branch bankers, advisors, relationship managers, investment bankers, traders, marketing, AI, and cyber. A new lever is also opening: the risk and regulatory infrastructure built over seven years of consent orders can now itself be made efficient.

“As we completed the work and moved past the consent orders that we had in place, you will see us continue to make those processes that we put in place more efficient… that will be a very methodical approach… but it is certainly part of some of the efficiency that you are seeing come through in the last couple of quarters.” — Mike Santomassimo, CFO

Assessment: Self-funding investment is the most underrated line in this story. Wells is paying for its entire growth program out of efficiency savings and still shrinking absolute non-revenue costs, which is why 9% revenue growth converts to 20% PPNR growth. The risk-and-regulatory cost base becoming a lever rather than a tax is the compounding benefit of the consent-order closure, and it arrives gradually rather than as a headline.

Guidance & Outlook

The shape of the guide tells you more than the level. Wells held $50B rather than raising it into a 16% EPS beat, which is the opposite of what a management team chasing a quarter would do, and it is the proximate reason the stock fell. Underneath, the drivers rotated in both directions: loan growth ran materially ahead of plan (a positive), the rate path flipped from cuts to a possible hike (a modest positive for ex-markets NII), and noninterest-bearing deposits came in below plan (the negative). Those offset to the same total.

Implied second-half ramp: Total NII must reach ~$25.59B in H2 against $24.41B in H1, a 4.8% increase. Wells delivered 4.6% on the same measure last year, with the same extra-day benefit in Q3 and the same earning-asset compounding. On the ex-markets line the required ramp is ~4.9% against last year's 3.5%, which is the tighter test.

Street versus the guide: Consensus had been positioned for a raise given the loan and markets momentum, and did not get one. The gap between a maintained guide and a hoped-for raise is most of the 2.7% decline. A sell-side price-target trim to $94 on margin concerns, against an aggregate Street target that has drifted up to roughly $98, captures the split: the sell-side is lowering near-term margin expectations while still carrying targets 15% above spot.

Guidance style: Consistent with the four-quarter pattern of underclaiming. This management has now declined to raise NII guidance into a beat twice, declined to date the ROTCE target four times, declined to reset the CET1 target ahead of rule finalization twice, and volunteered its own flattering item once. That is a posture, not an accident, and it is worth a premium rather than a discount over time.

Analyst Q&A Highlights

The $50B Bridge and What the Second Half Requires

The call opened on the arithmetic of the guide, specifically whether roughly 3% earning-asset growth is the load-bearing assumption. Management confirmed the shape and pointed to last year's pattern as the template, adding the Q3 day-count benefit and fixed-asset repricing as the supporting mechanics.

Q: “To get to $50 billion I think we need to assume that the average earning assets continue to grow at around this 3% pace. And given your comments about loan growth and the deposit growth, is that kind of what we need to foot forward to get there? Any other things we need to think about?”
— Ken Usdin, Autonomous Research

A: “When you look at what is going to progress for the second half of the year, it is very similar to what we saw last year in terms of the step-up as we went through each of the quarters. You do benefit from an extra day as you go into the third quarter… You get the benefit of the fixed asset turnover given where rates are. So it is not a bad assumption relative to what to expect, but we still feel very good about getting to that $50 billion in total.”
— Mike Santomassimo, CFO

Assessment: Confirmation that the guide rests on balance-sheet compounding rather than margin recovery, which is the right place for it to rest given the margin is still compressing. The “very similar to last year” framing is checkable and roughly correct on the total-NII line. It is less comfortable on the ex-markets line, where the required ramp exceeds last year's. Management was not asked to reconcile that, and did not volunteer it.

What Actually Stabilizes the Margin in the Fourth Quarter

A follow-on pressed for the mechanism behind the promised Q4 stabilization rather than the assertion. The answer identified a single decelerating variable: the markets balance sheet lapping its post-cap catch-up growth.

Q: “On that NIM stabilizing in the fourth quarter, what are the pieces that get there? Is it that one piece slows relative to the growth rate? Is it that you lap some comps? What are the helpful things underneath that can give us the confidence that stabilization happens?”
— Ken Usdin, Autonomous Research

A: “We do not expect to see the markets balance sheet to grow at the same pace. And so the impact that we have seen over the last few quarters moderates. And that is certainly part of the story as you get into the latter part of the year… we expect just a small decline potentially in the third quarter. You know, hopefully, it ends maybe even being better than that, and then we sort of stabilize from there.”
— Mike Santomassimo, CFO

Assessment: A specific, falsifiable mechanism rather than a hope, and it is internally consistent with the compression already decelerating from 13 bps to 4 bps. The dependency is that management controls the pace of the markets book, which they repeatedly stressed they do. This is the cleanest single item to grade next quarter: if Q3 compression lands near 4 bps and Q4 flattens, the margin debate ends.

Separating Structural From Cyclical Margin Pressure

The most conceptually sharp question of the call asked whether the NIM pressure is secular rather than cyclical, given two internal forces pulling opposite ways: markets growth diluting the margin while a maturing card book should lift it. The answer reframed the balance sheet as two businesses, and conceded the dilution is real while arguing it is purchased rather than suffered.

Q: “You have two strategies that are competing factors on the NIM. One is this great growth in markets, which is obviously NIM dilutive. And also strong momentum in card, which in theory could be NIM accretive, especially once the accounts mature… how should we think about whether or not we should expect more secular pressure on the NIM beyond the macro factors?”
— Erika Najarian, UBS

A: “The way we think about this is to separate out our balance sheet into a couple of different components. One is just the business that we have, which generates the majority of NII. That is very stable. And then we have these businesses we are looking to grow… it is those additional businesses that have narrower margin that is bringing down the NIM. But we are looking at it in terms of what it means in the shorter term for profit growth and for returns… over time, that should also help us grow NIM as we attract more noninterest bearing over a period of time.”
— Charlie Scharf, CEO

Assessment: The two-book framing is the correct lens and management should lead with it. The stable core generates most of the NII; the growth overlay is margin-dilutive by design and revenue-accretive in aggregate. What went unanswered is the timing of the noninterest-bearing follow-through, which is exactly the assumption that got trimmed this quarter. The framework is sound; the dated proof is missing.

Is the Markets Cross-Sell Real, or Just Balance Sheet Rental?

The central bear challenge, posed directly: the Street cannot see the low-margin balance sheet converting into fee growth, and suspects it may not. This drew the quarter's most specific rebuttal, including the admission that the sell-off was margin-driven.

Q: “The stock obviously sold off when you started talking about your margin outlook… one of the pushbacks this morning has been the ex-markets revenue growth predominantly NII driven. So the street is struggling to see the cross sell of deploying that market's balance sheet into lower NIM and then that translating into better fee growth on the markets side.”
— Ebrahim Poonawala, Bank of America

A: “You look at the financing side of the business… it is not quite a double, but it is pretty close when you look at the year on year performance in the overall financing revenue within the business. And then you saw roughly a 20 plus percent increase in the trading related revenue off the back of that… when you start looking at the individual clients where we are deploying some of this incremental financing balance sheet to, every single one of them, all but a couple, have done significantly more business with us than they did just a year ago.”
— Mike Santomassimo, CFO

Assessment: The most decision-relevant exchange on the call, and management had receipts: financing revenue nearly doubled, trading revenue up more than 20% behind it, tracked client by client, corroborated externally by 24% markets revenue growth and rising league-table share. The bear premise that the trade is unproven is now factually weaker than it was in April. What remains fair is that the proof arrives in fees while the cost arrives in NII, so anyone anchored on NII ex-markets will keep seeing only the cost.

Dating the ROTCE Target, One Quarter After Printing It

A recurring line of questioning noted that while few doubt Wells can reach the high end of the target eventually, the timing has grown more uncertain over the past six months. Management declined again, but made the refusal itself the argument.

Q: “No one doubts when they think about can Wells achieve higher end of your 17% to 18%… I think as the street is trying to digest what the execution around this growth strategy may imply, the timing of that has become a bit more uncertain over the last 6 months… how do you think about when you could achieve that target?”
— Ebrahim Poonawala, Bank of America

A: “What we do not want to do as a company is give you a definitive date and then have the interest rate environment change, the markets environment change, credit change, and then you believe that we have not actually delivered on something, because the fact is we are subject to those things… our intention is to get there and then raise the bar higher for the future. So if we did not have the kind of confidence that we can get there in a reasonable period of time, we would not be saying that… our confidence is higher, not lower, as each quarter goes by.”
— Charlie Scharf, CEO

Assessment: The fourth consecutive refusal, and the most defensible version of it yet, because it arrived in the same quarter management disclosed the venture gain that produced the 17.7%. A team willing to date the target would have declared victory this quarter. The intention to reset the bar higher afterward is new and is a quiet signal that management views 17–18% as a waypoint. The undated target remains a real sentiment cost and a reason the multiple has not moved.

The Noninterest-Bearing Deposit Downgrade

The one lowered assumption in the outlook drew a probe on cause: is this the checking franchise failing, or customers chasing yield? Management rejected both.

Q: “The change in noninterest bearing from what you were expecting earlier in the year, is that related to any developments in your checking account growth, or is it more attributable to rate seeking behavior on customers and just the rate environment?”
— John McDonald, Truist Securities

A: “It is actually not related to the checking account growth. That is actually progressing quite well… we expected a little bit more growth than we are seeing… We are seeing really good success in growing interest bearing deposits and growing other business with clients in the payments space and the treasury management space. And so those things will bring noninterest bearing deposits with them over time. It just takes a little bit longer for that stuff to get onboarded… we are not seeing pricing pressure or client behavior drive any of the results.”
— Mike Santomassimo, CFO

Assessment: The distinction matters and is credible: 13 consecutive quarters of primary checking growth is inconsistent with a franchise problem, and flat year-over-year deposit costs are inconsistent with a pricing problem. What is left is a lag argument, and lag arguments are unfalsifiable until they resolve. This is the quiet counterweight to the bull case on ex-markets NII, and it deserves more attention than it got.

Whether the NII Guide Had Quietly Shifted

A pointed challenge suggested management had softened from firm confidence in $50B at mid-quarter conferences to a hedged “plus or minus” on the call. The pushback was immediate and unusually flat.

Q: “Both you and Mike said at conferences in the second quarter, you were very confident about the $50 billion NII. And today, gone to 50 billion plus or minus. Seems like a little bit of a shift. Any color on what is driving that?”
— Vivek Juneja, JPMorgan

A: “No shift at all. The 50 billion plus or minus is exactly what we said in January and exactly what we said at the end of the first quarter. And what I said at the Morgan Stanley conference and others. And so I think no shift at all, and we are very confident. No intention to shift anything, our guidance is the same, and we feel confident about it.”
— Charlie Scharf, CEO

Assessment: The record supports management: “$50B plus or minus” is verbatim what was guided in January and reaffirmed in April. That an experienced analyst heard a downgrade anyway is the more telling datum, because it shows how primed the market is to read hedging into this particular line. When the same guide restated identically reads as a cut, sentiment rather than arithmetic is doing the work, which is precisely the setup that creates entry points.

Competitor Underwriting and Where the Cycle Is Loosening

The closing exchange asked whether rivals are reaching for risk, and if not, what evidence would show it. The answer split the market cleanly in two and put Wells on the record about where the excess is.

Q: “Are you seeing any signs of risk taking by your competitors in terms of underwriting in the commercial loan area or it could be in consumer? If not, what are you looking for as we go forward for some aggressive underwriting that could lead to issues in the next credit cycle?”
— Gerard Cassidy, RBC Capital Markets

A: “On the consumer side, we would say, not really. What we see is consistent underwriting versus the people that we compete with… I think on the wholesale side, it is a very, very different story. And that is where you see the deployment of significant amounts of capital, not just from banks and from non-banks. And there is a wide range of risk that people are taking in the lending activities… there are others that are willing to take more risks than we are.”
— Charlie Scharf, CEO

Assessment: A bank telling you which half of the market is getting sloppy, while its own charge-offs improve 10 bps, is doing the analyst's job for them. The consumer-versus-wholesale split is the single most useful macro observation on the call and inverts the risk map from three months ago, when the worry was households. The unresolved tension is that Wells grew CIB loans 26% into the environment it just described as opaque, and management was not pressed on that.

What They're NOT Saying

1. The size of Wells' own data-center and AI-adjacent exposure. Management described its underwriting framework for data-center credit in genuine detail, distinguishing core-and-shell from power from chips by payback period and credit support, and then declined to attach a single number to any of it. Having volunteered late-cycle wholesale risk as the environment's defining feature, the omission of their own quantum is conspicuous. Q1 produced a granular NDFI disclosure table under similar pressure; the data-center equivalent has not arrived.
2. A NIM excluding markets. An analyst explicitly requested it on the call, noting that Wells already discloses the pieces. Management did not respond to the request. Wells now publishes NII excluding markets but not the margin excluding markets, which is the one number that would settle whether the core spread business is healthy underneath the markets dilution. The reluctance is understandable and the disclosure gap is real.
3. Any update on the NDFI fraud loss or the European portfolio review. Last quarter's single fraud-related loss in real-estate finance prompted independent reviews and a conclusion that it was isolated. No follow-up was offered and none was requested. Silence probably means the conclusion held, but a proactive confirmation would have cost nothing.
4. A second-half venture-gain assumption. Management flagged the $847M as lumpy without indicating what a normal quarter looks like, leaving every model to guess at the run-rate. Given the gain effectively equals the revenue beat and ~1.1pp of ROTCE, the absence of a framing range is the single largest source of dispersion in second-half estimates.
5. A ROTCE timeline, for the fourth consecutive quarter. Now notable specifically because the target was printed this quarter. Declining to date a goal you just hit tells you management does not regard it as durably hit.
6. What happens to the buyback if the RWA relief lands. Management confirmed the ~7% RWA reduction still stands and confirmed they will not act until finalization, but offered no framework for what roughly 80 bps of freed CET1 would do to the repurchase pace. The largest capital lever in the story remains deliberately unmodeled.

Market Reaction

• Pre-print setup: WFC closed at $87.67 on July 13, down 5.9% year-to-date against an S&P 500 up 9.8%, but up 4.7% over the trailing 30 days and 5.1% over the trailing twelve months. The 52-week closing range entering the print was $73.42 to $96.39, leaving the stock roughly 9% below its high.
• Reaction-day move: The stock gapped down 3.2% at the open ($84.85), traded a $83.83 to $88.68 range, and closed at $85.29, down 2.7% ($2.38) on the session, against an S&P 500 up 0.4%.
• Volume: 38.9M shares against a 15.2M 30-day average, 2.6x normal and the heaviest repositioning of the four earnings-day declines.
• Read-through: The decline was an outlook event, not a print event. Management held $50B rather than raising it into a 16% beat, quantified the ex-markets line at ~$48B for the first time, guided to further Q3 margin compression, and cut the noninterest-bearing deposit assumption. The causation is on the record from the call itself, where an analyst opened by observing the stock “sold off when you started talking about your margin outlook.”

The pattern is now the story. This is the fourth earnings-day decline in five quarters (the exception being the +7.1% Q3 2025 print that prompted our upgrade), and it is the mildest of them at 2.7%, on the heaviest volume. Over the same five quarters ROTCE moved from 15.2% to 17.7%, tangible book per share compounded 7% to $46.13, the share count fell 6%, and average loans grew 12%. A market that sells a bank four times while its returns, book value, and balance sheet all compound is not reacting to results; it is expressing a persistent doubt about their composition.

That doubt is coherent. It says: your growth is fee-and-markets driven, your core spread engine is flat, and your best return quarter was purchased with venture gains you told us not to count. Every clause of that is defensible on this print. What the doubt keeps missing is the other side of the same trade. The margin Wells is spending is buying record investment banking fees, 64% equities growth, and four places of M&A league-table advancement, and management has retained the ability to reverse the spend if the pull-through stops. Selling the cost while refusing to price the benefit has been the market's posture for four quarters, and it has coincided with 21% first-half EPS growth and a 1,570 bps year-to-date underperformance. We think the gap closes toward the fundamentals rather than the sentiment.

Street Perspective

Debate: Is the earnings beat real, or is it venture gains in a trench coat?

Bull view: Strip every flattering item and the quarter still works. Clean EPS near $1.84 beats consensus by roughly 7%, revenue ex-venture-gains still grows about 6%, PPNR grows 20%, the efficiency ratio improves four points, and non-revenue expenses fall outright. Beats built on expense discipline and fee breadth across every segment do not require a gain to be impressive.

Bear view: The revenue beat is arithmetically the venture gain, the ROTCE headline is roughly 1.5 points inflated, and the tax line chipped in another $0.04. A bank that needs an $847M equity mark to touch its own return target has not reached that target. Take those away and this is a good-not-great quarter being marketed as a breakout.

Our take: The bear has the arithmetic and the bull has the conclusion. Roughly 60% of the beat does not repeat, which is exactly why the stock fell and exactly why management said so first. But the durable half of the print is the half that compounds: 20% PPNR growth, a 60% efficiency ratio, 24 straight quarters of headcount reduction, and improving credit. H1 ROTCE of 16.1% against 14.4% a year ago is a clean two-quarter number with no venture help in Q1, and it is the correct thing to extrapolate.

Debate: Does core NII ex-markets at +2% invalidate the growth story?

Bull view: No, because it measures one side of a deliberate trade. Wells is deploying low-spread financing balance sheet to win prime and flow business, and the returns land in fees rather than spread. Markets revenue up 24%, equities up 64%, record investment banking fees, and rising league-table share are the proceeds. The trade is also reversible at management's discretion, which is a genuine risk control.

Bear view: Two consecutive quarters of 2% growth in the core spread engine, at a level below the Q4 2025 peak, is a franchise treading water while accounting mix flatters the optics. The ~$48B guide now needs a step-up of nearly 4% per quarter to clear a peak set eighteen months prior, with the margin still compressing and noninterest-bearing deposits revised down. If the second-half ramp misses, the whole $50B narrative goes with it.

Our take: The bull case is stronger than the Street currently credits, but the bear case earned a real point this quarter with the noninterest-bearing revision. Anchoring on NII ex-markets while the markets book converts to fees books the cost and ignores the benefit, and the benefit was independently visible this quarter for the first time. That said, the ~$48B guide is the tighter of the two hurdles and it needs a steeper climb than Wells produced last year. We carry this as the live bear pillar, and it is a growth-rate question rather than an earnings-power question: NII grew year over year and sequentially through 25 bps of margin compression.

Debate: Is the late-cycle wholesale environment a risk to Wells or an opportunity for it?

Bull view: Wells is the disciplined actor in a loosening market, and said so on the record while its charge-offs fell 10 bps, commercial charge-offs hit 10 bps, and nonperformers dropped $824M. It is declining the parts of data-center financings it does not like and taking the parts it does. Loose competitors create share for disciplined ones when the cycle turns.

Bear view: Every bank narrates discipline at the top of a cycle. Wells grew CIB loans 26% and total average loans 12% into precisely the environment its CEO called opaque, and it released $181M of CIB provision while doing so. Pristine credit is a lagging indicator, and a bank that just told you leverage is building in places that are hard to see has told you something about its own book too.

Our take: Genuinely unresolved, and it replaces the energy-price consumer worry as the thesis's live risk. Three things tilt us toward the bull. The growth is concentrated in investment-grade financing and new client acquisition rather than yield reach; the warning was unprompted, which is not how a bank behaves when it is rationalizing its own book; and the data-center underwriting framework Scharf described is the correct one. But 26% wholesale loan growth into an acknowledged froth deserves quarterly scrutiny, and the refusal to size the AI-adjacent exposure is the gap we most want closed. This is now the item that would move us toward Hold.

Debate: Why has the multiple not moved for four quarters?

Bull view: It has not moved because the market is anchored on a regulatory-discount story that no longer exists and a margin optic that management controls. At 1.85x tangible book and ~11.7x forward EPS, against JPMorgan near 3x, WFC is priced as though the asset cap and consent orders were still in place. Returns are up 250 bps year over year, the book compounds, and the count shrinks. The re-rating is deferred, not cancelled.

Bear view: Four quarters is long enough to stop calling it a delay. The market has repeatedly examined this franchise and concluded the growth is lower-quality than the headline, the ROTCE target is undated, and the capital relief is hypothetical until the rules finalize. A discount that persists through improving fundamentals may be information rather than error.

Our take: The bear's challenge is fair and we take it seriously; a discount that survives four quarters of good results deserves an explanation better than “the market is wrong.” The explanation we find persuasive is that Wells has been converting a regulatory-constraint story into a growth story, and those re-rate on evidence rather than announcement. The evidence is now accumulating in a specific, checkable place: the margin should stop compressing in Q4, the ex-markets line should clear its old peak, and the RWA relief should land. Each is falsifiable within four quarters, which is the horizon on which we are underwriting. At 1.85x tangible book we are paid to wait, and the July 2025 initiation at 1.9x with far more uncertainty is the reminder that this stock does not stay cheap for a reason forever.

Model Implications

Valuation framing: At the $85.29 reaction-day close, WFC trades at 1.85x tangible book value of $46.13 and roughly 11.7x our raised FY2026 EPS estimate of ~$7.15–7.45. (Our Q1 note cited 1.65x tangible book; the correct multiple at that quarter's $81.70 close and $44.98 tangible book was 1.82x. The peer discount is real but narrower than we previously represented, and this note uses the filed figures.) JPMorgan trades near 3x tangible book on roughly 20% ROTCE, so a portion of the gap is earned rather than anomalous. The question is whether a bank compounding tangible book 7%, shrinking its count 6%, growing loans 12%, and printing 16.1% first-half ROTCE deserves 1.85x. We think it deserves roughly 2.05–2.25x on a year-end tangible book near $47.50, which frames fair value at ~$98–108, or about 13.5–15x our FY2026 estimate. That implies roughly 15% to 27% upside from the reaction close. We have modestly raised the range from Q1's $95–105, reflecting the resolved consumer-credit risk and the higher EPS base, tempered by the new wholesale-credit watch item.

Thesis Scorecard — Post-Earnings

Overall: The thesis strengthened, with its risk map rotated rather than reduced. The bear point we had flagged as the one that would move us toward Hold, the energy-driven consumer, resolved about as decisively as a risk can: charge-offs improved 10 bps, every cohort held, and management's macro language inverted. In its place a new and better-founded risk arrived, volunteered by the CEO rather than extracted by an analyst, concerning late-cycle wholesale underwriting and AI-adjacent credit. Meanwhile the core bull pillars each advanced: growth broadened to all four segments on both revenue lines, efficiency reached 60%, and the returns engine printed 17.7% (roughly 16.2–16.5% clean, 16.1% for the half). The one pillar that did not advance is the one that moved the stock: core NII excluding markets is stalled at +2% and now carries a quantified second-half hurdle, with a genuine negative revision to noninterest-bearing deposits underneath it.

Action: Maintaining Outperform. Wells passed the specific test we set for it last quarter and failed none of the others, while the market sold it for a fourth time in five prints. At $85.29 the stock trades at 1.85x tangible book and ~11.7x our raised ~$7.15–7.45 FY2026 estimate, having underperformed the S&P by roughly 1,570 bps year-to-date in a period when ROTCE rose 250 bps, tangible book compounded 7%, and the share count fell 6%. We see fair value of ~$98–108. Three checkable events sit inside our horizon: margin stabilization in Q4, the ex-markets line clearing its Q4 2025 peak, and RWA relief on rule finalization. The watch item has changed address: it is no longer the consumer, it is the wholesale book. Triggers that would move us to Hold: evidence that the markets balance sheet is no longer pulling through fee revenue, a second-half NII miss that breaks the $50B framing, or any sign that 26% CIB loan growth is buying the risk Scharf just warned everyone else about. None is in view today.