A $4 Trillion Quarter and 2,600 Fewer Employees: Visa Raises Both Guides but Won't Price Its Own AI Dividend

Results vs. Consensus

Vendor consensus exists for the financial lines only. The three business-driver rows have no published consensus, so they are measured against the prior quarter's reported growth rate and scored as acceleration rather than as a beat. Consensus for the top two lines is the tight cluster across four independent vendors ($3.22 to $3.23 on EPS, $11.35B to $11.40B on revenue). One vendor carried a materially higher bar ($3.29 / $11.61B), which would have made this a $0.03 beat rather than a $0.09 one. We use the widely cited cluster and flag the range.

Year-over-Year Comparison

Prior-year dollars for the five revenue-component lines are shown as n/a because the release discloses those components for the current quarter only, with growth stated as a rounded percentage. Every other figure in the table is as printed in the current and prior-year statements of operations. Personnel expense excluding severance removes the $563M charge disclosed in the non-GAAP reconciliation; the prior-year quarter carried no severance add-back.

Sequential Comparison

The sequential table is where the quarter's tension is most visible. Revenue rose 3.6%, and non-GAAP EPS was essentially flat at $3.32 against $3.31. GAAP operating income fell 4.9% and GAAP net income fell 6.5%, both because the severance charge landed inside personnel expense. Sequential comparisons of GAAP profit for this quarter carry little signal about the underlying business; the $563M is the whole story.

Revenue: the beat is broad, the yield is not

Every revenue line except one grew at or above the pace of its underlying driver. Service revenue grew 14% against the 9% constant-dollar payments volume growth from the prior quarter on which it is recognized, with pricing and card benefits bridging the gap. Data processing grew 17% against 10% processed-transaction growth on pricing, value-added services performance, and a richer cross-border transaction mix. Other revenue grew 45% on advisory and marketing services plus pricing. Pricing appears as a named contributor in three of the four lines, which is the H2 pricing benefit management flagged a quarter ago arriving on schedule.

The exception matters. International transaction revenue grew 6% while the cross-border volume that drives it grew 12% in constant dollars. A year ago that relationship ran the other way.

"International transaction revenue was up 6%. Below the 12% increase in constant dollar cross border volume growth excluding intra Europe, primarily due to lapping the currency volatility peak last year and mix." — Chris Suh, CFO

Assessment: The volatility component is arithmetic and will anniversary out; last year's Q3 was the volatility peak, and lapping it mechanically compresses this year's Q3. The mix component is the one to watch, because Visa Direct carries a structurally lower yield than card transactions and Visa Direct is growing faster than the network. A revenue line that converts high-value cross-border volume at a declining rate is a slow leak in the highest-margin part of the model, and management has now cited the same two factors for three consecutive quarters.

Margins: the compression is deliberate, the payback is undisclosed

Non-GAAP operating margin of 66.7% is down 83bps Y/Y, and management was explicit that the overspend was not planned.

"This was above our expectations, primarily due to larger than expected FX impact from balance sheet remeasurement and higher than expected personnel expense." — Chris Suh, CFO

Underneath the aggregate, the composition is instructive. Marketing expense grew 54% Y/Y to $649M, which is the World Cup showing up as cost before it shows up as revenue. Personnel expense grew 40.5% as reported, but excluding the $563M severance it grew 8.3%, roughly in line with revenue. General and administrative grew 4.4%. This is not a company losing cost discipline broadly; it is a company spending hard against one quadrennial marketing event while simultaneously taking a restructuring charge.

Assessment: The 83bps of compression is defensible for a quarter that carried peak World Cup activation spend and a workforce action. What is not yet defensible is the absence of any forward margin math. A company that removes 7% of its workforce is entitled to be asked what that does to the expense base, and Visa has declined to answer.

EPS: clean at the headline, thinner underneath

Non-GAAP EPS of $3.32 grew 11.4% and cleared both consensus and, more meaningfully, management's own mid-to-high-single-digit guide for the quarter. GAAP EPS of $2.97 grew 10.4% despite absorbing $563M of severance and $237M of litigation provision, helped by an easier prior-year comparison that carried $615M of litigation. The non-GAAP tax rate of 18.4% was 110bps above last year and in line with guidance.

Assessment: The beat against the company's own guide is the more relevant fact than the beat against the Street, and it was substantial: roughly three to four points of EPS growth above what management told investors to expect in April. Visa guided the trough quarter and then materially outran it, which is the pattern that has defined this management team's cadence for two years. The caveat is that the outperformance came disproportionately from revenue rather than cost, since opex ran roughly two points hot on a like-for-like basis.

Revenue Composition

Read down the "Read" column and the shape of Visa's revenue model in FY26 becomes clear. Two lines, service revenue and data processing, together 67% of gross revenue before incentives, are being lifted five to seven points above their volume drivers by pricing. One line, international transaction revenue at 24% of gross revenue, is being dragged six points below its driver by a comparison effect and a mix shift. And the fastest-growing line by far is the one most exposed to a tournament that will not recur until 2030.

Client incentives deserve their own note because they behaved exactly as management said they would three months ago.

"Client incentives grew 18%, a step up of 4 points from Q2 primarily due to lapping low incentive growth last year and strong client performance." — Chris Suh, CFO

Assessment: Last quarter management flagged that Q3 would see incentive growth step up as the company lapped the prior year's trough, and it did, by precisely the magnitude implied. Incentives are now guided slightly higher again in Q4 on renewals plus new business won, with roughly 20% of payments volume expected to have been renewed by fiscal year end. A renewal cycle that heavy is a near-term revenue headwind and a medium-term moat reinforcement, and it is being executed while reported growth still accelerates. Credit management for calling this one accurately a quarter in advance.

Growth Engine Performance

Consumer Payments: the best US quarter in seven years, assembled from parts

US payments volume grew 10% Y/Y in constant dollars, up about 2 points from Q2, which management characterized as a rate not seen since fiscal 2019 excluding the post-COVID recovery period. Both card-present and card-not-present accelerated. Credit rose 11%, up more than a point from Q2; debit accelerated more than 2 points to 9%. International payments volume grew 10%, consistent with recent quarters. Total payments volume crossed $4 trillion in a quarter for the first time in the company's history.

The spend-band data continues to reject the consumer-bifurcation narrative that drove the Hold ratings on this name in FY25.

"Growth across consumer spend bands saw incremental improvement from Q2, with the highest spend band continuing to grow the fastest. Across our volume, both discretionary and nondiscretionary spend remains strong. We do not see signs of the lower spend consumer weakening in our volumes." — Chris Suh, CFO

That is now the third consecutive quarter with the same message, and the underlying data has improved each time rather than deteriorated. It is a genuine rebuttal, not a talking point.

Assessment: The engine is healthy and the consumer signal is clean, but the acceleration is not fully repeatable and the company effectively said so by publishing July. Quarter-to-date through July 21, US payments volume ran +9% with credit and debit both at +9%, and processed transactions at +9%. Management attributed the step-down to retail and promotional-event timing, the loss of a favorable day-mix that helped June, and the change in fuel cost. Model the underlying US run-rate at 8% to 9%, not 10%.

Commercial & Money Movement: volume accelerated while revenue decelerated

CMS revenue grew 17% in constant dollars, a seven-point deceleration from Q2's 24%. Management attributed the step-down to the absence of the performance-adjustment benefits that flattered Q2 and to lapping pricing impacts that began in FY25 Q3. Both are comparison effects rather than demand effects, and the volume data supports that reading: commercial payments volume grew 13% in constant dollars, accelerating 2 points from Q2, with a point of acceleration from each of the US and international. Visa Direct transactions reached 4 billion, up 21%.

Management was unusually direct that the volume strength is not attributable to a handful of large wins, telling investors that while a few client wins have contributed since FY25 Q4, the bulk of the strength comes from the underlying domestic and cross-border portfolios and is expected to continue well into the future. New wins in the quarter spanned a fleet agreement with a European payments platform, an inaugural B2B travel portfolio with a major Saudi bank, and a multi-country renewal with a fast-growing B2B travel issuer.

Assessment: Read the volume, not the revenue, this quarter. Commercial payments volume accelerating 2 points to 13% while overall payments volume runs 10% means the commercial mix is still enriching, which is what the thesis requires. The revenue deceleration to 17% is the honest normalization of a Q2 print that carried one-time benefits management flagged at the time. Visa Direct decelerating from 23% to 21% is worth watching but is at a scale (4 billion transactions) where some deceleration is arithmetic.

Value-Added Services: the acceleration is real, the attribution is muddier

VAS revenue grew 34% in constant dollars to $3.8B, roughly a third of net revenue, accelerating seven points from Q2's 27%. Management named three drivers: underlying business drivers including marketing-services engagements related to the World Cup, pricing, and the Pismo acquisition. Two of those three are not organic and one is not recurring.

What is durable sits one level down, and management gave the right disclosure to see it.

"Looking collectively at issuing solutions, acceptance solutions, and risk and security solutions, the revenue has grown more than 20% year over year every quarter over the last 12 months in constant dollars." — Chris Suh, CFO

That is the number to carry into a model. Three of the four VAS portfolios, excluding the advisory-and-other portfolio where marketing services and the World Cup live, have compounded above 20% for four straight quarters, and management stated that all four portfolios have individually grown faster over the trailing twelve months than the historical rates disclosed at the February 2025 Investor Day. Product evidence supports it: subscription manager and stop-payment services now carry 2 billion enrolled credentials; Unified Checkout, launched globally in March, has been enabled by more than 4,500 sellers and acquirers; and the advisory business delivered 1,200 consulting projects for over 700 clients across more than 100 countries in the quarter, which management said exceeds what it delivered in all of 2019.

Assessment: Underwrite VAS at low-to-mid 20s, not 34%. The 20%-plus growth across issuing, acceptance, and risk-and-security is the structural rate, it is transaction-and-credential-linked so it compounds mechanically with the network, and it has now held for four consecutive quarters. The advisory-and-other portfolio is genuinely the fastest grower and genuinely the most cyclical, and this year it is carrying a World Cup. When FY27 laps this, VAS growth will optically decelerate several points on no change in the underlying business, and that will be a buying opportunity rather than a thesis break if the other three portfolios hold their 20%-plus.

Key Business Drivers

The July column is the most useful disclosure on the page, and it splits in two directions. Domestic metrics decelerated: US payments volume from 10% to 9%, processed transactions from 10% to 9%. Cross-border metrics accelerated above the full-quarter average, with total cross-border at 14%, e-commerce at 18%, and travel at 12%. Management cautioned that June and July both look unusually high on the cross-border e-commerce line and said it expects that line to settle back toward a more typical relationship with travel.

The World Cup Quarter

Visa is the official payment technology partner for FIFA tournaments, and it extended that agreement during the quarter. The tournament ran from June 11 through the round of 32 on June 30, meaning FY26 Q3 captured roughly three weeks of it and FY26 Q4 will capture the knockout stages. Management gave more granular event data than it usually does, which is itself a signal about how much of the quarter it explains.

The commercial side matters more than the volume side. Over the trailing twelve months Visa delivered more than 300 World Cup engagements to more than 140 unique clients across 70 markets, and about 20% of those clients were using Visa's marketing services for the first time. That first-time-user statistic is the one with a tail: a client that buys marketing services for a tournament and sees a return is a candidate to buy them again for something else.

Two client outcomes were disclosed with enough specificity to be checkable. A leading Brazilian bank ran campaigns from February to June that drew a million cardholder participants, produced an 8% lift in card activation, and generated $400M of incremental payments volume. In Mexico, a new tournament-branded card drove nearly 400,000 new debit cards and more than 250,000 new credit cards for one issuer, and five joint campaigns tied to national-team match days produced more than a 5x increase in average transaction size.

Assessment: This is the clearest demonstration yet that Visa's marketing services line is a revenue business rather than a brand-cost line, which is the argument management made last quarter and has now evidenced twice. The honest accounting is that the tournament is inflating three things simultaneously: US and cross-border volume in Q3 and Q4, VAS revenue through marketing services, and marketing expense (up 54% Y/Y). The revenue and the cost both normalize in FY27. The residual asset is the 20% of engaged clients who had never bought these services before, and the extended FIFA partnership that keeps the asset available.

Key Topics & Management Commentary

Overall Management Tone: Confident on execution and conspicuously non-committal on economics. Prepared remarks were the most product-dense of the past four quarters, running through partnerships, launches, and client wins at a pace that left little room for framing, and the operational metrics offered were specific and checkable. But on the three forward questions that would let an investor convert this quarter into an FY27 model, the cost savings from a 7% workforce reduction, whether the medium-term revenue algorithm is now stale, and how much of the volume acceleration was event-driven, management acknowledged each question and declined to answer any of them. The posture on the restructuring in particular was notably light: the largest workforce action in the company's recent history was delivered as one sentence inside a discussion of engineering productivity.

Topic 1: A 7% Workforce Reduction, Announced as a Footnote to an AI Story

The most consequential disclosure of the quarter received the least airtime. In prepared remarks, after several minutes describing how AI has changed Visa's internal engineering practice, McInerney delivered the action in a single sentence.

"Today, we announced that we are eliminating roles with the majority being in our technology and product teams, to ensure that we are continuing to position Visa for future growth." — Ryan McInerney, CEO

Neither the headcount nor the percentage was stated on the call. Both are confirmed elsewhere: approximately 2,600 roles, equal to 7% of the workforce. The $563M severance charge appears in the 8-K non-GAAP reconciliation, where it is the single largest add-back of the quarter, larger than the $237M litigation provision and roughly seven times the amortization and acquisition-cost line. For scale, Visa's entire severance charge across the first nine months of FY25 was $213M.

The framing is the AI restructuring, and management supplied real numbers for it.

"As a result of the unlocks we can realize with this new tooling, we are reforming our product development teams that used to be 10 or more into smaller more nimble Agentic squads of 2 to 4." — Ryan McInerney, CEO

The claimed productivity gains for teams on the agentic tool chain: 80% more code commits, an 80%-plus improvement in requirement definition (from 30 days to 5), and 65%-plus faster feature development. Alongside it, Visa says it now runs more than 150 AI-powered applications and shipped more than 300 major product releases over the trailing twelve months. The savings are earmarked for reinvestment in consumer and commercial payments, blockchain and stablecoin technology, digital wallet technology, payment orchestration, and agentic commerce applications.

Assessment: Visa is the first payments network of this scale to explicitly restructure its engineering organization around AI coding agents and to size the action in the thousands. If the productivity claims are even directionally right, this is a structural change to the cost base of the single largest expense line, and personnel is a bigger lever at Visa than at most networks because the company has been growing headcount to build the VAS franchise. The problem is that management has given investors the charge without the benefit: no annualized savings figure, no expense-base baseline for FY27, no split between what is reinvested and what is retained. Until that arrives, the restructuring is a $563M certainty against an unquantified maybe, and no responsible model can capitalize it.

Topic 2: The US Acceleration and What Is Left After the One-Offs

The +10% US payments volume print is the strongest domestic quarter Visa has produced since fiscal 2019 outside the pandemic rebound, and it was the most-probed item in Q&A. Management named five contributors: higher tax refunds, the cost of fuel, retail including promotional-shopping-event timing, Visa Direct strength, and World Cup spend. It declined to size any of them individually or in aggregate, describing the strength instead as broad-based across credit and debit, discretionary and non-discretionary, card-present and card-not-present, and across spend bands.

Both readings are defensible, and the July data is what arbitrates. US payments volume at +9% quarter-to-date through July 21, with credit and debit converged at +9%, sits one point below the Q3 average, and management attributed the step-down to the reversal of the same retail-timing and day-mix effects plus fuel cost. That is an internally consistent story: the one-offs added roughly a point, and the underlying rate is 9%, itself up from the 8% Visa printed in Q2.

Assessment: The correct conclusion is less exciting than +10% and more encouraging than "it was all the World Cup." Underlying US volume improved roughly a point sequentially to about 9%, and event effects added another point on top. A point of genuine sequential acceleration in a $4-trillion-a-quarter network is a good outcome. Investors extrapolating 10% into FY27, however, are extrapolating tax-refund timing and a tournament.

Topic 3: The International Transaction Revenue Gap

International transaction revenue is the highest-yield line in Visa's P&L and it grew 6% while its driver grew 12%. Management has now attributed the gap to the same two factors for three consecutive quarters: lapping last year's currency-volatility peak, and mix, specifically that yields vary by client, product, and region, and that products like Visa Direct carry structurally lower yields than card transactions.

The volatility factor is finite. FY25 Q3 was the peak volatility quarter, so this is the hardest comparison of the cycle, and it eases from here. But management then guided Q4 volatility to Q1 levels and described that as implying more of a drag than was previously incorporated, which means the tailwind investors might have expected from an easier comparison is being partly offset by a lower absolute volatility environment.

Assessment: The mix factor is the durable one and it cuts against a piece of the bull case. Visa Direct is a strategic priority, is growing 21%, and dilutes the blended cross-border yield as it scales. That is an acceptable trade if Visa Direct's volume growth outruns its yield discount, which at 4 billion transactions it currently does. But it means cross-border volume growth is no longer a clean read-through to international transaction revenue, and anyone modeling that line off volume will be too high. Model international transaction revenue in high single digits for FY27, not low double digits.

Topic 4: Value-Added Services at a Third of Revenue

VAS reached $3.8B and roughly a third of net revenue, growing 34% in constant dollars. Management disclosed that all four VAS portfolios individually grew faster over the trailing twelve months than the historical growth rates published at the February 2025 Investor Day, and that the three non-advisory portfolios collectively exceeded 20% growth in every one of the last four quarters. The advisory-and-other portfolio remains the fastest-growing and management expects that to continue.

The product cadence behind it was heavily detailed: an AI financial assistant that banks can white-label using their own data plus Visa's network data; DPS full-service credit, an integrated debit-and-credit issuer-processing solution combining DPS and Pismo, piloting in Q4 with a first US client secured and generally available next year; Unified Checkout, live globally since March and enabled by more than 4,500 sellers and acquirers; and the Visa Vulnerability Agentic Harness, a security orchestration layer published to clients on GitHub.

Assessment: The strategic case is now observable rather than aspirational, which is a change from a year ago. A third of revenue growing north of 20% structurally, linked to transactions and credentials so that it compounds with the network rather than independently of it, is the strongest argument for Visa carrying a premium multiple to a pure toll-road network. The discipline required is to underwrite the 20%-plus, not the 34%, because the difference between them is a tournament and an acquisition.

Topic 5: Stablecoins Move from Settlement Rail to Issuance Stack

Visa's stablecoin posture escalated materially this quarter. Last quarter the company was a settlement participant and a validator on two chains. This quarter it joined Open Standard, a consortium that plans to issue a new stablecoin, OpenUSD, designed for global money movement, and it launched the Visa Stablecoin Platform for minting, movement, and management, designed to let partners settle with Visa in stablecoins, provide on-chain wallet-as-a-service infrastructure, and move money between fiat and stablecoins. That platform will be integrated with Pismo, which management said can enable tokenized deposits for financial institutions.

Asked whether OpenUSD is a competitive attack on the incumbent stablecoins, McInerney declined the frame.

"our role is not to pick winners. Our role is to help clients connect to the stable coin ecosystem securely and at scale." — Ryan McInerney, CEO

He also delivered the most candid line of the call on the category's actual state, noting that stablecoins have yet to scale beyond a few use cases, principally the stablecoin-linked cards Visa has issued in various markets.

Assessment: Moving from settling in stablecoins to operating the minting and orchestration infrastructure, and to helping issue one, is a genuine escalation in commitment and capital at risk. The multi-coin, multi-chain neutrality is the right competitive posture for a network whose value is universality. The honest caveat is management's own: this is still pre-scale. The strategic value here is optionality and disintermediation insurance, and it should be valued as such rather than modeled as revenue.

Topic 6: Agentic Commerce Gets Two Very Large Partners

A quarter ago Visa's agentic-commerce evidence was a command-line proof of concept and a network protocol. This quarter it is partnerships with OpenAI, under which Visa provides its network, credentialing capabilities, and security infrastructure to support agentic commerce experiences, and with Meta, enabling new ways to pay across Facebook and Instagram powered by Visa Intelligent Commerce using Visa tokens. Supporting infrastructure shipped alongside: an agent score, an agent directory, and a token assurance framework intended to make agent-initiated transactions transparent and trusted.

McInerney's framework is that this follows the adoption curve of every prior payments cycle, standards first, then early adoption, then scale, and that trust is the rate limiter.

"Agentic commerce is a when not an if. We are building, like, the products, the services, the protocols, ensuring that the ecosystem has what it takes, and this will happen. And it will be a positive it will be a positive tailwind for Visa. Once we get through those steps of the journey." — Ryan McInerney, CEO

Assessment: Landing the two most-trafficked consumer AI and social surfaces in the same quarter substantially de-risks the distribution question, which was the sharpest open item on this topic three months ago. The unresolved question is unchanged and remains the right one to track: whether agent-mediated transactions carry the same economics as human-initiated ones at scale. Management has consistently asserted similar economics and has never disclosed a number. Watch international transaction revenue yield and the incentive line for the first evidence either way, and treat the current absence of disclosure as neither confirmation nor refutation.

Topic 7: Pismo and DPS, With the Large-Bank Ambition Clarified Downward

The most useful strategic clarification of the call came on the issuer-processing and core-banking stack, and it was a narrowing. Visa is combining DPS and Pismo into an integrated debit-and-credit issuer-processing solution aimed explicitly at small banks, mid-sized banks, and fintechs, piloting in Q4 with a first US client. Asked whether the target is large banks or smaller ones, McInerney separated the two products cleanly and ruled the large-issuer processing market out.

"it is not a market need we have seen from large issuers." — Ryan McInerney, CEO

Pismo separately continues to serve banks of all sizes for core-banking cloud migration, which is the lane the recently announced large US bank agreement sits in. Outside the US, Pismo is the single go-to-market platform and has now entered 19 new markets since acquisition, up from 15 a quarter ago.

Assessment: This is a more precise strategy than the one implied last quarter, and precision is worth more than ambition here. But it is also a walk-back of the most exciting version of the Pismo story. Three months ago management teased more large-financial-institution core-banking wins to come; this quarter no new large-FI win was named, and the announced product expansion targets the small-and-mid-market. That migration remains the proof point, and it remains a multi-year program that has not yet gone live. Treat additional large-FI wins as upside, not as base case.

Topic 8: Europe and the Credential Engine

Europe continues to be the geography where the payments-nationalism bear case is supposed to bite, and continues not to. Visa grew credentials in Europe by more than 40 million over the last twelve months, which management said is over 70% faster than the annualized credential growth the region produced from FY2019 to FY2024, and expects more than 30 million additional credentials from wins already secured. The quarter's marquee win was the entire consumer credit portfolio of a major UK retail bank. Global credentials grew 8% and tokenized penetration is approaching 60% of e-commerce transactions worldwide.

Elsewhere, Visa renewed a 55-year relationship with a leading Brazilian bank across consumer and commercial credit and debit plus several value-added services; signed a four-bank group in Colombia for domestic processing, cross-border Visa Direct, and small-business commercial card issuance, in a market where Visa's processing penetration has gone from single digits five years ago to above 90%; and won a US community bank's consumer debit portfolio. Client satisfaction was quantified as well.

"For the third consecutive year, we received a score of 76 in our annual global client engagement survey an enviable number in any industry." — Ryan McInerney, CEO

Assessment: The Colombia data point is the most instructive on the page: single-digit to 90%-plus processing penetration in five years is what winning a market looks like, and it happened in a region where local-rail alternatives are actively promoted. Europe adding 40 million credentials at a rate 70% above its own recent history, with 30 million more already contracted, makes the local-scheme displacement thesis hard to sustain on current evidence. This bear point remains permanently live and permanently unrealized.

Topic 9: Capital Return Slowed as the Price Rose

Visa repurchased approximately 14.5 million class A shares at an average $330.71 for $4.9B, and paid $1.3B in dividends, for $6.2B returned. It also deposited $250M into the litigation escrow, which reduces the as-converted class B share counts at a volume-weighted average price of $333.42 and has the same economic effect on EPS as a buyback. The board declared a quarterly dividend of $0.670 per class A share. Remaining authorization stands at $28.4B. In July the company expanded its commercial paper capacity to $7B.

The sequential comparison is the interesting part. In Q2, Visa bought $7.9B of stock at an average $320.66, the largest quarterly repurchase in its history. In Q3, with the average price 3.1% higher at $330.71, it bought $4.9B, a 38% reduction. Including the escrow deposit, buyback-equivalent capital deployed fell from roughly $7.9B to roughly $5.15B.

Assessment: Management is being price-sensitive with the buyback, which is exactly what shareholders should want and exactly what most managements do not do. It is also information. The team that leaned in hardest in the company's history at $320 stepped back by more than a third at $330. That is not a bearish signal about the business, but it is a data point about where insiders see value, and it deserves weight from anyone adding at $366 or above. Note also that nine-month operating cash flow of $16,342M is down 2.8% Y/Y while nine-month buybacks of $16,430M are up 22.7%, a gap funded partly by the balance sheet.

Guidance & Outlook

All guidance is on an adjusted basis, defined as non-GAAP results in constant dollars excluding acquisition impacts.

How the guide changed, in plain terms. Both full-year lines moved up. Revenue growth went from a range whose bottom half was low-double-digit to a point estimate at the low end of low teens, and EPS growth went from low teens to the low end of mid teens, which is a raise of roughly one to three points depending on where inside each band you sit. That is the second consecutive quarter Visa has raised the full-year guide, and it did so while absorbing a heavier incentive step-up, a lower assumed volatility environment, and an opex quarter that ran roughly two points above plan on the guide's adjusted basis.

Implied Q4 shape. Q4 revenue growth is guided similar to Q3 on an adjusted basis, and Q4 adjusted EPS growth is guided to the low end of mid teens against roughly 10% adjusted in Q3. So EPS accelerates while revenue holds flat, despite non-operating expense more than doubling to ~$80M and the tax rate rising to ~19%. The arithmetic requires operating leverage: Q4 opex growth guided to low double digits against Q3's actual 17%. Some of that is the marketing that shifted out of Q3 landing in Q4, which cuts the other way, so the underlying implication is a meaningfully lighter cost quarter than Q3 delivered.

Street position. Published targets clustered near $400 entering the print, with three separate raises in the three weeks beforehand. Against the post-print close of $368.73, that cluster implies roughly 8.5% of upside, which is a Street that likes the business and does not think the stock is cheap.

Guidance style. Consistent with the pattern: guide the near quarter conservatively, beat it, raise the year. Q3 was guided to mid-to-high-single-digit EPS growth and delivered 11%. The Q4 guide of low-end-of-mid-teens adjusted EPS growth is the most aggressive quarterly EPS guide this management team has given in the current cycle, which is either genuine confidence or the arithmetic consequence of a full-year raise. Given the cost-shift disclosure, we read it as the latter with a modest cushion.

On FY27. Management declined to guide and framed the process rather than the outcome.

"We have clear line of sight into our expected renewals. Product pipeline, and the expected pricing impact across our solutions which, as you know, has tended to be similar in its contribution the past few years." — Chris Suh, CFO

The load-bearing clause is that pricing contribution has been similar for the past few years and is expected to be similar again. Pricing was a named driver in three of four revenue lines this quarter. If that repeats in FY27, the pricing lever alone supports several points of growth before any volume contribution.

Analyst Q&A Highlights

Separating event-driven US spend from the underlying trend

The dominant line of questioning on the call was how much of the US acceleration is repeatable. The question was put directly, naming the specific one-offs management had itself disclosed and asking for the split between event-driven and underlying. Management restated the list without sizing any component, and pivoted to the breadth of the strength across credit and debit, discretionary and non-discretionary, and spend bands.

Q: "On the U.S. side, it looks like volume up 10%. You called out that is the highest in quite a bit of time here. So can you do you mind just going back and maybe unpacking the growth a little bit more across the items you called out. I heard tax refunds, fuel prices, Visa Direct. I think there is some timing and World Cup, of course. Just trying to separate the event driven spend versus the strong underlying trends and maybe some wins in there as well."
— Tien-Tsin Huang, JPMorgan

A: "There was a number of factors that contributed to that acceleration that we saw in Q3. Go through the whole list. It was higher tax refunds. The cost of fuel, retail, which included the timing of some promotional shopping events, strengthened Visa Direct, excuse me, and, of course, the enthusiasm around FIFA that we have seen. Those are all contributors."
— Chris Suh, CFO

Assessment: A list is not a decomposition. Management had the numbers and chose not to share them, which is a defensible choice for a company that does not want investors anchoring on a normalized figure it would then have to defend. The July disclosure supplies the answer anyway, and it is roughly one point of event effect on top of a nine-point underlying rate.

Whether the medium-term revenue algorithm still applies

The sharpest question of the call laid out the Investor Day framework explicitly, 9% to 11% net revenue growth built from 15% to 17% combined commercial and value-added services and 5% to 7% consumer payments, and observed that actuals have run well above it for roughly two years. It asked, carefully and without requesting formal guidance, whether the framework should now be considered stale. Management did not engage with the framework at all.

Q: "It does now appear for, probably, the best of 2 years that has been quite different with CMS and VAS growing probably close to mid twenties. And consumer payments probably growing more like low-double, low-single digit. And, obviously, the 13%-14% revenue growth well ahead of 9% to 11%."
— Fahed Kunwar, Rothschild & Co Redburn

A: "We are focused on running the company, We are focused on executing our strategy, driving product innovation, delivering for our clients, and taking all the steps I talked about to continue to run the company better and position us for sustained long term growth that is our focus right now."
— Chris Suh, CFO

Assessment: The cleanest non-answer of the quarter, and the most consequential. A company outrunning its own published medium-term framework by 300 to 400 basis points for two years has exactly two honest options: raise the framework or explain why the outperformance is temporary. Declining both preserves the option to beat a low bar, but it also means investors have no sanctioned basis for FY27 and beyond. The next Investor Day is where this gets resolved.

Where the restructuring savings go

The workforce reduction drew one question, and it was the right one: reinvest all of it, or let some reach the bottom line, and over what period does the return arrive. The response ran long on the breadth of available investment opportunities across consumer payments, value-added services, commercial, stablecoins, agentic commerce and the brand, described the efficiency-and-reinvestment process as continual, and never addressed the drop-through. The financial addendum committed to a level of margin, not a direction.

Q: "I wanted to just touch on the workforce reductions and the associated cost savings. I am just curious is the expectation to reinvest all of those savings? Or actually sort of to drop some of those down to the bottom line? And then as we think about the reinvestment of those savings over what period of time do you guys anticipate a return?"
— Sanjay Sakhrani, KBW

A: "we have grown and diversified our business. And we have done that while maintaining our industry leading operating margins. Looking forward, we do expect that we will continue to be able to deliver strong margins into the future as well."
— Chris Suh, CFO

Assessment: "Maintaining" and "strong" are the operative words, and neither is "expanding." The most natural reading is that essentially all of the savings are earmarked for reinvestment and margin stays roughly flat. If that is the plan it is a reasonable one for a business with this many growth vectors, but investors should stop modeling the 7% reduction as a margin event and start modeling it as a funding mechanism for the product roadmap.

Why cross-border revenue is lagging cross-border volume

An early question pressed on the six-point gap between international transaction revenue growth and the cross-border volume that drives it. The answer was the most complete technical explanation management gave on the call, separating a finite comparison effect from a structural mix effect.

Q: "I wanted to ask about the delta between cross border and international revenue, the volume versus revenue mix. I think you alluded to some of this Christopher, in your prepared remarks, but if you could elaborate on that. It would be helpful."
— James Friedman, Susquehanna

A: "The difference that you point out between international transaction revenue and the volume the cross border volume I talked to is really related to 2 things. And they are similar things that I have spoken about in previous quarters. 1 is volatility. And volatility primarily related to the fact that last year in Q3, we are lapping the peak of the of the year. The highest volatility quarter was Q3 of last year, and we are lapping that in this quarter. Then the second factor is mix of the business and the composition of our yields within international transaction revenue can and does vary across our business."
— Chris Suh, CFO

Assessment: Straight answer, and the disaggregation is useful. The volatility half self-corrects; the mix half does not, and management named Visa Direct as an example of the lower-yield product diluting the blend. That is the strategy working as designed and compressing the yield as a side effect. It is a real cost of the money-movement expansion and should be modeled as permanent.

Issuer processing and core banking: which banks, and for what

A detailed question separated two commonly conflated issues, whether the processing strategy is about new revenue, negotiating leverage, or relationship depth, and whether the target customer is large banks or smaller institutions. The answer was the quarter's clearest strategic disclosure, and it narrowed the addressable market for the newly announced integrated product.

Q: "And on the bank-size focus, it is often is it about large banks, like the big names we have seen announced, like Wells Fargo and Citi, or is it about smaller banks and fintech? And, again, the answer might be both."
— Timothy Chiodo, UBS

A: "increasingly, in The US, what we are seeing is smaller banks and midsized banks as well as fintechs are looking for an integrated debit and credit processing solution that will help them simplify their operations and accelerate product innovation. And you know, that is the market need that has led us to the DPS-Pismo solution, you know, that we have talked about. it is not a market need we have seen from large issuers."
— Ryan McInerney, CEO

Assessment: Credit for precision over hype. Management explicitly declined to claim the large-issuer processing market and confined the integrated product to small, mid-sized, and fintech issuers, while keeping large banks in scope for core-banking cloud migration only. That is a smaller near-term revenue pool than the framing three months ago implied, and a more credible one.

Reconciling a strong Q4 guide against named headwinds

One question pushed on the apparent contradiction between a strong Q4 guide and the headwinds management had itself flagged, including tough FX, regional conflict, and less tournament revenue. The response identified the two technical offsets rather than disputing the premise.

Q: "Your guidance for fiscal fourth quarter calls for low double digit to low teens, including M&A, and that is incorporating, I think, tough levels of FX, while you mentioned it also likely incorporates uncertainty around The Middle East and less FIFA revenue. So we are trying to just reconcile the strength we are seeing in that kind of guide for fiscal fourth quarter to despite those headwinds."
— Darrin Peller, Wolfe Research

A: "From a technical perspective, are 2 offsetting factors as you think about the models. 1 is volatility. Which as we exit Q3, we think Q4 will be closer to the Q1 levels. Obviously, that could change, but that is the level that we are anticipating. And the second is incentives. Which we talked about, which is really a reflection of our new and renewal business."
— Chris Suh, CFO

Assessment: Both named offsets are revenue headwinds, not tailwinds, which means the Q4 guide holds up despite them rather than because of anything new. The implicit support is value-added services momentum and pricing. This is a guide that depends on VAS continuing to run hot into a quarter that still carries the tournament's knockout stages.

Agentic commerce as addressable-market expansion

The closing question asked how agentic commerce expands Visa's opportunity beyond simply tokenizing agent-initiated volume, specifically whether it creates new services to sell and new economic relationships such as agent-to-agent. The answer stayed at the framework level, mapping the adoption curve against prior payments cycles and identifying trust as the rate limiter, without addressing the agent-to-agent economics.

Q: "Ryan, I wanna ask about your updated thinking related to Agentic Commerce and Visa. There were quite a few announcements coming out of your payment forum on OpenAI partnership, new capabilities."
— Harshita Rawat, Bernstein

A: "Agentic commerce is a when not an if. We are building, like, the products, the services, the protocols, ensuring that the ecosystem has what it takes, and this will happen. And it will be a positive it will be a positive tailwind for Visa. Once we get through those steps of the journey."
— Ryan McInerney, CEO

Assessment: The conviction is total and the specificity is nil, which is appropriate for a category management repeatedly described as being in its very early stages. The agent-to-agent economics question went unanswered, and it is the one that determines whether this is addressable-market expansion or a defensive necessity. The OpenAI and Meta partnerships are the substantive news; the framework is unchanged from last quarter.

What They're NOT Saying

1. The size of the workforce reduction. Neither the 2,600 roles nor the 7% figure was stated on the call. Prepared remarks said only that roles are being eliminated with the majority in technology and product. For an action of this scale at a company of this profile, delivering the number outside the call and the earnings release body is a deliberate choice about where the conversation happens.
2. Annualized savings from that reduction. Asked directly, management gave no figure, no timeline, and no split between reinvestment and drop-through. A 7% headcount reduction against a personnel line running roughly $1.9B a quarter excluding severance is a material number that investors have been left to estimate themselves.
3. Whether the Investor Day revenue algorithm still stands. Actuals have exceeded the 9% to 11% framework for roughly two years. Asked whether the framework should be updated, management redirected to execution. The framework is now either wrong or being deliberately under-claimed, and investors cannot tell which.
4. Any quantification of the World Cup contribution. Extensive city-level, category-level and country-level colour was volunteered, but no aggregate revenue or volume contribution was given, and management explicitly deflected when asked to normalize for it. The most-cited driver of the quarter is the least measurable.
5. The expense attached to marketing-services revenue. One question asked specifically about the cost side of the marketing-services line that is driving the other-revenue and VAS acceleration. The answer addressed VAS growth durability and never returned to the expense question. With marketing expense up 54% Y/Y, the incremental margin on this fastest-growing revenue stream remains undisclosed.
6. Any new large-financial-institution core-banking win. Last quarter management teased more to share over time on large-FI Pismo wins. This quarter none was named, and the announced product expansion pointed at small and mid-sized institutions instead. The absence is not a negative in itself; the contrast with the prior quarter's framing is worth noting.
7. Agent-to-agent transaction economics. Asked whether agentic commerce creates genuinely new economic relationships, management described the adoption curve and did not address the economics. The claim that agent-mediated transactions carry similar economics to existing products has been asserted for two consecutive quarters without a supporting number.
8. Any FY27 framing beyond process. Management confirmed it is running scenarios and has line of sight on renewals, product pipeline and pricing, and committed to nothing. Given that the pricing contribution was described as similar to prior years, that is a soft positive, but it is the only forward-looking substance offered.
9. Free cash flow trajectory. Nine-month operating cash flow of $16,342M is down 2.8% Y/Y while nine-month buybacks rose 22.7% to $16,430M. Neither the cash-flow decline nor the funding gap was addressed in prepared remarks or Q&A.

Market Reaction

• Pre-print setup: V closed at $366.59 on July 28, the top of its 52-week closing range of $295.52 to $366.59. The stock entered the print +4.5% YTD (from $350.71 at 2025 year end), +3.1% over trailing twelve months (from $355.47), and +9.0% over the trailing 30 days (from $336.23 on June 26). The S&P 500 was +8.5% YTD over the same period, so V entered roughly 400bps behind the index for the year but running hard into the print.
• After-hours move: Modestly negative on the July 28 print, trading down about 1% to roughly $362.72 before and during the call.
• July 29 session: Gapped down to a $357.00 open, a 2.6% decline from the prior close, and traded as low as $355.48 (-3.0%) before recovering through the day to close at $368.73, up 0.6% (+$2.14) and above the pre-print close, having touched an intraday high of $373.97 (+2.0%). Volume was 11.3M against an 8.7M 30-day average (1.3x).
• Index context: The S&P 500 fell 1.5% on the same session, so V closed roughly 2.1 points ahead of the index on a day the broad tape was down.

The shape of the reaction is more informative than the net move. A beat on both lines, a raise to both full-year guides, and a first-ever $4 trillion volume quarter produced an opening gap down of 2.6% and an intraday low 3.0% below the prior close. That is not a market disputing the quarter. It is a market that had already paid for it: three sell-side target raises in the preceding three weeks, a 9.0% run in the preceding month, and a close at the very top of the 52-week range going in. The round trip completed: the stock closed the session up 0.6% at a fresh 52-week closing high, on 1.3 times average volume, while the S&P fell 1.5%.

The recovery through the session suggests the initial reaction over-weighted two things. The first is the optics of a 7% workforce reduction landing on the same day as a record quarter, which reads as a defensive signal on first pass and as a margin lever on second. The second is the six-point gap between cross-border volume and international transaction revenue, which looks alarming in isolation and considerably less so once the volatility comparison is understood as a lapping effect that eases from here.

What the tape is not yet paying for is the restructuring. With no savings figure disclosed, there is nothing for the market to capitalize, and the $563M charge is a certain cost against an uncertain benefit. That asymmetry resolves at the FY27 guide next quarter, which is the near-term catalyst worth positioning around.

Street Perspective

Debate: Is the medium-term revenue algorithm permanently reset higher?

Bull view: Visa has beaten its published 9% to 11% net revenue framework for roughly two years, and the drivers are structural rather than cyclical: value-added services at a third of revenue growing above 20% organically, commercial payments volume growing 3 points faster than total volume, and a pricing contribution management describes as consistent year after year. The framework was set in February 2025 and the business has outgrown it.

Bear view: The outperformance is heavily assisted. This quarter alone carried a World Cup, favourable tax-refund timing, promotional-event shifts, acquisition contribution of nearly 1.5 points, and half a point of FX. Strip those and growth is close to the framework rather than far above it. Management's refusal to raise the framework when asked directly is the tell: they know the composition better than anyone and are not willing to underwrite it.

Our take: The bears have the better read on this quarter and the bulls have the better read on the trajectory. Roughly two points of the current growth rate is genuinely non-repeatable. But the VAS engine, at 20%-plus across three portfolios for four straight quarters and mechanically linked to transactions and credentials, is a permanent addition to the algorithm that did not exist when the framework was written. We model FY27 net revenue growth at low double digits, above the old framework and below FY26's reported rate.

Debate: Does the AI restructuring reach the margin line?

Bull view: A 7% headcount reduction concentrated in the most expensive functions, justified by measured productivity gains of 65%-plus faster feature development, is a structural reset of the largest expense line. Personnel expense excluding severance ran about $1.9B this quarter; even partial retention of the savings is worth meaningful margin. Visa is early to a change that every network will eventually make, and early movers on cost structure compound the advantage.

Bear view: Management was asked point blank and committed only to maintaining strong margins, not expanding them. The named reinvestment list is long and expensive: blockchain, stablecoin infrastructure, digital wallets, payment orchestration, agentic commerce. This is a funding mechanism for a product roadmap, not a margin event, and the $563M charge is the only number investors actually have.

Our take: The bears are right about FY27 and the bulls may be right about FY29. Take management at its word: margins stay roughly flat as savings fund the roadmap. The genuine value is optionality, because a company that has demonstrated it can remove 7% of headcount without impairing output has a lever it can pull again if growth disappoints. Do not model margin expansion. Do note that the downside case for this stock just got a floor under it.

Debate: How much of the consumer strength survives the comparisons?

Bull view: Three consecutive quarters of the same message, that the lowest spend bands are not weakening and the highest are growing fastest, backed by improving rather than deteriorating data each time. US volume accelerated 2 points with both credit and debit participating, card-present and card-not-present both improving, and July still running 9%, itself a point above what the company printed two quarters ago. This is a healthy consumer, not a decorated one.

Bear view: The +10% is a confection of tax refunds, fuel prices, retail-event timing and a World Cup, and it decelerated to +9% within three weeks of the quarter closing. Management could quantify the decomposition and would not. Cross-border e-commerce at 16% accelerating to 18% in July was described by the CFO himself as unusually high and expected to settle back.

Our take: Both sides are describing the same 9% and disagreeing about the framing. The underlying US rate improved about a point sequentially, which is genuine, and event effects added another point, which is not. A 9% underlying US run-rate against a 5% to 7% consumer-payments framework is a good outcome that does not need the tenth point to support the thesis. The risk is investors who anchored on 10%.

Debate: Are agentic commerce and stablecoins expansion or insurance?

Bull view: Partnerships with the largest consumer AI platform and the largest social commerce surface, in the same quarter, resolve the distribution question that was the sharpest objection three months ago. On stablecoins, Visa moved from settling in them to operating minting and orchestration infrastructure and helping issue one. In both cases Visa is positioned as the trust and credentialing layer, which is the part of the stack that is hardest to replicate and easiest to monetize.

Bear view: Neither generates material revenue today, and the CEO conceded stablecoins have yet to scale beyond a few use cases. The claim that agent-initiated transactions carry similar economics to existing products has now been asserted for two quarters without a single supporting figure. Building the rails your potential disintermediators will run on is defensive spending dressed as strategy.

Our take: This is insurance that may become expansion, and it should be valued as insurance until the economics are disclosed. The correct posture for a network facing a technology shift is to be the trusted layer inside it rather than to contest it, and Visa is executing that well. But an investor paying 28x for this stock is paying for the core network and the VAS franchise. Treat agentic and stablecoin optionality as free, and be pleasantly surprised.

Model Update Needed

Valuation impact. On our revised FY26 non-GAAP EPS of roughly $13.15, V trades at about 28x the current fiscal year against the post-print close of $368.73. Rolling forward at low-double-digit EPS growth gives an FY27 estimate near $14.90, or roughly 25x. That is a full but not extended multiple for a business compounding revenue at low double digits with a third of revenue in a 20%-plus services franchise and a 3% annual share-count reduction.

Target price: $405, implying +9.8% from the post-print close of $368.73. The target applies roughly 27x to our FY27 estimate, a modest premium to the current multiple justified by the VAS mix shift and the optionality on restructuring savings that we have deliberately not modeled. We would revisit upward if the FY27 guide next quarter quantifies the savings, and downward if the underlying US run-rate slips below 8% or the international transaction revenue gap widens beyond the volatility explanation.

Thesis Scorecard Post-Earnings

Overall: thesis strengthened operationally, unchanged on risk/reward. Nine of eleven thesis points came out of this quarter confirmed or challenged in Visa's favour, and the two that moved against it are both price-and-yield items rather than franchise items. The business is executing better than the standing thesis assumed. The stock has already been paid for that.

Action: hold; add on weakness rather than strength. The FY27 guide next quarter is the catalyst that resolves the two open questions, restructuring savings and whether the revenue framework gets re-baselined. Entering that print at 28x with the Street's targets clustered only 8.5% above the price is a thinner setup than the one that existed in April at $320, which is precisely the level at which management was buying its own stock most aggressively.

Bottom Line

Visa printed $11,633M of net revenue, up 14%, and non-GAAP EPS of $3.32, up 11%, beating consensus on both and comfortably clearing its own mid-to-high-single-digit EPS guide. Quarterly payments volume crossed $4 trillion for the first time in the company's history. Processed transactions reached 71.7 billion. US payments volume accelerated roughly 2 points to 10%, the best domestic quarter since fiscal 2019 excluding the pandemic rebound. Value-added services grew 34% in constant dollars to $3.8 billion, roughly a third of net revenue. Both full-year guides moved up, for the second consecutive quarter. There is no line in this print that a shareholder should be unhappy about.

What makes the quarter genuinely interesting is not the beat. It is that Visa removed roughly 2,600 roles, about 7% of its workforce, took a $563M severance charge, explicitly attributed the reorganization to AI-driven engineering productivity, and then declined to tell investors what any of it is worth. Management named neither the headcount nor the percentage on the call, and when asked directly whether savings reach the bottom line, committed only to maintaining industry-leading margins. That is the largest single-quarter restructuring in the company's recent history delivered as a productivity anecdote.

The standing thesis is that Visa is a quality compounder positioned as the trusted bridge between every emerging payment rail and real-world spend, compounding revenue at low double digits and EPS at low-to-mid teens while returning roughly 5% of market capitalization annually. This quarter validates every pillar of that and adds a new one: a demonstrated ability to take meaningful cost out of the largest expense line without apparent damage to output. It also escalates one bear point, because cross-border volume is converting to revenue six points worse than it did a year ago and half that gap is structural mix rather than a passing comparison.

We are maintaining Outperform and holding conviction where it was rather than raising it again. The reason is not the business, which is performing better than we underwrote three months ago. It is the setup. The shares entered the print at the very top of their 52-week range after a 9% month, the Street's target cluster sits about 8.5% above the post-print close, and a beat-and-raise with a first-ever $4 trillion volume quarter opened 2.6% lower. Management's own capital allocation says the same thing more precisely than we can: $7.9B of buyback at an average $320.66 in the March quarter became $4.9B at $330.71 in the June quarter.

What would change the rating: a quantified restructuring benefit at the FY27 guide that supports genuine margin expansion, or a re-baselining of the medium-term revenue framework, either of which would argue for upgrading conviction. In the other direction, an underlying US run-rate slipping below 8%, an international transaction revenue gap that persists after the volatility comparison anniversaries out, or evidence that agent-mediated volume carries structurally worse economics than management has asserted. Absent those, this is a high-quality compounder executing well at a price that already reflects it.