TechFlow Logo
Login/ Sign up
ETH Gas
Gwei
Fear
gas
Stablecoins and Agentic Can't Make It into Visa's Financial Report

Stablecoins and Agentic Can't Make It into Visa's Financial Report

2026.08.14
Share
The more completely the product is presented, the more conspicuous the absence of money becomes.

By: Will Awang

On July 16, Visa launched the Visa Stablecoin Platform, a platform for minting, circulating, and managing stablecoins. Its inaugural coin is OpenUSD, a stablecoin jointly initiated by an alliance of 140 major institutions. That week, the news dominated feeds in the payments and crypto circles—an network that clears nearly half of global card transactions announced it would build the operational foundation for stablecoins.

Twelve days later, during the Q3 FY2026 earnings call, Visa CEO Ryan McInerney presented the platform to global investors with remarkable completeness: its features, positioning, launch coin, and future integration with Pismo to support tokenized deposits. Yet, across the entire segment, there was not a single word mentioning how it would be priced, which accounting line item would record the revenue, when it would go commercial, or who the first customers would be. Throughout CFO Christopher Suh’s financial section, the three letters "VSP" never appeared once.

The more completely the product was explained, the more conspicuous the absence of revenue figures became.

The Q&A session made this even more explicit. Of the 13 analysts, only Jeff Cantwell from Seaport asked about stablecoins, and his question was still whether "OpenUSD is targeting Circle and Tether"; the remaining questions focused on cross-border growth, one-time FIFA tailwinds, value-added services, client incentives, and that income formula asked about every year. Among the two analysts giving Visa the highest target prices, one asked about the stage of internal AI adoption, while the other asked about disruptions in promotion timing.

The market does not treat VSP as headliners, nor do those pricing Visa consider it a core issue. Moreover, McInerney’s very first sentence answering Cantwell poured cold water on the stablecoin narrative. He did not address competition directly; instead, he noted something else: stablecoins have not truly scaled beyond a handful of use cases—for example, our stablecoin-linked cards issued globally.

This article does only one thing: place every move Visa makes related to stablecoins and agentic technologies onto its earnings revenue table, one by one, to measure them. Once measured, you can see their strategic position within Visa, and how much the narrative around these two terms has been inflated.

I. Visa’s Profit Ruler

The CFO only talks about one thing during earnings calls: the revenue line. How much each segment grew, what drove it, and which metrics it ties to. Stablecoins did not appear in his script for a single reason: they do not belong to any of them.

So this chapter establishes a ruler first: Who is Visa, how many charging points does it have, and how much does it ultimately take per hundred dollars?

If you think of Visa as a highway, everything that follows becomes clear. It does not manufacture cars, drive them, nor own any warehouses at the start or end points. It builds roads, sets rules, and collects tolls at exits. A transaction traveling on the road involves four roles: cardholders, issuing banks (the cardholder’s bank), merchants, and acquirers (service providers on the merchant’s side). Visa is none of them, but it is the road all four must traverse together.

How are the tolls split? Look at the overall ledger first. The total cost a merchant pays for a transaction is called the Merchant Discount Rate (MDR), roughly 2% to 3% in the US, split into three parts: the largest share is the interchange fee, going to the issuing bank; a portion is the acquirer’s markup; and only a tiny remainder is the network fee, going to Visa. Although Visa sets the interchange fee, it keeps none of it: the fee flows from the acquirer to the issuer to compensate the issuing bank for bad debt, fraud, and capital costs—a reason why global regulators keep circling back to it relentlessly.

Merchants bleed the money, while cardholders reap the incentives.

Card networks essentially organize issuers and acquirers into a coalition of shared interests—standards are merely the connection method; what binds everyone together is profit distribution.

The entire road operates on the same logic: let others take the lion’s share. That "beep" in a phone sees the wallet provider pocket a cut from the issuer; Stripe and Adyen leverage merchant relationships to run their own fleets along the route; those fintechs proclaiming they are "disrupting payments" are actually running under Visa’s license plates.

Visa turns every potential disruptor into a paying driver on its own road.

1.1 Four Charging Points for a Cup of Coffee

A user in Hong Kong buys a coffee at a convenience store in Tokyo for $10, swiping a Visa card issued in Hong Kong.

Assuming the store’s MDR is 2.5%, the merchant nets $9.75 and pays out $0.25. The bulk of this $0.25 is the interchange fee, flowing to that Hong Kong issuer—which bears the user’s credit and fraud risk; a portion goes to the Japanese acquirer; and Visa takes approximately $0.09. (Visa does not disclose line-item fees; actual pricing varies by region, card type, and merchant category. The figures below are reverse-engineered from public totals. With all assumptions stated, you can recalculate them yourself.)

That $0.09 is not collected in one lump sum; it stops and charges separately at four distinct charging points.

First point: Service Revenue. What it charges for is not this specific swipe, but the fact itself that "this card can be accepted." The same card works at a Tokyo convenience store, a New York website, and a Shenzhen subway turnstile—the existence of this acceptance network is the product itself, settled based on the issuer’s transaction volume. It charges by amount, as a percentage: the pricier the coffee, the more this point collects.

Second point: Data Processing Revenue. Within that second or two of swiping, the authorization request travels from POS to the acquirer, to VisaNet, to the issuer, and back along the same path; clearing and netting also happen overnight.

This point charges per transaction, a fixed amount annually. With roughly $20 billion in annual data processing revenue across ~290 billion processed transactions, each transaction averages about $0.07. Whether this coffee is $10 or $1,000, that $0.07 remains identical.

This point also has a unique source of growth found nowhere else: penetration rate—diverting vehicles already in motion onto Visa’s own roadway. Colombia’s processing penetration was single-digit five years ago and now exceeds 90%; this growth requires consumers to spend not a single extra cent.

Third point: International Transaction Revenue. This coffee was bought with a HKD card in a JPY country, triggering the most expensive point. It costs more because it bundles both cross-border fees and foreign exchange: Visa must process the transaction and convert between HKD and JPY; FX is the primary source of gross margin here. This point also charges by amount, so within that $0.09, this point contributes significantly more than a local transaction of the same amount.

Fourth point: Other Revenue. If this transaction runs through risk controls, if the issuer’s card accounts run on Pismo, if Visa runs this card’s marketing campaigns: this point charges by service, neither by amount nor per transaction, unrelated to the coffee itself. Risk & security (Featurespace), issuer processing (Pismo, DPS), acceptance solutions, consulting, and marketing services all fall into this category.

Briefly clarify an easily confused metric: Value-Added Services (VAS) is a business-line metric spanning four accounting categories (risk control counts toward data processing, consulting and marketing count toward other revenue), whereas "Other Revenue" is merely one financial statement line item; the two are not equivalent. Both sets of figures are examined together in section 1.3.

Now looking back at that $0.09: the $0.07 from the second point is fixed, leaving roughly $0.02 from the amount-based points. By contrast, Visa’s average transaction is $56; calculated against the company-wide net rate of 0.29%, it yields about $0.16, following the exact same split: $0.07 fixed, $0.09 amount-based.

Under the same rate structure, a $56 transaction charges 0.29%, while the $10 coffee effectively charges 0.9%. The smaller the transaction, the heavier the weight of that $0.07 fixed fee.

1.2 Tokens Riding Along with the Cars

There is another element that occupies no charging point yet rides along with the cars: Tokenization. Called Visa Token Service, this system replaces the 16-digit number on the card face with an alternative number valid only in specific scenarios.

Take this user. He links the card to Apple Pay; upon linking, Visa issues a token valid exclusively for that device. He saves the card on an e-commerce site later; that generates another token. Three months later, the card expires and is replaced; the physical card number changes, but both tokens update automatically without him needing to re-link. If that e-commerce site suffers a data breach, what leaks out is a string of dead numbers useless anywhere else.

Every issuance, update, validation, and routing constitutes an extra passage maneuver on the road, billed under the second point. Approximately half of global e-commerce transactions are now completed via tokens.

Credentials issued by Visa to AI agents travel over this same infrastructure.

1.3 Client Incentives, and What This Quarter’s Ruler Measured

After collecting through the four points, one item must be deducted: Client Incentives—money returned to issuing banks and acquiring institutions.

Rebates negotiated during large issuer renewals and onboarding terms for new clients all sit here. Crucially, incentives are treated as revenue contra-items, not costs, so Visa reports "net revenue." This quarter, incentives grew 18%, jumping 4 percentage points higher than last quarter.

After collecting through the four points and deducting incentives, divide net revenue by payment volume:

0.29%, or 29 basis points. For every hundred dollars passing through, Visa finally pockets 29 cents. Subsequent sections will uniformly refer to this as 29 basis points.

Matches perfectly with the last full fiscal year: net revenue of roughly $40 billion against $14.2 trillion in annual payment volume, equating to 28 basis points.

The striking feature of this number is that it hasn’t budged in a decade. Over a longer horizon, Visa’s posted gross toll rate has remained stable around 40 basis points of traffic volume; meanwhile, the proportion returned as incentives rose from roughly 17% to nearly 29%, almost doubling. Pinched from both ends, the net toll rate stays virtually stationary. In other words, the uptick generated over the years from two high-price lanes—cross-border and value-added services—did not become Visa’s extra profit; it turned into rebates paid to traffic generators.

This is the key to understanding Visa: it is not raising toll rates; instead, it uses shifts in traffic composition to pay for ever-increasing customer acquisition costs, keeping the blended rate steady.

A company that has survived decades by maintaining equilibrium will not dismantle the balance itself for a new gimmick.

II. Stablecoin-Related Business

With Visa’s ruler established: one road, four charging points, 29 basis points.

Up to today, Visa has only engaged in scale-level contact with stablecoins across two areas: U-cards on the consumer side, and how institutions settle accounts on the backend. The two operate differently. More critically, they land in entirely different places.

2.1 The Essence of U-Cards Remains Card Transactions

What happens in that instant a Visa card bearing a stablecoin balance—colloquially known in the industry as a U-card—is swiped?

The network reads the BIN, routing the authorization request to the corresponding issuer. The issuer queries the underlying crypto platform: does this user have sufficient balance? If yes, the platform instantly calculates how much stablecoin needs to be sold to cover the purchase and executes the conversion within the authorization window. Once the fiat amount is confirmed, the issuer replies with authorization, and the merchant receives approval.

The entire process completes in milliseconds. To the merchant, however, this differs not at all from any ordinary Visa transaction—they receive fiat currency, requiring zero line-of-code changes on their acquiring end.

This is the most critical feature of a U-card: the on-chain segment occurs entirely before entering Visa’s road.

A U-card cannot be built by a single company. It requires assembling at least several components: card network membership and a BIN, on-chain custody, real-time conversion liquidity, and a full suite of KYC/AML and compliance capabilities. The traditional approach secures a licensed institution as a BIN sponsor, which provides membership and card ranges, allowing crypto projects to white-label issuance on top. This route bypasses the years-long process of applying for licenses independently.

In recent years, a more advanced model has emerged: full-stack issuers like Rain and Reap have directly become Visa Principal Members—bypassing bank intermediaries entirely, acting as the issuer of record, and settling directly with Visa. Just Rain alone sustains over 200 card programs.

So how is the money on this chain divided?

Issuers take the interchange, typically 1% to 2% of transaction volume; this either becomes cashback subsidies for users or retained profit. Convertors take FX and spread margins. Project operators may also charge cardholder annual and withdrawal fees. Visa takes that 29-basis-point slice outside of all of these.

So where does a U-card land? It passes through all four charging points because it fundamentally is a standard card transaction. Once conversion finishes, what enters the network is fiat; the charging points have no idea the stablecoin ever existed.

Visa itself has documented two U-card settlement models: under the traditional model, project operators first swap stablecoins for fiat before settling with Visa, using blockchain solely as a ledger; under the emerging model, Principal Members settle directly with Visa in USDC, after which Visa’s digital custodian converts it to fiat to pay the acquirer. Either way, only the latter segment changes—how institutions clear balances among themselves.

This serves as the heaviest counterargument in the entire piece: the only stablecoin action generating revenue is precisely the one converting stablecoins back into card transactions.

2.2 The Backend Road: Settlement

Visa cards solve consumer payments; stablecoins solve fund movement. Two layers of infrastructure, each handling a segment.

The two roads do not intersect. The manner in which stablecoins appear on these two paths also differs: on the front-end consumer route, it must first convert back to fiat to enter VisaNet; all four charging points established in Chapter 1 lie on this route. On the backend settlement road, it can act directly as a clearing asset, settling the netted payable/receivable between issuers and acquirers.

For a card transaction, consumers experience a one- or two-second authorization, but that is merely an information flow. Real money must move between the issuer and acquirer, with Visa’s treasury and settlement system standing in between. Each day, it nets all payables and receivables for a given member institution into a single figure, then settles it in one go. This net figure is what Visa officially calls VisaNet obligations.

Traditionally, this net amount travels through fiat systems—ACH, Fedwire, SEPA, or local clearing—and these systems only open during banking hours. Stablecoin settlement changes only one thing: which asset is used to clear this net amount, enabling near real-time, round-the-clock processing.

Demand exists for this because time gaps cost money. Transactions authorized on Friday night settle as early as Monday, requiring someone to front the capital during the interim. Not everyone fronts it—Visa’s rule mandates collateral for clients failing credit standards. Those with good credit generally don’t need to lock funds, but crypto issuers with substantial volume yet mediocre credit must deposit money there, unable to deploy it for operations.

Collateral exists because Visa guarantees to its clients: if any party fails to fulfill settlement obligations, Visa covers it. Its average daily settlement exposure reached $91.2 billion in FY2025. On this backend road, Visa does not charge tolls, but it shoulders the guarantee.

2.3 From Stablecoin Settlement to VSP

Visa began stablecoin settlement pilots as early as March 2021. The first pilot was with Crypto.com, solving a very specific pain point: originally, selling crypto assets to USD and wire transferring to Visa to cover card program debt required swapping twice. Switching to USDC allowed direct transfers to Visa’s custody wallet, eliminating one conversion step and a wire cutoff delay. Pilots subsequently expanded from the issuer side to the acquirer side (Worldpay, Nuvei), launched domestically in December 2025, and supported nine blockchains by April 2026.

Visa’s benefit list for this initiative is straightforward: funds accessible within seven days, uninterrupted weekends/holidays, reduced collateral requirements enabled by seven-day settlement cycles, automated treasury management, and simplified reconciliation. Most critically comes that liability-limiting caveat—the consumer’s card usage experience remains completely unchanged.

What about scale? Annualized at $3.5 billion in November 2025, $4.5 billion in January 2026, and $7 billion in April. Doubling in five months—that is the growth trajectory.

Meanwhile, Sheffield provided a contrasting denominator during a January Reuters interview: Visa handled $14.2 trillion in payment volume the prior year. Dividing April’s $7 billion by that baseline yields just 0.05%.

Swapping the asset clearing VisaNet obligations from fiat to stablecoins leaves Visa’s collected fees entirely unchanged. Benefits accrue to issuers—shortened exposure windows, less locked collateral, weekend transfer capability. The MDR does not change due to faster settlement, and consumers feel nothing at checkout.

This represents an operational improvement, not a revenue recognition event. And exactly this segment is what VSP aims to commercialize.

Settlement pilots were built for Visa’s own use; VSP sells this capability to others—a newly constructed charging point on the backend road. Theoretically, Wallet-as-a-Service should follow Service Revenue logic, while minting and transfers align with Data Processing logic. Yet to date, Visa has disclosed no billing methodology, accounting classification, or commercial rollout timeline. The platform is in beta, and clients remain unnamed.

It is not that they cannot charge; it is that they have not started charging yet.

Calculated comprehensively, the only money Visa earns from stablecoins today consists of two streams: the one converting stablecoins back into card transactions, and the consulting/advisory services sold regarding stablecoins.

This playbook is far from fresh. Looking back at Chapter 1: stablecoins are simply the newest name on the paying drivers list. And to date, not a single name on this list has opened a new charging point for Visa.

Mapping the three actions back onto the two roads: U-cards reside on the front-end road; settlement and VSP operate on the backend road. Since charging points exist only on the front-end, this entirely explains why U-cards generate revenue while settlement and VSP currently do not.

III. Agentic Commerce-Related Business

Once stablecoins are measured, the front-end half turns out simpler—it lands entirely on the front-end road.

3.1 Front-End: Devices Enabling Agents to Enter the Road

Nothing novel on the front-end: every agentic product Visa builds serves as a mechanism allowing agents to access the legacy road.

Visa itself drew this division. During the July VSP launch, Chief Product & Strategy Officer Jack Forestell stated that AI transforms the front-end of commerce, while stablecoins reshape the backend. Three months later on the earnings call, McInerney repeated this phrasing almost verbatim. This is not PR spin; it reflects internal strategic allocation.

The product list is brief: Agent Score, Agent Directory, Token Assurance Framework, plus partnerships with OpenAI and Meta. All funnel through the fourth point—billing methodology undisclosed, but it sells scoring, directory, and credential management capabilities, charging by service rather than transaction amount. They utilize the exact same token infrastructure from Chapter 1, merely extending it from mobile phones to agents.

3.2 Displacement vs. Creation: Visa Only Sees Half

There is only one scenario that fits neither front-end nor backend: agent-to-agent transactions.

Bernstein analysts asked this directly—how agentic technology expands the Total Addressable Market (TAM), and whether agent-to-agent commerce could forge a new economic structure.

McInerney addressed the first half, using historical analogies: e-commerce, mobile commerce, tokenization, tap-to-pay. Each underwent a sequence of "establish standards, early adoption, consumer momentum buildup, and finally scaling." His conclusion was that agentic commerce is a question of "when," not "whether." He omitted answering the second half.

Yet that series of analogies itself answers the question. E-commerce, mobile payments, tokenization, tap-to-pay—the common thread across all four precedents is that they altered "how people pay," not a single one changed "who is buying."

Agentic commerce is actually two distinct phenomena, a divergence we repeatedly emphasized in our prior agentic payments research:

One category involves agents purchasing on behalf of humans. An IT Agent automatically adjusts SaaS seats based on employee onboarding/offboarding; a Travel Agent books flights and hotels according to budget and policy. These transactions carry higher amounts, with merchants and channels resembling traditional e-commerce—it is displacing existing human shopping behaviors, carving market share.

The other involves machine-to-machine procurement. A Research Agent pays mere fractions of a cent to multiple data providers for financial reports to cross-verify; a Primary Agent outsources subtasks to a Retrieval Agent, which subsequently purchases dashboards from a Data Agent. These transactions are high-frequency, low-value, with neither party holding accounts or contracts—it is creating entirely new demand; these transactions previously did not exist.

Visa’s analogy sequence covers only the first category from beginning to end. Sell-side analysts wield the exact same ruler: Morgan Stanley’s projection of $190B to $385B by 2030 frames it as "representing 10% to 20% of US e-commerce," measuring AI-assisted online shopping, covering merely the first half.

Visa is not building a new road for agentic commerce; it is merely swapping out the old road’s drivers for a new batch.

(Agentic Payments Through Visa’s Lens)

IV. Strategic Positioning of Stablecoins and Agentic Commerce at Visa

4.1 Tier Four, Rivaling Brand Advertising

First, examine how high it ranks Visa’s priorities. Answer: Tier Four, listed alongside brand advertising.

This quarter, Visa laid off roughly 2,600 employees—about 7% of its workforce—primarily in tech and product roles, recording $563 million in severance charges. KBW’s Sanjay Sakhrani asked whether savings from these cuts flow to profits or get reinvested.

McInerney replied that all savings would be reinvested, then listed investment directions sequentially: Consumer Payments (cash-dominant market acceptance expansion, premium demographics, cross-border e-commerce), Value-Added Services (risk & security, marketing services, Pismo, Featurespace), Commerce & Funds Movement (unified B2B, embedded finance, cross-border remittances), before concluding with "on top of that, there are stablecoins and agentic."

While earnings transcripts vary slightly in ordering the top tiers, stablecoins and agentic consistently rank last in every document. This ad-hoc ranking by the CEO under pressure regarding capital allocation holds more credibility than any keynote address.

Two corroborating factors are equally clear. First, they do not constitute growth engines—Visa’s 10-K filings and quarterly earnings calls frame strategy around only three pillars: Consumer Payments, Commerce & Funds Movement, and Value-Added Services. Stablecoins and agentic are merely engine components, diffusely embedded within these three legs, leaving no dedicated charging point in the strategic architecture.

Second, they represent the only verticals lacking a defined TAM. Visa heavily relies on TAM narratives: Consumer Payments cites annual addressable consumer spending exceeding $40 trillion (excluding China/Russia), with "underserved" portions surpassing $20 trillion; Commerce & Funds Movement points to $200 trillion in annual payment volume opportunity; Value-Added Services targets $520 billion in annual potential revenue. Each pillar pairs with a trillion- or hundred-billion-scale denominator, while stablecoins and agentic boast zero concrete numbers.

4.2 Triple Jump: Settlement, VSP, and OUSD Form a Single Thread

Tier Four does not equal absent strategy. Conversely, Visa’s moves along the backend road follow strict sequencing.

Level One: Settlement Pilot. That five-year-old track from Chapter 2. It generates no revenue, but constructs an entire logistical dedicated corridor: custody, multi-chain transport, fiat/stablecoin inflows & outflows, and reconciliation.

Level Two: VSP. Opening this dedicated corridor for public use, selling it externally to clients. As outlined in Chapter 2, these clients are precisely companies that have minted coins but cannot enable end-users to spend them. Visa’s Head of Global Growth, Rubail Birwadker, articulated it more plainly:

"Rather than acquiring stablecoins, the focus is how this interoperates with client treasury settlements, funds movement workflows, and existing banking setups."

This is a confession of treasury business philosophy, not a manifesto for crypto ventures.

Level Three: OUSD.

Two details regarding OUSD. First, Visa secured a seat on the distribution channel, not ownership of the alliance—the alliance is managed by Stripe-affiliated personnel, with its most direct competitor sitting at the same table. Second, the inaugural coin supported by VSP is precisely OpenUSD.

The loop closes: the roadway Visa built for its own settlement opens up as platform VSP; and the initial fleet designated to traverse this route bears the license plate of governance participation and that exact stablecoin.

4.3 Next Door: Buying a Bridge for $1.8 Billion

Facing the identical challenge, Mastercard delivered a vastly different answer sheet.

On August 3, six days post-Visa earnings call, Mastercard closed its acquisition of BVNK. Founded in 2021, this firm processes approximately $30 billion in annual stablecoin payment volume across 200 countries and regions. The deal carries a maximum consideration of $1.8 billion—$1.5 billion base price, plus up to $300 million in performance-linked contingent payments; originally slated for year-end closing when unveiled in March, execution advanced nearly five months early. An additional detail: Visa Direct appears on BVNK’s client roster—Visa’s cars were already driving on the bridge Mastercard purchased.

Placing these figures side-by-side: the stablecoin traffic Mastercard acquired in a single stroke exceeds Visa’s own five-year cumulative volume by more than fourfold.

The nature differs sharply too: BVNK’s $30 billion represents client payment traffic generating revenue, whereas Visa’s $7 billion stems from backend settlement self-processing, yielding zero incremental revenue.

Mastercard’s Chief Product Officer Jorn Lambert’s framing warrants comparison: stablecoins are actively resolving genuine needs across cross-border B2B, remittances, pass-through payments, settlement, and treasury. One enumerates concrete use cases; the other emphasizes lack of scaling—occupying identical temporal windows and industry standings.

Settlement prioritizes compatibility; new platforms dictate alignment.

One spent capital on its balance sheet; the other allocated funds along a product roadmap.

(Mastercard Acquires BVNK for $1.8B: Buying Time, Not Technology)

V. Why Still New Cars on Old Rails?

By now, the strategic shape is clear: mint no coins, assume no liabilities, capture no float. Build a dedicated lane on the backend road, and construct narrow slip roads at every potential bypass junction.

The question is, on what grounds does this clearly conservative playbook suffice?

5.1 Sufficient Moat Width

The sell-side consensus delivers a remarkably clean response. Third Bridge experts distilled it into a single sentence:

"Stablecoins threaten the economics of the four-party model, not its infrastructure—it lacks consumer protection, fraud management, and trust layers, which precisely form the sturdiest positions held by Visa and Mastercard."

So why is consumer protection a moat? Because it is costly. A single chargeback triggers a complete evidence-gathering, investigation, adjudication, and fund-recall process involving all four parties: cardholder, merchant, issuer, acquirer, demanding manual intervention at every link. Visa has operated these rules for decades, accumulating precedents, SLAs, evidentiary standards, and appeals mechanisms—competitors aren't incapable of replicating them; rebuilding costs simply outweigh the benefits.

This moat’s width suffices, yet goods generating zero disputes have already departed the card rails. Categories like A2A machine transactions purchasing tokens, invoking APIs, or renting compute power exemplify this—delivery equals immediate consumption, leaving no room for reversal. On-chain trades lack authorizations, recourse, or dispute channels; for such commodities, being "naked" is precisely an advantage.

They depart not because on-chain is cheaper (though it undeniably is), but because they never utilized the specific value proposition cards provide. Conversely, what remains on cards today are precisely those transactions where dispute costs haven't sufficiently compressed.

Thus, the question isn't "will stablecoins be cheaper," but rather where the cost curve for dispute resolution will bottom out. If an agent can autonomously retrieve order logs, logistics proofs, product imagery, and chat histories, automatically cross-check merchant refund policies, draft dispute claims, and track them to closure, the marginal cost of handling a dispute drops from "dozens of dollars in manual labor" to "mere cents in computation." At that juncture, frameworks and rules persist, but the cost barriers sustaining them thin considerably.

This constitutes a bet wagering on declining dispute costs, rather than systemic collapse. Misjudge it, and Third Bridge’s thesis holds true, rendering Visa’s position even more entrenched than it appears today.

5.2 Explicitly Stating Load-Bearing Assumptions

All preceding analyses—two roads, four charging points, 29 basis points—rest on an unwritten premise.

Could a single stablecoin payment render any card network obsolete?

If merchants collect digital dollars directly from consumer or corporate wallets, Visa isn't underpaid; it is entirely absent. The income statement wouldn't show a shrinking figure; it would simply show nothing.

This scenario pilots in Japan. Effective January 26 this year, shops like "Edo Shokuhinkan" at Haneda Airport Terminal 3 began accepting USDC settlement trials, co-initiated by payment gateway operator NETSTARS and Japan Airport Building Co., targeting inbound travelers who scan shop QR codes and pay directly with USDC stored in private wallets like MetaMask; previously, SBI VC Trade and Aplus announced spring 2026 launches for similar merchant settlement trials, while Lawson tests deploying stablecoin settlement atop existing POS terminals without dedicated hardware.

So why hasn't this scaled broadly today? Four barriers hold it back: habit, rewards subsidized by interchange fees, dispute rights, and credentials already tucked in wallets. Deconstruct them, and none prove unreplicable—rewards can be funded by wallets or merchants themselves, dispute mechanisms can be rebuilt on new rails, and credentials embed directly into apps. Only habit bears the structural load.

Where habits guard nothing, leakage has already begun—cross-border B2B, dropshipping. This traffic lacks cardholders, rewards, or chargeback expectations; here, stablecoins' advantage boils down purely to cost efficiency.

Visa’s stablecoin strategy appears composed precisely because the traffic it guards overlaps maximally with entrenched habits. But habits are not eternal.

VI. What Exactly Does the Market Buy?

Measuring to the very end, one question remains unanswered: if stablecoins and agentic touch no charging points, what exactly does the market's premium valuation multiple buy?

Scanning this year's sell-side literature reveals an answer so frequent it borders on tedious.

January earnings: McInerney described the quarter as "very strong," attributing 15% gains across revenue and EPS to resilient consumer spending, robust holiday seasons, and value-added/commercial products. April results: headlines proclaimed "Visa's Strong Performance Driven by Resilient Consumer Spending." July quarter: leads reiterate sustained strength in consumer and commercial expenditure despite macro uncertainties.

During March's stock pullback, this was stated most bluntly. The prevailing debate condensed to: "payment stocks undergo reassessment, gauged by whether resilient consumer spending can continue propping up premium valuations." Visa dispatched CPO Forestell to Wolfe's FinTech forum to defuse tensions, carrying data showing January US payment volume grew 8%—credit 9%, debit 6%.

For a company dubbed "future payment infrastructure," the answer handed down under stock pressure is simply how many cards Americans swiped in January. This forms Visa's valuation anchor.

Observe the second anchor: mix shift. Value-added services grew 27% at constant currency last quarter, hitting 30% of net revenue; this quarter it jumped to 34%, nearing one-third. Sell-side growth models basically posit "consumer payments provide the base, while VAS and commercial payments deliver acceleration." The third anchor is buybacks—$20 billion newly authorized in April, with $4.9 billion repurchased this quarter alone.

Three anchors, none named stablecoin.

So what role do stablecoins and agentic play in sell-side prose? The most archetypal formulation reads: "Bullish cases for Visa encompass 'robust consumer spending, steady cross-border activity, expanding value-added services, and growing blockchain initiatives.'".

The first three carry hard numbers; the final term is merely an adjective dangling at the sentence's end. It is not absent; it exists, but it bears no structural weight.

Hence, the complete picture stands thus: markets purchase current global consumption paired with an untethered, unpriced future.

The former boasts figures, quarterly validations, and models; the latter merely requires presence, proving unequivocally that this company will not miss the next technological cycle.

And exactly what Visa delivers is this—it builds slip roads at every conceivable bypass junction, though constructed extremely narrowly. Issuance captures an alliance seat; settlement pilots navigate nine blockchains; operations deploys VSP; issuer processing leverages Pismo to interface with tokenized deposits; front-end extends token infrastructure to agents.

Slip road density sharply mismatches capital prioritization, and this mismatch inherently defines its nature: insurance and options, not directional bets.

Insurance and options demand early and broad accumulation, but require no premium pricing.

Returning to the opening. That headline-grabbing announcement sits firmly as Visa's Tier Four capital allocation. The disconnect never lies between "Visa's rhetoric" and "Visa's ledgers"—every utterance from McInerney maintained restraint; we simply overinterpreted.

Inflation resides strictly on the narrative side, not within the corporate entity.