Visa: How Bank of America Solved Payments
On September 18, 1958, Joseph Williams, a manager at Bank of America, mailed 60,000 credit cards to residents of Fresno, California. No one had asked for them. The recipients, housewives, mechanics, shopkeepers, found plastic cards in their mailboxes embossed with their names, granting them $300 of revolving credit at local merchants who had not yet agreed to accept the card. It was an audacious solution to the classic chicken-and-egg problem of payments: consumers would not adopt a payment method merchants did not accept, and merchants would not install infrastructure for consumers who did not exist. Bank of America solved this by owning both sides temporarily, everaging its deposit base to create consumer demand ex nihilo, then strong-arming Fresno merchants into compliance.
What began as BankAmericard would eventually become Visa, but the essential economic architecture was established that afternoon in Fresno. Visa is not a bank. It does not lend money, bear credit risk, or hold deposits. It is a protocol: set of technical and commercial rules that allows fragmented, competing financial institutions to route payment information between one another. Like the internet protocol TCP/IP, Visa provides the lingua franca that allows Chase (an issuer) to communicate with First Data (an acquirer) when a consumer buys groceries. For this routing service, Visa captures roughly 7.5 cents per transaction. The remaining economics flow to the issuing bank, which uses the bulk of the 2-3% interchange fee to fund consumer rewards, cash back, airline miles, purchase protection, that drive card adoption.
This arrangement has made Visa the most valuable financial intermediary in history, with a market capitalization exceeding that of Bank of America itself. Yet the model rests on a peculiar historical foundation. Visa operated as a nonprofit consortium owned by banks for approximately fifty years before its 2007 initial public offering. This nonprofit incubation was not a gimmick; it was the essential condition that allowed Visa to accumulate the network effects and switching costs that would be illegal to build today. Any attempt to coordinate pricing across competing banks in the current regulatory environment would trigger immediate antitrust action. Visa’s dominance is, in effect, grandfathered in: structural advantage bequeathed by a half-century of benign neglect.
The Interchange Engine
The economic engine of Visa is deceptively simple: revenue scales linearly with transaction volume, while fixed costs remain minimal. Once the core switching infrastructure is built, adding volume requires negligible marginal investment. The protocol rarely changes; the 20-character limit on merchant names and the T+2 settlement cycle are legacies of 1960s mainframe computing that persist today. This creates operating leverage of exceptional power. When a consumer swipes a Visa card in Mumbai or Miami, the company collects its few cents without bearing the credit risk of default, the cost of customer acquisition, or the expense of merchant servicing. Those burdens fall to the issuing and acquiring banks.
What makes this sustainable is a delicate alignment of incentives centered on the interchange fee. When a consumer spends $100, approximately $2.50 to $3.00 is extracted from the merchant as interchange. Of this, Visa keeps roughly 7.5 cents. The issuing bank captures the remainder, $2.00 or more, which it uses to fund rewards programs, fraud protection, and marketing. This creates a virtuous cycle that Visa does not control but absolutely depends upon. Issuing banks compete aggressively to issue Visa cards because Visa’s interchange rates (particularly for premium credit products) are higher than those of debit networks or competing card schemes. Higher interchange funds richer rewards, which drives consumer preference for Visa cards, which forces merchants to accept Visa, which increases transaction volume.
The genius of the model is that Visa has outsourced both consumer acquisition and merchant acquisition to parties with deeper pockets and stronger relationships. Banks market Visa cards because the economics are compelling; merchants accept them because consumers demand them. Visa sits in the middle, insulated from competitive pressure on either side. Consumers do not choose Visa; they choose Chase Sapphire or Capital One Venture. Merchants do not negotiate with Visa; they negotiate with acquiring banks. Visa is the invisible protocol layer, the railroad tracks upon which bank marketing machines run.
This insulation is structural. Because Visa operates as an “open loop”, connecting thousands of independent issuers and acquirers, it achieves scale that American Express, with its “closed loop” model (owning both consumer relationships and merchant contracts), cannot match. But this openness comes at a cost. Visa cannot unilaterally improve the customer experience. If Visa wishes to enhance fraud detection, reduce chargeback friction, or provide item-level purchase data to merchants, it must secure the cooperation of thousands of banks, each with competing interests and legacy systems. The protocol belongs to everyone and therefore to no one.
The Innovation Ceiling
The most underappreciated constraint on Visa’s business is not competition but rigidity. The protocol is encased in amber. Changing the technical standards, llowing, for instance, 200-character merchant names instead of 20, or transmitting item-level SKU data rather than aggregate transaction amounts, or settling transactions in real time rather than T+2, equires unanimous agreement from the member banks. Unanimity is nearly impossible to achieve when issuers and acquirers have divergent interests, when smaller banks fear disadvantage against giants like Chase and Bank of America, and when any change threatens to disrupt legacy mainframe systems.
This technical debt is not merely an operational inconvenience; it is a strategic vulnerability. Modern payment systems, buy-now-pay-later platforms, digital wallets, real-time payment rails, can offer merchants rich data, instant settlement, and integrated loyalty programs. Visa cannot match these features without breaking the protocol. If Visa were to unilaterally modernize its data fields or settlement timelines, it would alienate the thousands of banks whose cooperation is mandatory for network function. The banks own the customer relationships; Visa owns only the rails. If the banks decided that Visa’s modernization efforts threatened their economics or operational stability, they could migrate to MasterCard or attempt to build bilateral connections.
The result is an innovation ceiling that is invisible to consumers but increasingly apparent to merchants. A merchant processing Visa transactions today receives truncated, often cryptic merchant descriptors (twenty characters barely accommodates a business name, let alone location detail) and must wait two days for funds to settle. They receive no item-level data, isa knows the consumer spent $47.50 at a grocery store but not that they bought diapers rather than wine. Modern treasury management and inventory systems demand this granularity, but Visa cannot provide it without reconstructing the protocol from scratch.
Worse, Visa cannot reduce interchange rates to compete with alternative payment methods without collapsing the issuer economics that drive its distribution. If Visa were to lower the cost of acceptance for e-commerce merchants or small businesses, it would reduce the revenue pool available for consumer rewards. Issuing banks would then have less incentive to promote Visa cards, potentially shifting marketing budgets to MasterCard or to proprietary payment products. Visa is trapped: it cannot improve the merchant experience without angering the issuers who generate its volume, and it cannot modernize the protocol without unanimous consent from thousands of competing banks.
The Pricing Council Problem
The mechanism that sets interchange rates, the “pricing council”, represents the most fragile element of Visa’s architecture. In effect, competing banks convene through Visa’s governance structure to coordinate pricing. This is cartel behavior by another name, and Visa exists in a perpetual state of managed tension with antitrust authorities. The company maintains that interchange is necessary to fund network security and rewards, but merchants, and increasingly regulators, view it as a tax on commerce, a transfer from retailers (and ultimately consumers) to bank shareholders and affluent cardholders.
The regulatory threat is not theoretical. The European Union and Australia have already imposed strict caps on interchange rates, collapsing the economic model in those regions. In Australia, regulatory intervention reduced credit card interchange from roughly 0.95% to 0.50%, directly compressing issuer profitability and reducing the generosity of rewards programs. The result was not the collapse of the payment system, but a deceleration in card adoption and a shift toward debit and alternative payment methods. If similar caps were imposed in the United States, where Visa earns the majority of its profit, the consequences would be severe. Issuing banks might reduce card issuance, cut rewards, or migrate volume to networks with more favorable economics.
Visa’s public offering in 2007 was itself an attempt to address this fragility. As a bank-owned nonprofit consortium, Visa’s pricing coordination was transparently problematic under antitrust law. By going public, Visa could claim to operate as an independent entity setting prices for services rendered, rather than as a vehicle for bank collusion. But the fiction is thin. The pricing council still meets. The banks still effectively determine the rates that fund their own marketing. If political sentiment shifts, if merchants successfully frame interchange as a regressive tax that inflates prices for cash customers while subsidizing wealthy cardholders, Visa could face forced rate reductions that would unwind the entire incentive structure.
The Open-Loop Handcuff
American Express offers a revealing counterfactual. By operating as a closed loop, issuing its own cards and contracting directly with merchants, Amex maintains end-to-end control of the customer experience. It can offer rich transaction data, sophisticated fraud prevention, and premium customer service because it does not need permission from thousands of partner banks. When Amex detects fraud, it acts unilaterally. When it wishes to offer merchant financing or enhanced analytics, it builds the product itself.
Visa cannot do this. Its open-loop model creates network effects but prevents product differentiation. If Visa wishes to respond to the threat of buy-now-pay-later (BNPL) services, which bypass card networks entirely by embedding lending at the point of sale, it cannot simply launch a competing product. It must convince thousands of issuing banks to participate, integrate with their systems, and agree on economics. By the time this coordination is achieved, the market has moved on. The protocol’s strength, its universality, is also its weakness. It is a dumb pipe, and dumb pipes cannot pivot.
This limitation extends to data. In an era where merchants covet granular consumer insights to drive loyalty programs and inventory management, Visa sits on a goldmine of transaction data that it cannot fully monetize. The banks, not Visa, own the customer relationships and the associated data rights. Visa sees the transaction flow but cannot easily attach marketing messages, offers, or financial services to specific consumers without bank cooperation. The open loop diffuses value capture across thousands of institutions, leaving Visa with its meager 7.5 cents while others capture the customer lifetime value.
The Exogenous Growth Trap
Perhaps the most profound tension in Visa’s model is that growth is almost entirely exogenous. The company has no product lever to pull, no innovation cycle to accelerate, no customer segment to convert. It is a pure bet on the secular shift from cash to digital payments, on the expansion of cross-border commerce, and on the continued proliferation of credit card rewards. When these trends falter, Visa has no recourse.
The COVID-19 pandemic provided a natural experiment. Cross-border transaction volume, a high-margin segment where Visa earns significantly more than domestic payments, collapsed as travel ceased. Visa could not respond by lowering fees to stimulate volume, or by launching new services to capture alternative flows. It simply waited for airplanes to fill again. The stock price recovered as travel resumed, but the episode revealed the business’s essential passivity. Visa does not drive commerce; it taxes it.
In mature markets like the United States, the cash-to-digital transition is approaching its limit. Future growth must come from emerging markets, where banking infrastructure is weak and regulatory scrutiny of interchange is often severe. Visa’s expansion into unbanked regions is promising but uncertain; it depends on the construction of banking systems, the adoption of Western-style consumer credit, and the forbearance of regulators who may view Visa’s fees as an extractive burden on developing economies.
There is a scenario, distant but not impossible, where Visa becomes the victim of its own success. If regulators cap interchange, if merchants successfully organize boycotts or surcharges that steer consumers to alternative payment methods, if central bank digital currencies (CBDCs) offer instant, low-cost settlement that bypasses card networks entirely, Visa would have no defense. It cannot cut prices without losing bank support. It cannot improve the product without unanimous consent. It cannot pivot to a closed-loop model without sacrificing scale. It is a protocol optimized for a specific historical moment: moment when banks were fragmented, cash was king, and regulators tolerated coordinated pricing.
The Protocol’s Endgame
Visa represents a peculiar category of business: the incumbent protocol that captured the surplus from solving a coordination problem. By providing the standard that allowed competing banks to transact, Visa extracted a perpetual toll from global commerce. But the conditions that created this position, the fragmented banking system, the nonprofit incubation period, the political tolerance for interchange-funded rewards, are not permanent features of the landscape. They are historical contingencies.
The company now faces a future where the protocol itself becomes a liability. Legacy constraints, 20-character merchant names, T+2 settlement, the inability to transmit item-level data, make Visa increasingly anachronistic in a world of real-time payments and rich data exchange. Meanwhile, the pricing mechanism that funds the entire edifice rests on political forbearance that grows thinner as merchants and regulators recognize the regressive nature of the interchange subsidy.
Visa will likely survive and prosper for decades. The switching costs are immense, the network effects self-reinforcing, the bank relationships entrenched. But the model is brittle in ways that matter. It cannot evolve. It cannot cut prices. It cannot control its own destiny. It is, in the end, a tax collector waiting for the economy to grow, hoping that regulators do not notice the size of the levy, and praying that the banks it serves do not find a better way to route their payments.
The Tyranny of Unanimity
To understand why Visa cannot modernize, consider the specific gravity of the technical constraints. The 20-character merchant name limit is not an aesthetic quirk; it is hard-coded into the ISO 8583 messaging standard that Visa adopted in the 1980s, which itself was built atop the magnetic stripe specifications of the 1960s. When a consumer swipes a card, the transaction travels through the network as a fixed-format message where Field 43 (Card Acceptor Name/Location) is allocated exactly 40 bytes, nough for twenty Latin characters, or fewer if the system supports Unicode. This field cannot be expanded without reconfiguring the message parsers at every issuing bank, every acquiring bank, and every payment processor in the chain. A bank running 1980s-era mainframe COBOL systems, which describes a surprising number of regional and community banks in the United States, would require months of engineering work to accommodate a longer field. If even one critical participant fails to upgrade, the entire protocol breaks; truncated messages create reconciliation errors, chargeback disputes, and settlement failures.
This is why Visa still cannot transmit item-level purchase data. The protocol was designed to move money, not information. When a consumer buys groceries, the merchant’s terminal sends a total dollar amount ($47.50) but cannot append the SKU data (diapers, not wine) because the message format lacks the fields. Modern merchants treasure this granularity for inventory management and targeted marketing; they receive it from their internal point-of-sale systems but not from the payment network. Visa has spent years attempting to develop “enhanced data” products, but these require merchants to transmit information outside the standard authorization flow, usually through separate batch files that acquirers must agree to process. Adoption remains spotty because acquirers, who make their money on transaction volume, not data enrichment, have little incentive to invest in expensive infrastructure upgrades for marginal benefit.
The chip migration provides a sobering case study in protocol rigidity. EMV (Europay-MasterCard-Visa) standards were developed in the 1990s, yet the United States did not begin widespread adoption until 2015, wenty years later. The delay was not technological; the chip technology was mature. The obstacle was coordination. Shifting from magnetic stripes to embedded chips required new terminal hardware for millions of merchants, new card manufacturing processes for thousands of issuers, and new certification protocols for the processors in between. Because Visa could not force adoption, could not simply “upgrade” the network unilaterally, it had to rely on liability shifts. In October 2015, Visa changed the rules so that merchants who accepted fraudulent magnetic-stripe transactions bore the cost rather than the issuing bank. This created a financial incentive for adoption, but the transition still took years and cost billions. Gas stations, with their specialized pumps and regulatory complications, received an extension until 2020, five years after the initial deadline. If a simple security upgrade, one that reduced fraud for all parties, required two decades and regulatory threats to implement, the prospect of modernizing data fields or settlement timelines appears quixotic.
The settlement timeline presents an even more intractable problem. The T+2 cycle (transaction date plus two days) for merchant funding exists because the original system required physical transportation of paper drafts. Merchants deposited sales drafts at their banks; the acquirer shipped them to a processing center; Visa sorted and routed them to issuing banks; the issuer verified signatures and available credit; then authorization traveled back through the same channels. Even after electronic authorization replaced paper, the two-day float remained embedded in the economic model. Issuing banks profit from the float: he interest earned on funds held between the transaction and settlement. Acquiring banks have built treasury management products around the predictability of the cycle. Changing to real-time settlement would require unanimous consent to eliminate a revenue stream for issuers and to rebuild cash management infrastructure for acquirers. It would also expose the system to different fraud risks; the two-day window allows for provisional review of suspicious transactions. Visa has watched the rise of real-time payment rails, India’s UPI, Brazil’s Pix, the Federal Reserve’s FedNow, with a mixture of envy and dread. These systems settle instantly and carry rich data, but Visa cannot replicate their speed without destroying the economics that fund its bank partners.
Precedents in Price Control
The regulatory bear case is not speculative; it has already occurred in Australia and Europe, providing a laboratory for what happens when interchange rates fall. In 2003, the Reserve Bank of Australia (RBA) intervened in payment card pricing, concluding that the “honor all cards” rule and no-surcharge restrictions were anti-competitive. The RBA capped credit card interchange at 0.50 percent, down from an average of 0.95 percent. The results were immediate and instructive. Issuing banks, facing a 47 percent reduction in interchange revenue, slashed rewards programs. Australian consumers, who had enjoyed generous frequent-flyer linked cards, found that points accumulated at half the previous rate or disappeared entirely. Annual fees increased to compensate for lost revenue. Card issuance slowed; consumers shifted toward debit and alternative payment methods.
Critically, the sky did not fall. Payments continued; commerce persisted. The reduction in merchant costs did not translate into significantly lower consumer prices, merchants mostly pocketed the savings, but neither did the payment system collapse. What changed was the incentive structure for bank marketing. With thinner margins, banks reduced their aggressive pursuit of credit card customers. Visa’s transaction growth in Australia decelerated markedly in the years following the cap. The market became less “credit-centric,” reducing Visa’s effective tax rate on Australian commerce.
Europe followed a similar trajectory. The 2015 Interchange Fee Regulation (IFR) capped consumer credit card interchange at 0.30 percent and debit at 0.20 percent across the European Union. The impact on rewards was devastating. Premium travel cards that had offered lounge access and concierge services were devalued or withdrawn. Banks reduced credit lines and tightened underwriting standards, as the profitability of subprime cardholding evaporated. Visa’s European revenues per transaction declined, and the company was forced to rely on volume growth and value-added services to compensate, ervices that, as noted, are difficult to deploy without bank cooperation.
In the United States, the Durbin Amendment to the Dodd-Frank Act of 2010 provided a preview, though only for debit. The amendment, championed by Senator Richard Durbin, capped debit interchange for large banks at roughly 21 cents per transaction, down from an average of 44 cents. The immediate result was the collapse of debit rewards programs, ree checking accounts that had offered modest cash back disappeared overnight. Banks began charging for checking accounts or imposing minimum balance requirements. More significantly, large retailers like Walmart and Target invested in steering customers toward debit and away from credit, recognizing that the regulated rate was cheaper than the unregulated credit interchange. Visa lobbied ferociously against the Durbin Amendment, recognizing that if debit could be capped, credit was the next logical target.
The political arithmetic for credit interchange is more complex than for debit. Credit cards distribute benefits regressively; affluent consumers with excellent credit receive hundreds of dollars annually in rewards, funded by interchange fees paid by all consumers, including cash users who tend to be poorer. As this dynamic becomes more widely understood, through academic studies by the Federal Reserve and popular journalism, political pressure mounts. Senators from both parties have introduced legislation to allow merchants to route credit transactions over competing networks, which would drive interchange rates down through competition. The Credit Card Competition Act, introduced in 2022 and 2023, would require banks with over $100 billion in assets to enable at least two unaffiliated networks on their credit cards, otentially allowing a merchant to route a Chase Visa transaction through a lower-cost network like Pulse or Nyce, bypassing Visa’s interchange entirely.
If such legislation passes, Visa faces not a cap but a death-by-a-thousand-cuts. Merchants would invariably route transactions toward the lowest-cost option, forcing Visa to reduce its rates to remain competitive. The high interchange that funds rewards would compress toward the marginal cost of processing. Visa’s 7.5 cent take might survive, but the issuing bank’s $2.00 would shrink to pennies. Banks would then face a choice: accept lower profitability on credit cards, or reduce rewards and watch consumers migrate to debit, Buy Now Pay Later, or cash. Either outcome reduces Visa’s volume.
The Concentration Crisis
While Visa’s network comprises thousands of banks, the economic reality is concentrated power. In the United States, four institutions, JPMorgan Chase, Bank of America, Citigroup, and Capital One, issue roughly 70 percent of Visa credit cards by volume. This oligopoly creates a vulnerability that network effects alone cannot remedy. If one of these giants defected, building a bilateral payment system or migrating exclusively to MasterCard, Visa would face an existential revenue cliff.
Chase illustrates the danger most acutely. The bank is both the largest issuer and a sophisticated technology company with its own payment rail, ChaseNet. Originally built for private label and co-brand cards, ChaseNet processes transactions without routing through Visa or MasterCard. For years, Chase has used ChaseNet for specific merchant relationships, capturing the interchange revenue that would otherwise flow to the networks. If Chase were to expand ChaseNet to encompass its general-purpose credit cards, if it convinced major merchants like Amazon or Walmart to accept Chase cards bilaterally, Visa would lose a double-digit percentage of U.S. volume.
The bilateral threat extends beyond single banks. Amazon, which operates as both a massive merchant and a growing financial services provider (Amazon Pay, Amazon Lending), has the scale to negotiate direct connections with issuing banks. If Amazon told consumers, “Link your Chase account directly to Amazon Pay and save 2% on every purchase,” bypassing the card network entirely, both the merchant and the issuer would split the savings from avoided interchange. Visa would see nothing. This is not science fiction; it is the model that Alipay and WeChat Pay perfected in China, where card networks were marginalized by wallet-based direct debiting from bank accounts.
The concentration risk amplifies the protocol rigidity. If Visa possessed a truly superior, modernized protocol, rich data, instant settlement, superior fraud detection, it might tempt these large issuers to remain despite higher costs. But because the protocol is legacy-constrained, the large banks have less incentive to stay. They do not receive differentiated value from Visa’s rails; they receive commodity switching services that they could replicate or replace. MasterCard offers an identical product. Real-time payment rails offer a cheaper alternative for certain transactions. The only thing keeping Chase embedded in Visa is the ubiquity of acceptance and the inertia of reissuing millions of plastic cards with new numbers. But inertia is not loyalty.
The Fiction of Governance
Visa’s transition from nonprofit consortium to public corporation in 2008 was supposed to resolve the antitrust vulnerability while preserving the economic model. The IPO raised $17.9 billion, with proceeds used to establish litigation reserves and buy out the member banks. The narrative was that Visa would become an independent utility, setting prices based on market conditions rather than bank collusion. The reality is more cynical.
The pricing councils persist. Interchange rates are still set through committees dominated by large issuing banks. Visa’s board includes representatives from the major banks, maintaining the coordination mechanism that the Department of Justice has investigated repeatedly. The difference now is that Visa, as a public company, has a fiduciary duty to maximize shareholder value, which paradoxically conflicts with the interests of its bank partners. When Visa reports record profits, it does so by extracting value from the system that could otherwise flow to issuing banks as higher interchange. The tension is subtle but real: Visa wants to maximize its 7.5 cents; banks want to maximize their $2.00. In a capped or competitive environment, these interests collide.
The litigation history reveals the fragility of the “independent utility” story. In 1996, Walmart led a class-action lawsuit against Visa and MasterCard over debit card acceptance rules, settling for $3 billion in 2003. In 2004, American Express and Discover sued Visa and MasterCard over exclusionary rules that prevented banks from issuing competing cards, resulting in a $2.25 billion settlement and the end of the exclusivity agreements. These suits were brought when Visa was still a bank-owned entity; the banks were effectively suing themselves. The 2008 IPO was partly designed to wall off this litigation risk, to make Visa a separate legal entity that could bear the cost of antitrust judgments without bankrupting the member banks.
Yet the fundamental arrangement, competing banks coordinating pricing through a common platform, remains legally precarious. The Supreme Court’s 2018 decision in Ohio v. American Express provided temporary relief, ruling that Amex’s anti-steering provisions did not violate antitrust law because they preserved the two-sided market. But the ruling was narrow, and the dissent, authored by Justice Breyer, highlighted the danger of allowing networks to enforce rules that merchants and consumers cannot escape. A future court, or a future Congress, could easily view Visa’s pricing councils as horizontal price-fixing, per se illegal under the Sherman Act.
The Endogenous Limit
Visa’s defenders point to the network’s durability, its 99.999% uptime, its fraud rates that are lower than cash or checks. These are real advantages. But they are static advantages, defenses of a position rather than engines of growth. Visa cannot invest its way out of the protocol trap. It cannot acquire its way into modernity; buying a fintech startup does not change the ISO 8583 standard. It cannot price its way to growth; lowering fees collapses the bank ecosystem that generates volume.
The company has attempted to diversify. Visa Direct pushes payouts to debit cards, competing with ACH. Visa B2B Connect attempts to capture commercial payments. Visa Rewards seeks to monetize the data it does control. But these are marginal experiments, tiny streams feeding into a river whose course is fixed. The core business, consumer credit card routing, remains 80% of revenue, and that revenue is a function of global transaction velocity, over which Visa exercises no meaningful control.
In this light, Visa’s market capitalization represents a bet on regulatory forbearance and technological stagnation. The market assumes that governments will not cap interchange, that banks will not build bilateral rails, that consumers will not abandon rewards-funded credit for debt-free alternatives, and that the 1960s protocol will persist because coordination is too difficult. These are not unreasonable assumptions; they have held for decades. But they are assumptions about the stability of institutions and the slowness of change, not about the excellence of the product or the inevitability of the business model.
When historians look back at the payments industry, they may view Visa as a brilliant solution to a temporary problem: he need to coordinate trust between fragmented, competing banks in an era of analog commerce. The protocol worked because the alternative was cash, checks, and inefficiency. As commerce digitizes, as real-time settlement becomes standard, as central banks issue digital currencies that settle instantly with no intermediary, the 7.5 cent toll may come to look not like a natural monopoly but like a historical accident: tax levied by a committee of banks through a messaging standard designed when computers filled entire rooms. The tax has been collected for sixty years. The question is whether the next sixty will tolerate it.