This breaks down how Visa makes money as a middleman between banks, not consumers. The key insight: Visa's real moat is using high interchange fees (what merchants pay card-issuing banks) to fund consumer rewards, creating a self-reinforcing loop. Key holdings: Visa (earns ~$0.075 per transaction, market cap exceeds all its former owner banks combined), Chase (biggest Visa issuer and acquirer, could bypass Visa with its own network), Apple/Google (biggest threat—phone OS controls default card selection).
Visa began in 1958 as the BankAmericard credit card program, later evolved into a non-profit bank consortium, and ultimately restructured and went public in 2007. Today, its market capitalization exceeds the combined value of all its original member banks. The report’s core argument is that Visa is
Guest Alex Rampell (Partner at Andreessen Horowitz, founder of Affirm and TrialPay, which was acquired by Visa) deconstructs Visa's unique business model as a global interbank communication protocol layer. The most insightful takeaway of the episode: Visa is essentially a "protocol business," and its deepest moat is not merely network effects, but rather the use of high interchange fees to incentivize issuers to offer cashback rewards to consumers, creating a three-way alignment of interests among consumers, issuers, and the network. This structural dynamic means any alternative attempting to enter from the merchant side faces the dilemma of "solving only one side's pain point while failing to sway the other side's consumers."
Alex Rampell argues that most people misunderstand Visa because they have never directly dealt with it. A credit card transaction involves five parties: the consumer, the merchant, the issuing bank (e.g., Capital One), the acquiring bank (e.g., Bank of America Merchant Services), and the network (Visa/MasterCard) that connects them. Visa's role is purely that of an information router—when a consumer swipes a card, the acquiring bank sees the card number starts with "4" (Visa's identifier) and sends the transaction to Visa; Visa identifies the issuing bank based on the first six digits of the BIN code and forwards the request to that bank for approval. The issuing bank assumes the credit risk and decides whether to approve.
Key data: In a $100 transaction, the merchant receives approximately $97, with the remaining $3 distributed among the three parties. The vast majority goes to the issuing bank, while Visa earns only about $0.075 per transaction on average. The interchange rate is extremely complex—Visa's website features a 25-page fee schedule, varying by card type (debit/credit, signature card/standard card), transaction method (card-present/card-not-present), and merchant category (utilities/gas stations, etc.).
Mechanism breakdown: The higher interchange rate for Visa Signature Cards is essentially a product "designed" by Visa—using higher merchant costs to incentivize issuing banks to offer the card, with the banks then using this revenue to provide consumers with higher cashback (e.g., 2.5% instead of 2%). Alex points out that this creates a counterintuitive competitive logic: normally, one would expect fees to decline due to competition, but within the Visa/MasterCard duopoly structure, fees instead have upward momentum—because higher interchange fees mean stronger consumer cashback incentives, thereby driving more transaction volume.
Alex Rampell argues that Visa's origin story is key to understanding its business model—it was born out of a fragmented landscape where banks could not directly communicate with each other, rather than anyone's strategic design. On September 18, 1958, Joe Williams of Bank of America mailed credit cards (no application required at the time) to 60,000 people in Fresno and signed Florsheim Shoes as the first merchant. The initial fee was around 6%, but due to rampant fraud, Williams was fired. However, this aggressive "issue cards first, find merchants later" strategy solved the chicken-and-egg problem of network effects—Bank of America already had 45% of the city's residents as customers.
Over the following decades, multiple banks issued their own credit cards, leaving consumers with 10-20 different cards in their wallets, each usable only at specific merchants. This eventually led to the formation of two major bank alliances: Master Charge (later MasterCard) and BankAmericard (later Visa). Both were nonprofit organizations—jointly owned by member banks, with Visa extracting no economic profit and only charging fees to cover technology development.
Turning point: In 2007, Visa and MasterCard went public. Alex emphasizes that the motivation for the IPO was not to "seize an opportunity" but to avoid antitrust scrutiny—because the pricing committee jointly owned by banks set interchange fees annually, which essentially constituted price fixing. The IPO was primarily a secondary offering (banks selling their stakes), with minimal new capital raised. Today, Visa's market cap exceeds the combined value of all its original owner banks—Alex calls this "patricide": Visa was founded by Bank of America, yet its market cap now far surpasses that of Bank of America, because it has transformed from one of N participants into the central router through which all transactions must pass.
Alex Rampell argues that Visa’s greatest risk is not technological disruption, but the concentration of issuers and acquirers. Visa’s business lifeline is its issuer relationships—acquirers must accept all cards, but issuers can choose Visa or MasterCard. If Chase (both the largest Visa issuer and the largest online acquirer) says, “I don’t want to pay Visa for Chase card transactions at Chase merchants,” this is what is known as an “on-us transaction”—technically entirely feasible, requiring just a line of code to route the transaction to a different department in the same building. In Mexico, almost no transactions run over the Visa network because issuers and acquirers are one and the same. Alex warns: if the U.S. is left with only 10 large banks that all act as both issuers and acquirers, Visa could see an increasing number of transactions internalized.
Alex believes geopolitical risk deserves more attention than interchange fee regulation. He cites the Crimea incident as an example: after U.S. sanctions, Visa directly cut off transaction routing in the region. This exposed a fundamental issue—a country’s fiat currency now runs on networks controlled by foreign private entities. Alex draws an analogy to the U.S. establishing a Strategic Petroleum Reserve after the OPEC oil embargo: countries may be forced to build their own payment networks as a “strategic payment reserve.” China has UnionPay (CUP), and Russia has already begun building its own network. The trend of “networks becoming national security assets” may force Visa to set up entirely independent network instances in each country.
Australia was the first to cap interchange fees below 50 basis points, with Europe going even lower. Alex points out that the outcome has been detrimental to consumers—reward points have disappeared and annual fees have reappeared. This is because lowering interchange fees essentially transfers wealth from issuers (used for consumer rewards) to large merchants (such as Walmart), but consumers are insensitive to savings on small individual transactions, while the concentrated benefits for large merchants far outweigh the dispersed losses for consumers. The U.S. Durbin Amendment, which caps debit card interchange at 21 cents plus 5 basis points, has instead resulted in interchange fees for a $2 coffee exceeding 10%—more expensive than unregulated credit cards.
Alex Rampell argues that the greatest potential threat to Visa is not cryptocurrencies or new payment companies, but mobile operating systems. When wallets evolve from "dead cowhide wallets" to mobile software, the logic behind selecting a default card changes completely—Apple sorts alphabetically, benefiting American Express. More importantly, mobile operating systems introduce a "permission-based" mechanism: just as apps request access to contacts or location, future financial products will be discovered, authorized, and used directly on mobile devices. Consumers can instantly add new cards and set default cards, upending the traditional model where issuers build brand awareness through TV ads.
Regarding Stripe: Alex finds Stripe interesting because it is simultaneously one of the largest online acquirers and has begun issuing cards. If Stripe becomes both an issuer and an acquirer, it could self-clear—but Visa would still court it, as cards issued by Stripe may be used at merchants on the Visa network (e.g., Chipotle). Chase should have leveraged its dual identity (15-20% of Visa cards are issued by Chase) to dominate online acquiring, but failed to execute—Alex believes Stripe is more likely to succeed.
Regarding Plaid: Alex is a Plaid investor, but he believes Plaid does not directly compete with Visa—it focuses on data aggregation (e.g., Robinhood verifying users' bank account balances), not transaction processing. Plaid is more like "Visa in the non-commercial space," but it cannot solve the core pain point for consumer payments: consumers want cashback rewards, not direct payments from bank accounts.
Regarding Cryptocurrencies: Alex argues that if Visa were designed from scratch, a decentralized protocol (such as blockchain) would be a superior solution—all banks act as nodes hosting the ledger, no one can control it, and lawyers cannot turn it into a for-profit entity. But the window for "protocols over lawyers" closed 60 years ago. As a payment method, cryptocurrencies face a fundamental contradiction: they are deflationary assets, and consumers are reluctant to spend coins they expect to appreciate.
| Position | Guest Stance | Key Data |
|---|---|---|
| Visa | Bullish on moat, flags concentration and geopolitical risks | Earns $0.075 per transaction on average; market cap exceeds the combined value of its original owner banks |
| MasterCard | Comparative analysis, competitive relationship | Duopoly with Visa in open-loop networks |
| American Express | Comparative analysis, closed-loop model | Controls both issuing and acquiring sides |
| Chase | Risk warning (concentration) | Largest Visa issuer + largest online acquirer; ChaseNet can bypass Visa |
| Citi | Risk warning (concentration) | Largest MasterCard issuer |
| Stripe | Potential threat/opportunity | Engages in both acquiring and issuing, may self-clear |
| Plaid | No direct competition | Data aggregation, not transaction processing |
| Apple/Google | Biggest potential threat | Operating system controls default card selection and permissioning |
| Target | Case study (merchant perspective) | If interchange fees were eliminated, after-tax profit would double; Target Red Card offers 5% cashback |
| Walmart | Case study (most aggressive merchant against Visa) | Pays the most in interchange fees; MCX project failed |
| Dunkin' Donuts | Case study (small merchant pain point) | Interchange fee on a $2 coffee exceeds 10% under Durbin regulation |
| China UnionPay | Geopolitical case study | Domestic transactions in China do not go through Visa; cross-border transactions use the Visa network |
1. "Visa is essentially a protocol business, not a financial company." (Alex Rampell) — It does not bear credit risk, does not own consumer or merchant relationships, and merely serves as an information router between banks. Its revenue comes from a network fee of approximately $0.075 per transaction, not the bulk of interchange fees.
2. "Visa's moat is two-layered: network effects + economic incentive lock-in." (Alex Rampell) — Not only must merchants accept it because everyone uses it, but high interchange fees allow issuing banks to offer cashback, prompting consumers to actively choose cards, forming a self-reinforcing flywheel.
3. "Visa's IPO was not for fundraising, but to avoid antitrust — banks collectively pricing is price fixing." (Alex Rampell) — During the non-profit alliance era, banks negotiated interchange rates annually through a pricing committee, essentially constituting price fixing. Going public transferred pricing power from the bank alliance to an independent entity.
4. "Visa's market cap exceeds the combined value of all its original owner banks — this is 'patricide'." (Alex Rampell) — Visa was founded by Bank of America, yet its market cap now far exceeds that of Bank of America, because it evolved from one of N participants into the central router through which all transactions must pass.
5. "Networks are the new strategic petroleum reserves." (Alex Rampell) — After the Crimea sanctions, Visa cut off routing, exposing the fundamental vulnerability of sovereign fiat currencies running on foreign private networks. Countries may be forced to build their own payment networks.
6. "The biggest threat to Visa is not cryptocurrency or Stripe, but mobile operating systems." (Alex Rampell) — When wallets become software, the default card selection logic is controlled by the operating system; permission-based mechanisms allow consumers to instantly add/switch cards, upending the traditional brand mindshare model of issuing banks.
7. "If designing Visa from scratch, a decentralized protocol would be the better solution — but the window closed 60 years ago." (Alex Rampell) — Decentralization would mean no one can control the network, and lawyers cannot convert it into a for-profit entity. However, Visa's existing centralized architecture is deeply entrenched, and protocols cannot retroactively dismantle the legal structure built by lawyers.
8. "Lowering interchange fees is detrimental to consumers — rewards disappear, annual fees return." (Alex Rampell) — Regulatory experiments in Australia and Europe have proven: lowering interchange fees essentially transfers wealth from consumer rewards to large merchants, but consumers are indifferent to small per-transaction savings, while the concentrated benefits for large merchants far outweigh the dispersed losses.