Here’s a question that trips people up: How does Visa make money if it doesn’t actually lend you money or charge you interest? Most people assume Visa is a bank. It’s not. Visa doesn’t issue credit cards, doesn’t approve your purchases, and doesn’t carry the risk when you don’t pay your bill. Those are all handled by the bank whose name appears on your card next to the Visa logo.
What Visa does is run the network. When you swipe, tap, or insert your Visa card at a store, Visa’s systems route the authorization request from the merchant’s bank to your bank, get approval in milliseconds, and settle the transaction later. For this service, Visa collects a small fee. The
brilliance of this model is that it scales infinitely without the messy parts of banking like credit risk, loan defaults, or regulatory capital requirements.
By fiscal year 2024, Visa had built this into one of the most profitable businesses on earth:
- $35.9 billion in net revenue for FY2024, up 10% year-over-year
- 233.8 billion transactions processed on Visa’s networks globally
- $15.7 trillion in total payment volume flowing through the system
- 4.6 billion Visa payment credentials in circulation worldwide
- ~67% operating margin on net revenue, among the highest of any large company
Visa generates more profit per dollar of revenue than almost any business at scale. Banks that issue Visa cards? They operate on razor-thin margins, dealing with defaults, fraud losses, and regulatory costs. Visa sits in the middle, collects fees on every transaction, and bears almost none of that risk. The decision to be a network rather than a bank wasn’t obvious in 1958 when this all started, but it turned out to be one of the smartest strategic choices in business history.
The Strategic Context That Made Visa a Network
What Bank of America Was Actually Trying to Solve
The story starts on September 18, 1958, in Fresno, California. Bank of America launched something called BankAmericard by mailing 65,000 unsolicited credit cards to customers. This was the brainchild of Joseph P. Williams, who led the bank’s Customer Services Research Group. The goal wasn’t building a payments empire. It was solving a problem for Bank of America’s retail customers.
In the 1950s, the average middle-class American carried multiple revolving credit accounts with different merchants. You had a Sears card for Sears, a Mobil card for gas, maybe cards for local department stores. This system was inefficient for consumers juggling multiple payments and merchants dealing with separate billing systems. Williams saw an opportunity: a single all-purpose credit card accepted at many merchants, issued by one bank.
The early BankAmericard challenges:
- 22% delinquency rate in the first years as Bank of America struggled with fraud and credit management
- $20 million in losses that nearly killed the program and forced Williams’ resignation in December 1959
- Limited geographic reach because US banking laws prevented Bank of America from operating branches outside California
- Operational chaos when the bank began licensing BankAmericard to other banks in 1966, creating “interchange” nightmares between institutions
The program survived because Bank of America saw long-term potential despite near-term disasters. But by the late 1960s, the licensing model was breaking down. Multiple banks were issuing BankAmericards in their territories, but there was no coordinated system for settling transactions between banks. If a customer used a BankAmericard issued by one bank at a merchant served by another bank, figuring out who owed what became increasingly complicated.
When Dee Hock Transformed the Model
Enter Dee Hock. In 1968, Hock was a manager at National Bank of Commerce in Washington state, tasked with rolling out BankAmericard in the Pacific Northwest. He quickly realized the licensing system Bank of America had created was fundamentally broken. Banks didn’t trust each other, interchange fees were disputed constantly, and Bank of America’s attempts to control everything from California created resentment among licensee banks.
Hock proposed something radical: Bank of America should give up control entirely. Instead of one bank licensing its card to others, all the issuing banks should form an independent cooperative that owned and operated the system collectively. No single bank would control it, including Bank of America. This solved the trust problem and aligned everyone’s interests around growing the network rather than fighting over control.
The timeline of Visa’s reinvention:
- 1970: Bank of America relinquished control; issuer banks formed National BankAmericard Inc. (NBI) with Dee Hock as CEO
- 1973: Launched electronic authorization system, the precursor to VisaNet, enabling real-time transaction approvals
- 1975: Introduced first debit card through First National Bank of Seattle, expanding beyond just credit
- 1976: Rebranded from BankAmericard to “Visa” to eliminate Bank of America association and enable global expansion
- 2008: Went public with $17.9 billion IPO, largest in US history at the time, becoming Visa Inc.
Hock’s insight was that the value wasn’t in being a bank. The value was in being the network that connected all the banks. Banks would compete with each other on interest rates, rewards programs, and customer service. But they’d all use the same underlying infrastructure to process transactions. Visa would provide that infrastructure and stay neutral, earning fees from everyone without competing with anyone.
The Options Visa Actually Considered
Option 1: Becoming a Full-Service Bank
The most obvious path would have been for Visa to evolve into a complete banking operation. Instead of just processing transactions, Visa could have issued its own credit cards directly to consumers, set its own interest rates, taken on credit risk, and captured the full profit from lending.
This made intuitive sense. Banks issuing Visa cards were making significant money from interest charges, often 15% to 25% annually on outstanding balances. Visa was collecting tiny fractions of a percent on transaction volume while banks captured the lucrative interest income. Why not cut out the middleman and do it all?
What full-service banking would have offered:
- Direct customer relationships rather than depending on bank partners to distribute Visa cards
- Interest income from credit balances, often representing 70% to 80% of credit card issuer profits
- Complete control over underwriting standards, credit limits, and collections processes
- Ability to cross-sell other financial products like savings accounts, mortgages, and investment services
However, banking came with massive risks and costs Visa’s network model avoided. Credit card lending required maintaining regulatory capital to cover potential defaults. When economic recessions hit and unemployment spiked, banks faced waves of charge-offs as customers couldn’t pay. During the 2008 financial crisis, major credit card issuers saw charge-off rates exceed 10%, meaning one in ten dollars lent was never recovered.
Option 2: Vertical Integration with Select Bank Partners
Rather than going full bank, Visa could have acquired or merged with major card-issuing banks to capture more of the value chain while maintaining the network. This hybrid approach would give Visa some lending profits without fully transforming into a bank.
Several payment companies pursued this strategy. Discover Financial Services both operates the payment network and issues cards directly, capturing both transaction fees and interest income. American Express historically combined network operations with card issuance, though it has gradually opened its network to third-party issuers.
The vertical integration benefits:
- Capture lending profits from owned bank subsidiaries while still operating the broader network
- Ensure at least some guaranteed network volume from captive issuer operations
- Maintain flexibility to stay network-focused where it made sense while selectively owning issuance in key markets
- Provide reference implementation of best practices for independent bank partners to follow
However, vertical integration created conflicts with the banks that were supposed to be Visa’s partners. If Visa owned issuing banks that competed directly with independent member banks, why would those independent banks stay loyal to the Visa network? They might defect to Mastercard or build their own networks. The cooperative structure Dee Hock created depended on Visa staying neutral and never competing with its member banks.
Option 3: Pure Network Operation Without Banking
The path Visa ultimately chose was remaining a pure payment network that connected banks, merchants, and consumers without ever getting into the lending business. Visa would provide the technology infrastructure, brand, security, and transaction processing. Banks would handle everything else: issuing cards, underwriting credit risk, setting interest rates, and dealing with customers.
This model flipped traditional business logic. Rather than trying to capture more of the value chain, Visa deliberately stayed in one narrow layer and let partners handle everything else. The company’s entire business would be collecting small fees on transaction volume.
What the pure network model enabled:
- Zero credit risk exposure since Visa never lends money or carries balances
- Minimal regulatory capital requirements compared to banks needing reserves against potential losses
- Infinite scalability because processing one more transaction costs almost nothing once infrastructure exists
- Neutrality that kept all banks as partners rather than competitors
- Operating leverage where revenue grew with transaction volume but costs stayed relatively fixed
Why Visa Chose the Network-Only Model
The Economics of Zero Credit Risk
Visa’s decision to stay out of lending created a business model with profitability characteristics completely different from banking. Banks make money primarily through net interest margin, the difference between interest charged on loans and interest paid on deposits. This typically generates returns on assets of 1% to 2%, meaning banks earn $1 to $2 profit for every $100 in assets.
Visa doesn’t hold assets in the same way. The company doesn’t need loan portfolios, doesn’t need deposits, doesn’t need branches. It needs data centers, software engineers, and the Visa brand. The result is operating margins that banks could never achieve.
The margin comparison by FY2024:
- Visa operating margin: ~67% on net revenue in FY2024, meaning $0.67 profit for every dollar of revenue
- Typical credit card issuer: 15% to 25% operating margins when times are good, negative during recessions when charge-offs spike
- Visa profit margin: 55% net profit margin, up from 52% in FY2023
- Zero loan losses: Visa has no credit defaults because it doesn’t lend money, avoiding the write-offs that destroy bank earnings during downturns
When the 2008 financial crisis hit, credit card charge-offs at major issuers exploded above 10%. Banks lost billions. Visa’s business barely flinched. Transaction volumes declined temporarily as consumer spending fell, but Visa didn’t have loan portfolios imploding. This resilience through economic cycles made Visa far more valuable than comparable banks despite generating less total revenue.
The operating leverage was extraordinary. Once Visa built VisaNet’s infrastructure, processing additional transactions cost almost nothing. Going from 200 billion transactions to 234 billion transactions in FY2024 didn’t require proportional cost increases. Revenue grew while costs stayed relatively flat, expanding margins automatically.
The Network Effect That Made the Model Work
Visa’s value came entirely from network effects, something impossible to replicate if the company had tried competing with its bank partners. Every additional bank issuing Visa cards made the network more valuable for merchants because more customers could use Visa. Every additional merchant accepting Visa made the network more valuable for banks because customers wanted cards accepted everywhere.
This created a virtuous cycle. By 2024, Visa had 4.6 billion payment credentials in circulation accepted at tens of millions of merchant locations worldwide. No new entrant could build that overnight. Even if a competitor offered better economics, merchants and banks had already invested in Visa infrastructure. Switching costs were high.
How the network effects compounded:
- Merchants accepted Visa universally because the majority of customers carried Visa cards
- Banks issued Visa cards because merchants accepted them everywhere, making Visa cards most useful to customers
- Consumers preferred Visa cards because they worked anywhere, creating demand that banks had to satisfy
- The flywheel accelerated as each new participant made the network more valuable for everyone else
If Visa had become a bank issuing its own cards, it would have destroyed this dynamic. Independent banks would have viewed Visa as a competitor and potentially migrated to Mastercard or other networks. Visa’s neutrality was the foundation of its power. By never competing with partners, Visa ensured everyone stayed committed to growing the network.
The Regulatory Simplicity of Not Being a Bank
Staying out of banking meant Visa avoided the regulatory complexity that constrained banks. Banks face capital requirements mandating they hold reserves against potential losses. During the 2008 financial crisis and subsequent Basel III regulations, these requirements increased significantly, forcing banks to set aside more capital and limiting how aggressively they could lend.
Visa faced none of this. The company didn’t need regulatory capital because it didn’t take deposits or make loans. It needed operational infrastructure and brand investment, both funded easily through cash flow. This freed Visa to return more money to shareholders through dividends and buybacks rather than trapping capital in reserves.
The regulatory advantages:
- No Federal Reserve oversight as a bank holding company subject to stress tests and capital ratios
- No FDIC insurance requirements or assessments paid to protect depositor funds
- No restrictions on interchange fees in the same way banks faced Durbin Amendment caps on debit
- No geographic licensing limitations that constrained bank branch networks before interstate banking
This regulatory simplicity also made Visa far more valuable to investors. Bank stocks typically trade at 1x to 1.5x book value because of the risks and regulatory constraints. Visa traded at over 10x book value because the business model required minimal capital and faced fewer restrictions. Investors paid premium multiples for a payments company versus a bank.
What Actually Happened to Visa’s Network Model
The Scale That Justified Everything
By fiscal year 2024, Visa had proven the network-only model worked at a scale that justified Dee Hock’s original vision. The company processed 233.8 billion transactions across its global network, generating $35.9 billion in net revenue. This represented 10% growth over the prior year despite being one of the largest companies in the world by revenue.
The FY2024 operational results:
- $15.7 trillion in total payment volume, up from $14.8 trillion in FY2023
- $13.2 trillion in payments volume (excluding cash withdrawals), up 7% year-over-year
- $16.1 billion in service revenue, up 9% as payment volumes in the prior quarter drove current period recognition
- $17.7 billion in data processing revenue, up 11% from transaction growth
The revenue model was straightforward. Banks paid Visa fees based on payment volumes (service revenue), transaction counts (data processing revenue), and cross-border activity (international transaction revenue). None of this required Visa to take credit risk. The company earned fees whether customers paid their bills or defaulted, because Visa had already been paid for processing the transaction.
Cross-border transactions proved especially lucrative. When customers traveled internationally or made purchases from foreign merchants, Visa charged additional fees for currency conversion and cross-border processing. International transaction revenue reached $12.7 billion in FY2024, up 9% over the prior year. This segment carried higher margins because of the complexity and value of facilitating global commerce.
The Fintech Challengers and Visa’s Response
The 2010s brought new competitors that challenged Visa’s dominance, particularly in digital payments and peer-to-peer transfers. Companies like PayPal, Venmo, Square, and Stripe built payment experiences that often bypassed credit cards entirely, using bank transfers and digital wallets instead. These fintech companies threatened to disintermediate Visa by creating direct connections between consumers, merchants, and banks.
Visa’s response demonstrated the strength of its network model. Rather than fighting fintech companies, Visa partnered with them. PayPal, Venmo, Square Cash App, and most digital wallets ultimately connected to Visa’s network, issuing virtual Visa cards or linking to physical Visa cards on file. The fintech companies handled user experience and captured direct customer relationships, but transactions still flowed through Visa’s rails.
Visa’s adaptation to digital transformation:
- Visa Direct launched as real-time push payment platform, processing nearly 10 billion transactions in FY2024 for use cases like gig economy payouts and P2P transfers
- Visa token service created secure digital credentials for mobile wallets like Apple Pay and Google Pay, generating $500 million in annual revenue by 2020
- Partnership strategy embedded Visa into challenger banks and fintechs like Chime, Cash App, and Revolut rather than competing with them
- Maintained neutrality by enabling any company to build payment experiences on top of Visa’s infrastructure
The fintech era validated Hock’s original insight. Companies could innovate on user experience, customer acquisition, and value-added services. But the underlying transaction processing still needed the network that connected every bank and merchant globally. Visa owned that network, and rather than resisting change, the company made itself essential to the next generation of payment companies.
The DOJ Antitrust Case and Monopoly Concerns
On September 24, 2024, the US Department of Justice filed a lawsuit against Visa alleging violations of the Sherman Act. The complaint alleged Visa had monopolized general-purpose debit network services and engaged in anti-competitive practices that prevented smaller rivals from gaining market share. The case directly challenged whether Visa’s network dominance had crossed from competitive success into illegal monopoly.
The DOJ argued Visa controlled over 60% of debit card transactions in the United States, collecting more than $7 billion annually in fees from debit transactions alone. The complaint alleged Visa used exclusive agreements and penalties to prevent merchants from routing transactions to competing networks, and paid potential competitors like Apple and Square to not develop rival networks.
For Visa, the lawsuit represented the first major legal threat to its business model. If the DOJ prevailed and forced structural changes, Visa might lose the network effects that made the model valuable. The company maintained its practices were legal and that intense competition from Mastercard, American Express, PayPal, and fintechs prevented monopolistic behavior.
What If Visa Had Chosen Banking Over Networks
The Risk Profile That Would Have Changed Everything
If Visa had evolved into a full-service bank issuing its own credit cards and taking credit risk, the business would look completely different today. Instead of 67% operating margins and $20+ billion in net income, Visa would operate like a bank with 15% to 25% margins and exposure to credit cycles that periodically destroyed value.
The alternative scenario analysis:
- Bank-model Visa might generate $60 billion to $80 billion in revenue from interest income on credit balances
- Operating margins would compress to 20% to 25% in good times, 10% or negative during recessions
- Net income might be $12 billion to $15 billion in strong years, but losses during financial crises
- Market capitalization would likely be $150 billion to $250 billion instead of current $550+ billion
The math seems paradoxical. More revenue but lower value. The reason is that banking revenue carries risk. During 2008-2009, credit card charge-offs exceeded 10% at major issuers. Capital One, Discover, and American Express all saw massive losses. A bank-model Visa would have suffered alongside them.
The network model’s resilience through economic cycles justified premium valuation. When recessions hit, transaction volumes might decline 10% to 20%, temporarily reducing Visa’s revenue. But the business didn’t face existential credit losses. Investors paid 10x book value or more for Visa compared to 1x to 1.5x for banks because Visa’s earnings were far more predictable and protected from credit risk.
The Partner Relationships That Would Have Collapsed
Perhaps more damaging than credit risk, becoming a bank would have destroyed Visa’s cooperative network structure. The 14,500+ financial institutions that issued Visa cards did so because Visa remained neutral and never competed with them. If Visa started issuing cards directly, those banks would have migrated to Mastercard or tried building their own networks.
This isn’t hypothetical. Discover Financial Services both operates a payment network and issues cards directly. Its network is far smaller than Visa or Mastercard specifically because banks won’t issue Discover-branded cards when Discover competes with them. Discover is essentially locked into relying on its own issuance, limiting network growth.
If Visa had pursued banking, the company likely would have lost access to the global bank distribution that made the network valuable. Without universal acceptance, Visa’s brand would have eroded. Customers wouldn’t have carried Visa cards if they weren’t accepted everywhere. Merchants wouldn’t have accepted Visa if customers didn’t carry the cards. The network effects would have collapsed inward rather than compounding outward.
The Bottom Line: Why Visa’s Non-Banking Strategy Built the Perfect Business
Visa’s decision to stay out of banking and remain a pure payment network created one of the most profitable business models ever built. The company processes $15.7 trillion in annual payment volume across 234 billion transactions, earning $35.9 billion in revenue with 67% operating margins. This performance would be impossible if Visa competed with the banks it serves.
Dee Hock’s 1970 insight proved correct. The value wasn’t in being a bank. The value was in connecting all the banks through neutral infrastructure everyone could use without competitive concerns. Banks would compete on interest rates, rewards, and customer service. Visa would provide the network that made all their cards work everywhere.
The results by FY2024 validated the strategy completely:
- Operating margins of ~67% that banks could never achieve because credit risk and regulatory capital requirements limit bank profitability
- Network effects protecting Visa from disruption even as fintech companies innovated on user experience by building on top of Visa’s infrastructure
- Transaction volumes that grew through every economic cycle including 2008 financial crisis, because people still needed to make purchases even when credit markets froze
The genius was recognizing what not to do. Visa could have captured interest income from credit balances, but would have taken on credit risk that destroyed value during downturns. It could have competed with bank partners, but would have lost the cooperative network structure that made the business defensible. It could have resisted fintech innovation, but instead partnered with challengers and remained essential to digital transformation.
