How Banks Actually Make Money: The Business Model Explained

JPMorgan Chase made $49.6 billion in net income in 2023. That is more profit than Walmart, Coca-Cola, and McDonald's combined. JPMorgan employs 309,000 people, operates in over 100 countries, and processes tens of trillions of dollars in transactions annually.

Where does all that money come from?

Most people have a vague sense that banks "use your deposits to make loans" and charge interest. That is part of it — but it is a simplification that misses the full picture of how a modern bank actually generates revenue. Understanding the business model is essential for anyone working in fintech, evaluating bank stocks, or trying to understand where the disruption opportunities actually are.


The Core Machine: Net Interest Income

The fundamental business of banking is borrowing money cheaply and lending it expensively. The difference is the net interest margin (NIM), and it is the engine that everything else runs on.

When you deposit $10,000 in a savings account paying 0.5% interest, you are effectively lending that money to the bank at 0.5%. The bank takes that same money and lends it out as mortgages at 7%, car loans at 9%, credit cards at 21%, or invests it in securities yielding 4-5%. The spread between what the bank pays you and what the bank earns on its assets is the net interest margin.

Net interest income (NII) — the actual dollars of revenue generated by the spread — is by far the largest revenue line for most banks.

Revenue Category JPMorgan 2023 % of Total Revenue
Net Interest Income $89.3 billion ~57%
Non-Interest Revenue (fees, trading, wealth) $67.4 billion ~43%
Total Net Revenue $156.7 billion 100%

Source: JPMorgan Chase 2023 Annual Report

Net interest margin is heavily sensitive to interest rates. When rates rise — as they did aggressively in 2022-2023 — banks that are asset-sensitive (more of their assets re-price upward than their liabilities) benefit enormously. JPMorgan's net interest income nearly doubled from 2020 to 2023, largely because the Fed raised rates from near zero to over 5%.

When rates fall, the reverse happens. This is why bank stocks tend to move with interest rate expectations — the spread between short-term rates (what banks pay depositors) and long-term rates (what banks earn on mortgages and bonds) is the central driver of bank profitability.


Fee Income: The Business That Does Not Care About Rates

Non-interest revenue — fees — provides stability when interest rates compress margins. The major categories:

Service charges and account fees. Monthly maintenance fees, overdraft fees, wire transfer fees, ATM fees. This is the most visible and most controversial revenue category for retail banks. Overdraft fees alone generated approximately $8 billion for US banks annually before regulatory pressure began reducing them in the early 2020s.

Payment and card processing fees. When you use a debit or credit card, the merchant pays an interchange fee on every transaction — a percentage of the purchase price that flows back through the card network to your bank. Interchange is the hidden subsidy that makes credit card rewards programs possible. Your cash-back points are funded by the merchant who accepted your card. More on interchange below.

Investment banking and advisory fees. Large banks — JPMorgan, Goldman Sachs, Morgan Stanley — charge fees for advising on mergers and acquisitions, underwriting stock and bond issuances, and structuring complex financial transactions. These fees are lumpy (they depend on deal activity) and can be enormous — a single large M&A advisory engagement can generate $50-100 million in fees.

Asset management and wealth management fees. Banks that manage money — running mutual funds, managing private wealth, operating retirement accounts — charge fees as a percentage of assets under management. BlackRock, owned as a large stake by banks historically, manages over $10 trillion and earns roughly $15 billion annually in management fees.

Trading revenue. Large banks run trading desks that trade securities, currencies, and derivatives on behalf of clients and, within regulatory limits, on their own accounts. Trading revenue is volatile — it can be exceptional in volatile markets and poor in calm ones.


Interchange: The Most Misunderstood Revenue Line

When you pay $100 at a restaurant with your Visa credit card, the restaurant receives approximately $97-98. The other $2-3 disappears into a fee called interchange — and it is one of the most important revenue sources in banking that most people have never heard of.

The economics work like this: the merchant's bank (the acquiring bank) charges the merchant a processing fee, typically 1.5-3.5% of the transaction. The majority of that fee is passed through to the card issuer (your bank) as interchange. Visa and Mastercard take a small network fee. The acquiring bank keeps the remainder.

For a bank with millions of credit cardholders making hundreds of transactions per year, interchange accumulates into billions of dollars in annual revenue — with very little incremental cost per transaction.

Interchange rates are set by card networks and vary by card type, merchant category, and transaction characteristics. Premium rewards cards (the ones with airport lounges and cash-back programs) have the highest interchange rates. Debit cards in the US have regulated interchange (capped by the Durbin Amendment at $0.21 + 0.05% for banks with over $10 billion in assets). Credit card interchange is not federally regulated and can be significantly higher.

This is why banks issue premium credit cards aggressively and why they lobbied hard against the Durbin Amendment — interchange is highly profitable, and any policy that compresses it directly reduces bank revenue.

The fintech implication: companies like Apple Card (issued by Goldman Sachs), Marcus, and various BNPL providers have built businesses on interchange revenue. Buy-now-pay-later companies typically earn merchant discount rates (similar to interchange) of 3-8% on each transaction — significantly higher than card interchange — because merchants accept the cost in exchange for higher conversion rates.


How Fintech Is Attacking Each Revenue Line

Every fintech company is trying to disintermediate a piece of bank revenue. Understanding which piece helps evaluate which businesses have real potential and which are optimistic.

High-yield savings accounts attack net interest income. When Apple launched a 4.15% savings account through Goldman Sachs in 2023, it attracted $10 billion in deposits in the first four days. The proposition was simple: why let a bank pay you 0.5% when you can earn 4.15%? This is deposit disintermediation — pulling deposits out of traditional bank accounts and offering them back at more competitive rates.

Traditional banks responded slowly because raising deposit rates would compress their NIM. The result: $1.5 trillion in deposits left US banks for money market funds and high-yield accounts between 2022 and 2023. This is one of the most significant structural challenges facing retail banking.

Fee-free banking attacks service charges. Chime, which does not charge monthly fees or overdraft fees, was valued at $25 billion at its peak by building a customer base of 20+ million on the proposition that banking should not nickel-and-dime customers. Chime still earns interchange — but the absence of fee revenue forces a leaner operating model.

Embedded finance attacks payment revenue. When Shopify processes payments for merchants on its platform, it captures the economics that a bank or payment processor would otherwise earn. When Amazon runs Buy with Prime, it keeps the payment economics. The shift of commerce to platforms that embed financial services directly threatens bank payment revenue.

Roboadvisors attack wealth management fees. Betterment and Wealthfront charge 0.25% AUM fees versus the 1%+ that traditional wealth managers charge. Schwab offers index fund portfolios with zero management fees. The compression of wealth management fees is well underway — though high-net-worth clients remain relatively sticky with human advisors.

BNPL and alternative lenders attack lending revenue. Affirm, Klarna, and Afterpay originate consumer credit that would previously have sat on a bank's balance sheet as a personal loan or credit card balance. They typically underwrite differently (less reliant on FICO, more reliant on transaction data), move faster, and target the moment of purchase rather than requiring pre-approved credit lines.

Revenue Line Fintech Attacker Mechanism Threat Level
Net Interest Income (deposits) Neobanks, Apple/Goldman, money market funds Higher-yield accounts, deposit portability High — $1.5T already moved
Net Interest Income (lending) Affirm, Klarna, LendingClub, SoFi Alternative underwriting, embedded lending Medium — banks still dominate prime credit
Interchange / card fees Embedded finance platforms, BNPL Capture payment economics at point of sale Medium — requires scale to compete
Account / overdraft fees Chime, Current, Dave Fee-free accounts with interchange revenue High for lower-income segments
Wealth management fees Betterment, Wealthfront, Robinhood Low-cost index investing, zero-commission trading High for mass market, low for UHNW
Investment banking fees AngelList, Carta, SPV platforms Democratize access to private markets Low — relationship-intensive, regulatory moat


Why Banks Are Hard to Fully Disrupt

For all the fintech pressure, traditional banks retain structural advantages that make them durable.

Regulatory moat. A banking license — particularly a national bank charter or state commercial bank charter — takes years to obtain and requires substantial capital. Neobanks that want to offer FDIC-insured deposits must either obtain their own charter (expensive and slow) or partner with a chartered bank (which limits economics and creates dependency). This is why most neobanks are not actually banks — they are technology companies with banking partners.

Trust and brand. For large deposits — the kind where FDIC's $250,000 limit actually matters — customers choose established institutions. The flight to quality during the Silicon Valley Bank collapse in 2023 went to JPMorgan, Bank of America, and Wells Fargo. Trust is built over decades.

Corporate banking moat. The most profitable banking relationships — treasury management, trade finance, FX, custody for large corporates — are sticky, relationship-intensive, and difficult for fintech challengers to penetrate. The fintech disruption has been concentrated in consumer banking because that is where the switching costs are lowest.

The realistic outcome is not that fintech replaces banks — it is that banks become more focused on areas where their advantages are strongest (corporate, wealth, regulated activities) while fintech captures the mass-market, commodity, high-volume, low-margin business that banks never found particularly attractive anyway.


Key Takeaways

  • Net interest income — earning more on loans than you pay on deposits — is the core engine of banking, typically 50-60% of revenue for major banks. It is why banks care deeply about interest rate cycles.
  • Fee income diversifies revenue and provides stability. Service charges, interchange, investment banking fees, and wealth management fees reduce dependence on the interest rate environment.
  • Interchange is the largest revenue source most consumers don't know about. Every card transaction generates a payment to your bank. Premium credit cards are funded by this mechanism.
  • Fintech is attacking bank revenue from every direction simultaneously — deposits, lending, payments, fees, and wealth management. No single attack is existential; the combination is compressing margins.
  • Banks' most durable advantages are in corporate banking, trust for large deposits, and regulated activities — exactly the areas fintech has been least successful at penetrating.

Related Reading