Unbundling the bank: how fintechs attack the value chain

# Unbundling the bank: how fintechs attack the value chain

Open a traditional checking account and you get a bundle you never asked to buy. A place to store cash. A debit card. Bill pay. Overdraft "protection." International transfers. Maybe a small loan at the register. For decades, one account did all of this, and the bank made money on the parts you did not notice.

Then fintechs showed up and asked a dangerous question: what if we just sold the profitable part, by itself, better?

That is unbundling. This lesson traces how one checking account got sliced apart, and why the strategy only works in specific spots.

The bundle, and the cross-subsidy hiding inside it

First, one term. A cross-subsidy means using profit from one product to cover the cost of another. Banks do this constantly.

Your checking account is often "free." It is not free to run. So the bank earns money elsewhere and quietly uses it to cover the account:

• Overdraft fees: charged when you spend more than your balance.
• Interchange: a small fee (a fraction of each purchase) that a merchant's bank pays the card issuer every time you swipe.
• Net interest margin (NIM): the gap between the low rate the bank pays you on deposits and the higher rate it earns lending that money out.
• FX markup: a hidden margin baked into the exchange rate on international transfers.

The customer sees one friendly account. The bank sees a portfolio where a few lines carry the rest.

Unbundling works by attacking exactly those profitable lines, one at a time, and giving the boring parts away.

Slice 1: Chime attacks the fee-heavy account

Chime is a neobank: a digital-only banking app that partners with a chartered bank to hold deposits (Chime itself is not a bank). It went straight at the most hated cross-subsidy: overdraft and monthly fees.

The pitch was simple. No monthly fees. No overdraft fees. Get paid up to two days early.

How does Chime make money if it strips out the fees? Mostly interchange. Under US rules, smaller banks (below a size threshold set by the Durbin Amendment) can charge higher debit interchange. By partnering with a smaller sponsor bank, Chime's card swipes earn more per transaction than a big bank's card would.

So Chime kept the profitable slice (interchange on everyday spending) and deleted the slice customers resented (fees). The economics only work because incumbents were cross-subsidizing free checking with fees, leaving room to undercut.

Slice 2: Wise attacks the FX markup

Send money abroad through a traditional bank and you often pay twice: a stated fee, plus a worse exchange rate than the real one. That rate gap is the FX cross-subsidy, and it is easy to hide.

Wise (formerly TransferWise) unbundled just this slice. Its core move: show the mid-market rate (the real, midpoint exchange rate you see on a financial site) and charge a clear, separate fee instead.

The clever operational trick is that Wise often does not physically send your money across borders. It matches people sending money in opposite directions and pays out from a local pool of funds in each country. Less cross-border movement means lower cost, and Wise passes some of that on.

Wise did not try to be your bank. It took one profitable, opaque line item and made it transparent and cheap.

Slice 3: Affirm attacks point-of-sale credit

The final slice is credit. When a checkout screen offers "4 payments" or "pay over 12 months," that is BNPL (buy now, pay later): a short-term installment loan issued right at purchase.

Affirm unbundled the lending slice from the account entirely. You do not need an Affirm "account" the way you need a bank. You need a purchase to finance, at the moment you want it.

Affirm makes money two ways:

• Merchant fees: the store pays Affirm to offer financing, because it lifts sales.
• Interest: on longer loans, the shopper pays interest (Affirm emphasizes no late fees and clear terms).

This attacks the bank's lending cross-subsidy and its credit-card business, where the profit historically came from interest and fees on revolving balances.

Why unbundling works only where there is fat to cut

Here is the pattern. Each fintech above went after a slice where the incumbent was earning outsized, often hidden, margin, and used that margin to prop up a "free" or convenient bundle.

That is the rule: unbundling economics work only where incumbents cross-subsidize.

If a bank charged the true cost of each service transparently, there would be little room to undercut. The opportunity exists because the bundle hides who is paying for what. Overdraft-heavy customers subsidize fee-free ones. Travelers with bad FX rates subsidize the "free" transfer promise. Revolving borrowers subsidize the rewards points of people who pay in full.

A fintech finds an over-charged slice, isolates it, and offers a fair price. The customers who were subsidizing everyone else leave first.

The catch: unbundling is not the end of the story

Two forces push back.

One: the profitable slice is fragile. Chime's interchange advantage depends on a specific regulatory threshold. Wise's margin depends on FX opacity elsewhere in the market. Affirm's credit margin depends on interest rates and loan losses, which move with the economy. Pull one lever and the standalone product can wobble.

Two: unbundling tends to re-bundle. Once a fintech wins a slice and earns trust, it adds more slices. Chime added savings features and credit-builder products. Wise added multi-currency accounts and cards. This is natural: customer acquisition is expensive, so you sell more to the customers you already have. The industry calls this the "unbundle, then re-bundle" cycle. Andreessen Horowitz's often-cited framing is that every company will be a fintech company as embedded financial products spread.

So the endgame is not a thousand single-product apps forever. It is disruption of the old bundle, followed by new bundles built around the winning slice.

A quick way to spot the next unbundling target

You can apply the same lens to any incumbent, not just banks. Ask three questions.

1. Where is the margin hidden? Look for a "free" or bundled service. Something has to pay for it. Find the cross-subsidy.

2. Can the profitable slice stand alone? Overdraft cannot; it only exists because of checking accounts. But FX transfer, point-of-sale credit, and card issuing can each live as standalone products. Standalone potential is what makes a slice attackable.

3. Is the margin defended by opacity or by regulation? If customers overpay only because they cannot see the price (FX markup), transparency alone can win. If the margin is protected by a license or a rule, the fintech usually needs a partner bank or its own charter, which slows things down.

If a slice is profitable, separable, and defended mainly by opacity, it is a strong unbundling candidate. That single checking account failed all three tests at once, which is why it got carved into Chime, Wise, and Affirm.

Why incumbents do not just cut their own fees

If the fat is so obvious, why not remove it? Because the bank would be cutting its own profit to defend customers who might not leave anyway. The innovator's dilemma applies: protecting today's revenue looks rational right up until a focused competitor takes the whole slice. Incumbents move slowly, buy fintechs, or launch their own stripped-down brands, usually later than they should.

Key takeaways

• The checking account was a bundle held together by cross-subsidies. Free features were paid for by hidden ones: overdraft fees, interchange, net interest margin, and FX markup.

• Fintechs unbundle by isolating a profitable, opaque slice.