When Money Moves for Free, Who Gets Paid?
The $205 Billion Question: What Scalable Digital Cash Does to Financial Intermediaries
Every time you tap your card at a coffee shop, a small war is being fought over fractions of a percent. The coffee shop pays roughly 2% of the transaction to the card network and the bank. You don’t see it. It’s baked into the price of the coffee. But across the entire U.S. economy, that invisible toll adds up to tens of billions of dollars a year — paid by merchants, absorbed by consumers, and collected by a handful of companies that own the pipes through which money flows.
Now ask a different question: what happens when those pipes become essentially free to build and run?
That’s not a hypothetical. It’s the thing happening right now, slowly and unevenly, as scalable blockchain infrastructure matures. And the answer has implications that go far beyond cryptocurrency enthusiasts arguing on the internet. It touches every industry where someone earns a living by standing between a payment and its destination.
The Business of Moving Money
To understand what’s at stake, you need to understand where the money in money movement actually comes from.
Banks don’t make most of their money on loans anymore — at least not the way people imagine. A large chunk of what they earn comes from what economists call the deposit franchise: the gap between what they pay you on your checking account (usually nothing) and what they earn by investing that same money in government bonds or lending it out. When interest rates are high, that gap is enormous. In 2024, U.S. banks held about $3.86 trillion in accounts paying zero interest. With the Federal Funds Rate at 5.3%, the implicit yield on that money was over $200 billion. None of that went to depositors.
That’s not fraud. You agreed to a zero-rate account in exchange for a free debit card, a branch down the street, and the ability to pay your rent by mobile transfer. The bank earns the spread; you get the service. It’s a bundle — and the bundle has worked for decades because there was no unbundled alternative.
Card networks earn differently. Visa and Mastercard don’t take credit risk — they don’t lend you money. What they sell is certainty: the guarantee that when a merchant accepts your card, they will eventually receive the funds, no matter what happens to your relationship with your bank. That guarantee, and the network required to deliver it globally in under a second, is worth $64 billion a year in combined revenue for the two companies. Net margins above 40%.
Remittance companies occupy another corner of the same territory. Send $200 from the United States to Nigeria and you’ll pay, on average, around $33 in fees — over 16% of the transfer. The money makes its way through a correspondent banking chain: your bank to a U.S. intermediary, to a Nigerian correspondent, to a local bank, each one extracting a margin for the service of knowing someone on the other end of the chain. It’s slow, expensive, and the infrastructure underlying it hasn’t fundamentally changed in decades.
All of these businesses share a common structure: they earn rents by providing trusted transfer. Someone has to guarantee that money sent is money received. Building and maintaining that guarantee has historically required enormous fixed costs — compliance infrastructure, bilateral settlement agreements, legal frameworks, fraud systems, and the institutional trust that takes generations to build. The people who built it get paid for it. That’s the rent.
The Cost Is Dropping
Here’s what’s changing: the cost of producing trusted transfer is collapsing.
Measured on scalable blockchain infrastructure under controlled conditions, the per-transaction cost of settlement has been benchmarked at fractions of a cent so small they require scientific notation to express — in the range of $0.00000002 per transaction. That’s not a production figure accounting for all the overhead of running a global network. It’s a floor — the minimum cost of the core operation, the cryptographic verification that a payment happened and can’t be reversed.
The gap between that floor and what legacy systems charge is not a rounding error. It’s a structural wedge, measured in orders of magnitude. And where you find a structural wedge between the cost of producing something and the price charged for it, you have found economic rent.
The question isn’t whether that gap will be competed away. It’s how fast, through what mechanism, and — crucially — who loses and who gains in the process.
It’s Not Just Payments
The obvious application is payments. But the logic extends further than most people realize, because “trusted transfer” is the foundation of a surprising range of financial services.
Trade finance — the letters of credit and documentary collections that move goods across international borders — is expensive precisely because it requires trusted intermediaries to guarantee that the seller gets paid if the goods arrive and the buyer gets the goods if payment is released. That conditional structure (pay if and only if conditions are met) is exactly what programmable settlement can provide automatically. The intermediary that earns a fee for holding that guarantee in escrow is providing a service that, in principle, code can replace.
Treasury management — the corporate practice of pooling cash across subsidiaries, sweeping balances overnight to maximize yield, maintaining intraday liquidity — is largely an exercise in navigating settlement delays. Money that’s in transit isn’t earning interest. Money that hasn’t settled can’t be redeployed. Banks earn fees by managing that dead time. Faster settlement compresses the dead time and the fees along with it.
Collateral management in financial markets — the daily process by which banks and funds post margin, recall securities, and substitute collateral — runs on a system of bilateral agreements, custodians, and settlement cycles that can take days. The costs are enormous, and they fall on everyone in the system whether or not they’re aware of it.
In each case, the same pattern: someone earns money by managing a friction. The friction is the settlement delay, the uncertainty of finality, the need for a trusted third party to hold something in escrow while the other side of a transaction completes. Make settlement instant, cheap, and cryptographically certain, and the friction shrinks — and so does the rent extracted from it.
What Doesn’t Change
It would be a mistake to read this as a prediction that banks disappear or that financial intermediaries become irrelevant. That’s not what the economics says.
Credit is not a settlement problem. Figuring out whether a small business in rural Ohio deserves a loan, how much, at what rate, and with what covenants — that requires judgment, local knowledge, relationships, and accountability. A blockchain ledger provides none of those things. The kind of banking that depends on knowing your customer, building multi-year relationships, and making judgment calls about creditworthiness is not threatened by cheaper settlement. If anything, it becomes more valuable as the commodity functions of payment processing get competed down.
Liquidity transformation — the bank’s core magic trick of borrowing short-term from depositors and lending long-term to businesses — is not a payment function. It’s an economic function. Banks create liquidity by making illiquid assets (long-term loans) fundable with liquid liabilities (deposits). That’s genuinely valuable, and it’s unrelated to the cost of the wire transfer that moves money between accounts.
Fraud, disputes, consumer protection — these don’t go away either. A blockchain transaction that sends money to the wrong address is, absent specific design choices, irreversible. That irreversibility is a feature for some purposes and a catastrophic bug for others. The consumer protections embedded in the card system — the ability to dispute a charge, the fraud liability rules, the chargeback mechanism — exist because the world is full of mistakes, misrepresentations, and outright theft. Building the equivalent protections into a new system takes years and requires institutions.
So the honest version of the argument is not “blockchain replaces banks.” It’s that the parts of banking and financial intermediation that are really just rents on settlement friction — the parts where the fee exists because moving money used to be hard — those face structural compression. The parts that are genuinely about credit, judgment, relationships, and consumer protection do not face the same pressure and may actually gain value as the commodity functions clear out.
The Governance Problem Nobody Talks About
Here’s the catch that most analyses skip: cheaper settlement only reorganizes rents toward consumers and producers if the cheaper settlement layer is genuinely neutral infrastructure.
If the blockchain that processes your payments is controlled by a small group of developers who can change the fee structure, blacklist addresses, alter transaction rules, or fork the protocol whenever they want — then you haven’t escaped a rent-extracting intermediary. You’ve traded one for another, and the new one may be less accountable than the old one, because banks at least operate under regulatory oversight.
This is the governance problem, and it’s not theoretical. The history of major blockchain protocols is full of examples where the technical appearance of decentralization masked concentrated control: a handful of developers whose decisions determine protocol rules, a handful of mining pools whose hash power determines what gets confirmed, a handful of large holders whose economic interests shape what proposals get accepted. The fact that these decisions are made via consensus mechanisms doesn’t mean they’re made without concentrated power.
The condition for cheaper settlement to actually reorganize financial rents in a durable way is that the protocol rules are credible and immutable — meaning that no single party can change them in ways that recreate the choke point. Technical decentralization is necessary but not sufficient. Governance decentralization is what actually matters, and it’s much harder to achieve.
This isn’t an argument against blockchain. It’s an argument for being precise about which blockchain systems meet the governance standard and which ones don’t. The ones that don’t are at risk of producing a new class of rent-extractors operating under a decentralization brand but with concentrated power. The ones that do have the potential to deliver genuinely neutral settlement infrastructure — the TCP/IP of money, if you like: infrastructure that competition happens on top of rather than through.
What Happens to the People Left Holding Rents
The adjustment won’t be uniform. Banks and payment companies with different business models face different exposures.
A bank that earns most of its margin from the deposit float — collecting the spread between zero-rate checking accounts and invested assets — faces direct pressure as cheaper outside options develop. The logic is simple: when depositors have credible alternatives that offer yield, they move balances. The FDIC data already shows this in the 2022-2024 period, when rising interest rates caused non-interest-bearing deposits to fall by $1.66 trillion as depositors found money market funds and direct bond purchases. Blockchain-based accounts offering market yields would extend that substitution.
A remittance company operating in a high-cost corridor faces more acute pressure. When a technically sound, well-governed digital cash system can route a transfer from Houston to Lagos in minutes at negligible cost, the justification for a 16% fee evaporates. The question is only whether the regulatory environment in both jurisdictions allows the alternative to operate — which brings us back to the state’s role in maintaining or dismantling existing choke points.
Visa and Mastercard face a more complicated picture. Their revenue isn’t purely a settlement rent; it also reflects genuine network effects, brand trust, fraud infrastructure, and consumer protection systems built over decades. A challenger needs to replicate all of those things to fully contest the value proposition, not just the settlement layer. That’s a much harder task. But the direction of pressure is clear: as programmable settlement matures and as consumer comfort with digital-native payment systems grows, the premium for the legacy network shrinks.
The incumbents know this. The reason major banks are exploring digital asset custody, the reason Visa and Mastercard are both running blockchain pilot programs, the reason central banks in 130+ countries are developing CBDCs — none of this is because the industry thinks the technology is a fad. It’s because the industry understands that whoever controls the settlement layer in the next generation of financial infrastructure will inherit the rents from it.
The Bigger Picture
Strip away the technical details and the argument is simple.
Throughout history, anyone who controlled a chokepoint in the movement of money earned rent from that control. The Medici earned it by maintaining a network of correspondent relationships across European trading cities. Western Union earned it by owning the telegraph wire. Visa and Mastercard earn it by owning the authorization and settlement network. The rent is always justified — to some degree — by the genuine cost of building and maintaining the chokepoint. And it always gets competed away when technology reduces the cost of building the chokepoint to near zero.
We’re in the early stages of that competition for money movement. The cost of building the chokepoint — a global, cryptographically secured, always-on settlement network — is dropping faster than most financial incumbents want to admit. The rents being earned on the old chokepoints are large enough — hundreds of billions per year, globally — that the competition will be fierce, well-funded, and politically contested.
The outcome depends less on the technology, which is largely proven, than on the governance of the new infrastructure and the regulatory choices made by states that currently benefit from legacy financial arrangements. If the new settlement layer is genuinely neutral and trustworthy, rents reorganize toward users and toward credit functions that genuinely earn their keep. If the new layer is captured — either by new rent-extractors operating under a decentralization flag or by states engineering digital currency systems that maintain financial surveillance and control — then the technology changes the form of financial intermediation without changing its essential economics.
That’s the thing worth watching. Not the price of any token. Not the next upgrade to any protocol. Whether the plumbing of the financial system is becoming genuinely neutral infrastructure or just transferring control from one set of gatekeepers to another.