Visa's Flywheel Spins on Its Own. Except for the Part the DOJ Says It's Holding in Place.

Walk into almost any store on earth and the question isn't whether they take Visa — it's an assumption, like the floor being solid. That assumption is the whole business. Because every merchant who takes Visa makes a Visa card worth carrying, and every consumer carrying one makes Visa impossible for the next merchant to refuse. The loop has been spinning so long that the spin itself looks like the law of physics. In fiscal 2024 it turned 233.8 billion times, across 4.6 billion credentials and more than 150 million merchant locations, on $16 trillion of volume.¹ That is a flywheel at planetary scale. The interesting question is what keeps it turning.

The official story is that Visa is the textbook flywheel — a pure two-sided network where each side recruits the other and the wheel accelerates on its own momentum. That story is mostly true. But the U.S. Department of Justice has filed a live lawsuit arguing that the most profitable stretch of the loop isn't coasting on momentum at all. It's being held in place by hand.

How the loop actually compounds

Start with what Visa is, because almost everyone gets it wrong. Visa does not issue your card, does not lend you a dollar, and does not set your interest rate. It sells network access to nearly 14,500 financial-institution clients who do all of that and carry all the credit risk.² Visa's revenue — about $35.9 billion in fiscal 2024 — is fees on transaction volume, not spread on debt.⁶ So the asset Visa owns is not money. It is the agreement, shared by billions of people and millions of merchants, to route money through one set of rails. Every new cardholder makes those rails more valuable to merchants; every new merchant makes them more valuable to cardholders. That is the flywheel: each side is the reason the other can't leave.

This is also why nobody simply builds a second Visa. A challenger has to recruit cardholders no merchant yet accepts, and merchants no cardholder yet carries — solving both sides at once, against an incumbent that solved both before most of today's competitors existed. The network's age is part of the moat. The brand most people call 'Visa' didn't even arrive until 1976; the rails underneath it trace to a Bank of America card program launched in Fresno, California, on September 18, 1958.⁵ That is sixty-plus years of head start, compounding the whole way.

Where the spin stops and the hand starts

Here is the part the flywheel story leaves out. On September 24, 2024, the Justice Department sued Visa in the Southern District of New York, alleging it had monopolized general-purpose debit network services in violation of the Sherman Act.³ The complaint's core claim isn't that Visa is big. It's about how it stays big. The DOJ alleges that more than 60% of U.S. debit transactions run on Visa's network, generating over $7 billion in annual fees — and that Visa threatens merchants and acquirers with punitive rates if they route a 'meaningful share' of their debit transactions to a competitor, insulating roughly three-quarters of its debit volume from competition.³ Read that mechanism slowly. A pure flywheel doesn't need to punish anyone for leaving, because leaving is its own punishment — you give up the network. The DOJ's theory is that Visa adds a second lock: a contractual penalty for routing away. The wheel keeps turning, but part of it is being held.

This is not a press release the company can wave away. In June 2025 the judge rejected Visa's motion to dismiss, ruling the government's allegations plausible and calling the dismissal bid a 'premature resolution of factual issues' — which means the case proceeds to discovery rather than dying at the pleading stage.⁴ A surviving complaint is not a verdict. But it does mean a federal court found the coercion theory credible enough to test. And it isn't Visa's first brush with the limits of expansion: when it tried to buy the fintech Plaid for $5.3 billion, the DOJ sued, and Visa and Plaid abandoned the deal in 2021 before any court ruled on the merits.⁸ The pattern the government sees is a flywheel that, where momentum slows, reaches for something firmer.

But isn't a dominant network supposed to look like this?

The fair objection is that allegations are not facts, and that a winning network inevitably looks coercive from the outside. Of course merchants feel trapped — that's what a strong network is: the point you can't route around. Visa's defense, in effect, is that its volume reflects a genuinely better product, and that even economists skeptical of the company concede debit networks carry real, legitimate network effects and economies of scale that naturally raise barriers to entry.⁷ By that reading, the loop is doing exactly what loops do, and the DOJ is mistaking gravity for a grip. There's force in this. Network dominance and anticompetitive conduct can produce identical-looking market shares, and telling them apart is precisely what discovery is for.

But notice the tell. The same analysis that grants Visa real network effects also notes that meaningful competition is possible — from payment apps, real-time payment rails, and even cryptocurrency — making this a contested market, not a sealed one.⁷ If competitors can in principle break in, then a flywheel running purely on its own gravity would simply let them try and fail. The DOJ's claim is sharper: that Visa doesn't let them try, because the contractual penalties make routing away too expensive to attempt. A truly self-reinforcing loop needs no such clause. The very existence of the alleged threats is the evidence that, on the most lucrative stretch, the spin alone wasn't trusted to do the job.

Visa's flywheel is one of the most powerful loops ever built — six decades of compounding, $16 trillion a year flowing through rails everyone simply assumes will be there.¹ Most of it spins on real gravity: two sides that cannot afford to leave each other. But the most profitable stretch, the DOJ argues, isn't coasting — it's clamped.³ And that's the quiet lesson hiding inside the world's tidiest network story. A flywheel that truly runs on its own momentum never needs a penalty for stopping. The clause is the confession.