American Express: The Rival That Chose Economics Over ReachNarrow moat

American Express issues its own cards, acquires its own merchants and keeps the entire fee rather than a slice of it. That produces much better economics per transaction and a much smaller network. It has spent two decades closing the acceptance gap and has still not closed it, particularly outside the United States and particularly among small merchants who object to the rate and can afford to.

The comparison is valuable because it isolates the variable. Amex proves that a payment network *can* charge far more than Visa does — the constraint on Visa's pricing was never that a higher price is impossible. It also proves what the higher price costs: a merchant base that can and does refuse you, and a business that must compete for the cardholder directly with rewards it funds out of its own margin rather than out of the issuer's.

Visa took the opposite trade — a tiny fee, near-universal acceptance, and somebody else funding the rewards. Fifty years later one of these companies moves $13.9 trillion¹ and the other does not, and the one that moves less is considerably more profitable per transaction.

Neither model is obviously wrong. What the contrast establishes is that Visa's low price is a strategic choice that bought ubiquity, not a competitive concession extracted from it.