The greatest credit companies of all time
Seven companies, seven different reasons they became legendary. Each solved exactly one hard problem and built a durable moat around it. None of them were great at everything, they were great at one thing and adequate everywhere else. We graded Iron against each of the seven, then closed the highest-leverage gaps.
By Iron, an internal gap analysis
The seven, and what each was actually great at
Frank McNamara forgot his wallet at dinner and built “pay later, no cash” from that single friction. First year: 10,000 members, 28 restaurants, 2 hotels, on nothing but a cardboard card and a signature.
Being first to a real, felt problem. The wedge was radical simplicity, not technology.
Rival banks were bricking their own network fighting over BankAmericard’s rules. Hock built National BankAmericard Inc. as a member-owned, non-stock “chaordic” organization: cooperate on the rails, compete on the product.
Neutral infrastructure nobody owns and everybody trusts. Universal acceptance became the product.
Closed-loop network: Amex is the issuer, the acquirer, and the processor at once, so it owns both sides of every transaction.
Owning the full loop turns a payment network into a status and data platform. Premium pricing to merchants funds premium service to members.
Every bank charged the same 19.8% flat rate to every customer. Fairbank and Morris priced risk per person with an Information-Based Strategy, and shipped thousands of live pricing tests instead of one product.
Underwriting-as-R&D. The data compounds; competitors copying the price cannot copy the pricing engine.
Launched into a market of confusing annual fees and hidden terms with the opposite: no annual fee, and the first mass-market cashback (1% Cashback Bonus).
Radical transparency as a growth lever, not a compliance cost.
Instead of buying attention with mass marketing, MBNA co-branded with 1,400-plus existing communities (alumni associations, AAA, professional groups) who already trusted each other. 25% average annual earnings growth through the 1990s on this alone.
Distribution through borrowed trust. CAC near zero because the affinity group did the vouching.
Two structural tricks, a century apart. In 1799 Aaron Burr got a charter approved for a company to pipe clean water to Manhattan, then slipped in a clause letting it deploy “surplus capital” however it liked. Six months later the water company quietly opened a bank, a charter smuggled past the legislature as a rider clause. Then in 1907, with no central bank yet in existence, J. Pierpont Morgan personally organized the response to a national panic by locking the country’s leading bankers in his library until they committed their own capital to the bailout. One man acted as the central bank, out of his own house.
Find the structural trick nobody else has, and be willing to personally underwrite the system when it breaks. A century of consolidation (Chase National, Chemical Bank, J.P. Morgan & Co., Bank One) built the balance sheet that made the second half of that promise credible at scale.
The gap table
Graded against iron.credit as it exists today (site, deck, and strategy docs), not against the eventual company.
Two are strong (the founding wedge and the transparency positioning), four are real but half-built (sovereignty, the data moat, affinity distribution, and being the balance-sheet backstop, all planned but not claimed), and one was a true blind spot: Iron had a risk ladder but no membership. That was the single highest-leverage gap, the one lesson none of the others substitute for. Visa, Capital One, Discover, and MBNA all sell you a better transaction. Amex is the only one that sold you who you become by carrying the card, and it is why Amex commands premium pricing none of the transaction-first players ever matched.
Closing the gaps
The grade rubric on the Score page was accurate but cold: letters, LTVs, and APRs with no reason to want to climb it beyond the numbers. Added a membership framing directly above the rubric: what each grade means as a level of trust extended, not just a rate. No invented tiers or privileges Iron cannot yet deliver, that would break its own no-predatory-tricks rule. It names the identity that already exists in the mechanics.
The go-to-market was described as a beachhead but never connected to the proven playbook it actually is. Iron’s distribution is an affinity strategy in MBNA’s exact shape (a trusted community vouches, CAC approaches zero), applied to Network School and the global bitcoin community instead of alumni associations.
All three share one root cause and already have real plans (the sovereignty trigger for rails, the Data phase for underwriting). Chase’s “be the balance-sheet backstop” and Iron’s “own the licence” are one capital-and-licensing reality, not three separate problems. Closing them further means public commitments Iron cannot yet back with a live licence, a signed facility, or a live loan book. Never claim ahead of the fact.