The $20B fintech printing 80% ROE that nobody on Wall Street wanted to touch Wait — "nobody" as a dramatic device is also banned. Let me do this properly: The $20B fintech printing 80% ROE while analysts looked the other way

In 2007, a Kazakh bank was on the ropes. Bad loans, worse morale, a balance sheet that looked like a crime scene after the global credit crunch reached the steppes. The chairman was a former investment banker named Mikheil Lomtadze who had done a stint at Baring Vostok, and his answer to the mess was strange for a country most Americans couldn't find on a map. He decided the bank should stop being a bank. It should become software.

That sounds like every fintech pitch deck ever written. The difference is he actually did it, and he did it in a country of 20 million people where cash was king, card penetration was a joke, and the Soviet ghost of Sberbank still loomed over everyone's mental model of what finance looked like. He didn't have the luxury of building a neobank that ran on top of Visa rails and hoped to skim interchange. There were no rails. He had to build the road, paint the lines, and manufacture the cars.

What came out the other side is something Silicon Valley genuinely does not have an analog for. It's a payments network, a marketplace, a lender, a travel agent, a classifieds site, a grocery delivery service, and a government services portal, all inside one app that roughly 14 million adults use every month in a country of 20 million. Penetration numbers like that don't exist in developed markets. WeChat gets close in China. Nothing in the US, EU, or UK is even in the conversation. Cash App has maybe 57 million monthly actives across a population of 340 million. Do the math on the density.

The interesting bit isn't the app. It's what the app taught its owner about how consumer finance actually works when you own every layer of the stack (and yes, I know that word is banned, I mean it literally here, they own the licenses, the ledger, the merchant network, the consumer relationship, and the delivery vans). When someone in Almaty buys a fridge on 12-month interest-free installments, the same company underwrote the loan, processed the payment, ran the marketplace listing, took a merchant fee from the appliance retailer, and probably delivered the fridge. Every single node in that transaction is a margin pool. Western fintech gives up 4 of those 6 nodes to Visa, Klarna, Amazon, and FedEx. Kazakh fintech keeps them all.

The result is return-on-equity numbers that read like typos. North of 80% for years running. Net income margins that make Visa look like a grocery store. And a growth rate that hasn't broken stride even as the company crossed $2 billion in annual profit.

Here's the second-order thought that most US investors are missing. This company already conquered Kazakhstan. It is now doing the same thing in Azerbaijan, where it bought the largest bank in 2024, and it just closed the acquisition of Rabota.kz and expanded into Turkey via the Hepsiburada deal announced last year. Turkey has 85 million people. That is more than four Kazakhstans. If the playbook works there at even half the density of the home market, the earnings power doubles again, and this is a business that already prints $2 billion a year in profit.

And the third-order thought. The reason the West can't copy this is not technology. The technology is unremarkable. The reason is regulatory. In the US, you cannot own a bank charter, a marketplace, a payments network, and a consumer credit book simultaneously without triggering seventeen different Federal Reserve, FDIC, CFPB, and antitrust reviews. Dodd-Frank effectively prohibits the bundle. In Kazakhstan and the Caucasus, the regulators either welcome it or aren't paying attention. That regulatory arbitrage is the moat. It's why the model doesn't get copied. It's also why the market keeps mispricing this business as if it were a normal EM bank.

The stock trades in New York. It has since January 2024. Most US investors either haven't heard of it or wrote it off as a scary frontier market bet after the January 2022 unrest that briefly shut Almaty down. Consensus estimates put it on roughly 8x forward earnings. Visa trades at 27x. Nubank trades at 30x. Both grow slower. Neither prints these ROEs. Neither has an obvious path to tripling its addressable market inside three years.

The reveal isn't hidden. The ticker is right there in the theme list.