There are many ways to come across new ideas. We often get recommendations from our Mastermind members. Other times, you literally open the Wall Street Journal, and an interesting setup jumps up at you. Admittedly, though, that has only happened to me once or twice in my “investing career.”

And then there’s the option of following superinvestors and their most recent buys. Today’s pitch was inspired by a mix of both. As I mentioned in one of my last newsletters, I recently visited a friend in Malta, and he told me about Kaspi and what a phenomenal business it is.

Being obsessed with e-commerce players, I obviously couldn’t help but look at it. Turns out, Mohnish Pabrai is a big fan as well. He recently bought a stake in Kaspi and slotted it in his “heads I win, tails I don’t lose much” bucket.

Let’s dive in!

— Daniel

A good deal is a good deal, and as the team behind this newsletter, we’re obligated to let you know when one arises.

That’s why we’re interrupting your regularly scheduled company breakdown to let you know about something special: On September 19th, Shawn, Daniel, and Kyle will be hosting an all-day conference in Midtown Manhattan dedicated to Intrinsic Value.

Blocks away from the financial center of the world, we’ll be reigniting the spark behind fundamentals-based investing, uniting value-driven investors from all walks of life and corners of the planet. And we want YOU to be there alongside us.

Learn more and claim your spot while tickets are still available:

Some of you may wonder why you’ve never heard of a company I’m framing as a tech monopoly with, maybe, the most attractive valuation I’ve ever seen for a business of this quality. Well, the reason is that a few of you live in Kazakhstan. At least that’s my best guess.

In Kazakhstan, Kaspi is about the strongest monopoly you could imagine – it owns e-commerce, payments, consumer lending, and it stretches into a long list of things no business we’ve ever touched had anything to do with.

It's a super-app, in the WeChat sense of the word (and probably even more entrenched in daily life), and I'm careful with that term because most companies that get called a “superapp” don't really earn that label. Before we go any deeper, though, let’s first understand how Kaspi got to where it is today.

Twenty years ago, Kaspi was nothing more than an ordinary Tier 2 bank in a Kazakhstani banking market dominated by Halyk, the old Soviet-era savings bank and the largest in the country to this day. That changed in the early 2000s when two people stepped in and took over the business.

Today, they are widely considered the founders of Kaspi since they completely restructured the business. One of them is Vyacheslav Kim, a well-known Kazakh businessman with a retail background; the other is Mikheil Lomtadze, a Georgian tech entrepreneur with a Harvard MBA and a background at Baring Vostok – one of the largest private equity funds in Russia.

Lomtadze still runs the company, and it’s not a stretch to say that few ever imagined his ambitious vision could become reality.

What was that vision? Well, he wanted to turn Kaspi into an app that every Kazakhstani would use daily. An app that not only covers banking and payments but also e-commerce, logistics, delivery, and government services.

After all the ecosystem businesses we’ve looked at (Sea Limited, Mercado Libre, Amazon, Uber, etc.), I guess I don’t have to tell you how powerful the connections among them can be. A user can order groceries on Kaspi’s marketplace, process the payment directly through Kaspi, file taxes with Kaspi, apply for a driver’s license, and register a car. You can even apply for marriage registration on Kaspi!

The point is, Kaspi knows just about everything about its users, and there are few things more important than data in today’s world.

And the reason Kaspi could take over all of this territory is similar to what we discussed in our Nubank newsletter. The legacy players, and mostly the banks, didn’t serve the average citizen. They focused on serving the wealthy class, where they could earn a good margin with little risk. That’s why Halyk Bank initially thought the winning strategy was to double down on banking products and relationships only with the top 10% of customers.

Twenty years later, it’s easy to see that that bet didn’t work out. Kaspi now controls the consumer relationship, and that turns out to be the most valuable part of the ecosystem. For example, Kaspi is not the only one offering all the government services discussed above. However, it’s not nearly as popular to download Halyk’s newly started superapp when everything they need is already accessible on Kaspi.

There are three core functions of the Kaspi app: the marketplace, fintech, and payments. Each of these pillars is so dominant that it would be a market leader even as a standalone entity.

To give you an impression of market penetration, more than 70% of the country's population actively uses Kaspi, and "active" here means interacting with the app more than 77 times a month, or roughly two and a half times a day! For comparison, the average American uses a debit card about 25 times a month, so Kaspi's users are transacting three times as often. I think Shawn, though, swipes his credit card much more frequently. I hope he will forgive me for all my jokes about his credit card behavior!

Since Kaspi started as a bank, payments were the most natural expansion of the business model. Kaspi runs about 18 million transactions a day – QR at the register, peer-to-peer, bill pay, and some B2B – moving close to $100 billion in volume last year.

But it’s certainly not the largest business unit by revenue, representing only 16% of sales. However, it accounts for roughly 40% of net income, resulting in a net margin of close to 66%. Visa is at “only” 51%…

Now, to be fair, Visa pulls these numbers on revenue of over $40 billion; Kaspi’s payment business is still fighting to reach $1.5 billion. Nevertheless, the margin profile is impressive.

The reason for this massive margin is that Kaspi owns the entire transaction loop. When you scan a QR code at a café, the money moves from account to account across Kaspi's own network. There's no interchange fee paid out to Visa or Mastercard, because there's no Visa or Mastercard in the loop. Kaspi is both the network and the processor. So the marginal transaction costs are essentially nothing.

The downside of the payments business is that it’s relatively mature. It’s still growing revenue in the mid-to-high single digits, but it’s the slowest-growing business in the Kaspi ecosystem.

If you read through my past newsletters on Sea Limited, you might be familiar with the e-commerce ladder concept, which I use as a framework when evaluating companies like Sea Limited, Meli, Amazon, and Alibaba.

The idea is that Level 1 is the entry tier where companies focus on cheap, unbranded goods, slow third-party shipping, and no ecosystem wrapped around it. TikTok Shop is the clearest example. You might argue that Pinduoduo also fits this category, although there’s certainly a case for PDD as a Level 2 marketplace. Generally, the Chinese e-commerce market has its own rules, but I digress.

And I should say that Level 1 isn’t necessarily “worse.” It serves price-first shoppers, often in smaller cities without access to other marketplaces. It’s yet another case of legacy players ignoring a part of the market that opens up opportunities for new entrants. In the long run, though, most companies wanna move up the ladder.

Level 3 is where you serve customers who arrive with clear purchase intent and high expectations for quality and speed. But they are also willing to pay for that comfort and quality.

Reaching Level 3 has historically cost a fortune. Mercado Libre, for example, poured billions into fulfillment centers, delivery fleets, and even its own planes.

What makes Kaspi really interesting is that, judged by flywheel and moat, it belongs unambiguously in Level 3, and yet it has never spent billions on logistics. Fulfillment runs asset-light, through third-party carriers and a grid of more than 10,000 self-service lockers they call Postomats.

And it works quite well. 84% of orders ship free, and about half arrive within 48 hours. Still, on its own, that asset-light setup would read to me as the weak point in the thesis. But for Kaspi, the customer lock-in was never about logistics. Kaspi is so entrenched in the everyday life of Kazakhstanis that logistics only need to be “good enough.”

In our podcast episode, I mentioned to Shawn that this dynamic reminds me of an old Munger quote: “The smart man knows the rules, the wise man knows the exceptions.”

Anyway, let’s get to the numbers. The marketplace accounts for about 47% of revenue but only 26% of net income. So, basically, the payment dynamic is reversed. GMV also looks quite small, at $19 billion, compared to the other marketplaces we discussed. And that $19 billion includes $4 billion of GMV from Hepsiburada, a Turkish marketplace that Kaspi acquired recently (we’ll get to that).

The local e-commerce growth story is all about expanding take rate. Amazon pulls 30–40% all-in; call it 30 cents on every dollar crossing the platform. Kaspi takes 12% on the marketplace, or 16% after adding advertising and delivery.

So the merchant commission level is roughly the same. The major difference is in the maturity of the ad and logistics business.

So I expect some significant take-rate expansion in the next few years. But it will be a steady increase. You don't close that gap by squeezing sellers. I mean, technically, Kaspi could, since the merchants have nowhere else to go, but a monopoly that farms its own captives eventually pays for it, and you don't want customers who resent you.

The better way to close it is by selling things merchants actually want. The advertising product, which is on-platform promoted listings and the highest-margin lever there is, is compounding over 70% a year and still reaches only about 7% of merchants. That's a model proven at every dominant marketplace on earth, so it just needs time to play out.

Delivery is slightly different. Kaspi’s postomats and third-party courier model are naturally less monetizable than Amazon’s or Meli’s logistics networks.

The fintech business is really interesting! Although I feel like I say this with every business unit today. When I was first pitched on Kaspi, I wasn’t sure what I thought about their lending approach. And I’ll get to that in a minute. Before that, let us quickly go over the key numbers.

Fintech accounts for 38% of revenue and 33% of net income, with roughly $24 billion in loans across four lines: buy-now-pay-later at checkout, plain consumer credit, financing for merchants and small businesses, and auto loans. In this business, it becomes clear that Kaspi started as a bank.

While Meli and Sea are still scrambling to gather deposits at scale to fund their credit books, Kaspi already funds its lending directly from more than 6 million depositors and roughly $14 billion sitting in the app. This is an advantage because customer deposits are the cheapest source of funding for the loan book.

But the real differentiator is the data, and I hate to admit it, but it makes even Meli’s data look second-class. By the time a customer applies for a loan, Kaspi already knows that person’s income – their salary lands in their Kaspi account, spending habits from Kaspi Pay being used dozens of times a month, and potentially the repayment character (likelihood to pay back debts) from prior Kaspi loans. It even captures the customer's government services footprint.

Perhaps there’s even data on the correlation between paying back loans and divorces, car accidents, or whatever else Kaspi knows about you. I wouldn’t be surprised.

One thing I do know is that all that data and Kaspi's entanglements in daily life seem to be reasons to prioritize Kaspi loans over other loans. I mentioned at the start of this section that I was a bit skeptical about the lending business. The reason was that Kaspi’s non-performing loan (NPL) coverage sits far below 100% – 78% to be exact. For comparison, Meli runs above 150% on 90-day NPLs, meaning they’re much better prepared to absorb losses.

In theory, that means that Kaspi is 22 cents underprovisioned for every bad dollar. Now, if you have a lot of asset-backed loans, that might be fine, since you can recover more of the lost loan amount by reclaiming the asset. But upon checking, that wasn’t the case for Kaspi either. The book is about as unsecured as Meli's.

What ultimately reconciles this is the so-called loss given default. Turns out, Kaspi claws back far more of a defaulted unsecured loan than a normal lender ever could, because it can see precisely when and how much money is landing in a borrower's account, and, importantly, walking away from the app that runs your entire financial life isn't a realistic option for most Kazakhs. Kaspi collects late payments months after most others would have already written the loan off. So the thin coverage ratio isn't the red flag it looks like at a glance.

After covering Kaspi’s business at home, which is undoubtedly one of the best I’ve yet looked at, let’s get to a part of the thesis that I’m a bit more torn about.

In late 2024 and early 2025, Kaspi paid $1.12 billion in cash for a 65% stake in Hepsiburada, one of the leading e-commerce platforms in Turkey. Hepsiburada runs at 16–20% market share, and Turkey has 85 million people to Kazakhstan's 20 million, so in a single move, Kaspi roughly quintupled the size of the market it's playing for, with a serious asset in hand.

You don’t need much imagination to see how this could become a major growth engine for Kaspi going forward. Turkey is culturally close to Kazakhstan; the market is five times the size; Hepsiburada is already No. 2; and Kaspi has more than 20 years of experience and a track record of running a dominant e-commerce player.

And the early success might support that view. Purchase activity is up 19% in Q4; GMV grew in the low teens, and revenue is up in the high teens, so operating leverage is already showing.

What makes me pause a little is my experience looking at other marketplaces that tried to expand internationally. Very few were successful. Coupang, one of the first companies we ever looked at in this newsletter, has largely seen its success confined to South Korea, for example. Sea Limited, on the other hand, is one of the few with success beyond its home market. And to be honest, the jury is still out on whether this Turkey initiative is only successful at grabbing market share elsewhere or will actually become a profitable business for Kaspi in the long run (although I’m optimistic it will).

Besides, expansion is also a sign that the home market seems mature. And that is where the bulk of Kaspi’s high-quality business sits.

What amplifies my concerns about the Turkish market is that it’s known to be tough to operate in. Something we haven’t yet discussed is the currency headwind that investors will face when investing in Kaspi. As many of you know, Turkey has endured hyperinflation for several years now, and while the worst might be behind it, the country’s economy is hardly a beacon of stability.

Beyond that, Turkey’s fast-growing e-commerce market makes it an interesting target for many e-commerce players. The biggest player in the market is Alibaba-backed Trendyol. Alibaba is known (as a Level 3 marketplace) to spend heavily on logistics. It’s very reasonable to assume it would do the same in Turkey to defend Trendyol’s market position.

So Kaspi might be pulled into a capex war against a much bigger competitor. And remember, one reason for the great margins at home is that Kaspi was never forced to burn through billions to build a logistics network. Many of the other advantages, like the customer lock-in through government services, don’t apply to Turkey either.

And looking at Hepsiburada’s margins, we can already see the impact of the investment cycle that kicked off when Kaspi took over. Now, to be fair, I believe Mikheil Lomtadze when he says the goal is to manage Hepsiburada at EBITDA break-even, meaning Kaspi doesn’t need to burn its own cash flows to support the Turkey business.

That said, I’m skeptical of how successful the Turkey expansion can be under those circumstances. It's fair to argue, though, that there’s a lot of potential upside with limited downside risk as long as Kaspi sticks to its strategy of not burning money in Turkey. Easier said than done when you’ve already sunk large commitments into a new market.

I talked this through with one of our Mastermind members, a fund manager in Singapore who owns shares of the stock, and while he is more bullish than I am on the expansion, we agreed that Turkey is a difficult market to operate in, and the priority needs to be not to throw good money after bad.

One thing to know about Kazakhstan is that its economy is heavily reliant on oil exports. Therefore, the local currency, the tenge, is quite volatile. That’s not good for a company that earns in tenge and reports in USD. We must deal with a version of this in Nubank and Meli, too, but it's much more severe here.

At the moment, the tenge is quite firm, around 490 to the dollar, but that’s less an oil story and more of a reaction to the central bank holding the base rate at 18% to make holding the tenge worthwhile. I won't even pretend to have an idea of where oil prices will be in the short term (or the long term, honestly), but the general worry is that oil prices will return to low levels for an extended period, undermining the tenge’s exchange rate against major currencies.

Halyk's own estimate is 600-plus to the dollar by the end of next year, on softer oil and smaller National Fund transfers. And emerging-market currencies halving against the dollar over a decade is nothing out of the ordinary – the tenge has done it before.

In Kaspi's defense, over the past decade, it compounded revenue at more than 30%, operating profit at about 36%, and net income at more than 60% – all in dollars, after the tenge drag. So in the past, this dynamic didn’t stop it from generating incredible growth rates for foreign investors.

On geography, well, Kazakhstan has the longest border in the world with a country that is currently at war – Russia. The other border is China. So it’s safe to say there are calmer spots on planet Earth to operate from.

I guess the biggest risk isn’t Kazakhstan getting drawn into some form of military conflict, but that other people’s conflicts spill over and limit the transport capacity for Kazakh oil.

The elephant in the room for today’s letter is that a prominent short-seller issued a report on Kaspi, mainly pointing out its ties to Russia, even after Ukraine-war sanctions. Candidly, I didn’t find the short report too convincing. The one aspect of it that did give me pause was the allegation that Kaspi was used indirectly to launder money for a political leader. But regulators later found Kaspi was innocent and noted that Kaspi operates in full compliance with local anti-money laundering (AML) laws and international standards.

In case you wanna read more about it, here’s the short report and here’s a pretty good rebuttal.

You might notice that I’m a bit torn on Kaspi. The core business is phenomenal, and when we go through the valuation, you’ll see that the price you pay for it is extraordinarily low. However, there are serious concerns about matters that are largely beyond Kaspi’s control. And I remain doubtful about the expansion into Turkey.

Anyway, let’s get to the valuation.

One thing I should note: Kaspi is currently trading at just 7.5x earnings, but I will actually use that same exit multiple in my model, so I don’t assume any multiple expansion. I’d love to get that upside as well, but given the macro concerns I mentioned, I think it’s somewhat reasonable to have a multiple this low.

The same goes for Turkey; I don’t expect material profit growth coming from that expansion either. All of that is a free option on top (as long as it doesn’t turn into a money burner!). The model is bottom-up, built from three core segments, each carrying a 4% annual tenge-depreciation drag, so the numbers are already in dollars. For my 5-year estimates, I have payments compounding at 5%, the marketplace at 14%, and the fintech business at 10%.

Assuming the net income margin expands to 26% by the end of the decade due to better take rates, this would result in a net income CAGR of 11%. Combined with the dividend and some slight share count decline of just 0.5% per year, it’s not unreasonable to expect a 20% IRR over the next five years.

I’m really inclined to open a starter position, but I gotta admit that Shawn is right when he says that this is a company with a huge degree of not knowing what you don’t know. So I’m inclined to follow it for a while, speak to more people who are invested in it, and then revisit it soon. If we decide to invest, you’ll be the first to know.

To listen to our discussion of Kaspi, or for more company Deep Dives, check out our podcast here.

Updates on our Intrinsic Value Portfolio below 👇

To discuss stocks daily with Shawn, Kyle, Daniel, and all Intrinsic Value Mastermind Members, apply here.

"If it weren't for us, the market would be dominated by foreign trading platforms and sellers, not local businesses.”

— Mikheil Lomtadze, Kaspi CEO

📺 WATCH: Seth Klarman’s latest interview from Master’s In Business

🎧 LISTEN: William Green’s latest interview on investor psychology with Emily Haisley

📖 READ: Fund Smith’s semi-annual letter outlining their strategic shift

You can also read our archive of past Intrinsic Value breakdowns, in case you’ve missed any, here — we’ve covered companies ranging from Alphabet to FICO, Transdigm, Lululemon, PayPal, DoorDash, Crocs, LVMH, Uber, and more!

Enjoy reading this newsletter? Forward it to a friend.

Was this newsletter forwarded to you? Sign up here.

Apply to join our Intrinsic Value Community of investors

Learn how to join us in Omaha for the 2026 Berkshire Hathaway shareholder meeting!

Read our full archive of Intrinsic Value Breakdowns here

Keep an eye on your inbox for our newsletters on Sundays. If you have any feedback for us, simply respond to this email or message [email protected].