Kaspi.kz: A Deep-Value Investment Study

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it expanding an existing pie, or creating an entirely new market?5/10

  A

  The ceiling is high enough, but most of it comes from taking a deeper share of an existing pie, not creating a new market from scratch. The only real element of "new market creation" is the still-unproven curve in Türkiye.

  Start with Kaspi's home base. Its ceiling is not "how many people in Kazakhstan still do not use smartphones," but "how much more unit value can be extracted from people who already use Kaspi." The report shows that by the end of 2025, local Average MAU in Kazakhstan was about 15.7 million and Active Merchants were about 764,000, while Kazakhstan's total population is only a little over 20 million, and World Bank data put internet penetration at about 93% in 2024. In other words, there is very limited room left for user acquisition. This company stopped being a headcount-growth story long ago. Its ceiling sits in three second-stage monetization levers where penetration is still low: advertising, delivery, and merchant finance. The data make this clear. In 2025, Kaspi's official results showed advertising revenue up 64%, e-Grocery GMV up 53%, and Marketplace Take Rate up by another roughly 80 basis points year on year. What it is doing is turning the same user base from "payment-only" users into users captured across payments, shopping, lending, advertising, delivery, and government services.

  So in the domestic market, the proper characterization is "expanding an existing pie and raising the take rate," not opening a new market. The report's Marketplace Take Rate rising from 9.2% in 2023 to 10.5% in 2025, and e-Commerce Take Rate from 11.0% to 12.7%, is exactly the picture of "on the same GMV, I take more from each tenge." The advantage of this ceiling is high visibility and strong unit economics. The drawback is that it has an inherent limit: a country's total payments volume, e-commerce penetration, which the report says was about 17.1% in 1H2025, and household borrowing capacity all have boundaries.

  The only part with a "new market creation" flavor is Türkiye. Kaspi already owns about 86% of Hepsiburada, and the report says management stated in 1Q2026 that Türkiye already accounted for about half of the company's e-Commerce GMV. Türkiye has nearly 85 million people, more than four times Kazakhstan's population, so the theoretical ceiling was indeed multiplied several times in one step. But this is exactly where a Baillie Gifford-style blue-sky narrative should not be applied mechanically: for Kaspi, Türkiye is not "creating a new market," but entering a red ocean already filled with local players and regulation, where Kaspi does not yet dominate. The report explicitly says the company intends to manage Hepsiburada around Adjusted EBITDA breakeven and continue investing.

  The honest conclusion: the domestic ceiling is high but visibly finite, driven by monetization depth rather than market creation; the Türkiye ceiling has large imaginative room, but everything remains to be proven. Baillie Gifford would ask whether this is a market that can support a fivefold return over ten years. For Kaspi, the answer is: the domestic market can support steady compounding, but not a fivefold imagination; almost the entire fivefold wager rests on whether the Türkiye playbook can be replicated. This question gets 3/5: it is a good market, but the quality of "creating an entirely new market" is limited. The main story is taking a deeper share of the existing pie.

  Jun 11, 2026
• QCan its revenue at least double over the next five years? Will growth be driven mainly by volume, price, or new businesses?6/10

  A

  Revenue is very likely to double over the next five years, but the nature of that "doubling" needs to be understood: roughly half is real growth, and half is denominator expansion from consolidating Hepsiburada. The driver mix is "volume first, price second, and Türkiye as the new business that thickens the base," not a single explosive engine.

  First anchor the base. Under the report's definition, 2025 total revenue was KZT 404.6 billion (4.046 trillion), and stockanalysis' latest compiled TTM revenue had already risen to about KZT 4.30 trillion as of 2026 Q1, showing growth was still continuing into 2026. Doubling over five years implies a CAGR of about 15%. That bar is not demanding for Kaspi; its real growth over the past few years was far above that.

  Why the doubling case is credible, but must be disaggregated:

  First, the "real" growth of the domestic core is high single digit to double digit, driven by both volume and price. Kaspi's 2025 official results disclosed underlying revenue growth of 21% and underlying net income growth of 18%, excluding smartphone GMV and the effects of tax and base-rate disruptions, while reported growth was only 19% and 10%. This gap, where underlying is materially stronger than reported, is critical: it shows that the real momentum of the domestic business, with Payments TPV +19%, Marketplace revenue +23%, and Fintech revenue +20%, remains strong. The reported softness in 2025 was mainly caused by the 24% decline in smartphone sales, high interest rates, and tax reform, rather than by the business itself losing power. Rolling forward the domestic core alone at this pace makes a five-year doubling almost natural.

  Second, the "price" contribution comes from deeper monetization, but it has limits. In the report, Marketplace Take Rate rose 80bp and advertising revenue was +64%; both reflect the pricing logic of "I take more from the same transaction." This elasticity is real, but take rates cannot rise indefinitely. It is a booster, not the main engine.

  Third, the new business, Türkiye, is the largest variable in thickening the denominator and also the source of the discount to the quality of the "doubling." Hepsiburada's 2025 revenue was TRY 84,652 million, up 13.4% year on year, while GMV grew only 4.3%. Note two things: Heps' own growth rate, with GMV at 4.3%, is far below Kaspi's domestic growth, and a meaningful portion of its revenue is low-margin retail revenue. So consolidating Heps will make consolidated "revenue" expand quickly and easily create the visual effect of a "doubling," but that is denominator expansion, not quality improvement. The report's consolidated net margin falling from 41.7% in 2024 to 26.4% in 2025 is the cost.

  Map Kaspi to Baillie Gifford's question, "Can revenue double in five years, and is it driven by volume, price, or new businesses": yes, it can double, and all three engines are running. Domestic "volume" (TPV/GMV/TFV across multiple curves) is the base, "price" (Take Rate and advertising) is the booster, and Türkiye as the "new business" is the amplifier but drags margins. Honestly, the probability of doubling is high, but if the standard is a high-quality, high-margin doubling, it deserves a discount. The clean growth comes from the domestic business; Türkiye is "build scale first, prove profit later." Score: 4/5. Volume and price are both rising and a new business supports the base, so the doubling hurdle is not high, but Türkiye dilutes the "gold content" of the growth.

  Jun 11, 2026
• QFive years from now, what will take over as the next growth engine? Does this "second curve" exist today?5/10

  A

  The second curve does exist today, and there is more than one: near term, it is "second-stage domestic monetization" (advertising / delivery / merchant finance / government services); farther out, the bet is "replicating the playbook in Türkiye." But the honest distinction matters. The former is a natural extension of the core business with limited thickening; the latter is the true successor engine, and today it is still an unproven testing ground.

  The heart of this Baillie Gifford question is "who takes over five years from now, and has that curve already been planted today?" Kaspi's answer has two layers.

  First layer (already planted, high certainty, but limited ceiling): second-stage domestic monetization. The report repeatedly emphasizes several areas where penetration is still low: advertising, delivery, merchant finance, and merchant term deposits. The data already show them growing: Kaspi's 2025 official results showed advertising revenue +64% year on year and e-Grocery GMV +53%. These are not "new businesses"; they upgrade the 15.7 million MAU already in the app from "payments users" into "advertising + delivery + finance + government-services users." The advantage is certainty, high margins, and almost zero customer acquisition cost. The limitation is that it takes over from the core business by deepening the existing pie. It can support compounding, but not "rebuild another Kaspi in five years." Government Services, with cumulative visits from 12.2 million people, is the deepest mental anchor but the least directly monetized piece; it is more a moat of user stickiness than a revenue engine.

  Second layer (just planted, the real successor candidate, but the risk is all here): Türkiye. Kaspi already owns about 86% of Hepsiburada, and the report says Türkiye already accounted for about half of the company's e-Commerce GMV in 1Q2026. This is the only curve large enough to become a new main engine five years from now: Türkiye has nearly 85 million people, more than four times Kazakhstan's population. But its current state is "value creation unproven." Heps' 2025 GMV grew only 4.3%, and full-year net loss was TRY 5,699 million. Kaspi management explicitly intends to manage it around Adjusted EBITDA breakeven and continue investing in delivery, advertising, BNPL, and UX. In other words, the "seed" of this second curve has already been bought, at a cost of about USD 1.127 billion, but whether it can grow into the successor main engine depends on whether Kaspi can transfer its Kazakhstan closed-loop method of payments + finance + e-commerce into a more complex, more crowded, and more heavily regulated market. That is the largest uncertainty named by the report.

  There is one more potential variable worth mentioning: in April 2026, Tencent invested in Kaspi.kz, alongside Lomtadze and U.S. investors. Tencent is the originator of the global super-app model. This shareholder relationship could theoretically open new synergy around Kaspi's product methodology and internationalization, but it is not yet an "existing second curve." It is only an observable long-dated option.

  The honest conclusion: second curves do exist today and have been planted, which is a positive. But the near-term curve (domestic monetization) has limited thickening and is only an extension of the main curve, while the far-term curve (Türkiye) is the true successor but has not yet proven that it will not destroy value. Score: 4/5. There is a second curve, and it is not empty PPT talk, but successful succession is far from settled.

  Jun 11, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

  A

  The core competitive advantage is a compound moat of "brand + two-sided network effects + closed-loop data risk control + dominance in high-frequency use cases." Over the next three to five years, the direction must be split by geography: the domestic Kazakhstan moat is widening or at least solidifying; in Türkiye, it is not yet meaningful to say it is "widening," only that it remains to be proven. Overall, this is not a uniformly widening moat, but a combination of "strong domestic + weak overseas."

  Start with the quality of the moat. It is real. The report shows 2025 year-end local Average MAU of 15.7 million, Average DAU of 10.7 million, and DAU/MAU of about 68%. That stickiness is top-tier among super apps globally. The harder layer is data risk control: Kaspi's 2025 official results confirmed Cost of Risk of only 2.2% in 2025, while the report says its Fintech Yield stayed at 24%. Keeping the loss ratio at 2.2% in a high-growth consumer lending / BNPL business is the result of more than a decade of credit and collection data. It cannot be bought just by spending money. This is the hardest layer of the moat to replicate.

  Why the domestic moat will "widen or solidify" over the next three to five years:

  First, second-stage monetization keeps raising switching costs. Take Rate is still rising (Marketplace from 9.2% to 10.5% from 2023 to 2025), and advertising +64%, delivery, and merchant finance are still early. Every additional use case used by a user increases the cost of leaving. Every additional service adopted by a merchant (collection + marketing + delivery + financing) raises migration cost. This is a soft moat that can widen by itself.

  Second, license and capital barriers are strengthening. The report shows Basel III total capital adequacy and Tier 1 ratios both at 19.6% in 2025, and the company has obtained Fitch BBB- / Moody's Baa3 investment-grade ratings. In April 2026, it also successfully issued USD 600 million of notes due 2031. Building credit simultaneously in local financial infrastructure and international capital markets is itself a barrier new entrants struggle to cross.

  Third, the embedding of government services, with cumulative visits from 12.2 million people, turns it from "a financial app" into "digital infrastructure." This deep binding with state administrative use cases gives it a mental position that is extremely hard to shake.

  But two places where the moat is not widening, and may even face erosion, must be marked honestly:

  First, the domestic market is not a vacuum, and competitors are not weak. Halyk Bank is Kazakhstan's largest bank, with about 35% market share and long-term ROE above 30%, and it also operates a Super-App. Kaspi is currently clearly stronger in brand, user engagement, and cross-scenario integration across e-commerce and merchants, but "stronger" is not the same as "unchallenged." The report is right to list "platform entry point replacement" as a real threat.

  Second, the Türkiye moat starts from roughly zero. Hepsiburada will not naturally inherit Kaspi's local brand trust, risk-control data, or government interfaces. Türkiye is more open, more crowded, and more complex from a regulatory perspective. The report explicitly says this can only be considered "unproven." So on a consolidated basis, the "average purity" of the company's moat has actually been diluted by consolidating Türkiye. That is exactly why the report repeatedly stresses that Kaspi is more appreciable than investable.

  Baillie Gifford would ask whether this moat will widen or narrow over three to five years. For Kaspi: the domestic section is widening (deeper monetization + stronger licenses + government-service binding), while the Türkiye section has not yet been dug. Score: 4/5. The moat is real and widening domestically; the deductions all come from "the overseas section has not yet formed + the controlling-shareholder governance discount makes the moat less investable than admirable."

  Jun 11, 2026
• QIf its core business were disrupted, does it have the genes for self-reinvention? How does it handle mistakes and bad news?5/10

  A

  It has shown some genes for self-reinvention. It grew from a single bank into a two-sided super app and then proactively made a cross-border acquisition, showing that this company is not content to defend the status quo and is willing to rebuild a successful path. In its handling of mistakes and bad news, it has a record of "acknowledge + remediate" (the 2023 material weakness in internal control has been remediated), but it also has the flaw of selectively avoiding higher governance standards. Overall: strong willingness to reinvent, above-average honesty in correction, but discounted governance transparency.

  The implicit premise behind this Baillie Gifford question is: if the core business is disrupted, does the company have the genes for regeneration? The best window into those genes is how it has historically handled its own mistakes and bad news.

  Start with the positive evidence of "reinvention genes." Kaspi's history itself is one of self-reinvention: from a traditional bank, Kaspi Bank, it actively remade itself into a two-sided super app combining payments, e-commerce, and finance, and embedded government services, travel, classifieds, and e-Grocery into the same entry point. This was not passive defense; it was an active rebuilding of "who I used to be." In 2025, it spent about USD 1.127 billion to acquire Hepsiburada, in which it now owns about 86%, moving the battlefield from familiar home territory to unfamiliar Türkiye. This is a company willing to leave its comfort zone and place long-term bets. Add Tencent's investment in April 2026, and it shows Kaspi is still actively bringing in outside capabilities rather than becoming closed and complacent. A company that grew from a bank into infrastructure and then dared to go overseas has some muscle memory for self-reinvention.

  Now look at "how it handles mistakes and bad news." This is the honesty test, and Kaspi's record here is mixed.

  Positive: it did not hide bad news. The report discloses that the company had a material weakness in internal control in 2023, but completed remediation in 2025, and external auditors recognized its 2025 internal controls as effective. Faced with Hepsiburada's material weakness in internal control that still existed in 2024, the company also disclosed it truthfully rather than dressing it up. In performance communication, management repeatedly distinguished "reported" from "underlying" in the 2025 annual report and 1Q2026 results, and proactively laid out negative factors such as the 24% decline in smartphone sales, tax reform, and interest-rate disruption instead of only emphasizing good news. A company willing to admit that "my consolidated margin was diluted by Türkiye, and reported should not be conflated with underlying" is above average in correction honesty among emerging-market companies.

  Negative (and this must be deducted): in governance, it chooses to avoid higher standards where it can. The report points out that the company uses foreign private issuer exemptions, so its board is not required to satisfy Nasdaq's "majority independent" requirement (4 of 6 directors are independent, meeting the threshold voluntarily); it has related-party transactions with Kolesa and Magnum, and the two key individuals have ownership or control relationships with those entities. This shows that on the other side of "mistakes and bad news," namely governance transparency, its self-discipline is discounted. It tends to preserve operating room for controlling shareholders within what the rules allow.

  The honest conclusion: the reinvention genes, from bank to super app to going overseas, are real, and Tencent's investment supports its openness. It acknowledges and remediates financial-level mistakes. But it does not handle governance-level "bad news" cleanly enough. Score: 4/5. The regeneration genes are strong and correction has a record, but governance transparency and controlling-shareholder culture prevent full confidence in its "self-correction."

  Jun 11, 2026
• QDoes management, especially the founders, have a long-term view and deeply aligned interests with the company? Are they willing to sacrifice current profit for five to ten years from now?7/10

  A

  Yes, and this is one of Kaspi's strongest positives: the two founders together own more than 43%, making them genuine long-term owners rather than "salaried professional managers." Their behavior, choosing not to throw all profit into high-risk expansion but to go overseas while paying dividends, also proves a long-term view. The only deduction is not "insufficient alignment." It is the opposite: the governance and related-party discount created by alignment that is too deep and a controlling-shareholder culture.

  This Baillie Gifford question asks three things: long-term view, interest alignment, and willingness to sacrifice current profit for five to ten years out. Kaspi is nearly full marks on the first two, and "willing with restraint" on the third.

  Start with interest alignment. This is the hardest evidence. According to the report's disclosure as of the end of 2025, Mikheil Lomtadze owned 22.58% and Vyacheslav Kim owned 20.74%, together over 43%. I cross-checked this with a primary public source: Kazakhstan Stock Exchange (KASE) shareholder information as of July 1, 2025 showed Lomtadze at 22.51% and Kim at 21.32%, broadly matching the report, with small differences due to disclosure timing. Founder ownership at this scale means their personal wealth moves deeply with per-share value. This is not token "alignment" from a few options; it is true "the company is my net worth." The two founders have long co-led the company and maintained stable ownership, so the long-term view is not a PPT slogan.

  Now look at "willingness to sacrifice current profit for the long term." Kaspi offers a restrained yes, and the restraint itself is positive. On one hand, it dares to place a long-term bet: it spent about USD 1.127 billion to acquire Hepsiburada, in which it now owns about 86%, explicitly saying it would manage it around Adjusted EBITDA breakeven and continue investing in delivery / advertising / BNPL. That is exchanging current margins (consolidated net margin fell from 41.7% to 26.4%) for Türkiye's long-term possibility. On the other hand, it did not make an all-in gamble: Kaspi's 2025 official results show the company also maintained a dividend of KZT 850 per ADS and an approximately 64% payout ratio, corresponding to a current dividend yield of about 8.9%, and management said this level was sustainable at least for the rest of 2026. "Going overseas while paying dividends, without putting all profit into high-risk expansion" is exactly mature and disciplined capital allocation. It shows management has long-term ambition and also respects the risk of a single large bet. On capital allocation, the report discloses cumulative repurchases of about 4.5965 million GDR/ADS from 2022 to 2025, with spending of about KZT 149.1 billion. Repurchases basically offset share-based compensation dilution (the public share count was almost flat over two years), further confirming that management is rational about preserving shareholder value.

  But the deduction must be stated honestly. Its problem is not "insufficient alignment" but the side effect of "alignment that is too deep." The report points out that the company uses foreign private issuer exemptions, so the board does not have to meet Nasdaq's ordinary majority-independent-director rule; it has related-party transactions with Kolesa and Magnum, and the two key figures have ownership or control relationships with those entities. High control in an emerging market creates a reflexive discount. The market is not worried that founders are inattentive; it is worried whether minority shareholders will always be treated equally.

  Baillie Gifford would ask whether the founders are long-term owners and whether they are willing to sacrifice the present for the long term. For Kaspi: they are real owners (>43% ownership), they clearly have a long-term view, and they invest for the long term in a restrained way. Score: 3.5/5. Alignment and vision are top-tier; all deductions come from the minority-shareholder-protection discount created by controlling-shareholder culture + related-party transactions + governance exemptions. That discount is structural and will not disappear quickly.

  Jun 11, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its growth model sustainable and not dependent on harming society or regulators?6/10

  A

  If Kaspi disappeared tomorrow, people in Kazakhstan would miss it badly. It has already become national-level digital infrastructure, with daily activity penetrating almost every routine part of payments, shopping, borrowing, and government services. Its indispensability is extremely high. But there is a question mark around whether its growth model is sustainable and not harmful to society or regulators: a substantial portion of its profit comes from consumer lending in a high-interest-rate environment. With the central bank base rate at 18% and regulation tightening around consumer loans, the social and regulatory sustainability of this growth path is the soft spot that needs constant monitoring.

  This Baillie Gifford question has two cores: "indispensability" and "whether that indispensability is built on a sustainable foundation that does not exploit society or regulators." Both must be answered, and the second point matters especially for a financial platform involved in consumer lending.

  Start with indispensability. It is extremely high, and real. The report shows local Average DAU of about 10.7 million at the end of 2025 and DAU/MAU of about 68%, while Kazakhstan's total population is only a little over 20 million. This means more than half of the country's adults open Kaspi every day. Its "scenario depth" is even more telling: Kaspi's 2025 official results disclosed that in 4Q2025, each active user averaged 77 transactions per month. This is not "occasional use"; it is embedded into multiple must-use moments in a day. The report also discloses that Government Services had reached 12.2 million cumulative visits, meaning it even hosts functions that would normally be provided by the state. An entry point that simultaneously carries your salary card, grocery shopping, repayments, certificates, and rides, if it disappeared, would cost users not "an app" but "an entire life operating system." The degree to which it would be missed is comparable to a country suddenly shutting off its payments and government-service network. This indispensability is the full-mark version of the "how much would customers miss it" question that Baillie Gifford values most.

  But the second layer, the social and regulatory sustainability of growth, requires honest cold water.

  Risk one: profit dependence on high-interest consumer lending. The report shows Fintech is the main profit source (2025 net income of about KZT 819.1 billion, far above Payments and Marketplace), and Fintech Yield is as high as 24%. In Kazakhstan, where the central bank has maintained the base rate at 18.0% and March inflation was still 11.0%, the high yield of consumer lending naturally sits in the regulatory spotlight. The central bank's policy statement explicitly listed "slower consumer loan growth" as one factor in easing inflation, meaning regulators have both the willingness and the tools to step on the brakes. If regulators move on consumer-loan rate caps, reserves, or lending pace, Kaspi's most profitable engine would be hit directly. This is the core question around whether growth is sustainable and whether it depends on a regulatory opening that may tighten at any time.

  Risk two: the regulatory complexity of cross-border expansion. After entering Türkiye, the report specifically discloses Turkish e-commerce regulatory risk. This is a more complex regulatory market where Kaspi does not dominate, making it much harder than at home to "not harm regulators."

  To be fair to Kaspi: its Cost of Risk is only 2.2% and NPL is about 6%, which shows its lending is not a "predatory high-rate + high-bad-debt" model, but prudent lending driven by data risk control. That offsets part of the deduction on the "not harming society" dimension. It serves real household needs in payments, shopping, and borrowing; it is not manufacturing a bubble.

  Baillie Gifford would ask how much customers would miss it if it disappeared, and whether growth is sustainable without harming society or regulators. For Kaspi: the degree of being missed is full marks (national-level infrastructure), but sustainability deserves a deduction (profit heavily relies on high-interest consumer lending and sits within range of regulatory tightening). Score: 4/5. Indispensability is textbook-level; the deduction is that part of this indispensability rides on a consumer-credit tailwind that regulators may adjust at any time.

  Jun 11, 2026
• QHow are the unit economics of this business (gross margin, incremental returns)? Do they improve or deteriorate with scale? Where does the money it earns go?7/10

  A

  Domestic unit economics are outstanding: Payments is light-asset and high-margin, Fintech combines high yield with low losses, and overall return on equity has long been at globally rare levels. But on the question "does scale make it better or worse," Kaspi gives a forked answer: domestically, bigger is better through deeper monetization; overseas, Türkiye is pulling consolidated unit economics down. The money earned is allocated rationally: mainly dividends and buybacks, with overseas investment as a secondary use, and no reckless spending.

  This Baillie Gifford question has three parts: unit economics (gross margin / incremental return), direction of scale effects, and where the money goes. Take them one by one.

  First, the absolute level of unit economics: domestically, top-tier. The report discloses segment structure: in 2025, Payments revenue of about KZT 659.0 billion corresponded to net income of about KZT 433.0 billion (net margin about 66%); Fintech revenue of about KZT 2,481.4 billion corresponded to net income of about KZT 819.1 billion, with Cost of Risk only 2.2% and Fintech Yield stable at 24%. A Payments business with "almost zero marginal cost, where each additional transaction is high-margin" is the ideal form of unit economics. Fintech is the rare combination of "high yield + low losses." The best proof of incremental return is return on capital: the report's simplified ROE estimate using beginning and ending equity attributable to shareholders was about 88.7% in 2023, 80.1% in 2024, and 53.5% in 2025. Even after stepping down in 2025 due to Türkiye consolidation and a larger equity base, 53.5% ROE is top-tier among any global financial or platform company. This shows that each tenge retained and reinvested by Kaspi can create returns far above the cost of capital.

  Second, "does scale make it better or worse." This is the part that most needs disaggregation, and the answer is forked:

  Domestic: bigger is better. In the report, Marketplace Take Rate rose from 9.2% in 2023 to 10.5% in 2025, e-Commerce Take Rate rose from 11.0% to 12.7%, and advertising revenue was +64%. This is classic positive scale effect from deeper monetization. More users and merchants -> denser network -> more value-added services can be taken from each transaction -> better unit economics. This is the virtuous cycle of "the bigger it gets, the more profitable it becomes."

  Consolidated: pulled down by overseas. This fact must be marked honestly: Kaspi's consolidated net margin fell from 41.7% in 2024 to 26.4% in 2025 because Hepsiburada consolidation brought in a large amount of low-margin retail revenue. Heps' own 2025 full-year net loss was TRY 5,699 million, and GMV grew only 4.3%. So from the "whole company" perspective, scale in the short term is "larger but worse unit economics." This is not deterioration in the domestic business; it is the company actively consolidating a low-margin expansion vehicle into a high-quality parent. Baillie Gifford would be alert to this: if scale expansion dilutes unit economics, the question becomes "is the incremental return on this expansion worth it?" That is the core of the unproven Türkiye case.

  Third, where does the money go? Rationally. The report discloses three uses: first, dividends, at KZT 850 per ADS and about a 64% payout ratio, corresponding to a current dividend yield of about 8.9%; second, buybacks, with cumulative repurchases of about 4.5965 million GDR/ADS from 2022 to 2025 and spending of about KZT 149.1 billion, broadly offsetting share-based compensation dilution; third, overseas investment, namely the approximately USD 1.127 billion acquisition of Hepsiburada. "Return most of it to shareholders, invest part of it for the long term, and do not pour all profit into high-risk expansion" is mature and restrained capital allocation, not cash-burning.

  Baillie Gifford would ask how unit economics are, whether scaling makes them better or worse, and where the money goes. For Kaspi: domestic unit economics are top-tier and improve with scale; consolidated economics are temporarily pulled down by Türkiye; capital allocation is rational. Score: 4/5. Domestic unit economics and capital returns are world-class; the deduction is that consolidated scale effects are currently negative, and whether Türkiye can eventually restore unit economics remains unproven.

  Jun 11, 2026
• QWhat conditions must all hold for it to rise fivefold in ten years? Are those conditions realistic? What expectations are embedded in today's share price?3/10

  A

  For it to rise fivefold in ten years, two things must both happen: "profit rises about 2.5x + the valuation multiple rises about 2x." The latter, discount narrowing, is precisely the link Kaspi cannot count on. Honestly, a fivefold return over ten years is not fantasy, but it will not happen automatically from "cheap + compounding." It must be layered with an optimistic scenario in which the market is willing to remove the regional / governance discount. Conversely, today's share price of about USD 78.6 implies very low expectations. The market is barely pricing in success in Türkiye or any discount narrowing. That is both the risk and the source of asymmetry.

  First quantify the target. Baillie Gifford's "fivefold in ten years" corresponds to a total return CAGR of about 17.5%, including dividends. Starting from the current market cap of about USD 1.493 billion and share price of about USD 78.58 (June 10, 2026 close), fivefold would mean a market cap of around USD 7.5 billion ten years later, if relying only on the share price. This target can be broken into two conditions that must both hold:

  Condition one: sustained high-quality profit growth (relatively feasible). Current TTM revenue has already risen to about KZT 4.30 trillion, and 2025 underlying revenue / net income grew 21% / 18%. If the domestic business maintains high single-digit to double-digit growth and Türkiye gradually moves from losses to contribution, profit rising 2-2.5x over ten years is realistic. This half, Kaspi is likely to deliver.

  Condition two: a significant expansion in the valuation multiple (most doubtful, and the true swing factor for the fivefold outcome). Kaspi currently trades at a trailing P/E of only about 6.7x. If profit is 2.5x higher in ten years but P/E is still 6.7x, the share price would only rise 2.5x, far short of fivefold. To close the gap, P/E must expand from 6.7x to around 13-14x. The problem is that this discount is structural. Compared with global peers, MercadoLibre's current P/E is about 43x, Nu Holdings is about 20x, and Block's forward multiple is about 20x. Kaspi's ROE and scenario density are no weaker, but it receives only a single-digit P/E. The gap is the market's pricing of Kazakhstan / Türkiye sovereign risk, regulation (central bank base rate of 18%), and controlling-shareholder governance discount. Getting P/E to double is equivalent to asking the market to believe these risks will ease for a long time. That cannot be delivered unilaterally by company operations; it depends on perceived institutional quality in a country, proof of successful Türkiye integration, and substantive improvement in governance transparency.

  Put the two conditions together: a fivefold return in ten years = profit up 2.5x (feasible) x P/E doubled through discount narrowing (highly uncertain). This is the logic behind the report's "optimistic scenario of USD 130-170"; it explicitly says the key to the optimistic scenario is "partial narrowing of the valuation discount." So the honest conclusion is: a fivefold outcome requires the optimistic scenario to hold, and the life-or-death variable in that scenario is not growth, but whether the discount can narrow. That is not fully under the company's control.

  Now answer Baillie Gifford's key sub-question: what expectations are embedded in today's share price? The answer: expectations are so low that they are close to "just do not collapse." A 6.7x P/E and 8.9% dividend yield mean the market is effectively pricing Kaspi as "a company whose profit may stagnate, whose Türkiye expansion may fail, and whose discount may never disappear." It gives no credit to Türkiye success, no credit to discount narrowing, and even doubts steady domestic compounding. This is the essence of the report's "obvious discount relative to global peers." Such low implied expectations are a double-edged sword. The bad side is that the market may keep seeing it this way forever, the fivefold case may fail, and returns may come only from earnings plus dividends. The good side is asymmetry: an investor does not need the optimistic scenario to get a reasonable return (8.9% dividends plus domestic double-digit growth is already decent), and if the discount narrows even partially, the fivefold upside option ignites.

  Baillie Gifford would ask what conditions must all hold for a fivefold return in ten years, whether they are realistic, and what is priced in today. For Kaspi: profit must rise 2.5x (realistic) and the P/E must double through discount narrowing (uncertain); today's share price implies extremely low expectations and prices in almost no optimistic scenario. Score: 4/5. The fivefold path is clear but depends on uncontrollable discount narrowing, so the deduction is that the swing factor is not in the company's hands. Still, low implied expectations provide real downside protection and upside optionality, which makes it more durable than a pure growth stock.

  Jun 11, 2026
• QWhy has the market not realized all this yet? Is it because it does not understand, looks down on it, or cannot look far enough? What would become the "narrative inflection point"?3/10

  A

  The market does not "fail to understand" Kaspi's business quality. Its ROE, stickiness, and moat are already understood by professional investors. Nor does the market "look down on it"; sell-side coverage and index inclusion are already there. The issue is "it cannot look far enough + it does not dare to believe." The market knows this is a good business, but does not dare believe that Kazakhstan / Türkiye sovereign, regulatory, and governance risks will ease. So it prefers to assign a permanent discount rather than pay for long-term imagination. The narrative inflection point will not come from another strong earnings report, because that will not change the discount logic. It will come from structural events that make the market believe the regional / governance discount should narrow.

  This Baillie Gifford question requires distinguishing three reasons why "the market has not realized it": does not understand, looks down on it, or cannot look far enough. For Kaspi, the first two basically do not apply. The real answer is the third, with a more precise phrase: "does not dare to believe."

  Why it is not "does not understand." Kaspi's business quality is visible and almost undisputed: 2025 underlying revenue / net income grew 21% / 18%, ROE remained about 53.5% even after falling, Cost of Risk was only 2.2%, and DAU/MAU was about 68%. Any professional investor can understand these figures. The report itself also says, "if one looks only at business quality, it may be one of the rare high-quality platform fintech assets in global public markets in recent years." The "does not understand" argument does not hold.

  Why it is not "looks down on it." It is already listed on Nasdaq, covered by mainstream sell-side analysts, and has obtained Fitch BBB- / Moody's Baa3 investment-grade ratings. Its April 2026 bond issuance was about 3.5x oversubscribed, and in April 2026, Tencent even participated in the investment. Even a top global strategic investor "respects" it. The "looks down on it" argument also does not hold.

  The real reason is "cannot look far enough + does not dare to believe." The market's logic is: I accept that you are highly profitable today, but I cannot reliably predict the policy environment in Kazakhstan five years from now, the integration outcome in Türkiye, or the direction of controlling-shareholder governance. So I would rather pay a 6.7x P/E only for the "certain present" and not pay a premium for the "uncertain future." The rationality of this discount is clearest in the peer set: MercadoLibre is about 43x, Nu is about 20x, and Block is about 20x forward, while local peer Halyk Bank is even lower, at only about 3.95x P/E. The entire Kazakhstan asset belt has been given a "country discount." This is not the market failing to understand Kaspi; it has understood Kaspi and applied a jurisdictional risk label to it. The report puts it precisely: "the market clearly does not price Kaspi as a pure high-quality platform, but continues to apply regional and governance discounts."

  So what would become the narrative inflection point, the key sub-question Baillie Gifford adds? The key insight is that a better domestic earnings report is not the inflection point, because it does not touch the root cause of the discount. The true narrative turn would come from structural changes that make the market "dare to believe." I see four categories:

  First, Türkiye turns from "value unproven" to "value proven." If Hepsiburada clearly turns profitable over the next few years and proves that Kaspi's playbook can migrate, the market will rewrite "Türkiye is a drag" into "Türkiye is the second growth pole." This is the single largest inflection point.

  Second, macro normalization. Kazakhstan's central bank has hinted that if inflation (11.0% in March) continues to fall, it may consider rate cuts in the second half. Once the 18% base rate enters a downward path, it benefits Kaspi's funding costs and consumer lending, and it also lifts overall market risk appetite for Kazakhstan assets.

  Third, substantive improvement in the governance discount. If the company proactively reduces related-party transactions, improves board independence, or if external strategic shareholders such as Tencent bring stronger governance constraints, the "governance" part of the discount would have a reason to narrow.

  Fourth, inclusion in more mainstream indexes or endorsement by benchmark long-term capital could change the marginal buyer base.

  The honest conclusion: what the market has not realized is not "the business is good," but "whether the discount should be this deep." The essence of the inflection point is "sovereign / regulatory / governance risk is proven to be narrowing," not another high-growth quarter. Score: 4/5. The mispricing logic is clear and the inflection points are identifiable, but the deductions are that these inflection points, especially Türkiye proof and macro normalization, take time and are not under the company's sole control. The discount may remain more stubborn and last longer than expected.

  Jun 11, 2026