Kazakhstan's dominant super app — a WeChat-PayPal-Amazon hybrid at a fraction of their valuation, with a Turkish gamble attached.
Kaspi.kz is the rare genuinely original business model in emerging markets — a vertically integrated super app combining payments, e-commerce, and consumer lending into one mobile ecosystem. The closest analogues are WeChat Pay (payments), MercadoLibre (marketplace), and a sub-prime consumer bank, all rolled into a single app with 25 million active users across Kazakhstan and, since 2025, Turkey. There is no meaningful Western equivalent.
The company's insight was simple but powerful: every financial transaction in Kazakhstan was a pain point. By owning the digital payment rail (Kaspi Pay, Kaspi QR, peer-to-peer transfers) and using that data to underwrite consumer credit, then building a marketplace on top of the same app, Kaspi created a closed loop that gets stronger with every new user and merchant.
| Segment | Q1 2026 Revenue | YoY Growth | Adj. EBITDA | Key Driver |
|---|---|---|---|---|
| Marketplace | $1.1B (KZT 520B) | +49% | $247M (+12%) | e-Commerce GMV +41%; Hepsiburada (Turkey) now 50% of GMV |
| Fintech | $897M (KZT 430B) | +25% | $334M (+12%) | Loan portfolio +23% YoY; pivot to longer-duration higher-yield loans |
| Payments | $331M (KZT 158B) | +7% | $187M (flat) | TPV +14% to $23.7B; mix shift to zero-fee QR compresses take rate |
| Total | $2.3B (KZT 1.1T) | +31% | $768M (+9%) | Net income $526M (flat YoY) |
Kaspi's revenue is a hybrid of transactional and quasi-recurring streams. Payments fees recur predictably with user behaviour (77 monthly transactions per active user — an extraordinary engagement metric). Fintech interest income is contractually sticky over loan duration. Marketplace GMV is volume-dependent. Overall, roughly 55–60% of revenue is quasi-recurring in nature, with high visibility from the existing loan book and embedded payment habits. No single customer represents even 1% of revenue.
Revenue scale: Full-year 2025 revenue was $7.7 billion (up 45% YoY), with TTM revenue now at approximately $7.9 billion. Market cap as of May 14 is $16.2 billion. Enterprise value is approximately $16.5 billion (near-zero net debt prior to recent bond issuance). The company employs approximately 16,000 people.
Pricing power: Kaspi has demonstrated consistent take-rate expansion in e-commerce and fintech, though payments take rates are structurally declining as volume migrates to Kaspi QR (a zero-fee product that benefits engagement but not near-term revenue). The company's ability to add advertising revenue (up 73% YoY in Q1 2026) atop its GMV base represents pricing power through new monetization layers rather than price increases.
Geographic concentration risk: This is the central bear thesis. Kazakhstan generated the overwhelming majority of earnings. Turkey (Hepsiburada) now represents 50% of e-commerce GMV but near-zero EBITDA. Kazakhstan's GDP is approximately $260 billion, tightly correlated with oil prices and geopolitically proximate to Russia.
Revenue CAGR 2021–2025 approximately 55% in USD terms; note 2023–2025 includes Hepsiburada consolidation effect.
| Metric | Q1 2026 | FY 2024 | FY 2023 | Trend |
|---|---|---|---|---|
| Net Income Margin | ~23% | ~41% | ~44% | ↓ Dilution from Hepsiburada losses |
| Adj. EBITDA Margin | ~34% | ~44% | ~46% | ↓ Marketplace cost weight |
| Fintech EBITDA Margin | ~37% | ~45% | ~50% | ↓ Funding cost pressure |
| Payments EBITDA Margin | ~57% | ~60% | ~65% | ↓ QR product mix shift |
| ROE | 51.1% (TTM) | Elite-tier; exceptional | ||
| ROIC | 84.2% (TTM) | Far above cost of capital | ||
Net margins have compressed from ~44% in 2023 to ~23% on a blended Q1 2026 basis. This is almost entirely explainable by two factors: (1) Hepsiburada runs near EBITDA breakeven and dilutes consolidated margins; (2) Kazakhstan's base interest rate remained elevated (NBK rate ~14–15%), compressing Fintech EBITDA as deposit funding costs rose ~220bps YoY. Management expects these headwinds to normalize as rates come down. The Kazakhstan core business remains extraordinarily profitable. This is the critical distinction between temporary margin dilution and structural impairment.
| Item | Value (Q1 2026) | Commentary |
|---|---|---|
| Total Assets | $23.2B | Primarily loan book ($15.2B) and customer accounts ($15.5B) |
| Customer Loans (Net) | $15.2B | Average portfolio grew 23% YoY; cost of risk 0.7% — excellent |
| Customer Deposits | $15.5B | Primary funding source; creates natural asset-liability match |
| New Bond Issuance | $600M at 5.9% | 5-year senior unsecured notes (2031); general corporate purposes / Turkey |
| Tier 1 Capital Ratio | 20.1% | Basel III; well above minimums. KZ regulator Tier 1 at 13.5% |
| Debt/Equity | 0.14× | Very low leverage for a bank-holding company |
| Free Cash Flow (FY2024) | ~$2.0B | FCF consistently exceeds reported net income pre-Turkey integration |
Kaspi's business is inherently self-funding: customers pay interest on loans, deposits provide cheap funding, and payment fees come in daily. The company historically generated $1.5–2.0B in annual free cash flow against a market cap of ~$16B, implying an FCF yield of approximately 9–12%. The Hepsiburada integration requires incremental capital investment (~$300M for Rabobank Turkey acquisition) but does not threaten the core cash-generation engine. Working capital is positive: Fintech earns interest before paying depositors on a net basis.
"My own investment, together with those of key members of our senior management team, reflects our strong belief in Kaspi.kz's Super App business model, long-term strategy and future growth opportunities."
— Mikheil Lomtadze, Co-Founder & CEO, April 2026
Mikheil Lomtadze is the co-founder and CEO, a Georgian-born entrepreneur who co-founded Kaspi.kz in 2008 alongside Vyacheslav Kim. Lomtadze previously held senior roles at Baring Vostok Capital Partners and has been the architect of Kaspi's super app strategy from inception. This is not a hired-gun CEO parachuted into a troubled business — he built the business from zero. He has been in the role for 17+ years, making him among the longest-tenured founder-CEOs at a publicly listed company of this size.
Track record at Kaspi: Under Lomtadze's leadership, Kaspi grew revenue from ~$500M in 2019 to $7.7B in 2025 — a 15× increase in six years — while maintaining net margins above 40% in the core Kazakhstan business. Lomtadze's key strategic decisions — building the super app flywheel, resisting geographic over-expansion for over a decade, and the eventual deliberate move into Turkey — have all been executed with discipline. The one live unknown is Turkey: that bet is still unproven.
Skin in the game: Critically, in April 2026, Lomtadze personally purchased additional ADSs from Baring Fintech Venture Funds, alongside Tencent and senior management. This is a textbook high-conviction signal. He co-owns a meaningful stake in a $16B company; his personal wealth is substantially tied to KSPI's performance. The Tencent investment — acquiring a block from an early fund exiting — also validates the business model from one of the world's most sophisticated tech investors.
David Ferguson serves as head of investor relations and is the external voice of management strategy. The CFO and COO structure within Kaspi is less prominently disclosed than is typical for Nasdaq-listed companies, which is a minor governance concern — investors must rely more heavily on Lomtadze's vision and judgment than would be ideal. The management team has been stable, with low visible turnover at the senior level.
Board independence: Kaspi is listed on Nasdaq and therefore subject to its governance requirements. The board includes independent directors, though like most founder-led emerging market companies, the power sits firmly with the founder-CEO. Deloitte LLP serves as external auditor — renewed at the April 2026 AGM. This is a credible audit firm for cross-border IFRS reporting.
Capital allocation philosophy: Management has consistently returned capital while funding organic growth — a rare combination. Dividends have been quarterly and substantial (64% payout ratio in Q1 2026). The company has not pursued empire-building acquisitions; Hepsiburada is their first and only major external deal, and it was pursued only after exhausting Kazakhstan's organic opportunity set. This is founder discipline, not financial engineering.
Kaspi's moat is genuine, multi-layered, and reinforcing. This is not a company that claims a moat because it has a brand or a patent — it has structural competitive advantages that would take a well-funded competitor 5–8 years and billions of dollars to replicate, if they could at all.
Network effects: Every additional merchant on Kaspi Pay makes the payment network more useful to consumers; every additional consumer makes the marketplace more attractive to merchants; every transaction generates underwriting data that improves Fintech credit decisions. These are three interlocking network effects operating simultaneously. The company's 77 monthly transactions per active user is extraordinary by global standards — it reflects daily habitual use.
Switching costs: A user with their salary deposited at Kaspi Bank, their mortgage through Kaspi Fintech, their utility payments on Kaspi Pay, and their household shopping through Kaspi Market faces enormous friction to switch. Each product individually might have low switching costs; the bundle creates a nearly unbreakable relationship.
Scale and data advantage: Kaspi processes more transactions annually than Kazakhstan's entire GDP multiple times over. The resulting credit scoring model — built on real payment behavior, not just credit bureau data — is not replicable without the same transaction volume. This creates a permanent underwriting advantage over any entrant.
The Kazakhstan core moat is wide and durable. The company controls approximately 90%+ of digital payments volume in Kazakhstan and is classified as a systemically important financial institution. Regulatory barriers alone deter entry. The Turkish moat is thin to non-existent — Hepsiburada competes against well-funded local and international rivals. The moat building in Turkey hasn't started yet; that's the risk.
Evidence of moat in the numbers: 51% ROE, 84% ROIC, 34%+ EBITDA margins on a consolidated basis despite Hepsiburada drag, and 15 years of uninterrupted market-share growth. These fingerprints confirm a genuine wide moat in the core Kazakhstan business.
Kazakhstan's digital economy is structurally underpenetrated and growing rapidly. The country has a population of ~20 million with a median age of ~30, high smartphone penetration, and historically low but rapidly rising e-commerce adoption. The IMF projects Kazakhstan GDP growth at 3–4% annually over the next five years, underpinned by oil revenues and a nascent non-resource sector. The digital payments and lending TAM in Kazakhstan alone is estimated at $50–80 billion in annual transaction volume; Kaspi already processes ~$166 billion — indicating it has moved from serving the TAM to defining it.
Turkey is the strategic prize and the strategic risk. A market of 85 million people with GDP of ~$1 trillion, significant e-commerce penetration but fragmented payments and underdeveloped BNPL lending infrastructure. If Kaspi can replicate even a fraction of its Kazakhstan playbook in Turkey, the incremental revenue opportunity dwarfs the current business. Turkey's e-commerce market is growing ~25% annually. The challenge: established competitors include Trendyol (Alibaba-backed), Amazon, and local fintech players with deep pockets.
Competitive intensity in Kazakhstan: Essentially nil. Kaspi's 90%+ payment market share and its regulatory status as a systemically important bank create near-insurmountable barriers. Halyk Bank is the nearest banking competitor but lacks Kaspi's technology infrastructure. Global players like Visa and Mastercard are payment scheme partners, not direct competitors.
Regulatory environment: Kazakhstan's fintech regulation acts primarily as a moat — the banking license is extremely difficult to obtain, and Kaspi's designation as a systemically important institution comes with preferential treatment (and added regulatory obligations). The one live regulatory risk: Kazakhstan periodically adjusts smartphone registration requirements and payment framework rules, each of which caused short-term GMV disruption in 2024–2025. These are friction, not structural threats.
Cyclicality: The 2020 COVID period saw brief disruption followed by rapid acceleration (digital adoption surged). Kazakhstan experienced an oil-driven downturn in 2015–2016 that severely impacted consumer lending — a reminder that the Fintech segment is vulnerable to macro deterioration. The company did not exist publicly through 2008–2009.
The stock peaked at $131.76 in July 2024 — it currently trades at $88.65, a 33% decline from the all-time high. Several factors explain the discount:
1. The Culper Research short-seller attack (September 2024) accused Kaspi of Russian sanctions exposure. Kazakhstan regulators defended Kaspi; the allegation was not substantiated but permanently raised geopolitical risk perception.
2. Hepsiburada consolidation diluted consolidated margins and introduced Turkish macro exposure (lira volatility, high inflation, complex regulatory regime).
3. Kazakhstan interest rate pressure compressed Fintech EBITDA margins in 2025.
4. Structural emerging market discount — Western investors apply a 40–60% multiple discount to Kazakhstan-domiciled companies regardless of business quality, due to political risk, currency risk, and liquidity risk.
| Assumption | Input |
|---|---|
| Revenue growth (5-year CAGR) | 15% (vs historical 40%+) |
| Terminal EBITDA margin | 32% (vs current 34%) |
| Discount rate | 13% (EM-adjusted WACC) |
| Terminal growth rate | 3% |
| Implied intrinsic value | $115–140 per ADS |
| Current price | $88.65 |
| Implied upside | 30–58% under conservative assumptions |
Even using a discount rate 3 percentage points above what you'd apply to a US fintech, and assuming revenue growth falls to less than half the historical CAGR, the stock appears meaningfully undervalued. Simply Wall St's DCF model arrives at a fair value of ~$205 per ADS. The consensus analyst average price target is $104–122.
Value trap risk: Low. The business is not in structural decline — it's in an investment phase. Insider buying (Lomtadze's own capital in April 2026) and Tencent's entry are strong signals against value trap interpretation. The low share count growth (0.25% dilution in one year) confirms this is not a dilution story masquerading as growth.
Dividends: Kaspi pays quarterly dividends of KZT 850 per ADS, implying a ~64% payout ratio at current earnings and an approximately 8% annual yield at the current share price. The dividend has been consistent and growing, representing an unusually generous capital return policy for a company still investing aggressively in international expansion. The board approved another KZT 850 dividend in Q1 2026, demonstrating confidence despite net income pressure.
Share buybacks: In 2025, management conducted an approximately $100M ADS buyback program, and share count has remained nearly flat — dilution from stock-based compensation is well-contained (only 0.25% dilution in 12 months). This is excellent discipline: they are not issuing equity to fund empire-building and not diluting shareholders through excessive SBC.
M&A track record: Kaspi's M&A history is short but consequential. The Hepsiburada acquisition ($1.127B for 65%+, now 86% stake) was the first major external deal, completed in 2025 after years of organic-only growth. It is too early to judge the outcome. However, the strategic logic — replicating the super app flywheel in Turkey's 85-million-person market using an existing e-commerce platform as the anchor — is coherent and mirrors how WeChat Pay and AliPay built dominance. The risk: Turkey is not Kazakhstan, and competitive dynamics are far more complex.
Organic reinvestment: R&D and technology spend is embedded throughout the business model rather than broken out separately (Kaspi reports as a financial institution). The consistent pace of product launches — Alaqan palm payment, B2B payments expansion, travel booking, government services integration — indicates high organic reinvestment in product capability.
Debt management: The April 2026 $600M bond issuance at 5.9% (5-year) is the first significant external debt raise at the holding company level. It is credit-positive (oversubscribed, investment-grade pricing) and adds financial flexibility for Turkey. Leverage remains modest.
Full-year 2026 guidance (unchanged): GMV growth of 20%; TPV growth of 15%; TFE growth of 5%. These are constant-currency, pro-forma targets that include Hepsiburada. The guidance was maintained in full after Q1 — a signal that management has high confidence in the trajectory.
Turkey: the central narrative. Kaspi's ambition is to replicate its super app flywheel in Turkey. The immediate plan: run Hepsiburada near EBITDA breakeven while expanding user base; close the Rabobank A.Ş. banking license acquisition (mid-2026 expected); launch Kaspi fintech products (BNPL, savings, consumer loans) via HepsiPay mobile channel; and target 100 million total users across both markets long-term. Hepsiburada was near EBITDA breakeven in Q1 2026 — this is early evidence of execution progress.
Turkey already represents 50% of e-Commerce GMV. Orders grew 43% pro-forma. Purchases per consumer in Kazakhstan reached 15 per quarter, up 44% — the engagement deepening thesis is intact. Advertising and delivery revenue grew 73%, the highest-margin revenue stream within Marketplace. These are measurable early signs the strategy is working, not vaporware.
Kazakhstan product innovation: The palm-payment system (Alaqan) positions Kaspi in biometric payments. Government services integration (bill payments, tax filings, pension contributions via the app) deepens the civic utility of the super app in ways that are nearly impossible for a private competitor to replicate — these are exclusive digital service contracts.
Management credibility on guidance: Historically strong. Kaspi has beaten consensus estimates consistently. Q1 2026 results beat analyst revenue expectations by 4.7% and EPS by 6.76%, continuing the pattern. The one credibility dent: full-year 2025 net income guidance was initially set at 18–20% and revised down to 10–12%, primarily due to Hepsiburada integration investment and higher-than-expected Kazakhstan funding costs. This was flagged proactively, not as a surprise miss.
Uzbekistan interest: Kaspi has filed a formal letter of interest to participate in the privatization of Humo, Uzbekistan's payments system. This suggests the Turkey bet — if successful — would be followed by further Central Asian and Caucasus expansion. This is a very long-dated option, but it signals management's thinking about the platform's geographic potential.
AI as a threat to Kaspi: Low to moderate. Kaspi's competitive advantage lies in its payment infrastructure, banking license, and network effects — none of which AI can replicate. AI-powered fintech entrants could theoretically improve credit underwriting for competitors, but Kaspi's underwriting data moat (billions of real transactions) gives it an AI training advantage, not a disadvantage.
AI as an internal tool: Kaspi uses machine learning extensively in its credit risk models, fraud detection (real-time transaction monitoring across 77 monthly transactions per user), personalization of marketplace recommendations, and merchant analytics. These are not disclosed in detail but are standard applications at this scale of transaction data. The company's engineering team has expanded significantly to support the Turkish platform integration.
Tencent's investment as an AI signal: Tencent is one of the world's most sophisticated AI and super app operators. Its decision to become a significant shareholder in Kaspi — described as a "primarily financial investment" — nonetheless brings strategic optionality for technology collaboration. Tencent's super app expertise (WeChat has executed across 1.3 billion users what Kaspi is attempting across 25 million) represents potential product knowledge transfer.
Data assets: Kaspi owns one of the most comprehensive individual-level financial behaviour datasets in Central Asia. Every purchase, payment, loan application, and travel booking creates a layered credit and behavioural profile. This proprietary dataset becomes more valuable as AI underwriting becomes more sophisticated. It is not licensed, sold, or shared — it is exclusively Kaspi's competitive weapon.
R&D posture: Kaspi classifies itself as a technology company operating under banking regulations, not a bank with a technology department. The distinction matters: engineering talent is treated as a core competitive resource, not a cost centre. The Hepsiburada acquisition was partly motivated by access to Turkish engineering talent to support platform expansion. Technology investment is growing faster than revenue — a deliberate choice to build out the international infrastructure ahead of the revenue it will eventually generate.
| Holder Category | Approx. Stake | Trend | Comment |
|---|---|---|---|
| Mikheil Lomtadze (Founder/CEO) | ~25-30% | ↑ Bought more Apr 2026 | High conviction; aligned with shareholders |
| Vyacheslav Kim (Co-Founder) | ~25-30% | Stable | Non-executive; long-term holder |
| Tencent | ~5% (new) | ↑ New position Apr 2026 | Strategic validation; long-term financial investment |
| Institutional Investors | ~35-40% | Mixed (91 added, 97 reduced Q last) | Neutral on net; uncertainty around Turkey thesis |
| Short Interest | 0.89% of float | Very low | Not heavily shorted; Culper attack did not create lasting short position |
Analyst consensus: Buy consensus with an average price target of $96–122 depending on source. Susquehanna recently raised its target to $95 (Neutral rating). The spread between high ($140) and low ($96) targets is wide, reflecting genuine uncertainty about the Turkish expansion trajectory rather than disagreement about the core Kazakhstan business.
Baring Fintech Venture Funds had been a major early investor and had been a source of share supply for several quarters. The April 2026 transaction in which Baring sold 6 million ADSs to Tencent, Lomtadze, senior management, and US endowments (Washington University, University of Wisconsin) effectively cleared this overhang. The buyers were long-term fundamental investors, not hedge funds. This is a structurally positive event for the shareholder register.
Bear case: $38–50 per ADS. This assumes Kazakhstan sanctions risk materializes (50% discount to current price); or Turkish expansion is abandoned after destroying $1.5B of capital AND Kazakhstan interest rates remain elevated indefinitely, compressing earnings to ~$1.3B annually on a ~9× distressed P/E = ~$62 per ADS. The extreme tail risk (secondary sanctions) has been modeled by short sellers at near-zero. We assign this a ~5% probability.
Key assumption: 15× forward P/E on $11+ EPS by 2028. Timeline: 24–36 months.
Probability assigned: ~15%. Most likely not zero earnings, but severe impairment.
Kazakhstan core business continues compounding at 15–20% revenue growth. Fintech margins recover partially as interest rates ease. Turkey reaches EBITDA breakeven in 2026 and begins modest profitability by 2027. No sanctions action. Multiple expands from 7× to 10–11× as uncertainty resolves. Expected annualized return: 20–30% over 24–36 months, inclusive of the ~8% dividend yield.
Approximate upside-to-downside ratio: 1.6:1 at current price. Adequate but not compelling — a better entry below $78 improves this to approximately 2.5:1, which crosses our threshold for high-conviction positioning.
Kaspi.kz is a genuinely exceptional business — a dominant, deeply-moated super app with 84% ROIC, 51% ROE, and a founder-CEO who is buying alongside you. The core Kazakhstan franchise is one of the most profitable financial platforms in the emerging world relative to its size, trading at 7× trailing earnings while growing revenues at 30%+. At the right price, this is a compelling investment. At today's price of $88.65, the risk/reward is not quite compelling enough to act decisively. The Turkish expansion introduces binary outcomes that are not yet resolved; geopolitical risk from Kazakhstan's Russian proximity remains a tail that, while unlikely, could be catastrophic; and the asymmetry at current prices (1.6:1) falls short of our 2:1 threshold. A pullback to $75–80 would price the Turkish risk more fully, improve dividend yield to 9–10%, and push the asymmetry to an attractive 2.5–3:1. The trigger: either wait for a better entry price, or wait for the Rabobank Turkey banking license approval (expected mid-2026) to de-risk the execution thesis before initiating a full position.