Finvest
KSPI Fintech · Super app · Kazakhstan · E-commerce · Thesis updated July 17, 2026

Kaspi still compounds, but rules now bite

01 Running thesis

A great app meets tighter rules

Kaspi is still one of the cleaner super app stories in public markets. It has millions of consumers and merchants using the same app to pay, shop, borrow, save, and sell. That gives Kaspi many chances to earn small fees from the same customer without needing a classic card-fee model.

The bull case is that Kaspi keeps adding useful products to an already busy app. e-Grocery GMV rose 53% year over year. Kaspi AI helps merchants improve listings. Kaspi Alaqan, its pay by palm system, reached almost 0.5 million users in Almaty and about 10% of store transactions there in under 90 days. Management also showed discipline by ending the capital-heavy 1P part of e-Cars in Q4 2025.

Turkey is the big swing factor. Hepsiburada has moved fast enough that Turkey is now roughly 50% of Kaspi e-Commerce GMV. Management plans to run the Turkey business around EBITDA breakeven in 2026. The pending Rabobank purchase would give Kaspi a Turkish banking license, which is the missing piece for launching fintech products there.

The bear case is not about product demand alone. Kazakhstan has higher interest rates, higher reserve needs, a 25% bank corporate tax rate from 2026, and a new 2% capital buffer on most retail loans. Those rules tie up capital and pressure near-term net income. The stock can work if the app keeps compounding and Turkey opens a new profit pool, but the next year has real drag.

May 2026Q1 2026 added evidence that the Turkey push is moving fast. Management said Turkey is now roughly 50% of Kaspi e-Commerce GMV.
Mar 2026The 2025 20-F confirmed two key points. Kaspi exited the capital-heavy 1P part of e-Cars, while Kazakhstan added a new banking law and a 2% capital buffer on retail loans.
Mar 2026Q4 2025 showed both relief and pressure. Dividends resumed at 850 KZT per ADS and smartphones returned to growth in January, but higher taxes and Turkey breakeven plans limit near-term upside.
Nov 2025Q3 2025 showed strong product pace with Kaspi AI, pay by palm, and third-party advertising tools. Smartphone shortages were a clear drag, but merchant monetization kept improving.
Aug 2025Q2 2025 confirmed Marketplace revenue was growing faster than GMV because advertising and delivery were scaling. Fintech profit growth stayed held back by higher deposit costs.
May 2025Q1 2025 brought a guidance cut from smartphone import rules, higher reserve needs, and a new tax on government securities. The Rabobank agreement still improved the long-term Turkey option.
Mar 2025The 2024 20-F clarified that e-Grocery had moved to a first-party model, adding inventory and logistics risk. B2B Payments also showed major scale, with TPV up 81% in 2024.
Feb 2025Hepsiburada closed in January 2025 and was funded from operating cash flow. Rising rates changed the Fintech setup, making funding costs a bigger headwind for 2025.
02 Business model

Fees, spreads, and more merchant tools

Kaspi makes money from three platforms. Payments earns fees from merchant and consumer payments, plus some interest income from balances. Marketplace earns seller fees, delivery fees, advertising fees, and retail revenue from first-party areas like e-Grocery. Fintech earns interest and fees from BNPL, consumer loans, car finance, merchant finance, and deposits.

The core trick is cross-selling. A shopper who uses Kaspi QR can also buy goods, take BNPL, book travel, save money in a deposit, and use government services in the same app. A merchant can accept payments, advertise, deliver items, borrow, take business deposits, and manage taxes through Kaspi Pay.

This model can be very profitable when transactions grow faster than costs. The weak spot is that fintech uses a balance sheet. When rates rise, deposits cost more. When regulators raise reserves or capital buffers, Kaspi must hold more low-return assets or capital. First-party grocery also adds inventory and logistics risk that the classic third-party marketplace did not have.

Turkey adds both upside and mess. Kaspi controls Hepsiburada, but the bank deal is separate and still needs approval. Building a marketplace plus bank model in a new country could create a second growth engine, or it could dilute margins while management learns a new rulebook.

03 Product portfolio

What users actually do in the app

Cash cow

Kaspi Pay and QR payments

Payments are the daily habit layer. Kaspi QR, card payments, bills, P2P, and B2B payments keep consumers and merchants active.

Growth engine

Marketplace

Marketplace covers e-Commerce, in-store mobile commerce, Travel, e-Cars, and e-Grocery. Revenue is growing faster than GMV because advertising, delivery, and travel fees are rising.

Cash cow

Fintech

Fintech includes BNPL, general loans, car finance, merchant finance, and deposits. It is profitable, but higher deposit costs and new capital rules are pressuring margins.

Option

Kaspi Alaqan pay by palm

Alaqan lets shoppers pay with a palm scan. Fast early use in Almaty could make checkout easier and deepen merchant lock-in, but rollout cost and margin impact are still open questions.

Growth engine

Merchant services and advertising

Kaspi sells merchants delivery, classifieds, brand ads, and tools to advertise on platforms like Facebook and TikTok. Kaspi AI Assistant helps improve product listings.

Option

Hepsiburada and Turkey fintech

Hepsiburada gives Kaspi a large Turkish marketplace base. Rabobank would give it a banking license to test Kaspi-style fintech products in Turkey.

Steady

Deposits

Kaspi is growing customer deposits, including high-yield fixed-term products for consumers and business deposits for merchants. This supports lending, but it costs more when rates are high.

04 Business segments

Three engines, close in size

Payments34%modest
Marketplace34%growing fast
Fintech33%flat

The mix uses 2025 segment net income from Kaspi’s latest 20-F and the company view that Payments and Marketplace together are about 68% of the mix. Hepsiburada is included in Marketplace, so Turkey now matters more to the segment.

05 Risk factors

What could break the thesis

Kazakhstan capital rules squeeze returns

High impact · High odds

A new 2% sectoral countercyclical buffer applies from April 1, 2026 to retail loans, which make up 83% of Kaspi Bank’s risk-weighted assets. That means more capital must sit behind the loan book. It can slow growth or lower returns even if credit quality stays fine.

We watchKaspi Bank capital ratios, retail loan growth, and management comments on the 2% buffer.

Higher rates hurt fintech margins

High impact · High odds

Kaspi funds loans mainly with customer deposits. When market rates rise, it must pay more to keep deposits. The 2025 20-F says higher than normal rates directly reduced Fintech profitability.

We watchNational Bank of Kazakhstan base rate, deposit rates, Fintech yield, and interest expense growth.

Taxes and reserves keep net income muted

Medium impact · High odds

Kazakhstan added a 10% tax on revenue from government securities and higher reserve requirements. Banks also face a 25% corporate tax rate from 2026. These items can make profit grow slower than revenue.

We watchEffective tax rate, required reserves, and net income growth versus revenue growth.

Turkey integration gets messy

High impact · Medium odds

Kaspi is trying to manage Hepsiburada near EBITDA breakeven in 2026 while also buying a separate Turkish bank. A marketplace and a bank do not automatically fit together in a new country. Political consumer boycotts could also hurt Hepsiburada GMV.

We watchHepsiburada GMV, Turkey EBITDA, Rabobank approval, and any disclosed fintech launch dates.

Smartphone supply shocks return

Medium impact · Medium odds

New smartphone registration rules and iPhone shortages caused a sharp GMV hit in Q4 2025. Demand returned to growth in January 2026, but the event showed how one high-ticket category can move Marketplace results. A repeat would hurt GMV and merchant fees.

We watchSmartphone GMV growth, iPhone availability, and Kazakhstan import or registration rule changes.

First-party grocery lowers marketplace margins

Medium impact · Medium odds

Most of Kaspi’s marketplace is third-party, where merchants hold inventory. e-Grocery is first-party, so Kaspi handles more logistics and inventory risk. Fast growth is good, but the margin profile is less clean.

We watche-Grocery GMV growth, Marketplace margin, delivery costs, and inventory write-downs.
06 Quick answers

In one breath

What does Kaspi.kz actually do?

Kaspi runs a super app in Kazakhstan. Consumers use it to pay, shop, borrow, save, book travel, and use government services, while merchants use Kaspi Pay to accept payments, sell goods, advertise, deliver, and borrow.

Why is Turkey important for Kaspi?

Kaspi bought control of Hepsiburada, a Turkish e-commerce company. Turkey is now roughly 50% of Kaspi e-Commerce GMV, and the pending Rabobank deal could let Kaspi launch banking and fintech products there.

What is Kaspi Alaqan?

Kaspi Alaqan is a pay by palm system. In Almaty, it reached almost 0.5 million registered customers and about 10% of store transactions in under 90 days, but the cost and profit impact of a nationwide rollout are still unclear.

Why is the Finn performance score not higher?

Kaspi is still growing, but several outside pressures are hitting near-term profit. Higher rates, higher reserves, new taxes, and new capital buffers all make the business harder to run at its old level of profitability.