VEON's Financial Services Comes Together

VEON is an emerging-markets digital operator with telecom and digital businesses in Pakistan, Ukraine, Kazakhstan, Bangladesh and Uzbekistan. It serves 151.5m mobile customers and 227.7m digital users. Group revenue was US$1,271m in 2Q26, of which digital was US$342m, or 26.9%. Digital is reported in three verticals: Financial Services, Digital Life and Digital Enterprise. Management runs the first of the three through a dedicated unit, VEON Digital Financial Group, which spans the JazzCash wallet and Mobilink Microfinance Bank in Pakistan, Simply in Kazakhstan, Beepul in Uzbekistan and the Mukto Pay launch in Bangladesh.

Together they take the Pakistan financial services model into a second and a third market.

• On 15 September VEON Digital Financial Group and Beeline Kazakhstan signed a memorandum of cooperation with the National Bank of Kazakhstan. The release describes a five-year framework covering payment and money transfer processing, settlement between financial market participants and connection to the national payments infrastructure.
• Beeline intends to pursue first-category payment organization status, subject to regulatory approval. According to the Kazakh business daily Kursiv, that status would let Beeline open accounts, issue payment cards and connect to the interbank payments system.
• On 14 September, according to The Business Standard, Group CEO Kaan Terzioglu and Banglalink CEO Johan Buse wrote to the Bangladesh Bank governor to reaffirm their interest and engagement with regards to the digital-bank license application Banglalink first filed in 2023.
• The letter follows the payment service provider license Bangladesh Bank granted Banglalink on 12 August, under which the Mukto Pay wallet is targeted to launch commercially in September. Money transfers, merchant and e-commerce payments, bill payments and salary disbursements are the first use cases.

We read the sequence as the replication stage of a model VEON has spent ten years proving in one market.

• The JazzCash consumer app in Pakistan marked its tenth anniversary on 7 September. In Pakistan the wallet came first, lending followed once payments data existed, and a bank charter and then an insurer were added.
• The result is a financial services business that we expect to exceed US$500m of revenue in Pakistan alone this year.
• Bangladesh and Kazakhstan are now at the wallet-and-license stage (that Pakistan passed through years ago), with a group unit, a Mastercard technology partnership and a shared AI credit-scoring capability that did not exist the first time.

Financial services revenue was US$151m in 2Q26, up 48.5% year on year, and US$285m in the first half, up 45.0%.

• That made it about 44% of digital revenue and roughly 12% of group revenue.
• The line has compounded steadily: US$316m in FY24, US$425m in FY25, and a first half of 2026 that already equals two thirds of the FY25 total.
• Growth accelerated through the year, from 28.1% in 4Q25 to 41.0% in 1Q26 and 48.5% in 2Q26, driven by higher lending disbursements at JazzCash and broader merchant and SME services, according to the company.

Financial services appears to carry a lower EBITDA margin than telecom and needs far less capital.

• VEON does not disclose EBITDA for the financial services segment per se, but we note overall digital EBITDA margin was 36.1% in 2Q26 against 46.1% for telecom.
• Pakistan's digital margin rose 8.1pp year on year to 32.3% for 2Q26, which the release attributes to lower JazzCash cash-in commissions after stricter customer verification and to better bad-debt provisioning.
• Digital as a whole runs at about 7% capex to revenue against about 20% for telecom, so the financial services line converts a lower EBITDA margin into a comparable share of cash.

Pakistan is where the model has run longest, and its transaction volume now equals about a sixth of the economy.

• JazzCash had 27.7m active users in 2Q26 and 62m registered customers at its tenth anniversary.
• Trailing twelve-month gross transaction value grew 67.5% to PKR19.6tn, which management says is close to 16% of Pakistan's GDP.
• JazzCash issued more than 224,000 loans a day in the quarter, and the number of merchants transacting on the platform rose above 735,000 active merchants and 1.7 million total merchants. Most of the 1.7 million merchants are Raast QR users.
• Raast is the State Bank of Pakistan's instant payment system, and VEON's FY25 release said JazzCash processed over 80% of Raast payments.

Pakistan financial services revenue grew 53.3% year on year to PKR39.0bn in 2Q26.

• Pakistan digital revenue reached 38.7% of total Pakistan revenue in 2Q26.
• VEON does not disclose the cost of risk on the loan book, the loan losses charged against it each year, and we look for more detail at the Capital Markets Day.

Three businesses sit around the wallet: a bank, an insurer and an investment product.

• Mobilink Microfinance Bank holds the deposits and the lending book, launched Islamic banking in December 2025 and is acquiring Apna Microfinance Bank, subject to approval.
• The July acquisition of 76.33% of TPL Insurance for about US$16.4m adds underwriting to a base that already includes 17.9m FikrFree policyholders.
• In June JazzCash and Mobilink Bank opened Treasury Bill investing inside the app, developed with the State Bank of Pakistan and the Ministry of Finance.

Three things in the Pakistan model travel to other markets and one does not.

• The sequence of payments first and lending second travels, because it is how underwriting data is built.
• The regulatory pattern travels, because in each market the wallet license has preceded the bank license.
• The distribution edge travels, because Jazz, Banglalink and Beeline are all the largest or second-largest operators in their markets.
• What does not travel automatically is the ten years. Pakistan's numbers took a decade of iteration to reach, and we would not model Bangladesh or Kazakhstan on the Pakistan curve.

Organizational change came before the license moves, and it is the reason we treat the September announcements as a program rather than two local initiatives.

• In April VEON appointed Ahmet Kayhan as Chief Payments and Digital Assets Officer, from the payments groups DgPays and NEOPAY. The 15 September release is issued in the name of VEON Digital Financial Group, the unit he leads.
• In July the group signed a partnership with Mastercard covering AI-based credit scoring, embedded finance, digital wallets, merchant services, loyalty and remittances across Ukraine, Kazakhstan, Pakistan and Uzbekistan, with the option to extend to other markets.
• Terzioglu described the mechanism on the call in one sequence: payments create daily engagement, engagement creates data, data improves underwriting, and underwriting enables lending, insurance and wealth management.
• Pakistan's 224,000 loans a day show that sequence working at scale. The Mastercard agreement, the group's 77 live AI initiatives and the AI Foundry that JazzCash opened in Islamabad on 7 September are the tooling to run it again faster.

The distribution advantage is that the customer already exists.

• VEON's operators bill 151.5m mobile customers every month and hold the identity and spending data that a standalone financial services provider has to acquire through marketing and then verify at its own cost.
• Multiplay customers, those who use at least one digital service alongside voice and data, reached 45.3m in 2Q26 and generate 4.0x the ARPU of voice-only users.
• Terzioglu put the reach in one comparison on the call: across the five countries, 288m people are connected to the internet and 228m are already VEON digital customers.
• The financial services line therefore does not need to win a whole market to be large. It needs to convert its own base, and the test in each new market is what share of that base opens a wallet in the first year.

Bangladesh is the first full replication, and it starts from a stronger position than Pakistan did a decade ago in every respect except competition.

• Banglalink has 35.0m mobile customers and 6.7m multiplay customers.
• The payment service provider license was granted on 12 August, Mukto Pay is targeted to launch commercially in September, and, according to The Business Standard, interest and engagement with regards to the digital-bank application were reaffirmed with the central bank governor on 14 September.
• Behind the license moves sits the June framework under which VEON committed US$250m to connectivity, digital infrastructure, financial services and AI in Bangladesh, as the anchor for a US$1bn foreign direct investment target agreed with the government.

The market is large and already digital because mobile money reached Bangladesh a decade before VEON's wallet.

• According to The Daily Star, the country has more than 200m registered mobile financial services accounts across 14 licensed providers.
• Two providers dominate: bKash reports about 42m active users and Nagad about 28m. That is the difference from Pakistan, where JazzCash built the category.

Banglalink's entry case rests on its own base.

• Banglalink has 35m customers who already transact with it for airtime and data, 20.6m Toffee entertainment users after the World Cup, and a digital revenue line that grew 256.6% in local currency in 2Q26 to 13.9% of segment revenue.
• The wallet launches into that base rather than into an open market.
• On the July call management described a Bangladesh launch that begins with payments and then moves into a full financial ecosystem, with a digital-bank license as the second step.

Kazakhstan is the market where the model faces the strongest incumbent and where the 15 September step is therefore the most telling.

• Beeline Kazakhstan has 11.7m mobile customers and about 1m fixed customers, and its Simply wallet and virtual card had 5.3m active users in 2Q26.
• The 15 September memorandum with the National Bank of Kazakhstan sets a five-year framework for payment processing and settlement through the national payments infrastructure.
• Beeline intends to pursue first-category payment organization status, which Kursiv reports is the license class that permits account opening, card issuance and interbank connection.

The incumbent is Kaspi.kz, and Beeline is not going to displace it.

• Kaspi.kz is a Nasdaq-listed super-app that combines payments, marketplace and financial services and processed US$26bn of payments volume in 2Q26 alone.
• Kaspi also shows that a payments-led financial services business in one of VEON's own markets can reach a market value of about US$19bn.

Beeline's case is narrower: payments and cards sold into a base whose acquisition cost the telecom relationship has already paid.

• Beeline brings 11.7m telecom customers, a wallet already at 5.3m users, and a regulatory path to run payments and cards on the national payments infrastructure.
• Kazakhstan is also the segment where telecom margin compressed in 2Q26 because of a VAT increase and equipment-heavy family bundles, so the pace of financial services revenue there is the more important variable for the country's profit line over the next two years.

The remaining two markets are at opposite ends of the regulatory queue.

• In Uzbekistan the Beepul wallet had 1.0m active users in 2Q26 on a revised counting method, and the 21.0% rise in Uzbek digital revenue was driven by mobile financial services and peer-to-peer payments, according to the release.
• Terzioglu named Uzbekistan alongside Bangladesh as a market where the Pakistan success is repeatable, and the Mastercard partnership covers it.
• In Ukraine, Kyivstar has the largest digital ecosystem in the group outside Pakistan, with Helsi in healthcare, Uklon in mobility and Tabletki in pharmacy, but no financial services line yet.
• Management said it is waiting for the regulatory environment to allow it to proceed, and Ukraine's digital revenue growth of 83.0% in 2Q26 came without a financial services contribution.
• If a license route opens in Ukraine, it would add a fifth market with 21.8m mobile customers already in place, although the economics would be shared with Kyivstar's minority holders, since VEON owns 83.6% of the separately listed company.

VEON trades at 4.3x forward EV/EBITDA and 1.7x forward EV/Sales on consensus, against medians of 4.9x and 1.9x for emerging-market telecom operators and 8.8x and 1.9x for listed digital financial services companies.

• The table sets the two peer groups side by side because VEON's financial services line is now large enough, at about 12% of group revenue with revenue growing near 50%, that the telecom group alone no longer describes the business.
• Our target price rises to US$98 from US$82 on a sum of the parts that values the Kyivstar stake at market and at our own target, the digital financial services line at its peer median and the other four operators at the emerging-market telecom median. The forecasts are unchanged from our 2Q26 results note.

The digital financial services peers are not one group, and the nearest comparable carries no premium.

• Kaspi.kz, the Kazakhstan incumbent and the closest comparable for a profitable emerging-market payments ecosystem, trades at 4.8x forward EV/EBITDA, close to the telecom median, so payments volume on its own earns no premium in this group.
• The premium sits with the lending-led digital banks, Nubank and SoFi. The Brazilian acquirers PagSeguro and StoneCo, at 6.6x and 3.7x, earn a thin fee on each transaction and are priced for it.
• The 2Q26 release attributes the line's growth to higher lending disbursements at JazzCash, which places it nearer the digital banks than the acquirers.
• The line is valued today inside a telecom multiple. We hope to see more detail at the Capital Markets Day.

Valued part by part, VEON is worth US$88 per ADS with Kyivstar at its market price and US$108 with Kyivstar at our US$20.30 target, and our US$98 target is the midpoint.

• The parts are our unchanged FY26E group EBITDA of US$2,235m allocated in proportion to reported first-half segment EBITDA: US$759m for Kyivstar, US$220m for the digital financial services line and US$1,256m for Pakistan telecom, Kazakhstan, Bangladesh and Uzbekistan after HQ costs.
• VEON's 83.6% of Kyivstar is worth US$3.9bn at our US$20.30 target and US$2.5bn at the market price of US$13.02, which is 6.3x and 4.1x Kyivstar's FY26E EBITDA with its leases counted as debt. Kyivstar is listed, so a reader can take either anchor, and Exhibit 10 carries both.
• The digital financial services line is worth US$1.9bn at 8.8x, the digital financial services peer median, on our estimate of US$220m of FY26E EBITDA. VEON does not disclose financial services EBITDA, so the 35% margin on about US$620m of revenue is our assumption, between the 36.1% group digital margin and Pakistan digital's 32.3%. Each margin point moves the value by about US$0.7 per ADS and each turn of multiple by about US$2.8.
• We use the middle of the peer group because the line sits between its two ends. The payments-led names, Kaspi.kz at 4.8x and the acquirers at 3.7-6.6x, price businesses whose revenue is transaction fees; the lending-led digital banks at 13-18x price loan books with published cost of risk. The median itself is where Remitly and dLocal sit, cross-border payments companies with growth but no loan book. VEON does not disclose the split between lending and payments revenue; the 2Q26 release attributes the line's growth to higher lending disbursements at JazzCash, and revenue grows 48.5%, but its cost of risk is not disclosed and its revenue is earned in frontier currencies, so the middle of the group is a fair place for it.
• EV/EBITDA is a blunt measure for a lending business because funding cost sits below the line; we use it because VEON reports the segment on that basis and the loan book is funded by Mobilink Microfinance Bank's deposits rather than by group debt. At Kaspi's 4.8x the line would be worth US$1.1bn and the target about US$87.
• Pakistan telecom, Kazakhstan, Bangladesh and Uzbekistan are worth US$6.2bn at 4.9x, the emerging-market telecom median in Exhibit 9, on EBITDA that already carries the group's HQ costs. The four are a mixed set rather than a uniformly weak one: telecom EBITDA in the first half grew 22% in Pakistan and 7% in Uzbekistan and fell 7% in Kazakhstan and 37% in Bangladesh, so the median fits them better than a discount to it. Three of the seven forward multiples behind that median, for MTN Group, Millicom (Tigo) and Airtel Africa, sit above their trailing figures, and without them the median would be 4.0x. We hold 4.9x because after the 10% holding-company discount, the four operators are carried at an effective 4.4x, close to the 4.1x at which Kyivstar, in a war, trades, and still below the median.
• We apply a 10% holding-company discount to the digital financial services line and the four operators, the parts without a market price, for the friction of moving cash out of frontier markets to a holding company that carries the group debt. The rate sits at the bottom of the 10-25% range usual for a listed holding company because the bridge already counts every lease liability as debt and the operators' EBITDA already carries the group's HQ costs at the peer multiple. The Kyivstar stake carries no discount because it is listed. The bridge deducts US$3.6bn of net debt including leases outside Kyivstar, the group's US$3,675m less Kyivstar's US$123m, because the peer multiples in Exhibit 9 are struck on enterprise values that include lease liabilities and the parts must be measured the same way. Our 2Q26 note struck the US$82 target at 3.4x on net debt excluding leases; US$98 is 3.9x on that older basis.
• We set the target at US$98, the midpoint of the two anchors, because we hold a target on Kyivstar and the market does not, and the target should not rest wholly on either view. At US$98 VEON would trade at 14.8x our FY26E EPS against the 15.9x telecom median, and at 4.8x FY26E EV/EBITDA with leases counted as debt against 3.9x at the current price on the same basis and the 4.9x telecom median. Exhibit 9 shows VEON at 4.3x because the vendor uses consensus EBITDA and its own share count; our figures use our forecasts and the 70.92m ADS outstanding after buybacks. Published sell-side targets run from US$70 to US$108 with an average of US$86, so ours sits above the average and below the highest. At the current price, once the Kyivstar stake is taken out at market, the market pays about 4.1x EBITDA for the digital financial services line and the four operators combined, below the telecom median and with no premium for the digital financial services line.

Three of the items above are applications or intentions rather than approvals.

• The Kazakhstan first-category status is something Beeline intends to pursue under the memorandum.
• The Bangladesh digital-bank license application, first filed in 2023, continues to be under review.
• Ukraine has no license route yet, and Mukto Pay's September launch is targeted and subject to remaining regulatory and technical requirements.
• Each of those is a date to check at the 3Q26 results on 6 November and at the Capital Markets Day on 16 November.

The other two risks are familiar to VEON investors: currency and the Bangladesh incumbents.

• Financial services revenue is earned in rupees, taka, tenge and som and reported in dollars, and the group's revenue line already absorbs several points of currency depreciation a year.
• In Bangladesh, bKash and Nagad have the acceptance networks and the habit, so Mukto Pay's early metrics could look small against them for several quarters even if the conversion of Banglalink's own base is on plan.
• We would judge the Bangladesh launch on wallet activation inside the Banglalink base, and the Kazakhstan step on whether the license application is filed, rather than on market share against the incumbents in the first year.

The tripwire for this thesis is the Capital Markets Day.

• The 16 November event is the first occasion on which management can disclose Mukto Pay's launch metrics, the Kazakhstan license timetable and the financial services revenue and profit split that management said in May it would provide within six to nine months.

The following tables are extracted from our Zero One Financial Model. Forecasts are unchanged from our 2Q26 results note.