Euronext Amsterdam: ADYEN (ISIN NL0012969182). Run September 17, 2026. Built from shareholder letters H1 2018 to H1 2026, ESEF annual report data FY2020 to FY2025, earnings call, Investor Day and M&A transcripts through August 13, 2026, and cited Adyen press releases and news. Events swept through September 17, 2026; most recent event checked: CFO Ethan Tandowsky's departure on August 31, 2026 and interim CFO Hwa Tsao from September 1, 2026. Not a valuation and not a recommendation.
Adyen is a Dutch licensed bank that runs a single, in-house global payments platform for large merchants. It sells gateway, risk management and acquiring in one stack, online and in store, so an enterprise can accept payments in dozens of countries through one integration and one contract (H1 2026 Shareholder Letter). It holds a credit institution licence from De Nederlandsche Bank, a US federal branch licence and a UK branch licence, which lets it connect directly to card schemes and local rails rather than renting a bank (H1 2026 interim statements, general information note).
It makes money on each transaction. Merchants pay a fixed processing fee per transaction plus, where Adyen acquires, a settlement fee priced as interchange plus scheme costs plus Adyen's mark-up (FY2025 annual report, revenue accounting policy). Adyen manages the business on net revenue, which strips out interchange and scheme costs; net revenue was EUR 2,364.2 million in FY2025 on EUR 1,394.3 billion of processed volume, a take rate of about 17.0 bps (H2 2025 Shareholder Letter; take rate inferred from those two figures). In H1 2026 net revenue was EUR 1,302.9 million, up 19% (21% constant currency), on EUR 803.8 billion of volume at 16.2 bps (H1 2026 Shareholder Letter).
| Year | Net revenue (EUR m) | EBITDA margin | Volume (EUR bn) | Take rate (bps) | Net income (EUR m) | FTE (year end) |
|---|---|---|---|---|---|---|
| 2020 | 684.2 | 59% | 303.6 | 22.5 | 261.0 | n.d. |
| 2021 | 1,001.7 | 63% | 516.0 | 19.4 | 469.7 | 2,180 |
| 2022 | 1,330.2 | 55% | 767.5 | 17.3 | 564.1 | 3,332 |
| 2023 | 1,626.1 | 46% | 970.1 | 16.8 | 698.3 | 4,196 |
| 2024 | 1,996.1 | 50% | 1,285.9 | 15.5 | 925.2 | 4,345 |
| 2025 | 2,364.2 | 53% | 1,394.3 | 17.0 | 1,062.5 | 4,771 |
| H1 2026 | 1,302.9 | 49% | 803.8 | 16.2 | 544.1 | 5,020 |
Sources: H2 shareholder letters 2020 to 2025 and H1 2026 letter (net revenue, EBITDA, volume, FTE); FY ESEF xBRL facts (net income). 2021 to 2023 net revenue are H1 plus H2 letter figures. Take rates for 2022 to 2025 are inferred as net revenue divided by volume. 2025 take rate likely flattered by lower eBay volume (inferred: take rate rose while volume growth fell from 21% to 8%). FY2019 net income and EBITDA were later restated upward (net income EUR 204.0 million in the H2 2019 letter vs EUR 234.3 million in FY2020 ESEF comparatives), so pre-2020 figures are not used for trend math.
Share-of-wallet compounding inside an enterprise base that barely churns. Adyen lands a large merchant on one flow or one region, then takes more of its volume as the merchant adds channels and countries. Management discloses that typical share of wallet rises from under 20% in years 3 to 7 to over 40% after year 12, and that about two thirds of H1 2026 growth came from customers onboarded in 2024 or earlier (H1 2026 Shareholder Letter). Economically this means most growth needs no new customer acquisition cost: it is sold into an existing integration, which is why growth has held near 20% even as new-logo wins get harder to find at the very top of the market.
Operating leverage from a single platform. One codebase serves every region and channel, so the cost of serving the next euro of volume is small. Between FY2020 and FY2025 net revenue rose by EUR 1,680 million and EBITDA by EUR 843 million, an incremental EBITDA margin of about 50%; the same ratio holds for 2022 to 2025 (shareholder letters; inferred arithmetic). A market-average business does not turn half of each new revenue euro into EBITDA while also growing headcount by 2.3 times (2,180 FTE in 2021 to 5,020 in H1 2026).
A runway that is large relative to current share, and widening. Management sizes addressable commerce at EUR 26 trillion of a EUR 34 trillion market (Investor Day November 2025). FY2025 processed volume of EUR 1.39 trillion is about 5% of that (inferred). Unified Commerce (in-store volume up 28% in H1 2026) and Platforms (volume up 42%) push into segments where Adyen is under-represented, and the Talon.One and Orb deals open marketing and billing budgets that sit outside the payments line (H1 2026 letter; M&A call April 23, 2026).
Pricing power does not reach owners. Adyen's model deliberately passes scale back to merchants through volume tiering: take rate fell from 22.5 bps in 2020 to 16.2 bps in H1 2026 (H2 2020 and H1 2026 letters). Growth therefore depends on volume and wallet share outrunning price concessions; when management was asked in August 2026 whether competitive pricing had intensified it answered only that large deals carry sophisticated pricing (H1 2026 call). Growth has also decelerated in steps, from 46% in 2021 to about 20% now; H1 2023 growth slowed to 21% with EBITDA down 10%, and 2025 guidance was cut in August 2025 (H1 2023 and H1 2025 letters). Finally, about 19% of FY2025 pre-tax profit was finance income on cash and float (EUR 267.6 million of EUR 1,394.0 million, FY2025 ESEF), which moves with policy rates and already held back 2025 net income growth to 15% (H2 2025 letter).
The market can extrapolate the 2020 to 2021 growth phase and the new 'financial operating system' framing. Management's own guidance is only 'around 20%' a year for the next few years with 6 to 12 months of visibility (Investor Day 2025 transcript; H2 2025 letter), and part of 2026 and 2027 growth is now bought: about 1 point in 2026 and 1 to 2 points from Talon.One in 2027 (H1 2026 letter; M&A call). The embedded finance and agentic commerce lines are real options but are not yet material; management said it expects no material agentic impact for 8 to 12 months (Q1 2026 prepared remarks).
Strong
Yes. Roughly 20% organic growth converted at about 50% incremental margins, on single-digit share of a growing market, produces clearly more future owner cash flow than a market-average business that grows with nominal GDP. The grade stops at Strong, not Exceptional, because the growth is volume-led with falling unit price, has decelerated in steps with a guidance cut in 2025, and a fifth of pre-tax profit is rate-sensitive interest income.
Integration and operational switching costs for global enterprises. An enterprise that runs online, in-store and platform payments in many countries through one Adyen integration would have to re-build integrations, re-certify terminals, migrate tokens and re-negotiate local acquiring to leave. That cost is highest for Unified Commerce merchants (486 merchants processing at scale across channels, H1 2026 letter) because store hardware and reconciliation are tied to the platform. This lengthens the life of the relationship itself: longstanding customers such as Google, Microsoft, Spotify and Uber are still expanding after a decade (H1 2026 letter).
Mission-criticality with performance data the merchant cannot replicate. Authorization rates, fraud losses and payment costs are revenue lines for a merchant. Adyen Uplift raised conversion by an average 0.9 percentage points by end H1 2026 (H1 2026 letter). When a payments provider visibly adds revenue, the customer's cost of a switch includes the risk of losing that uplift, which favours renewal.
Licensed, direct infrastructure that is slow to copy. Banking licences in the EU, US and UK, direct scheme membership and new local licences (UAE Category II licence, direct access to France's STET clearing, H1 2026 letter) are years of regulatory work. They lengthen duration at the industry level by keeping the set of true global full-stack competitors small.
The same multi-provider set-up that lets Adyen win share of wallet lets merchants take it away. Adyen wrote in H1 2023 that North American enterprises prioritised cost over functionality and that 'online volumes are easiest to transition back and forth' (H1 2023 Shareholder Letter). In 2025 a single large customer (eBay) shifted volume, cutting total FY2025 volume growth from 21% to 8% (H2 2025 letter). The volume churn metric counts lost merchants, not lost wallet share inside retained merchants, so it overstates captivity. There are no long-term minimum volume contracts disclosed for the base as a whole, and technology shifts (account-to-account rails, agentic checkout, wallets and orchestration layers that route between acquirers) could lower the cost of switching further. Adyen's own responses (local rails, Adyen Agentic) are plausible but unproven over a full technology cycle, and the company's record is about 20 years with only one global recession as a scaled business.
'Single platform moat' and 'sticky enterprise relationships'. The mechanism is real for store and platform customers, but for Digital, still 55% of H1 2026 net revenue, much of the volume sits with merchants that deliberately run two or more acquirers and move volume on price. The stickiness that is disclosed is relationship stickiness (low merchant churn), not revenue-per-merchant stickiness. The new loyalty and billing products are designed to create harder lock-in, but they closed on July 1, 2026 and have no retention record yet (Adyen press release July 1, 2026).
Moderate
Partly. Relationships last longer than for a market-average supplier and the licensed global stack keeps competitor numbers low, which does lengthen the cash-flow stream. But the revenue attached to a relationship is contestable at the margin, as 2023 and 2025 showed, so the stream is durable in direction rather than locked in size. Moderate, graded down one step from where low churn alone would place it because the churn metric misses wallet-share loss.
Low physical capital per unit of growth. The platform runs on Adyen's own private data centres and software; capex has run at about 3% to 7% of net revenue, with a stated sustainable level of up to 5% (H2 2020, H2 2022, H2 2025 letters). A business growing 20% a year that needs only about 5 cents of capex per euro of net revenue keeps most of its growth funded by a small slice of cash flow.
Customers fund the working capital. Merchant funds pass through the balance sheet before payout: payables to merchants and financial institutions were EUR 8.1 billion at June 30, 2026 against EUR 0.6 billion of receivables from merchants (H1 2026 balance sheet). Growth in volume therefore releases cash rather than consuming it, and the float earns finance income. A market-average business must fund receivables and inventory as it grows.
Earnings are mostly cash, with low equity dilution. FY2025 FCF (EBITDA less capex and lease payments) of EUR 1,081.3 million was about 102% of net income of EUR 1,062.5 million (H2 2025 letter; FY2025 ESEF). Equity-settled share-based compensation was EUR 23.7 million in FY2025, about 1% of net revenue (H2 2025 letter), so the reported cash is not quietly paid for with dilution.
The real reinvestment is people, and it is expensed. Employee benefits were EUR 431.1 million in H1 2026, about 33% of net revenue (H1 2026 letter), and headcount more than doubled from 2021 to 2026. In 2022 to 2023 a hiring push ran ahead of revenue and cut the EBITDA margin from 63% (2021) to 46% (2023); H1 2023 EBITDA fell 10% year on year (H1 2023 letter). FCF already nets this cost, so it does not break the bucket, but it shows growth needs continued opex reinvestment that management can accelerate at will. Second, M&A has started: about EUR 1.0 billion for Talon.One (EUR 750 million) and Orb (USD 335 million; euro total approx., unverified), roughly one year of FCF, with the first deal described as a build-or-buy exception (M&A call April 23, 2026; Investing.com, 2026). Third, as a bank Adyen must hold regulatory capital; the EUR 4.9 billion of own cash is not being returned (no dividends or buybacks in the FY2025 or H1 2026 cash flow statements), and management ties capital allocation choices to DNB processes (M&A call). Idle, regulator-bound cash does not raise owner returns. SBC also rose to EUR 29.6 million in H1 2026 alone (H1 2026 cash flow statement).
'Asset-light, 50%-plus margins, so all earnings are free.' The capex and working-capital picture is exceptional, but a buyer of the earnings stream also pays for a growing engineering and sales organisation, rising data-centre capex in 2026, and now acquisitions. And retained cash sitting in a bank balance sheet is not the same as cash owners can take out.
Strong
Yes. Capex near 5% of revenue, merchant-funded working capital, FCF roughly equal to net income and small SBC give Adyen clearly lower reinvestment drag than a market-average business, evidenced across 2020 to 2025. It is Strong rather than Exceptional because the opex-heavy growth model has already produced one margin reset, M&A has entered the reinvestment mix, and the accumulated cash stays on the balance sheet.
Diversified, largely non-discretionary transaction mix across channels and regions. Adyen processes for digital services, retail, food and beverage, travel, mobility and platforms across four regions (H1 2026 letter). Diversification means a slump in one vertical is diluted by others, and structural share gains (e-commerce, unified commerce, platforms) cushion cyclical volume dips.
Fortress balance sheet with no financial leverage. No borrowings, A- S&P rating, EUR 4.9 billion of own cash at June 30, 2026 (H1 2026 letter). A downturn cannot force a refinancing or a dilutive raise, and the company can keep investing through it.
Growth above market as a shock absorber. Because Adyen grows several times faster than underlying market volume through share-of-wallet gains, a market volume slowdown lowers its growth rate rather than its revenue level. Management stated in August 2026 that it grows about three times the market (H1 2026 call).
The public record contains only one recession, and a short, e-commerce-favourable one. EBITDA has fallen: H1 2023 EBITDA dropped 10% year on year as hiring outran a growth slowdown tied to price competition in North America (H1 2023 letter). Growth is exposed to consumer spending and trade: the 2025 guidance cut came from US tariffs hitting APAC online retailers (H1 2025 letter). Single-customer volume shifts matter (eBay, 2025). About a fifth of pre-tax profit is finance income that falls with rates (FY2025 ESEF). As a bank and acquirer, Adyen carries merchant liability risk (for example refunds and chargebacks if a merchant fails before delivering goods, a travel-sector risk seen in 2020) and regulatory risk across DNB, US and UK supervisors. Operationally everything runs on one platform: a DDoS attack on April 21, 2025 caused transaction failures in Europe (Adyen incident update, April 2025). And governance is in transition: CFO Ethan Tandowsky left on August 31, 2026 with an interim successor and an external search under way, while co-CEO Ingo Uytdehaage is directing integration of the first two acquisitions (Adyen press releases May and July 2026).
'Net cash and a bank licence make it bulletproof.' Net cash protects the balance sheet, not the revenue base. Revenue is tied to volumes that respond to consumer spending, trade policy and a small number of very large merchants who can shift volume, and profit carries rate sensitivity. The low-fragility reading also leans on a 2020 test in which e-commerce was the beneficiary.
Moderate
Somewhat. Adyen is clearly less fragile than a market-average business in balance-sheet terms and has never shown a revenue decline, which is worth something. But its earnings have already fallen once, its growth rate has been hit twice by external shocks, it has single-customer and rate sensitivity, and it has not been tested by a long consumer recession. Moderate, graded down from Strong for thin downturn evidence.
Bucket 1 (more future owner cash flows) and Bucket 3 (lower reinvestment drag) justify a clear premium; together they describe a business that grows net revenue around 20% a year, turns about half of incremental revenue into EBITDA, and funds that growth with about 5% capex and customer float. Buckets 2 and 4 add a smaller premium: relationships are long and the balance sheet is unleveraged, but revenue per relationship and growth rates are exposed.
Bucket 1. The premium is load-bearing on sustained 20% organic growth at about 50% incremental margins. Bucket 3 amplifies it, because almost none of that growth has to be paid for in capital, but low drag on a slow-growing business would be worth far less.
Bucket 2. The 'single platform lock-in' narrative is stronger than the evidence. Merchant churn is below 1%, but enterprises multi-source acquiring, Adyen itself says online volume moves easily, and one customer's shift cut 2025 volume growth by more than half.
Because two of four buckets are only Moderate, which fails the Extreme rule that no bucket may be dominant while others are merely moderate. The underlying mechanism is contestability: large merchants can re-route online volume and extract tiered pricing, so Adyen's excellent cost position and product do not translate into locked-in volume or rising unit price. That caps both duration and resilience.
Observable evidence that the broader stack creates hard lock-in: disclosed net revenue retention by cohort that holds through a price-led competitive push; net revenue take rate stable while volume grows 20% or more; no further single-customer volume shocks for several years; Talon.One, Orb and financial products reaching a disclosed, material share of net revenue with high attach rates; and a consumer downturn in which net revenue and EBITDA still grow. A policy that returns surplus bank cash to owners would also strengthen Bucket 3.
Classification. Large premium. The grade profile is two Strong buckets (owner cash flows, reinvestment drag) and two Moderate buckets (duration, fragility), which maps directly to Large under the framework: Strong across two buckets with at least one real constraint. It is above Medium because the two Strong grades rest on multi-year primary evidence (2020 to H1 2026 shareholder letters and ESEF data), not on labels. It is below Extreme because Extreme condition 4 holds and condition 2 is avoided only with a caveat.
The blocker: Share of wallet is contestable. A large merchant can re-route online volume to a cheaper acquirer without re-platforming, and volume tiering hands scale gains back to customers (take rate 22.5 bps in 2020 to 16.2 bps in H1 2026), so neither duration nor pricing power is exceptional.
What would need to improve: Evidence that Adyen's broader stack (Talon.One loyalty, Orb billing, Intelligent Money Movement, issuing, capital) converts share of wallet into hard lock-in: disclosed revenue retention of top cohorts through a price-led competitive push, net revenue take rate stabilising while volume grows 20% or more, and a multi-year record of no large-merchant volume losses. Plus a disclosed downturn in consumer spending in which net revenue and EBITDA still grow.
Extreme tests: Case relies mainly on duration, brand or pricing power: Does not hold. The case rests on growth plus low capital intensity, not on duration or pricing. | Owner cash-flow growth is structurally capped: Does not hold, with a caveat. No hard cap, but take-rate tiering and market volume growth set the pace; management guides only 6 to 12 months out. | Reinvestment drag meaningfully above average: Does not hold. Capex about 5% of net revenue, FCF roughly equal to net income, negative working capital. | One bucket dominant while others only moderate or constrained: Holds. Duration and fragility grade Moderate. This alone rules out Extreme.
| Bucket | Key support | Key limit | Premium contribution | Final takeaway |
|---|---|---|---|---|
| More owner cash flows | About 20% organic growth, about 50% incremental EBITDA margin 2020 to 2025, roughly 5% share of addressable market | Volume-led with falling take rate; a 2025 guidance cut; about 19% of pre-tax profit is interest income | Strong | The core of the premium: a long, profitable growth runway. |
| Longer duration | Volume churn below 1% for years; wallet share rises past 40% after year 12; licensed global stack | Online volume moves easily between acquirers (H1 2023); eBay shift in 2025; churn metric ignores wallet loss | Moderate | Relationships last; the revenue attached to them is contestable. |
| Lower reinvestment drag | Capex about 5% of revenue; EUR 8.1 billion merchant float; FCF about equal to net income; SBC about 1% | Opex-heavy growth caused a 2023 margin reset; about EUR 1 billion of M&A (approx.); cash retained in a bank | Strong | Growth costs little capital, which multiplies the value of Bucket 1. |
| Lower fragility | No debt, A- rating; no revenue decline on record; 2020 net revenue up 28% | H1 2023 EBITDA down 10%; tariffs and single-customer shocks; rate sensitivity; CFO transition | Moderate | Balance sheet is safe; the revenue growth rate is not. |
| Overall | Fast, capital-light compounding on a large runway | Contestable share of wallet limits duration and resilience | Large premium | Deserves a Large premium, not Extreme. |
This is the premium the economics justify. What to do about the price the market asks is the reader's decision.