Klarna Group plc KLAR

Key data

The business

Klarna is a payments network becoming a bank. A shopper at Apple, Uber or one of 1.2 million merchants picks Klarna at checkout and splits the purchase, usually four interest-free installments, sometimes a longer Fair Financing plan. Klarna collects a fee from the merchant, about 2.8% of what is bought, plus interest on the financing products. Merchants pay because Klarna lifts conversion and basket size. The moat is a two-sided network: merchants pull shoppers, 120 million shoppers pull merchants, and underwriting tiny short-duration balances is the skill rivals cannot cheaply copy. The newer engine is everyday money: the Klarna Card, $11.7 billion of consumer deposits, and paid Memberships, now 2 million subscribers with revenue up more than 600% year over year. That is the shift from a checkout button to a banking relationship, and the July 2026 US industrial-bank charter application is the tell for where management is steering.

The numbers

Read the quarterly line as a turn, not a level. Diluted EPS crossed zero in two quarters.

Q2 revenue grew 27%, transaction margin dollars grew 42% to $446 million, and operating profit swung to positive $27 million from a $46 million loss a year earlier. They can now be separated, and the fear of hollow, take-rate-driven growth broke the right way: strip out an accounting presentation change and the comparable take rate rose to 2.84-2.85% from 2.80% guided, so Klarna earns more per dollar shopped, not less.

Revenue compounded 24% a year from 2023 to 2025, and the 2026 guide of $4.08-4.16 billion is only about 17% growth, a real deceleration: GMV growth halved from 33% in Q1 to 18% in Q2. But the unit economics moved the other way. Transaction margin dollars are guided to $1.62-1.65 billion, 1.09% of GMV against 1.04% before. That is the normal shape of a lender scaling into profit, growth slowing as margin widens. What this memo believes that the tape does not: the turn to operating profit is real and roughly half-priced against a broken-BNPL narrative, and the print that settles it is next year's operating margin holding as growth cools.

The lender's line is credit, and it looks benign.

Management

Sebastian Siemiatkowski, co-founder, still runs it, and the record is founder-aggressive: a US charter push, a EUR 900 million forward-flow deal that sells receivables off the balance sheet for 24 months, and Memberships built from scratch. No insider bought or sold in the open market over the past year, unsurprising for a company one year public and still inside lockups, so there is no signal to read there. Directors and officers took $21.4 million in Fixed Equity in 2025, sizeable against a $273 million net loss. The honest mark against management is the August guidance cut, from above $4.34 billion to $4.08-4.16 billion, the first real forecast miss since listing, even as the company beat its own loss estimates every quarter. They spend where they see engagement, which is either discipline or a founder chasing the next network.

How it fails or surprises you

US charter and 30 million users (right tail). Klarna already touches 30 million Americans a year but monetizes them lightly. If the industrial-bank charter clears and even a slice deepen toward the Swedish cohort's 33 purchases a year, ARPAC re-rates hard. The market pays nothing for it today because the application is pending. First tell: US ARPAC and charter approval.

The credit cycle no one prints in advance. Provisions at 0.52% of GMV sit near a cycle low, and $124 short-duration balances mean losses surface within a quarter or two, not years, if US employment cracks. The forward-flow sale moves risk off the book but does not remove it from the model. First tell: provision above 0.65% of GMV and Stage 3 rising.

The growth the read explains least. GMV growth halved to 18% while merchants grew 54%. If volume, not just optics, is decelerating, the "scaling into profit" story becomes "maturing into a low-growth lender at 5x book." First tell: GMV growth and reported take rate next quarter.

Closing thoughts

This is one clean uncertainty, the US charter, sitting on top of one exposure no print resolves in advance, the consumer credit cycle. The charter is a dated call option the market is not paying for. The credit book is where the left tail lives, and short-duration lending means you find out fast, which cuts both ways: a downturn bites within two quarters, but so does the recovery. The fatter tail is up, because the operating turn already happened while the stock fell 70%, and the thing that would break it, a US credit shock, would show in the provision line before it shows in the equity. What is at risk if credit rolls over is real, a lender at 5x book with thinning margin. What the charter is worth if it lands is a different company.

The bet is still that Klarna keeps adding merchants and consumers and gets each consumer to spend more, and the consumers keep paying it back on $124 balances. It breaks if the provision rate climbs through 0.65% of GMV while GMV growth keeps halving. Watch those two numbers together next quarter, because either the margin turn is durable or it was borrowed from a benign credit window that is closing.