Summary

  • Polestar’s H1 retail-sales metric counts end-customer handovers across channels and sale types; the company says it may or may not directly generate revenue.
  • The reported gross-margin improvement removes a prior US$724 million impairment from the comparison. The disclosed adjusted margin instead moved from 1.4% to −8.5%.

The seductive reading of Polestar’s first-half release is a recovery story. Gross margin moved from negative 49.4% in H1 2025 to negative 8.4% in H1 2026. Operating loss narrowed by 43%. Its retail network grew and retail sales reached a record 30,423 cars.

That reading puts four different clocks on one dashboard and calls the result momentum. Polestar’s own filings keep them apart. Retail sales are end-customer handovers, not a revenue-recognition schedule. Gross margin is a reported accounting ratio, not a clean current model-margin signal when the comparison year includes a large impairment. Operating cash and financing are a separate liquidity clock. And the company’s U.S. restructuring is an active exposure, not a completed cost or a forecast.

Start with the 134-car difference. H1 retail sales were 30,423, up 0.4% from 30,289. Revenue was US$1.360 billion, down 4.4% from US$1.423 billion. Dividing either revenue figure by retail handovers would manufacture a number that Polestar has not disclosed as price, revenue per car or margin per car.

The reason is explicit in the H1 results and more fully explained in the 2025 Form 20-F. Retail sales include new cars handed over through all sales channels and sale types, including internal, fleet, retail, rental and leaseholder channels, irrespective of market model. The release says the figures may or may not directly generate revenue. The annual report adds that a vehicle is included once invoiced and registered to an external or internal counterparty, irrespective of revenue recognition.

That is not an accounting footnote that can be ignored after the headline has been made. Vehicle arrangements may include transfer of the vehicle, connected services, roadside assistance and free service maintenance. Polestar says it recognises revenue when the customer obtains control of the goods or services. It also identifies residual-value guarantees in some financial-service-partner vehicle sales as variable consideration. A handover is therefore useful operating evidence: it speaks to customer-facing distribution and physical activity. It does not establish the amount, timing or gross basis of revenue that will appear in the period.

The composition of the retail measure reinforces the point. The H1 release identifies 1,384 external vehicles with repurchase obligations and 2,166 internal vehicles. These figures do not prove that the rest of the 30,423 creates matching revenue, nor that either subset should be valued at an assumed average selling price. They establish that the handover count spans arrangements that require a different accounting question.

The margin comparison has an equally important base problem. H1 2026 reported revenue of US$1.360 billion produced a US$115 million gross loss and a gross margin of −8.4%. H1 2025 reported a US$703 million gross loss and a gross margin of −49.4%. Read literally and in isolation, the improvement is 41.0 percentage points.

But Polestar discloses that the earlier cost of sales included US$724 million of net impairment expense. Its own reconciliation removes net impairment from both periods: H1 2026 adjusted gross loss was US$116 million, while H1 2025 adjusted gross profit was US$20 million. The corresponding adjusted margins were −8.5% and positive 1.4%.

This is not a licence to declare that the adjusted figure is the only real one. The reported impairment was real in the reported accounts. The right inference is narrower: the 41-point reported swing is not a same-for-same measure of current production economics. It cannot support a claim that vehicle margin recovered by 41 points. On the disclosed adjusted basis, the business moved in the opposite direction, from a small positive gross margin to a negative one.

Polestar provides several current-period pressures without quantifying each into one clean bridge. It cites pricing pressure, residual-value-guarantee costs mainly in the United States and related restructuring measures, lower carbon-credit sales, duties, higher battery raw-material costs and carline mix. Carbon-credit sales fell to US$57 million from US$90 million; some were booked in other operating income. The company also identifies a growing Polestar 4 contribution as a positive mix effect. The disclosure does not allocate the US$136 million adjusted-gross-profit change among those items. A precise attribution would be invented.

The U.S. position should stay in its own ledger. Polestar says it was not granted authorization under the current Connected Vehicle Rule to sell vehicles in the United States from model year 2027 onward. It expects to sell prior model-year inventory and then support existing customers, including service and warranty commitments. The company estimates that U.S. operations increased H1 operating loss by about US$211 million, compared with about US$110 million in H1 2025, and says further negative adjustments should be expected as the restructuring continues.

Those are management estimates and an uncertainty statement, not a total cost of exit or a projection of future revenue.

The cash clock is different again. Ending cash was about US$888 million after an H1 operating outflow of US$850 million, investing outflow of US$211 million and financing inflow of US$769 million. The financing line may improve liquidity; it does not convert operating loss into internally funded economics. Nor does it answer whether the next product launches, sales-point expansion and customer-support obligation will improve the adjusted gross margin.

The practical conclusion is deliberately unfashionable. Treat retail handovers as an operating distribution signal. Treat recognised revenue as an accounting outcome. Treat reported and impairment-adjusted gross margins as different views with different questions. Treat cash and financing as separate. A company can expand its sales footprint and hand over more cars while revenue falls, and it can report a dramatically better gross margin while the comparable adjusted margin worsens. The investor who keeps those surfaces separate has not missed the story; the separation is the story.

Sources: Polestar H1 2026 results; Polestar 2025 Form 20-F; BIS Connected Vehicle Rule portal; final rule.