ASOS’s interim release on 23 April 2026 was sold as evidence that the new commercial model is working. The mechanism is inventory release, not customer recovery. The brand relaunch that CEO José Antonio Ramos Calamonte is still preparing for later this year is being credited for results that arrived before any of it shipped.

Stock is where the cash came from. Inventory is down roughly 60% against the FY22 peak, with a further reduction in H1 FY26. That drawdown is working capital ASOS no longer has to fund. It is also the reason every other line on the income statement now looks better.

The balance-sheet repair followed mechanically. ASOS used the FY25 cash unlock to execute a convertible bond refinancing and to bank the Topshop joint venture proceeds from Heartland, which brought around £118m in net receipts. Both transactions were priced against the leaner balance sheet the inventory drawdown had produced. Neither was priced against a brand.

Gross margin expansion is downstream of the same exercise. The company posted its eighth consecutive quarter of margin improvement in H1 FY26, lifting 330 basis points to a 48.5% adjusted gross margin. Carrying less stock requires marking less down — the largest single lever in any apparel-margin story. Full-price mix rises mechanically when the markdown rail is starved, not because customers have suddenly chosen ASOS over Shein or Vinted. Profit per order rose 30% year-on-year against a smaller, cleaner book of options.

The Test & React model is where the inventory framing is most clearly mistranslated. ASOS markets it as customer-led: small first runs, replenish what sells, drop what doesn’t. The operating logic is open-to-buy discipline at smaller intervals, an inventory-velocity gain sold as a brand-relevance number. The same reframing applies to the Microsoft AI stylist partnership: a tool running since FY25 now presented as evidence of digital transformation.

The customer story remains soft. Revenue declined 14% in H1 FY26, extending a multi-year run of top-line contraction. New customer acquisition turned positive in March 2026, up 9% for the group, for the first time since September 2021, but a single month at the close of the period is not enough to reverse four and a half years of cohort attrition. The statutory pre-tax loss stayed in nine figures.

A turnaround that does not yet show up in revenue is a balance-sheet exercise.

The strongest counter-argument is that brand work and inventory work are inseparable here. If ASOS’s buyers are choosing tighter ranges, faster Test & React replenishment, and a higher proportion of full-price options because the brand has been refocused on younger UK women, margin and stock turn are evidence of a brand thesis taking hold rather than a cost-cutting trick. There is also meaningful new-to-ASOS acquisition through Topshop.com, where the vast majority of customers arrive with bigger baskets than the ASOS average — a demand-side variable the revenue line does not yet capture. The condition for the broader brand reading to hold is that cohort revenue eventually follows the new-customer pulse, that the relaunch converts into UK share against Shein and Vinted, and that gross margin survives a half where stock has to build back. None of those have happened. Working capital release is doing the work; the brand is still being built.

The risk in this design is shrinkage as a business model. ASOS has executed a disciplined balance-sheet operation, lifted margins consistently, and bought itself the credibility to refinance — but it has done so against a top line that keeps falling. If the relaunch lands, the company will have built a smaller, cleaner ASOS with room to grow. If it does not, the working capital has already been released, the stock has already been written down, and the next round will have to come from operating leverage on a base that may keep contracting.