Abercrombie & Fitch

Abercrombie & Fitch is a specialty apparel retailer that designs its own clothes, has them made by contract factories in Asia, and sells them almost entirely through its own stores and its own websites. There is no wholesale to speak of. That matters: the company owns every touchpoint, which is why the rebrand could be executed at all. When you do not sell through department stores, you can turn the lights on, change the sizing and re-shoot the imagery in a single season and nobody else has to agree.

The group runs two brand families that now sit at roughly equal weight. Abercrombie brands — Abercrombie & Fitch and abercrombie kids — sell to a customer in her twenties at a mid-market price: a tailored trouser at $90, a jacket at $150. Hollister brands — Hollister, Gilly Hicks and the social-first label Social Tourist — sell to the teenager at a lower price point with a fictional Californian back-story, invented in 2000 and dated to 1922. The two brands are deliberately kept apart; a customer who ages out of Hollister is meant to arrive at Abercrombie as if it were a different company, and after the 2017 rebrand that is more or less true.

About 45% of sales come through digital, one of the highest shares in mall retail, and the store fleet has been shrunk and re-cut — the giant, dark, cologne-filled flagships of the Jeffries era closed between 2019 and 2021 and replaced by smaller, brighter boxes. The economics of the model are the economics of any specialty retailer: a gross margin in the low sixties, an expensive store base underneath it, and an operating margin that lives or dies on full-price sell-through. In FY2024 that margin reached 15%, an extraordinary figure for a mall apparel chain. In FY2025 it came back to 13.3% as tariffs and a flat Abercrombie brand took their share.

I want to be careful here, because the temptation with a turnaround this good is to call the moat wide. It is not. Abercrombie's advantage today is a brand that means something specific — well-cut, quietly expensive-looking basics for a woman in her twenties — and a distribution setup that lets the company control how that meaning is presented. That is a real asset. It is also an asset that On, Aritzia, Zara, Uniqlo and a hundred direct-to-consumer labels are chasing with the same customer in mind.

What the company does own, structurally, is the fully controlled channel. No wholesale means no discount partner to cheapen the brand, no department-store buyer choosing the assortment, and complete ownership of the customer data. Hollister adds a second, cheaper rung that catches the teenager early, and the two-brand ladder is genuinely hard for a single-brand competitor to copy.

The honest caveat is the history. This is a brand that was once the most desirable in the American mall and became, by 2016, the least-liked retailer in a national customer-satisfaction survey. It got there in under a decade. Brand equity in teen and young-adult apparel is a fast-moving thing, and Abercrombie has already demonstrated once how quickly it can be lost. A moat that eroded before can erode again, and the flat Abercrombie brand in FY2025 is the first data point I would watch.

The record divides cleanly at 2017. Under Mike Jeffries the company spent on scale and spectacle: enormous flagships in Fifth Avenue, Savile Row, Milan and Tokyo; new concepts that were launched and closed (Ruehl, the original Gilly Hicks stores); and a chief executive who was for years one of the highest-paid in American retail while returns fell. Shareholders revolted on pay in 2014, an activist pushed for his removal, and the board took the chairmanship from him before he left.

Under Fran Horowitz the allocation has been unglamorous and consistent. Close the flagships, shrink the boxes, remodel what is kept, spend on the digital stack, pay down the expensive debt, then buy back stock. There has been no acquisition, no new brand launched into a mall the company does not understand, and no dividend — a deliberate choice to return cash through repurchases while the share count falls by a tenth a year at current pace.

The question I would put to management is whether buying back $450M of stock at $150 a share is as good a use of capital as it was at $30. It is the same decision and a very different price. Everything else on the scorecard I find hard to fault.