Abercrombie & Fitch earned $4.17 per diluted share in the quarter ended 1 August. Its own guidance had been $1.80 to $2.00. The stock closed up 35.7%.
About $1.75 of that $4.17 came from roughly $100 million in refunded tariffs under the International Emergency Economic Powers Act. Not from selling clothes.
Two different quarters in one release
| Line | Figure |
|---|---|
| Reported EPS | $4.17 |
| Estimated tariff refund contribution | ~$1.75 |
| EPS excluding the refund | ~$2.42 |
| Prior guidance | $1.80-$2.00 |
| Net sales | $1.27bn, +5% |
Strip the refund and Abercrombie still beat its own guidance by something like 20-35%, on 5% sales growth and a fifteenth straight quarter of expansion. That is a good quarter for an apparel retailer in 2026, and it deserved a good reaction.
It did not, by itself, deserve 35.7%.
Why the distinction is worth the trouble
A tariff refund is cash. It clears, it funds buybacks, it is real money. What it is not is repeatable.
The refunds flow from the legal unwinding of IEEPA-based tariffs, which means the same line item shows up once, at whatever scale a given importer's back-payments happened to be. Next year there is no second refund — there is only whatever the tariff regime settles into.
So the market has to answer two separate questions and this print only answers one of them. Question one: is Abercrombie executing? Yes, clearly, and the raised full-year outlook of $13.10-$13.60 with a 14.5%-15.0% operating margin says management thinks so too. Question two: what is the run-rate margin once the windfall is out? That one the release does not resolve.
The other half of the retail week
Three sessions earlier Dick's Sporting Goods fell 30.7% on a guidance cut. Same consumer, same calendar, opposite outcome.
That spread is the actual state of US retail right now. There is no aggregate consumer story to buy or sell — there is a widening gap between operators who hold price and operators who discount to move inventory. Abercrombie raised guidance. Dick's cut it by more than $2 a share against consensus.
How I read it
I treat this as a good business inside an inflated headline, which is a more comfortable position than the reverse.
The underlying quarter was strong enough that the buyback — at least $500 million — is funded from operations rather than from the windfall alone. That is the detail I would check first in the 10-Q.
What I would watch: the next quarter, with no refund line. Consensus will anchor on a raised full-year number that contains a one-off, and the comparison in twelve months will be brutal for anyone who modelled $4.17 as a base. The gap between reported and underlying earnings is where retail investors most often overpay, and this is a textbook instance of it.