Torrid: The Q2 "Beat" Is a One-Time Tariff Refund; Proof Comes in the Back Half

Torrid's stock has more than doubled this year (up roughly 128% year to date) on the bet that its founder-led turnaround can salvage a plus-size apparel chain that has spent two years shrinking. The fiscal second quarter it reported Thursday evening looked like the bet paying off: net income swung to $5.2 million from $1.6 million last year, and adjusted EBITDA rose to $23.3 million. Spend two minutes on the press release, though, and the picture inverts. Almost the entire "beat" is a one-time tariff refund, not a business that got better.

The improvement is mostly a refund

The $11.1 million credit came from IEEPA tariff refunds, recorded as a reduction in cost of goods sold, plus $0.3 million in interest income. Management on the call urged investors to look through it, and the two views tell very different stories.

Reported gross margin rose to 38.7% from 35.6% a year ago. Strip out the refund and gross margin actually fell, to 33.9%. Adjusted EBITDA of $23.3 million looked like progress over $21.5 million last year; excluding the refund, it was $12.1 million — a roughly 44% drop, not the 8% gain the headline shows. The point of the "five on the quarter" narrative isn't that Torrid is suddenly profiting from demand; it's that a one-off accounting credit moved a small, still-weak number from red to black.

The top line is still falling

None of this is happening on the revenue side. Second-quarter net sales were $231.7 million, down 11.8% from $262.8 million a year ago, and comparable sales — the same-store measure retail investors should follow — fell 6.3%, about the same shortfall as the year-ago quarter's 6.9%. Total sales fell further than comps because the company is deliberately closing stores as part of its footprint-optimization plan: it ended the quarter at 457 locations versus 575 a year ago.

That explains the then-and-now of the profit story. Every extra point of margin and most of the earnings growth came from the refund and from stores removed from the cost base, not from shoppers buying more. The full-year sales guidance of $940 million to $960 million was left essentially unchanged, and the third-quarter guide of $230 million to $235 million is still down more than 12% from a year ago. On the revenue line, the "recovery" is not yet visible.

The bull case rests on a claim yet to be reported

What the market has really paid up for is an exit-rate story. Management said comparable sales turned positive in July and held into August after a weak start to the quarter — an "inflection point" in the founder's words. That is exactly the kind of forward proof that could justify the rally, and it may be real. But it has not yet shown up in a reported quarter, and the reported quarter contradicts it at the margin level: gross margin fell year over year once the refund is removed, which usually signals discounting to move inventory. Until July and August's positive comps appear as recorded revenue growth with gross margin that holds, the turnaround is a promised event, not a delivered one.

That distinction matters for price. On the company's raised full-year adjusted EBITDA guidance of $76 million to $86 million, the roughly $500 million enterprise value looks modest — about 6 times. But management was explicit that the raise exists solely to fold in the refund; the underlying profit outlook is unchanged. Excluding it, the multiple is closer to 7 times, put against a same-store sales number that is still negative. That is roughly where healthier specialty-apparel peers already trade — Torrid is not being handed a discount for its superior risk.

A thin balance sheet asks for proof

The last reason to wait rather than chase is that Torrid's balance sheet can't absorb a slip. Net debt sits near $278 million against a $222 million equity market value, shareholders' equity is negative, and the company ended the quarter with about $74 million in total liquidity. That is an operating runway measured in quarters, not years, for a retailer that still needs comps to merely reach zero. The tariff refund bought time, but a one-time credit can't fund a long slog.

The honest read on this quarter is "too early," not "buy the dip" and not "sell the rebound." The second quarter did not prove the turnaround — it booked a refund. The falsifiable test is three months away: when Torrid reports its fiscal third quarter in December, watch whether reported comparable sales turn positive for the first time and whether gross margin excluding any further refunds holds near or above the mid-30s. If that happens, the 128% run can keep paying for itself. If the July and August inflection fades into another negative quarter, the stock has rallied into a result the business hasn't caught up to. Neither case is decided tonight — and after a double in a year, paying full price for an unproven claim is the wrong side of the bet.