$0.86 of the $2.92 Was a Tariff Refund (3 September 2026): lululemon Cleared Its Own Guide, Missed on Revenue — and Cut the Year a Second Time

lululemon reported second-quarter fiscal 2026 results on 3 September, and the headline looks like a large beat: diluted earnings per share of $2.92 against a company guide of $1.76 to $1.81. Then read the footnote. $0.86 of that came from $134.5 million of refunded IEEPA tariffs booked into cost of goods sold and $4.1 million of interest, net of tax. Revenue actually missed the guide, comparable sales fell 9% with the Americas down 12%, and the full-year earnings guide was cut for the second consecutive quarter — from $10.95-$11.15 to $9.48-$9.73. The preview of this quarter argued that beating the quarter and holding the year were two different pieces of information. They were.

What actually happened

The quarter ended 2 August 2026. Every figure below is from the company's own release and quarterly filing.

Figures from the company's second-quarter fiscal 2026 earnings release and its quarterly report on Form 10-Q for the period ended 2 August 2026. The ex-refund lines are arithmetic on the company's disclosed $134.5m and $0.86 per share; they are not company-reported measures.

Note what the two ex-refund lines do and do not say. Underlying earnings of roughly $2.06 still cleared the $1.76-$1.81 guide, and that is real: selling, general and administrative expense grew just 5.7% this quarter against 12.4% in the first quarter, and the diluted share count fell 5.7% after the company repurchased 2.7 million shares for $330.0m. Cost control and buybacks did work. What they worked on was a revenue line that came in below plan and a comparable sales line that got materially worse.

The refund arrived, and it is smaller than the number everyone quoted

The preview of this quarter flagged $230m of IEEPA tariffs paid and $0 recognised on the balance sheet, and listed "any recognition of an IEEPA refund asset" as the thing to watch. The answer turned out to be more concrete than a recognised asset: the money actually moved.

lululemon disclosed that during the second quarter it received $134.5m of IEEPA tariff refunds, recognised in cost of goods sold, and $4.1m of associated interest, recognised in other income. Not accrued — received. Against $230m paid, that leaves about $95.5m outstanding, and the company states plainly that the ultimate additional amounts it may be refunded, if any, remain uncertain, and that as of 2 August 2026 it has not recognised an asset in relation to further IEEPA refund claims.

There is also a claim on the money that has grown since the preview. lululemon disclosed that its US operating entity is a defendant in two purported consumer class actions relating to alleged tariff-related pricing actions and potential governmental tariff reimbursements — Neuman v. Lululemon USA Inc. (No. 2:26-cv-11029, Eastern District of Michigan, filed 27 March 2026) and, newly, Alsaady v. Lululemon USA Inc. (No. 3:26-cv-05708, Western District of Washington, filed 30 June 2026), the latter asserting claims under Washington, Michigan and New York consumer protection statutes. Both seek damages or restitution for the alleged tariff-cost component of prices charged. The company states it intends to defend both. Nothing there is decided; it is simply why a refund flowing to a retailer is not an uncontested windfall. This site covered the general mechanics in how tariff refunds land in earnings and the largest example in Walmart's refund quarter.

The multiplier, again

The preview's central argument was about a ratio, not a direction: when this company revises revenue, earnings revise several times as hard, because the cost base is fixed. That ratio held.

A 5.9% revenue cut produced a 13.1% earnings cut — a multiplier of about 2.2 times, against roughly 3 times in June. Across both revisions, the revenue midpoint has come down 8.8% since March while the earnings midpoint has come down 21.3%.

Interim Co-CEO and chief financial officer Meghan Frank said in the release that the company was "taking a prudent approach with our revised full-year outlook", and that teams remain "focused on accelerating growth by strengthening our product offerings, increasing our marketing investments, and maintaining disciplined expense management." Those last two commitments point in opposite directions on the cost line, which is the tension the guide has to absorb.

Where the demand actually is

The segment table is where the quarter's real information sits, and it inverts the preview's framing of two businesses inside one income statement.

One quarter ago China Mainland grew 30% reported and 23% in constant dollars, and its $110.3m increase exceeded the entire company's $100.9m increase. This quarter it grew 4% reported and shrank 2% in constant dollars, with comparable sales down 8% on that basis — after opening 15 net new stores in the market since the year-ago quarter. Growth from new stores is masking a declining comp, and translation is masking the rest.

The composition of the Americas decline also changed, and this is the detail that matters most. In the first quarter the company attributed weaker comps to lower conversion and a lower average order value, partially offset by higher traffic. This quarter it cites lower conversion rates, reduced store traffic and a lower average order value, partly offset by higher e-commerce traffic. People arriving and buying less is a merchandising problem. Fewer people arriving is a brand-and-demand problem, and it is a harder one. The company describes an Americas action plan built on product creation, product activation and enterprise enablement, including greater reliance on full-price selling — which is a deliberate choice to trade volume for margin, and mechanically another reason the revenue guide came down. The same distinction between a headline comp and its composition ran through Home Depot's quarter, where ticket carried the comp and traffic fell.

The back half got lighter, and Q3 carries the reset

The preview's arithmetic showed the June guide asking the second half to deliver 68-69% of full-year earnings against roughly 57% in fiscal 2025. That burden has now been removed by cutting the year rather than met by improving the business.

Quarterly and half-year earnings per share do not sum exactly to the annual figure because share counts are weighted separately; treat these as close approximations. Two things stand out. The reported first half of $4.59 includes the $0.86 refund, so the underlying first half is nearer $3.73. And the implied second half of $4.89-$5.14 is 32% to 35% below fiscal 2025's back half — no longer a roughly flat-year assumption.

The third quarter is where that reset is explicit: revenue of $2.290bn to $2.320bn, a decline of 10% to 11% against $2,565.9m, and diluted earnings per share of $0.93 to $0.98 against $2.59 a year earlier — a fall of 62% to 64%. The revenue decline more than doubles from the second quarter's 4%, which tells you the company is not treating this quarter's comp as a trough.

A chief executive who inherits a reset guide

Heidi O'Neill becomes chief executive on 8 September 2026, five days after this print. Interim Co-CEO André Maestrini said in the release that the company looks forward to "welcoming our incoming CEO, Heidi O'Neill, next week as we begin an exciting new chapter."

The mechanical point is narrow and worth stating without drama: the interim team cut the year before handing it over. A new executive arrives with the second-half bar lowered from $7.45-$7.70 to $4.89-$5.14 and with the Americas action plan already named but not yet evidenced. Whether the next revision goes up or down is not something to guess at; what is knowable is that the guide she inherits embeds a 10-11% revenue decline in the current quarter, which is the shape of the year she will be measured against. Note too that head-office costs this quarter still carried $13.4m of expense associated with proxy-contest matters, inside a $45.8m net increase in head-office costs — governance friction showing up as an operating expense.

What it reads across to, honestly

lululemon is a mid-cap component of the US indices, not one of the mega-caps that move an S&P 500 or Nasdaq 100 session by itself. Treating one apparel result as a verdict on the consumer would be over-reading it. What the quarter does inform is narrower and more useful, in two directions.

First, on tariffs: this is a clean, documented case of how an invalidated tariff regime actually reaches an income statement — as cash received into cost of goods sold, in a partial amount, in one quarter, with the standing replacement duties and the de minimis change still in place underneath. Any importer reporting over the coming quarters runs the same plumbing, and reported margin will look better than underlying margin wherever it happens.

Second, on currency: the channel runs through translation, not footfall, and it is now working in the company's favour. Group revenue fell 4% reported against 5% in constant dollars, and China Mainland was +4% reported against −2% constant dollar — roughly six percentage points of that segment's reported change was the dollar rather than the business. A firmer dollar would strip that support out of the reported line without changing a single unit sold. That translation effect is downstream of the rate path, which is one of the five factors scored on the dollar's currency page, and more on how this site frames these channels is on the about page.

The single sentence to carry out of this quarter: the earnings beat was a refund of money already spent, the revenue miss and the 12% Americas comp were the business, and the guidance cut was the company telling you which of the two it believes.

Educational macro context only — not investment advice.