LULU — Q2 2026 Financial Report Analysis

A $134.5 million tariff refund carried the margin — underneath it, demand in the Americas got worse

lululemon's second quarter of fiscal 2026 (the 13 weeks ended August 2, 2026) shows revenue down 4% and gross margin up 200 basis points (a basis point is one-hundredth of a percentage point, so 200 bps = 2.0 points). Those two facts are close to unrelated. The margin improvement came almost entirely from a $134.5 million refund of import tariffs the US government had collected and a court subsequently struck down. Strip that one-off credit out and operating margin — the share of revenue left after all the costs of running the business, before interest and tax — was roughly 13.2%, against 20.7% a year ago, not the 18.8% the income statement reports.

The demand picture underneath is the more durable story: comparable sales fell 9% globally and 12% in the Americas, and management attributes that to fewer people visiting, a smaller share of visitors buying, and smaller purchases when they did.

Q2 2026 in figures

Prior-year store count is implied from the filing's disclosure of 16, 15 and 10 net new stores in the Americas, China Mainland and Rest of World since Q2 2025. The prior-year inventory figure is implied from management's statement that inventory "was $1.7 billion, a decrease of 1% from August 3, 2025"; the balance sheet itself compares to the February 1, 2026 fiscal year-end ($1,700.8 million).

What "comparable sales" is telling you that revenue isn't

Comparable sales counts revenue only from stores open at least 12 full months, plus all e-commerce. It deliberately excludes new stores, so it measures whether the existing business is selling more or less — not whether the company opened more doors. Reported revenue fell 4% while comparable sales fell 9%; the gap is the 41 net new stores opened over the past year filling in for sales the existing base lost.

"Constant currency" (the filing says "constant dollar") removes the effect of exchange-rate movements. A weaker US dollar this year makes foreign sales translate into more dollars, so several of lululemon's reported figures look better than the underlying trend. In the first half, currency added $69.6 million to revenue. China Mainland is the clearest case: reported revenue rose 3.6% to $407.1 million, but comparable sales fell 2% as reported and 8% in constant currency — the entire reported gain is currency plus $23.0 million from 15 net new stores and other channels. Traffic there was actually higher; conversion (the share of visitors who buy) was lower.

By segment, for the quarter:

The Americas' apparent margin improvement is the refund: the entire $134.5 million lands in Americas product costs, which fell 26.9% year over year. Americas product margin rose 600 bps, of which the filing says 830 bps came from the IEEPA refunds — meaning that excluding the refund, Americas product margin fell around 230 bps because of the replacement tariffs still being paid.

On the Americas decline itself, management is explicit about the mechanism: "reduced traffic and lower conversion rates, as well as a decrease in average order value." All three at once — fewer shoppers, a smaller share of them buying, and smaller baskets. That is a demand and brand problem, not a pricing or supply one.

Where the sales went, by product and by channel

Both channels lululemon controls directly shrank. The only one that grew is the one that includes outlets and wholesale — and the gross-margin bridge separately lists markdowns as a drag. Management says it has "a plan to increase the reliance of full price selling," which is an acknowledgment that the current mix is going the other way.

The tariff refund, in detail

During 2025 the US imposed higher tariffs under the International Emergency Economic Powers Act (IEEPA) and removed the "de minimis" exemption that had let low-value shipments enter duty-free. On February 20, 2026, the US Supreme Court invalidated the IEEPA tariffs. lululemon had paid $230 million under IEEPA and filed refund claims; in Q2 it received $134.5 million, booked as a reduction of cost of goods sold, plus $4.1 million of interest booked in other income.

Three things about that matter for reading the quarter:

1. It is not repeatable, and the rest is not banked. The filing states that further refund amounts "remain uncertain" and that as of August 2, 2026 the company "has not recognized an asset in relation to further IEEPA tariff refund claims." So the remaining ~$95 million of the $230 million paid is neither on the balance sheet nor assumed.
2. The tariff cost itself hasn't gone away. The administration has re-imposed tariffs "at different rates under alternative legislative powers," and the de minimis removal was explicitly unaffected by the ruling and upheld by the Court of International Trade on August 13, 2026. The gross-margin bridge confirms this: within the +430 bps of product-margin improvement, there is a −150 bps item "primarily from higher tariffs, as well as markdowns."
3. The headline and the driver point in opposite directions. Reported gross margin +200 bps = product margin +430 bps (IEEPA refunds +560, higher tariffs and markdowns −150, currency +20) less 230 bps of higher other cost of sales (occupancy and depreciation +180 bps from new and expanded stores; distribution centres and product departments +50 bps). Excluding the refund, gross margin was down roughly 360 bps to about 54.9%.

The filing also discloses that lululemon's US entity has been named in "purported consumer class actions relating to alleged tariff-related pricing actions and potential governmental tariff reimbursements" — i.e. claims that it raised prices for tariffs it is now getting back.

Costs: spending up on falling revenue

Selling, general and administrative expenses rose 5.7% to $1,006.3 million while revenue fell, taking SG&A from 37.7% to 41.7% of revenue — 400 bps of deleverage, and the single largest reason operating profit fell. Head office costs alone added $45.8 million:

• $14.6 million more in contractor, advisory and professional services, which includes $13.4 million of proxy-contest costs in the quarter ($24.8 million in the first half). This is a genuine one-off: it relates to the contest brought by founder Chip Wilson, settled by a cooperation agreement on May 26, 2026. At $13.4 million it is worth about 55 bps of operating margin.
• $14.1 million more in brand and community expenses.
• $7.6 million more in employee costs — but note the filing attributes this partly to "the reversal of stock-based compensation recorded in the second quarter of 2025," which made the prior-year base artificially low. That is a comparison distortion, not new spending. Bonus expense was actually lower "due to business performance."
• $6.9 million more in technology and $6.4 million more in depreciation.

Below the operating line

The effective tax rate (tax as a share of pre-tax profit) fell to 29.6% from 30.5%, cutting tax expense 15.1% and softening the net income decline. Other income rose $4.0 million — essentially the $4.1 million of tariff interest.

The gap between net income (−11.2%) and EPS (−5.8%) is buybacks: diluted share count fell 5.6%, to 112.9 million from 119.7 million. In the first half lululemon repurchased 4.9 million shares for $695.1 million, leaving $712.5 million of authorization. Cash stood at $1,389.7 million.

Operating cash flow for the first half was $589.3 million against $209.7 million a year ago — but that near-tripling is working capital and tax timing, not earnings. Net income over the same six months fell $161.2 million; the filing attributes the cash improvement to "an increase in cash flows from changes in operating assets and liabilities of $464.7 million, driven by changes in inventories and the timing of income tax payments."

On inventory, the discipline is real: $1,711.5 million, down 1% year over year against a 9% comparable-sales decline. A demand shortfall of this size usually shows up as a stock build that forces later markdowns; so far it hasn't.

First half, for context

Half-year revenue is flat only because currency added $69.6 million and new stores absorbed the comparable-sales decline. Operating profit fell nearly a quarter including the tariff refund.

Takeaway: Gross margin rose 200 basis points, but 560 of those points came from a one-time refund of tariffs a court invalidated — excluding it, gross margin fell roughly 360 basis points and operating margin was about 13.2% versus 20.7% a year ago. The quarter's real content is a 12% comparable-sales decline in the Americas caused by fewer visitors, a lower share of them buying, and smaller baskets, all at once. That is a demand problem, and a tariff refund does nothing to fix it.

What to watch

A 10-Q does not carry financial guidance, and this one does not. What the filing does give:

• The tariff position. $230 million paid under IEEPA, $134.5 million refunded, no asset recognized for the rest. Further refunds would be upside the company is not counting on; meanwhile replacement tariffs and the de minimis removal (upheld August 13, 2026) keep the underlying cost pressure in place. Management itself says the policy volatility "makes it more difficult to forecast costs... and provide reliable financial guidance."
• A new CEO. Heidi O'Neill took over on September 8, 2026, five weeks after quarter end, following the previous CEO's departure on January 31, 2026 and an interim co-CEO arrangement. The Americas turnaround plan is described only qualitatively — "product creation, product activation, and enterprise enablement," plus more full-price selling — with no quantified target in the filing.
• Q4 concentration. The filing notes roughly 37% of fiscal 2025 operating profit came in the fourth quarter. The holiday quarter, not Q3, decides the year.

My own read: the two things currently propping up the reported figures — buybacks shrinking the share count 5.6% and a weaker dollar adding $69.6 million of first-half revenue — are finite and do nothing for demand. More consequentially, opening 41 net new stores into a 9% comparable-sales decline is what produced the 180 bps of occupancy and depreciation deleverage in the gross-margin bridge: each new store adds fixed cost against a shrinking per-store sales base. Until the Americas comparable-sales number inflects, further store growth makes the margin arithmetic worse rather than better. In the Q3 report, the Americas comparable-sales figure and the full-price-versus-markdown mix are the numbers that matter; the headline revenue line will keep being flattered by stores and currency.