lululemon

IThe business

lululemon makes and sells athletic clothing for people who pay a premium for it. It opened its first store in Vancouver in 1998 and now sells leggings, jackets, shorts, tops, bags and some shoes under one brand. It designs the products and has other companies make them. It sells them mostly in its own stores and on its own website.² In fiscal 2025, the year to 1 February 2026, women’s clothing was 63% of revenue, men’s 24%, and accessories and other 13%.²

The design is a vertical one. lululemon owns the brand, the store, the website and the customer relationship, and it rents the factories. The 10-K says the company works with about 51 vendors, five of which made 47% of its products, and it names Vietnam (40%), Cambodia (18%), Sri Lanka (11%), Indonesia (11%) and Bangladesh (7%) as the main places.² It owns no factory and has no long-term contract with most of its suppliers.² That keeps capital light. It also means a tariff on those countries lands on lululemon and not on a supplier.

Where it sells

Company-operated stores brought in $5.05B, e-commerce $4.92B, and other channels, which cover wholesale to studios and universities, outlets, pop-ups, resale and licensed stores, $1.13B C (calculated).² So the website is almost as large as the whole store base, and it grew 8% in fiscal 2025 against 1% for stores.² lululemon ended the year with 811 stores, 44 more than a year before, plus 45 locations run by partners in the Middle East and elsewhere under licence.² Its own stores added 11% to square footage, and sales per square foot fell to $1,426 from $1,574, down 9% C (calculated).² New space grew faster than the sales it was meant to carry. That is a fact to keep in mind later, when I look at what the company spends on stores.

Three regions make up the company. The Americas, which is the United States, Canada and Mexico, brought in $7.85B, or 71% of revenue. China Mainland brought in $1.75B, 16%, and Rest of World, which is Asia-Pacific plus Europe and the Middle East, $1.50B, 13%.² The United States alone was $6.33B, 57% of the company, down 2% on the year.² That is the number to look at first, because the United States is where the damage is.

The Americas, where the trouble is

The Americas segment earned $2.56B of operating income in fiscal 2025 before corporate costs, a 32.6% margin, down from 38.0% the year before, and that is 116% of the company’s total operating income of $2.21B, because corporate costs of about $1.4B came off the top.² In other words, the Americas pay for the head office and for everything else, and when they shrink, everything else shows it.

The company is plain about the cause in the 10-K: lower conversion, less store traffic and a lower average order value, with some core products selling less.² In the second quarter of fiscal 2026 it said it again, and added that the brand itself was part of it: shifting consumer demand and brand sentiment.⁴ On the call, the interim chief executive and finance chief said negative commentary in the media and on social channels had hurt traffic in the quarter, and that some new launches got a softer response than planned.⁵ Leggings, the product the company is known for, fell about 20% in the quarter, and accessories fell 13%.⁵

The arithmetic matters, so here it is. Americas revenue was $3.24B in the first half of fiscal 2026, down 6% R (reported).⁴ The company now expects North America down by low double digits for the full year.⁵ If I take the Americas down 11% on fiscal 2025’s $7.85B, the year comes to $6.98B, which leaves $3.75B for the second half against $4.41B a year earlier, a fall of about 15% C (calculated).⁴ Take 10% and the second half is down 13%; take 12% and it is down 17% C (calculated). The screen note that started this page called the Americas decline “about 8%”. That was the second quarter. The guidance needs the decline to roughly double.

China Mainland and the rest

The other growth story has also cooled. For fiscal 2025 China Mainland grew 29% and Rest of World 16%.² In the first half of fiscal 2026 China Mainland revenue rose 16%, but 6 points of that was currency, so 10% at constant exchange rates; in the second quarter alone, reported growth was 4% and constant-dollar growth was a fall of 2%, with comparable sales down 8% at constant rates R (reported).⁴ The screen’s note that lululemon’s international business was being ignored leaned on the six-month figure. The quarter is what matters now. Interim co-CEO André Maestrini said the main issue was brand sentiment and the job was to restore consideration of the brand, and Q2 revenue missed the company’s own plan mainly because of China.⁵ Management still expects China Mainland up by high single digits and Rest of World by mid single digits for the full year.⁵

Tariffs, and a refund that flatters the year

Two forces cost lululemon gross margin in the last 18 months. The first is tariffs. In fiscal 2025, higher US tariffs and the end of the duty-free allowance for small parcels, known as de minimis, cut gross profit by about $275M before any offset.² The second matters here because the majority of lululemon’s US online orders are filled from distribution centres in Canada, and those used to qualify for the exemption.² The Supreme Court struck down the emergency-powers tariffs on 20 February 2026, but not the end of de minimis, and the administration started new tariffs under other laws.²⁴

In the second quarter, lululemon received a $134.5M refund of those tariffs plus $4.1M of interest, which lifted gross margin by 5.6 points and EPS by $0.86.³ It says it paid $230M in all and has filed claims for the rest, but books nothing until it gets cash.⁴ Two consumer class actions now claim that US customers paid a tariff-related price and that any refund belongs to them.⁴ So the refund is not a number to count on, and I take it out of earnings everywhere on this page, as the company’s own guidance does when it states it separately. Without it the second quarter’s gross margin was 54.9%, 3.6 points below a year earlier C (calculated).³

The things people forget

• Seasonality. About 37% of full-year operating profit came in the fourth quarter of fiscal 2025.² The third quarter is weak, and the company guides to an operating margin of about 6.5% for it against 17% a year earlier.⁵ That makes the holiday quarter carry the year.
• No debt. There is a $600M revolving facility with only $6.3M of letters of credit drawn on it, and $1.39B of cash at 2 August 2026.⁴ The lease liabilities of $2.14B are rent and sit inside operating profit already.⁴
• Real estate. The company owns its Ohio distribution centre and rents the rest. Minimum lease payments are $2.1B over the years ahead.²

I see a business with excellent unit economics, a brand under stress in its biggest market, and a cost base that was built for a growth rate it is no longer getting. The next section asks whether the brand can recover.

IIThe moat

A lasting advantage is something a rival with money and talent cannot copy in a reasonable time. lululemon has one source of it that matters, which is that people want the product and will pay for it. Everything else on the list below depends on that single fact.

The loop that made the company work goes like this. A product that people love sells at full price. Full-price sales produce a high gross margin. The margin pays for marketing, stores and design. The new product is better, and more people love it. The loop turns in both directions. When a product disappoints, sales at full price fall, the company marks the stock down, the margin shrinks, and there is less to spend on the next launch. The Americas are in the second half of that loop now.

Evidence the advantage is intact

• The Americas segment still earned about a 28.5% operating margin in the second quarter, even with revenue down 8%, once the refund is removed: 36.8% reported less the 8.3 points the refund added C (calculated).⁴ That is segment margin before corporate costs, but it is high for a business in decline.
• Inventory is clean. At 2 August 2026, dollar inventory was down 1% on a year earlier and unit inventory was down 7%, while the company chases 20% more volume in styles that sell.³⁵ A brand in trouble usually carries too much stock. This one does not.
• Gross margin without the refund fell 3.6 points in the second quarter. Tariffs explain 1.6 points of that, markdowns 0.7, and fixed-cost deleverage 2.3, and the company offset 1.0 point with savings C (calculated).⁵ Tariffs and deleverage are not signs the product is rejected.
• China Mainland and Rest of World are still profitable. China earned a 40.0% segment margin in fiscal 2025 and 37.2% in the second quarter of 2026.²⁴

Evidence it is narrowing

• Americas comparable sales have fallen for five quarters in a row: down 4%, 5%, 1%, 5% and then 12%.^252612133 Revenue fell 8% in the second quarter against a 1% rise a year earlier.⁴
• Leggings, the core category, were down about 20% in the quarter. The company says it is not yet able to fully offset that with new styles.⁵
• The Americas lost conversion, traffic and average order value in the same period. That is three signals pointing the same way, and the fall in average order value suggests customers are not paying more for the new items.²⁴
• Sales per square foot fell 9% in fiscal 2025.²
• The company is discounting more. Markdowns were up 70 basis points in Q2 and are guided up 60 in Q3 and 40 for the year, which it describes as end-of-season clearance and not promotions.⁵

On balance I think the advantage is narrower than a year ago, and possibly much narrower, in North America. Outside it, the evidence is mixed. I cannot see from a filing whether a damaged brand heals, and I say more about that in Section VI under what I cannot judge.

IIIThe competitive advantage

The honest comparison is with the athletic-wear brands that sell to the same customers: Nike, which is 4.2 times the size, Deckers, which owns HOKA and UGG and is half the size, and Under Armour, which shows what a brand in long decline looks like. Fiscal years differ. Nike’s ended 31 May 2026, Deckers’ and Under Armour’s 31 March 2026, and lululemon’s 1 February 2026.

The table says three things. First, lululemon’s gross margin is 13.7 points above Nike’s and 1.1 points below Deckers, which is the best measure that the product still commands a premium price C (calculated).²⁷⁸ Second, Deckers grew 9.8% with a higher operating margin, so the premium end of the market is not shrinking; the problem is lululemon’s, not the category’s C (calculated).⁸ Third, the market is not paying a premium for either: Deckers trades at 11.7 times its latest earnings, lululemon at 12.3 times my no-refund fiscal 2026 figure, and Nike at 18.4 times C (calculated).²⁸⁸⁷³

Under Armour is the cautionary case. Its revenue fell 3.8% and it made an operating loss of $163M C (calculated).²⁷ Its gross margin is 45.5%, 11 points below lululemon’s.²⁷ A brand can fall a long way from where lululemon is. The question is whether the Americas move toward that path or away from it, and no filing tells me that yet.

Where lululemon wins

• Margin per sale. With 56.6% gross margin, even a 20% fall in one product category leaves a profitable business.²
• Men’s and the new categories. Men’s revenue was down only 1% in the second quarter against 4% for women’s, and the 10-K lists golf, tennis and running among the activities it markets to.⁵²
• Outside North America, where the brand is younger. Rest of World grew 16% in fiscal 2025 and China Mainland 29%.²

Where it loses

• North America today. The Americas lost conversion, traffic and average order value, and the chief financial officer said that traffic is where the pressure shows.⁵
• Price discipline. Markdowns are up, and the 10-K lists new credit card affiliate programs, along with tariffs and markdowns, as a reason for lower product margin in fiscal 2025.²
• Fixed costs. Store space rose 11% in fiscal 2025 while sales per square foot fell 9%.²

IVManagement

The company has a new chief executive, a new head of product, a board enlarged to 13 directors and a founder with 8.7% of the stock. Almost every senior name changed in twelve months, so the useful questions are who holds the pen now, how they are paid, and what the people who ran it before did with the cash.

I count five senior departures or role changes in about ten months, including the chief executive and the president of the largest region. McDonald’s tenure had a long good stretch. Revenue rose 3.4 times and earnings per share 3.7 times under him C (calculated).²⁰² The board then chose to part with him by mutual agreement in December 2025 after a bad year, and ran a search of about four months.³¹²⁹ O’Neill ran Nike’s consumer, product and brand teams, which is the job lululemon says it needs done. Her filed biography shows no earlier chief executive role, and I treat that as an unknown and not a flaw.²⁹

Pay, and what it pays them to do

McDonald’s reported pay for fiscal 2025 was $24.0M. That is $1.4M of salary, $7.8M of stock, $13.8M of options and a $1.0M bonus.¹⁰ The options number includes $8.6M for the value of a change in his terms: when he left, his options kept vesting for 48 months.¹⁰ He also gets a $3.05M lump sum and $2.1M of severance paid over 18 months.¹⁰ He left after a year when earnings per share fell 9% and the 10-K’s return chart shows the stock down 58%.² Those payments are in his contract, so I do not read them as a scandal. I read them as the price of a board that had not planned a handover.

The incentive design is what I would follow. The annual bonus for fiscal 2025 paid 35.7% of target, based on what the company calls “modified operating income” of $2,498.0M, which excludes tariff costs, the removal of the small-parcel exemption, executive transition costs and proxy contest costs, and adds back savings from vendor negotiations. The proxy says that on reported operating income, which was $2,210.6M, the bonus would have paid nothing.¹⁰ The performance stock for 2023 to 2025 paid 134.4% of target using the modified figure, against 72.5% on the reported one.¹⁰ The gap is $287M of operating income the committee chose not to count. It is a defensible judgement, since tariffs hit everyone, but it moved real money to the same people whose strategy the shareholders were questioning in the same proxy.

The new cycle looks different, and harder. For fiscal 2026 to 2028, performance stock compares operating income with a baseline of $2,210.6M, the fiscal 2025 reported figure, pays between 0% and 250%, and is capped at 100% of target if the stock’s total return is negative over the period.¹⁰ My no-refund estimate of fiscal 2026 operating income is about $1.4B E (estimate). In my base case it does not get back above $2.21B until fiscal 2035 E (estimate). So these awards are likely to pay little or nothing. What is left to motivate is options and time-vested stock.

O’Neill’s package is exactly that. She gets a $1.4M salary and a bonus target of 200% of it, with a cap of 200% of target for fiscal 2026. The annual equity is worth about $10M, 60% in performance stock and 40% in options. On top she gets a one-time grant of $2.8M in restricted stock and $4.2M in options, vesting half in each of the next two years, and a $2.0M cash retention bonus.²⁹ At target that is up to about $23M in a full first year C (calculated). The grants of 9 September 2026 were 28,079 stock units and options on 105,295 and 40,223 shares, all with a strike of $99.72.⁹ The stock closed at $107.24 on 30 October, so those options are $7.52 in the money, and they pay for a rise from roughly the price of the day she started C (calculated).⁹¹ That aligns her with a recovery from here. It does not charge her for the fall that came before she arrived, which is fair. It also means she has put none of her own money in: the Form 4 shows grants and no purchase.⁹

Their own money

Sixteen directors and executive officers together owned 254,943 shares, including options that can be exercised soon, on 30 April 2026. That is less than 0.3% of the company, and about $27M at $107.24 C (calculated).¹⁰¹ Marti Morfitt held 84,632, worth $9.1M. The largest holder is Chip Wilson at 8.7%, with 9.90M shares, then BlackRock with 8.0% (from a January 2024 filing), Vanguard with 7.0% and a trading firm, G1 Execution Services, with 5.7%.¹⁰ Wilson is the only large holder with a personal stake in the brand, and the cooperation agreement of 26 May 2026 gives him three new directors, a vote for declassifying the board, and a standstill that runs until about the 2028 nomination deadline.³²³⁰

Three insiders bought stock on the market in 2026, which is the best sign I can find that someone inside thinks it is cheap. Director Chip Bergh bought 6,090 shares at $164.20 on 20 March and 4,275 at $117.05 on 15 June. Maestrini bought 3,275 at $151.02 on 1 April.³³ In all that is about $2.0M C (calculated). Each purchase is under water at $107.24, by 35%, 8% and 29% C (calculated).³³¹ I count the purchases and note that so far they have been poor timing. Sales since June 2025 were by McDonald in June 2025 (about 27,000 shares near $236), Burgoyne in December (about 13,500 near $204), Frank in December (2,658 at $211) and Neuburger (615 at $178 and 622 at $161).³³

What they did with the money, over a full cycle

I take fiscal 2016 to 2025 as the cycle. It covers a pandemic boom, a sharp inflation, a tariff shock and now a brand slump. Over those ten years lululemon generated $11.6B of operating cash. It spent $4.1B on stores, distribution centres and technology, and $5.6B on buying back its own stock C (calculated).¹⁷² It paid no dividend, borrowed nothing, and made one acquisition, MIRROR, for $452.6M of cash net of cash acquired.³⁴ Cash rose from $0.50B to $1.81B.²¹²

Two things stand out. The first is the buyback. Management spent most of its cash on its own shares, and it did so at the wrong time. The firm bought 20.1M shares for $5.33B from fiscal 2021 to the first half of fiscal 2026, an average of $265 a share. At $107.24 those shares are worth $2.16B C (calculated).^19141835241 That is a loss of about $3.2B, or $29 per share outstanding, from buying a growth stock at growth multiples that did not last C (calculated). The buyback in fiscal 2024 alone cost $1.64B at an average of $321. By contrast, the buyback in the first half of fiscal 2026 cost $0.70B at $142, and the second-quarter purchases averaged $120.⁵ Price discipline is better now. The program is expected to continue at the 2025 level.⁵

The second is MIRROR. In 2020 lululemon paid $452.6M for a connected home gym.³⁴ In fiscal 2022 it wrote down $407.9M of goodwill and other assets and set aside $62.9M against hardware inventory, and a further $23.7M in 2023.¹⁴¹⁸ The write-down was close to the whole price. A company that makes one acquisition in ten years and loses most of it is not an acquirer I would trust with a larger one. The board will have more cash than it needs. What it does with it is the capital-allocation risk on this page.

The return on that capital has been high. On my measure, operating income after tax over equity less cash, it was 48% in fiscal 2016 and 49% in 2025 C (calculated).¹⁷² Operating leases are not in the base, since rent is already deducted in profit, so the figure is high by design. The more useful fact is that the business needed almost no outside capital to grow from $2B to $11B of revenue.

VFinancials

The accounts are simple. There is one operating company, no debt, one class of tradable stock plus a block of Canadian exchangeable shares that convert one for one, and the main judgement in the numbers is the provision for inventory, $87.1M at the end of fiscal 2025 on $1.7B of stock.² The largest recent distortions are tariffs, a 53rd week in fiscal 2024, and the refund in fiscal 2026.

Margins

The gross margin was 48.4% in fiscal 2015 and 56.6% in 2025. It peaked at 59.2% in fiscal 2024 C (calculated).¹⁷² The operating margin went from 17.9% to 19.9%, after reaching 23.7%. SG&A rose from 30.5% of revenue to 36.6%. In other words, the business ran at a much higher gross margin and spent most of the gain on stores, people and marketing, so the operating margin rose by only about two points in ten years C (calculated).¹⁷² That is the point to take from the chart. The operating margin of 22% to 24% in fiscal 2023 and 2024 was a peak, not a base. My base case does not assume it returns.

The fiscal 2025 gross margin fell 2.6 points. The company puts 2.3 points on product margin, mainly higher tariffs, more markdowns and new credit card programs, partly offset by higher prices and lower product costs, and 0.3 points on occupancy and depreciation.² SG&A rose 8.1% against revenue up 4.9%. Marketing, store labour hours and technology spending all grew.² The guidance for fiscal 2026 has the gross margin down about 80 basis points including 130 basis points of refund, so it is about 210 points lower without it, and SG&A deleverage of about 450 points, for an operating margin about 530 points lower, 14.6%, of which 13.3% comes without the refund.⁵ I derive the 13.3% from the company’s numbers, and it is an estimate E (estimate).

This chart tells me more than any ratio. The company gave a revenue and earnings outlook for the year on 17 March. It cut it on 4 June. It cut it again on 3 September, and the second cut was the deepest. After the March outlook the stock rose 4%, from $159.27 to $165.39. After the June cut it fell 9%, from $124.92 to $114.23, and after the September cut it fell 17%, from $121.77 to $100.61 C (calculated).¹ Guidance is a forecast and not a fact. A series of misses by a management that is also the incumbent team is a fact about how well it can see its own business. I put weight on that in the odds below.

Earnings quality

Three items stop the reported numbers from being a clean guide. First, the refund: $134.5M of cost of goods sold and $4.1M of interest in the second quarter, worth $0.86 a share. Second, stock pay: it was $62M in fiscal 2025 after a $26.3M reversal of performance-stock expense in the second quarter of 2025, because the company judged that targets would be missed.² In the first half of fiscal 2026 it was $50.3M against $21.2M, so the run rate is about $100M a year C (calculated).⁴ I treat it as a real cost and it is already inside operating income. Third, a 53rd week in fiscal 2024 worth $163.2M of revenue, which flatters fiscal 2024 and makes fiscal 2025’s 5% growth look slower than the 7% underneath.²

Interest and other income was $28M in fiscal 2025, down from $70M, as cash fell and rates eased.² It is small. The effective tax rate was 29.5% and the guidance uses about 30%.²³ I use 30%.

From profit to cash

Operating cash flow was $1.60B in fiscal 2025, down $670M, mostly from the timing of foreign tax installments and inventory and receivable swings.² In the trailing twelve months to 2 August 2026 it was $1.98B, because first-half 2026 operating cash flow was $589M against $210M a year before, helped by a much smaller inventory build ($14M against $238M) C (calculated).⁴ Capital spending over the same period was $0.63B and stock pay $0.09B, so cash flow after stock pay is $1.26B, and $1.12B without the pre-tax refund C (calculated). That is the figure my screen used as its base.

The trailing figure looks good, and it is stale on two counts. The working-capital swing that helped the first half is one-off timing; prepaid and receivable income taxes grew to $480M from $352M at the year-end.⁴ And it describes twelve months that include four stronger quarters than the next four will be, since the company guides to a fiscal third quarter operating margin of 6.5%. I do not use it.

What an owner could take out

Here is the number I care most about. If lululemon opened no new store and spent only what keeping the present business as it is would cost, how much cash could it hand to owners in fiscal 2026? I start from guided earnings without the refund, add back depreciation, and deduct a range for the spending it would take to stand still. Stock pay is already a cost inside the earnings. lululemon does not say how its capital spending splits between upkeep and growth, so I try 0.95 to 1.25 times depreciation, with 1.10 as the middle.

The answer is about $0.93B, or $8.29 a share E (estimate). That is 6% below the guided net income, so for lululemon accounting profit is a fair guide to owner cash this year. It is 41% below reported fiscal 2025 net income of $1.58B C (calculated).² I cannot pin it down more tightly, because the company does not split its capital spending. Reasonable splits give $0.84B to $1.01B E (estimate). Either way, an owner at $107.24 gets about 7.8% of the price back from today’s depressed business, against 4.94% on a ten-year Treasury.³⁶ That is a good yield if fiscal 2026 is the trough. It is a poor one if the margin keeps falling.

The balance sheet

At 2 August 2026 lululemon held $1.39B of cash, no borrowings, and $593.7M unused under its revolver.⁴ Inventory was $1.71B, flat on the year in dollars and down 7% in units.³ Accounts payable were $347M and unredeemed gift cards $277M.⁴ Lease liabilities of $2.14B are rent, which is already deducted in operating income.⁴ In my valuation I count only the cash, $1.39B or $12.44 a share, and nothing for the $480M of tax prepayments, which could come back.⁴ I do not count the second-half cash flow of fiscal 2026 either, which is conservative by about $4 a share E (estimate).

VIValuation

Start with what would go wrong

Before any numbers, I ask what would most likely make this a bad investment from here, and work backward from that. For lululemon the business may be fine and the mistake is in what I assume about the brand. Here are the three ways it goes badly wrong, in the order I think most likely.

1. The Americas keep shrinking because the brand has lost its pull, not because of a cycle. Americas revenue is 71% of the company and earned a 32.6% segment margin.² If the Americas fall another 15% over three years, as in my bear case, and the margin troughs near 10%, the company earns about $1.02B of operating income in fiscal 2028, under half of fiscal 2025's $2.21B E (estimate). Under Armour shows that a brand can slide for a decade: its revenue fell 3.8% last year and it lost $163M at the operating line C (calculated).²⁷ I cannot rule that path out from a filing.
2. The margin stays near 13%. Tariffs at about 20% from October, the end of the small-parcel exemption, markdowns, and a store base that grew 11% while sales per square foot fell 9% all point to a margin well under the old 20%.⁵² If 13% is the new normal and revenue is flat, the price is about fair, as I show below C (calculated).
3. A new team spends the cash badly. The company has $1.4B of cash, a record of buying back stock at an average of $265 and of losing $0.4B on one acquisition, and a new chief executive who will want to make a mark.³⁴¹⁴⁵ The danger is a large deal or a store push at the bottom. It is a risk to value that I cannot size.

I put no odds on the first and third. I can show what the second does to value, and I do below.

What the price implies

On a trailing basis this looks like one of the cheapest large consumer brands I have written up, at 8.1 times fiscal 2025 earnings. That number is the trap. Earnings are falling by a third, so the right multiple is the forward one, 12.3 times, and the next year may be lower still. Deckers trades at 11.7 times and is growing 10% a year C (calculated).⁸²⁸ So the market does not pay a premium for lululemon’s margin today, and it is paying about what it pays for a business that is growing, which means it is already pricing a decline.

Over ten years the price rose 1.87 times, from $57.25 in October 2016. Earnings per share rose 7.0 times to fiscal 2025 and 4.6 times on my no-refund fiscal 2026 figure C (calculated).¹²³ On this measure the market paid between 24 and 104 times earnings from 2016 to 2024 and pays 12 now. The price has followed the multiple down more than it has followed profit.

Where the screen went wrong

This page began with a screen of mid-cap US companies that looked cheap. Its note on lululemon called it the weakest candidate, because the Americas were falling about 8%, but said the bear case sat above the price. That second claim is the one I tested. The screen’s method was a ten-year cash flow model on trailing owner cash flow of $1.128B, with a 9% discount rate and 3% growth after year ten. I reproduced its numbers to the dollar: $187 for 3% growth, $116 for a fall of 4% a year, $250 for growth of 7%. The table changes one thing at a time.

The base year does the work. Moving from $1.128B of trailing cash to $0.85B of guided fiscal 2026 cash cuts the screen’s bear value from $116 to $91, a number below the price, even at the screen’s own 9%. That is the finding. The bear case does not sit above the price. At 9% the screen’s own base falls from $187 to $145. My 10% moves it to $126 on constant growth. My explicit cases are different from a constant-growth line because they let margins and revenue recover, so they give higher values for the base ($148) and bull ($224) and a bear of $91 that matches the forward screen case.

So where the screen said “the price assumes cash falls about 5% a year for a decade”, the honest statement on a forward base is that it assumes cash stays level, roughly: -1.4% a year at 9% and +0.6% a year at 10% C (calculated). That is a much smaller gap. The shift from the screen’s trailing base to a forward one removes most of the margin of safety it found. What is left is real, since a flat business at today’s depressed margin is worth about the price and my base recovers, but it is a debate and not a gift.

Reverse DCF: what the price needs

With the margin stuck at today’s 13.3%, the price needs revenue to grow about 0.2% a year, which is flat, for ten years C (calculated). If the margin recovers to 15%, the price is consistent with revenue falling 1.6% a year; at 17%, falling 3.6% C (calculated). The other way to read it: on my base revenue path, which grows 3% a year, the price is right if the margin settles at about 12% and not the 16.5% of my base C (calculated). So the price already assumes the business never earns more than about 12% on sales, against 19.9% last year and 23.7% the year before. Put simply, the price is a bet that the margin does not come back.

There are two further ways to read the same price. At my base cash flows, the price equals a discount rate of 12.9%, which is 8.0 points above the Treasury. On my bear cash flows the price returns 8.7% a year C (calculated). I find the second number the more telling one, since it means that even if I am wrong in the direction I fear, the loss against a 10% requirement is small, about 15% of the price in value terms C (calculated).

The discount rate

I discount at 10%. The capital asset pricing model with the 4.94% ten-year Treasury and Damodaran’s 4.20% implied equity risk premium gives 9.5% on lululemon’s beta against the market over 119 months, 8.8% over 60 and 7.5% over 36 C (calculated).³⁶³⁸³⁹ I take the highest and add half a point, because the cash flows depend on a turnaround that nobody can be sure of, and a beta measured on past prices does not capture that. I use 10% where I used 8.5% for Costco and 9% for Visa.

The base is above the price in every row, but the base is my recovery case. A different table shows what matters more. Here is the base value if the final margin is lower than I assume, with revenue unchanged.

Three cases

My base case is a slow repair. Americas revenue falls 5% in fiscal 2027, because the second half of fiscal 2026 is still weak and the first half of 2027 faces tough comparisons, then is flat in fiscal 2028 and grows 2% to 3% a year after that E (estimate). It is back to its fiscal 2024 peak in about fiscal 2034. Outside the Americas, growth is 4% in the first year, 7% in the next two as China and the new stores recover, and settles at 4% E (estimate). That gives revenue of $14.1B in 2036 and operating income of $2.32B, about 5.3% a year from today’s depressed base C (calculated). The margin falls to 12.5% in fiscal 2027 before it recovers to 16.5% by 2031 E (estimate). In fiscal 2027 that is earnings per share of about $8.26 E (estimate), below the $8.75 I get for 2026.

The bull case has the new team deliver: Americas growth of 4% to 5% a year after one more soft year, China back to double digits, and the margin at 21% by 2032, back toward the 22% of the boom years E (estimate). The bear case has the Americas fall 9% in fiscal 2027 and 5% in 2028, the margin touch 10%, and revenue crawl back to $11.6B E (estimate). Even the bear is a business that makes $0.9B of cash a year at the end. It does not go to zero. It is also a long way from the price people paid in 2023.

What I can’t judge

Three inputs matter more than all the rest, and I cannot know them.

Where Americas margins bottom. The Americas segment earned 32.6% in fiscal 2025 and about 28.5% in the second quarter without the refund.²⁴ The company guides to a consolidated margin of 6.5% in the third quarter.⁵ I do not know how much of that is seasonal and how much is lasting. My range for the final consolidated margin is 12% to 21%, which turns a value of about $91 into $224. I show the range and not a point because the evidence cannot narrow it.

The refund and the class actions. The company paid $230M of emergency tariffs and has been refunded $134.5M, so up to roughly $95M before tax could still come, or $0.60 a share after tax C (calculated).⁴ Two consumer class actions ask a court to give some or all of such refunds to customers.⁴ The outcome could be plus $0.60 a share or minus an unknown amount in damages, and I put nothing in either direction. The size is under 1% of the market value either way.

Whether a damaged brand heals. I can observe sales, conversion and markdowns. I cannot observe sentiment, and the chief financial officer said on the call that sentiment was a main problem. The base case treats it as slow repair; Under Armour’s record says it can be slower.⁵²⁷

Multiply it out

My base needs several things at once. The Americas stop falling by fiscal 2028: I would give that 55%. The product reset works, with markdowns falling and full-price sales rising: 55%. China Mainland comparable sales turn positive and stay there: 60%. The operating margin gets back to 16% or more: 50%. No second shock from tariffs or consumers: 75%. Multiplied, that is about 7% E (estimate). The legs overlap, since a working product reset helps the Americas, China and the margin together, so the true figure is higher, perhaps one in five. The case against this base is as strong as the case for it. But the price does not need my base. It needs a flat business at a 13% margin. That is the main thing to take from the reverse DCF, and it is why I say the price reflects a real debate.

Quality and price are two questions

How good is the business? It was among the best I have studied, with a gross margin of 56.6%, returns on capital above 40% and no debt, and it is now a business whose largest market is shrinking and whose leadership has turned over.² I judge the quality as high in the past and unproven for the next five years. How good is the price? Reasonable but not obviously cheap: below my base and above my bear, with an owner cash yield of 7.8% on a trough year. A great business can be a poor purchase at the wrong price and a damaged one can be a fair purchase at the right one. This looks like the second, with the caveat that the damage is not yet measured.

The case that the market is right

I want to say plainly what the pessimists see, because they may be right. First, the trend. Americas revenue growth went from +3% to +1%, then −2%, −4%, −3%, −8%, and the company expects about −15% in the second half C (calculated).^4222325 Second, the guidance. Two cuts in three months mean that management has missed its own forecast twice, and the second leaves the third-quarter margin near 6.5%.¹²¹³³⁵ Third, structure. Tariffs at about 20% for the rest of the year and the end of de minimis are lasting costs.⁵² Fourth, the competition: Deckers is growing and earns a higher margin, and Under Armour shows what failure costs.⁸²⁷ Fifth, a leadership team that is new, with a record on capital that includes MIRROR and $5.3B of buybacks at an average of $265 C (calculated).¹⁴²⁴ If you add these up, a price of 12 times forward earnings is not obviously wrong. The market is not ignoring the risks. It is pricing them.

The next best use of the same dollar

A ten-year Treasury pays 4.94% with no risk to the dollars promised.³⁶ On my base cash flows lululemon returns 12.9% a year; on my bear, 8.7% C (calculated). That is a wide spread over the Treasury even in the bear case, which is a function of the low price. On this site, Uber is priced 24% below its base value, as lululemon is 28% below its own, but the Uber page’s bear value is 54% below its price, whereas lululemon’s bear value is 15% below. The loss if the worst happens is smaller here; the recovery needed for the base is larger C (calculated).¹ Deckers, which has its own page here, trades at 11.7 times fiscal 2026 earnings with growth of 10% and a 23% margin C (calculated).⁸²⁸ The two sit on similar multiples, and the trend in one is falling while in the other it is rising.

The margin of safety

My margin-of-safety price is $89, 40% below my base value, and about 2% under my bear value. At $89 even my bear case returns about 10.2% a year, close to the 10% I require, and the base returns about 15.2% C (calculated). I use a wider cushion than the 30% on my Costco page because the earnings here are not predictable. Two guidance cuts in three months and a new team are the reason. The price is 20% above it, and the stock last closed below $89 on 4 April 2018, at $88.24; its lowest close of the past year was $91.88 on 7 October 2026.¹

Where this could be in 5 and 10 years

My base holds at five years if the Americas have stopped shrinking by fiscal 2028 and the operating margin reaches 16% in 2031 on about $11.8B of revenue. It holds at ten if the margin stays near 16.5% on $14.1B in 2036 and capital spending stays near 5.6% of revenue. Owner cash then rises from $0.93B to $1.56B a year. If the price simply moved to the year 10 value, price growth would be 5.4% a year; the rest of the return is the $112 a share paid out on the way E (estimate).

The bear and the bull differ on growth and margin. The bear has $10.3B of revenue in 2031 and a margin that ends at 12%, so its owner cash stays under today’s until 2036 and its year 5 value, $88, is a little under today’s $91. The bull has $13.4B and a 20% margin in 2031, which is how it gets to $311 in year 10. The discount rate is the third number: at 12% instead of 10%, my base would be $141 in year 10 E (estimate).

The most likely way my base is wrong is the margin. The price already assumes about 12%. If the margin stops there, my base would be about $117 in year 5 and $132 in year 10, with $78 a share paid out E (estimate). I cannot judge whether a damaged brand heals, so I read these as a range and not as points.

These are estimates of value, not forecasts of the share price. The market price can sit above or below value for years.

VIIThe case for the price

What most people miss

First, the part of the Americas that people quote is the part that fell, but the segment still earned about a 28.5% operating margin in the second quarter once the refund is taken out: 36.8% reported less 8.3 points C (calculated).⁴ That is before $1.4B a year of corporate costs, but it is a high number for a business with revenue down 8%. Second, the stock is clean. Dollar inventory was down 1% on the year and units down 7% at 2 August 2026, while the company chases 20% more volume in styles that sell.³⁵ Brands in a slump usually carry too much stock and mark it down for years. Third, the markets outside the Americas are now 31% of revenue and the profit pool there is large: China Mainland earned a 40.0% segment margin in fiscal 2025.² The catch is that China comparable sales fell 8% at constant rates in the second quarter, so this fact cuts both ways.⁴

The bull case

The clearest version is the founder’s. Chip Wilson, who owns 8.7%, argues in his filed proxy statement that the company lost its creative and technical edge, made decisions that hurt the brand’s premium position, and needs directors who know product. He paid for a full-page Wall Street Journal ad titled “lululemon: in a Nosedive” on 7 October 2025, nominated three directors, and settled on 26 May 2026 for three seats and a plan to end the classified board.²⁴³² His case is that the problem is fixable because it is a product and brand problem and not a market one. Management’s is similar in tone. In April 2026 the new chief executive, Heidi O’Neill, said her job would be to accelerate product breakthroughs and unlock growth in markets around the world.²⁹ In December 2025 the board pointed to ten years in which revenue went from $2.1B to $11.0B and operating income grew nearly sixfold.⁴⁰

Among investors, I could verify very little. Giverny Capital held 128,803 shares at 30 June 2026, down from 142,600 at 31 March.⁴¹ Viking Global held 980,568 at 31 March and none at 30 June, which cuts against the bull case.⁴¹ None of Altimeter, Atreides, Duquesne, Appaloosa, Coatue, Tiger Global, Third Point, Pershing Square, ARK, Elliott Investment Management or Fundsmith reported any in its latest filing.⁴¹ A director and the president bought stock in 2026.³³

Against reality

Some of the bull case is true now. The product margin is high, the stock is clean, there is no debt, and the new product head, Maggie Gauger, started on 26 October.¹¹ If they do what the case needs, the Americas stop shrinking by fiscal 2028 and the operating margin climbs back to 16.5%, the price is worth $148 on my model, and if it gets to 21% and China re-accelerates, $224 E (estimate). That is the possible. Today’s numbers are a long way from it. The company guides to a third-quarter operating margin of 6.5% and an Americas decline that its guidance implies at about 15% in the second half, it has cut its guidance twice since March, and it has replaced much of its senior team.⁵³¹¹ The price assumes the margin never recovers beyond about 12%, so the argument for the price is that the market is too pessimistic about the repair, not that the repair has begun C (calculated).

What a 10x would take. Ten times $107.24 is $1,072 a share, about $119B of market value on 110.7M shares C (calculated). The highest close in my price record is $511.29 on 29 December 2023, which was about $65B on the then share count C (calculated).¹¹⁷ So a tenfold gain needs almost twice the market value of the peak. At 20 times earnings it needs about $5.9B of net income, and at a 20% operating margin and 30% tax that is about $42B of revenue, 3.8 times fiscal 2025’s $11.1B C (calculated). Over ten years from this year’s expected $10.4B that is revenue growth of 15% a year, with the margin back to its old level and a multiple of 20. lululemon did grow revenue 18% a year from fiscal 2015 to 2025, but from $2.1B, and its largest region is shrinking now. I did not run a study of other companies, and I know of none that has done this from a base this size in this condition. I think it is close to impossible and I would not build a case on it.

Where I come down

The forward base year moved me most: it removed the screen’s claim that even my bear case sat above the price. The clean inventory and the Americas segment margin moved me a little the other way. The two guidance cuts and the turnover of the team did not move me, and they are why my bear case is below the price. What would raise my values most is two things at once: Americas comparable sales turning positive while markdowns fall, and an operating margin without refunds above 15% by the second half of fiscal 2027. It would lower them if the second-half decline in the Americas runs worse than about 15% or the third-quarter margin comes in under 6%.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• Tariffs and trade policy. The guidance assumes a tariff rate of 10% to 12.5% through September and 20% after.⁵ The company says tariffs and the end of de minimis will keep hurting gross margin.² A court or an executive order could move this either way.
• China and brand perception. China Mainland was 16% of revenue and about 19% of segment profit before corporate costs, but comparable sales fell in the second quarter.²⁴ Negative social commentary in China hurt traffic, and a second round would hurt again.⁵
• Supply chain concentration. Five vendors make 47% of the product, and 40% is made in Vietnam. Fabrics come 34% from Taiwan and 29% from China Mainland.²
• Litigation. A securities class action filed in 2024 survived a motion to dismiss in part on 31 March 2026; six derivative suits are stayed; two consumer class actions over tariff pricing were filed in 2026.⁴ The company has recorded only immaterial provisions for legal matters.⁴
• Governance. The founder holds 8.7% and has a cooperation agreement that restricts him until about the 2028 nomination deadline. After that he can act again.³²¹⁰
• Leadership. A new chief executive, a new product head, and the departure of the brand and supply chain chiefs on 6 November 2026 mean the plan is being written by people who have not delivered a year yet.¹¹
• Currency. A fall in the Canadian dollar or yuan changes reported results; foreign exchange added $69.6M to first-half revenue.⁴
• Incentive design. Bonuses and performance stock have used an adjusted operating income that left out $287M of costs.¹⁰ Adjusted figures can disguise how the business is doing.

What would raise my base value: the fiscal 2026 third quarter and the holiday quarter showing Americas comparable sales turning less negative, the gross margin without refunds back above 55%, and a first guidance for fiscal 2027 that is not cut afterward. What would lower it: a third guidance cut, a second-half Americas decline worse than 15%, or a large acquisition.³⁵