MoatLedger Wonderful companies at fair prices. A ledger, kept patiently.
pass

Lululemon: When Full Price Stops Working

Before the gates: the numbers that make this hard

On paper, Lululemon looks like exactly the kind of business a value investor waits years to find. Five-year average ROE around 34.6%, a debt-free balance sheet, gross margins near 56.6%, and a trailing P/E of 9.5. Any screener worth its salt would surface it at the top of the list.

We passed on it anyway. The qualitative analysis — the four gates below — is where the story stops cooperating with the spreadsheet.

Gate 1: Circle of competence — knowable, with an asterisk

Lululemon designs its own premium yoga and athleisure apparel and sells it almost entirely direct: 811 owned stores as of the end of 2025, plus its own e-commerce. Little wholesale dependence means the retail margin stays in-house. The premium price is justified — or was — by “technical performance plus community,” a mix of product and ambassador-led store classes. The Americas is the largest region; mainland China is the growth axis, with FY26 Q1 revenue up 30% there.

Is it knowable? Yes. Filings, earnings calls (FY2025 results on 2026-03-17, FY26 Q1 on 2026-06-04), and steady coverage of competitors make the business reconstructible from public sources.

Is it predictable? Here we hesitate. A consumer brand is inside our circle as a type, but premium fashion does not have the demand persistence of a cola or a candy bar. The honest question — will the leading premium yoga brand still be the leader in ten years? — has moved toward being a probability bet over the past two years. This is the industry where Dexter Shoe’s lesson applies most directly: fashion moats are chronically overestimated. Gate 1 passes, but with confidence marked low, and the real judgment deferred to Gate 2.

Gate 2: Moat — the failure

The moat’s source is real enough: the brand, plus a partial cost advantage from vertical D2C integration. There are no switching costs and no network effects — the cost of changing your leggings brand is zero.

The pricing power evidence used to be excellent. Historically roughly 95% of product sold at full price, with industry-leading gross margins around 57–58%. But the litmus test has recently inverted. Through 2025–26, persistent markdowns across product lines became a recurring theme in coverage. FY26 Q1 gross margin fell sharply year over year to the mid-54% range (company figures, tariffs included), and operating income dropped 37%. A company that can raise prices without a prayer meeting is becoming one that cannot move inventory without a discount.

The direction is narrowing, on four counts:

  1. Share erosion. The US athleisure share of ~21% still leads, but Alo Yoga has grown rapidly in the premium D2C segment (to roughly 14%, with cultural dominance among Gen Z) and Vuori (valued at $5.5 billion, preparing an IPO) attacks precisely Lululemon’s weak spots: style, menswear, younger customers.
  2. Core market shrinking. Americas comparable sales fell 5% in FY26 Q1 in dollar terms, and the company itself guided to a high-single-digit annual decline in the region.
  3. Product misses. Management admitted on the call that new product launches “failed to resonate.” Fashion risk is currently beating the moat narrative.
  4. China’s +30% is market penetration, not moat. Real growth, but it is evidence of a new market, not of durable advantage — and reports suggest even that growth rate is decelerating.

A note for the AI era: neutral to negative. AI does not reinforce a brand moat here; it lowers the cost of influencer- and social-media-driven brand building, which makes Alo-style entrants more marketing-efficient. The barrier itself is thinning.

Verdict: fail. A brand moat can be identified, but its direction is clearly narrowing and the pricing-power litmus test has failed.

Gate 3: Management — noted, but moot

Capital allocation has been reasonably good. Debt-free operations; the only blemish is the ~$500 million Mirror acquisition and write-off, painful in principle but not fatal, and the loss was acknowledged and cleaned up. Buybacks grew from ~$0.56 billion in 2024 to ~$1.6 billion in 2025, with a further $1.0 billion authorized in December 2025 — buying during a falling stock price is the right direction, though buying during falling earnings still needs post-hoc verification.

The retained-earnings test, however, is turning into a failure: despite large retained profits over five years, the market cap is down 46% from its peak in 2026, breaking the dollar-for-dollar rolling test.

On candor and alignment: CEO Calvin McDonald resigned effective 2026-01-31 with no successor named (interim CEO Meghan Frank), and a proxy dispute is underway — a leadership vacuum that alone would warrant suspension of judgment. Credit where due: the interim CEO admitted the product missteps on the call.

Gate 4: Price — cheap, and not cheap enough

Trailing owner earnings are $1.39 billion, a 10.5% yield. But that is a rearview number. FY2025 net income was ~$1.6 billion (-13% YoY), free cash flow fell 42% to $0.92 billion, and FY26 guidance implies flat-to-declining revenue. A conservative normalized owner earnings estimate is $1.0–1.2 billion.

At a cautious 10x on those normalized figures — and even no-growth 10x may be generous in a declining-earnings phase — intrinsic value sits at roughly $10–12 billion. The current market cap of $13.3 billion is above the top of that range.

The P/E of 9.5 is an illusion built on trailing earnings. The 10.5% owner-earnings yield comfortably beats the 10-year Treasury at 4.3% today, but if earnings keep shrinking at double-digit rates, that advantage evaporates within two to three years. On normalized figures, there is no margin of safety. The price is not cheap; the earnings are falling — the textbook value trap setup.

Verdict

Rejected at Gate 2. The quantitative profile — 34.6% average ROE, no debt, 56.6% gross margin, P/E 9.5 — is exactly what a screen should find. The qualitative picture is what the screen cannot see: a full-price model breaking down, Americas comps at -5%, Alo and Vuori taking aim at precisely the company’s weak spots, product missteps admitted by management, and a CEO seat left empty amid a proxy fight. The failed pricing-power litmus test was decisive. A low P/E that front-runs declining earnings is not a margin of safety. (One data note: the zero debt figure is genuine — no borrowings — though operating lease liabilities of roughly $1.6–1.8 billion imply a lease-adjusted D/E near 0.35, still within bounds.)

What would change it: two consecutive quarters of positive Americas comparable sales, or a recovery of gross margin toward 57% alongside full-price selling and a permanent CEO with a credible product strategy — and this thesis gets re-reviewed as a solvable, one-to-two-year stumble rather than a structural break.


This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer