This piece was finalized before Airbnb’s May 7 earnings call; any relevant updates will be incorporated into Part II. I hope the wait proves worthwhile.

Introduction

“Certain people believe things for understandable reasons. But those reasons are nonetheless bad reasons.” — Sam Harris

Warren Buffett has often remarked that it takes years to build a reputation and only minutes to destroy one. The reverse, however, may be just as true.

Consider Marie Antoinette: for more than two centuries, she has been remembered, above all, for a line she almost certainly never uttered, “Let them eat cake.” The myth survived not because the evidence was strong, but because it was memorable, convenient, and easy to repeat. The truth, by contrast, required context—and context has never traveled quite as efficiently as a good accusation.

This phenomenon perfectly describes the lingering reputation of Airbnb co-founder and CEO Brian Chesky. For those unfamiliar, Chesky acquired the label of “overhyper” several years ago. While the charge was not entirely without basis, it has been repeated far more often than it has been critically examined.

A closer look at the record reveals a much more nuanced reality. Chesky’s track record does not show a pattern of grand promises collapsing upon contact with reality; rather, it shows an unusually animated CEO whose rhetoric occasionally ran ahead of his intended message. (That same zeal has undeniably contributed to immense wealth creation, but that is a separate matter.)

Since going public in 2020, Airbnb has met or exceeded consensus expectations for revenue and EBIT in all but one quarter. The supposed overhype, then, was rhetorical rather than financial—a crucial distinction.

Today, Chesky’s commentary is discernibly more restrained, suggesting he has learned from past scrutiny. He has also spoken openly about his suspected neurodivergence, which offers a more charitable—and likely more accurate—lens through which to understand what might otherwise be perceived as a genuinely disconcerting trait in a CEO.

While his reputation for overhyping no longer dominates investor discourse, it is still mentioned often enough to matter. It therefore remains, at minimum, a subtle interpretive filter through which Wall Street receives new information, causing current evidence to be discounted before it is properly weighed.

This lingering skepticism goes a long way toward explaining the striking divergence between Airbnb’s own medium-term financial outlook and the far more tempered expectations embedded in Wall Street estimates.

This mismatch is especially notable given that Chesky’s recent “implied guidance” has, in fact, been highly predictive. In September, for example, he expressed dissatisfaction with Airbnb’s roughly 10% growth rate—arguing that the company should be growing at least in the teens—before adding:

“I believe we are going to accelerate growth next year. I firmly believe that. I don’t want to overpromise what we’re going to do. But I don’t think people should take our deceleration and extrapolate that forward.”

Extrapolate they did.

In the weeks leading up to Q4 earnings, Chesky again signaled a reacceleration—both within the quarter and beyond. Even so, the stock fell more than 10% in the weeks that followed. Then came the print: just as Chesky had telegraphed months earlier, Airbnb reaccelerated in Q4 and guided to at least low-double-digit revenue growth for 2026.

Reputation, however, is only part of the story. Also relevant is what we might call a distribution problem—one I suspect will be resolved during the next earnings call or Summer Release.

Investors are conditioned to expect material information to surface through formal channels: filings, earnings calls, and press releases. Most of the time, that is exactly where it appears. Occasionally, however, important signals emerge in less obvious places, such as interviews, podcasts, or offhand remarks, where they are easier to miss and, consequently, easier to misprice. That appears to be the case here.

Before getting to the specifics, some context is needed. For those less familiar, 2025 was a pivotal year for Airbnb, arguably its most consequential from a product standpoint since the company’s founding. It launched three new businesses: Services, “reimagined” Experiences, and Hotels, specifically boutiques and independents. (While Airbnb has long had a small number of hotel listings on the platform, for reasons discussed later, they amounted to little more than a rounding error relative to what is now being built.)

Each of these new ventures, according to management, has the potential to generate billions in incremental revenue. Yet while their commentary has been highly bullish, it has also been carefully noncommittal.

Initially, Chesky suggested that these businesses could take three to five years to scale, which he effectively defined as reaching roughly $1 billion in revenue. He was also careful to add that not all of them would necessarily get there. By the Q2 2025 earnings call, however, his language had moved a notch higher, stating:

“I think each of these could easily be a multibillion-dollar business.”

The shift was subtle, but not trivial: the $1 billion benchmark had become multibillion, and the word “easily” nudged the implied floor higher still. The “could,” however, continued to leave room for ambiguity. CFO Ellie Mertz, meanwhile, remained more restrained in both scope and timing, using phrases like “significant contribution,” “material scale,” and “multiyear path.”

Beyond that, nothing in subsequent earnings calls—or in the many interviews and podcast appearances Chesky gave following the 2025 Summer Launch—strayed from this calibrated framing. (For readers who want the full chronology, I’ve compiled the verbatim remarks from both Chesky and Mertz [here], with timestamped links where possible.)

There was, however, one notable exception: an interview Chesky gave in January. The first real break came in his discussion of hotels. Noting that people often fail to realize the sheer size of the hotel industry, which is larger than the advertising industry dominated by Google and Meta, he stated:

“We do not need a large market share of hotels to have a huge business. Hotels will absolutely be a multibillion-dollar business.”

He offered a striking example, revealing that there were over 400 million searches for New York City on Airbnb last year. Given that Airbnb is “essentially banned” in New York, he pointed out that adding hotels means the company “can make quite a bit of revenue” there.

Two things stand out from this exchange.

First, the language was unequivocal. Gone were the earlier qualifiers and hedges; in their place was the definitive “will.”

Second, the anecdotal framing became hard data. Chesky had previously alluded to “millions and millions” of searches for New York City each year on Airbnb—a phrase that, though delivered with notable emphasis, dramatically understated the 400-million-plus reality.

Listeners do not infer magnitude from tone alone; they infer it from ordinary language conventions. Most would therefore have assumed a figure in the single-digit millions, reasoning that if the actual number were in the tens—or especially hundreds—of millions, he would have explicitly said so.

He also raised the implied floor on promoted listings, which are likely to roll out later this year or next, noting, “I think advertising is another multibillion-dollar business—it’s at least a billion dollars.” But the clearest break came when discussing the trajectory of these new businesses collectively:

“We basically put a horizon of three to five years for these to be multibillion-dollar businesses.”

Publicly, Airbnb had not quite done that—at least not explicitly. That suggests Chesky was either conflating internal assumptions with external commentary, or believed he had already been more precise than he actually was. Either way, the implications are significant.

Taken at face value, the comment implies that consensus revenue estimates for 2028 through 2030 may be understated by at least $3 billion to $8 billion, before assigning any additional value to growth in Airbnb’s core business. (For reasons discussed later, I believe further core growth is also highly likely.)

That estimate also excludes any contribution from additional billion-dollar businesses not yet disclosed, despite management’s expectation that Airbnb can launch at least one new billion-dollar business per year for the foreseeable future.

Now, to address the elephant in the room: while I clearly find Chesky far more credible than most, the investment case does not depend on taking his projections at face value.

Fortunately, we know enough about the underlying assumptions—and have ample supporting data—to stress-test the case ourselves. That includes the major sources of debate and concern, such as AI disintermediation, heightened competition, and regulatory challenges, all of which will be covered in depth in Part II.

Having done that work, my conclusion is that Airbnb is likely to substantially exceed topline estimates while also outperforming on operating margins. Before we can chart the path ahead, however, we first need to establish its starting coordinates.

Disclosure: I and/or accounts under my control are long ABNB. This is not investment advice, but my opinion as of today, which may change quickly and without prior notice. Readers should conduct their own due diligence. Some quotes may be lightly edited for clarity and relevance.

Brief History

Airbnb’s origin story has been told often enough that there is little value in revisiting it at length here. For readers interested in the fuller version, Acquired’s 2020 episode remains the best single treatment. What follows instead is a high-level sketch of how the business evolved.

Airbnb—then known as “Airbedandbreakfast.com”—began as a low-cost substitute for hotels, used mostly by younger, budget-conscious travelers. Early listings were often spare rooms in occupied homes, and demand skewed heavily toward cross-border travel. Over time, however, both the platform and its user base moved steadily upmarket—a shift helped, in no small part, by the fact that many early Airbnb users aged into higher income brackets.

As inventory moved increasingly toward entire homes and apartments, Airbnb’s value proposition also changed. The appeal was no longer simply lower prices. It was more space, more flexibility, and a different kind of travel experience: the promise, as the company once framed it, of “living like a local.” Today, private rooms account for less than 10% of listings.

By 2019, Airbnb had reached nearly $40 billion in gross booking value and almost $5 billion in revenue. It had also become one of the most recognizable consumer internet companies in the world. What it had not yet become was profitable.

Then came the pandemic. Within weeks, Airbnb lost roughly 80% of its business and cut about a quarter of its workforce. At the time, the shock looked potentially existential. Yet the business recovered with surprising speed. Cross-border travel gave way to domestic trips, and average stays lengthened as remote work untethered millions of people from the office.

The pandemic also spurred a sharp increase in family and group travel, contributing to higher average daily rates (ADRs). Some of these effects have since moderated, but several remain structurally above pre-pandemic levels. Stays of 30 days or longer, for example, have settled at roughly 17% to 18% of nights booked, up from about 14% in 2019, though below the 20%-plus peak reached during the pandemic. The average trip length is now roughly four days.

Airbnb went public in late 2020 at $68 per share. The stock reached more than $140 on its first day of trading, making it one of the year’s standout IPOs. At the time, the three co-founders—Chesky, Joe Gebbia, and Nathan Blecharczyk—collectively owned about 42% of the company, with near-equal stakes and near-total voting control.

As of April 2026, they still held roughly 27% of the economic interest—with Chesky and Blecharczyk together owning about 22%—while continuing to control approximately 81% of the vote. Joe Gebbia is no longer involved in day-to-day operations but remains on the board.

Airbnb turned profitable in 2021 and now generates some of the highest free cash flow margins in the S&P 500, though not without caveats—including a notably heavy stock-based compensation (SBC) load.

Business Model and Overview

“Our core business could easily be two to three times bigger—and we did $80 billion in bookings last year.” — Brian Chesky, June 2025

The beauty of Airbnb’s marketplace model is that, unlike hotels, it “doesn’t have to pour concrete,” as Chesky put it. Airbnb doesn’t own or operate the properties on its platform; it connects guests with hosts, facilitates the transaction, and takes a fee. The result is music to every investor’s ears: asset-light, scalable, and capable of producing robust margins.

That advantage is reflected in its market cap, which, until very recently, was exceeded that of any single hotel operator in the world.

And yet the hotel industry itself remains far larger than the short-term rental (STR) market. In the U.S., Airbnb’s largest market, the company still accounts for only about one in ten nights away from home. The other nine go to hotels. We’ll return later to why travelers choose one over the other.

Today, Airbnb has more than 9 million active listings across 220-plus countries. Despite that global footprint, the business remains concentrated in a handful of mature markets. Roughly 70% of revenue comes from just five countries: the U.S., U.K., France, Canada, and Australia.

By volume, EMEA accounts for the largest share of nights and experiences booked, at roughly 40%, followed by North America at about 30%, Latin America at 17%, and Asia-Pacific at 13%. From a revenue standpoint, however, North America remains disproportionately important, as shown below.

Airbnb now has more than 5 million hosts, around 90% of whom are individuals—which makes it the largest non-professionalized supply base in the world by a wide margin.

For most hosts, Airbnb remains a source of supplemental income rather than a full-time business. Around 55% are women, and common host occupations include schoolteachers, healthcare workers, and students. In the U.S., the average host now earns roughly $15,000 per year, up from less than $10,000 five years ago.

Historically, Airbnb’s supply growth has been largely organic. More than 30% of hosts were previously guests, while word-of-mouth continues to play an important role in bringing new supply onto the platform. This is made easier by the fact that roughly half of new hosts receive a booking within three days, and 75% do so within nine—significantly faster than on competing platforms.

Large events have also proved to be powerful supply catalysts. During the Paris Olympics, more than 700,000 guests stayed in Airbnb listings, while roughly 50,000 new listings were added to the platform. About 80% of those remained active afterward.

The vast majority of hosts still list only on Airbnb, though cross-listing has been rising. In the U.S., professional operators—defined by AirDNA as those managing 21 or more listings—cross-list at rates of roughly 60%, versus around 21% to 28% for non-professional hosts.

The mix between individual and professional hosts is shaped in large part by local regulation, a theme we will return to later. You can see this clearly in the chart below, which shows the supply mix across major European cities as of early 2024.

Competitive Landscape

The short-term rental market is effectively an oligopoly. Airbnb, Booking.com, and Expedia—which owns Vrbo—controlled roughly 71% of the market in 2024, up from 53% in 2019.

Booking and Expedia are traditional online travel agencies (OTAs), offering hotels, vacation rentals, flights, car rentals, and other travel products. Their inventory is also more professionalized, spanning hotels and large property managers. Airbnb and Vrbo are more narrowly focused on short-term rentals, though Airbnb’s expansion into hotels, services, and experiences is beginning to blur that distinction.

Accommodations account for the bulk of revenue at both Booking and Expedia, at roughly 90% and 80%, respectively, with the remainder tied to flights, car rentals, and other products. Booking also has the highest effective take rate, averaging roughly 14.4% over the past two years, compared with about 13.5% for Airbnb and 12.3% for Expedia.

Booking is particularly strong in Europe, where it holds roughly 48% share, versus about 40% for Airbnb and just 2% for Vrbo. In the U.S., the pattern reverses: Airbnb leads with roughly 43% share, followed by Vrbo at about 21%, while Booking trails at just 8%.

Vrbo occupies a narrower but still defensible niche. Its supply base skews toward entire homes, and its customer base over-indexes toward families, group travel, longer stays, and older travelers. That positioning gives Vrbo a clearer identity than its relative scale might suggest.

Still, the long-term trajectory is hard to miss. Despite a 13-year head start, Vrbo was overtaken by Airbnb within a few years. Today, Airbnb is roughly three times larger by revenue. Google Trends provides an imperfect but useful proxy for that shift.

Today, though, most bookings happen through apps—and Airbnb leads there as well. In 2025, roughly 64% of its bookings came through mobile, versus figures in the mid-50s for Booking and Expedia.

The Moat(s)

“Airbnb” has become nearly synonymous with short-term rentals. Guests rarely say they stayed in a vacation rental; they say they stayed in an Airbnb. Outside of Vrbo in the U.S.—and even there, to a much lesser extent—no competitor enjoys anything close to that linguistic hold.

Airbnb has even crossed the threshold from noun to verb: the pinnacle of brand strength. Naturally, it has leaned into this in its marketing: “Now you can Airbnb more than an Airbnb.” That brand strength shows up where it matters: since the pandemic, roughly 90% of Airbnb’s traffic has come through direct or unpaid channels—well ahead of peers.

The P&L implications are not subtle. In 2025, Airbnb spent about $1.6 billion, or 13% of revenue, on brand and performance marketing. Booking and Expedia, by contrast, spent roughly $8.2 billion and $7.3 billion, respectively—or about 30% and 50% of revenue—using the most directly comparable measures. This is sometimes called the “Google Tax,” and Airbnb pays far less of it.

What makes this brand moat especially notable is how it was built—without the the bundles and convenience offered by traditional OTAs. No flights. No car rentals. No loyalty program. And, until recently, virtually no hotels. That said, each of those categories will likely prove necessary if Airbnb is to become the “one-stop shop for travel” Brian Chesky envisions—a point we’ll return to later.

Beneath this moat sits a classic marketplace flywheel. Greater supply—especially when it is unique and high-quality—drives more demand. More demand, in turn, makes the platform more indispensable to hosts, drawing in still more supply. That deeper supply base improves price competition and, critically, gives Airbnb more flexibility to shape the marketplace itself.

The company has leaned into that advantage aggressively. Since 2023, Airbnb has been “elevating the top and cutting the bottom,” as Chesky put it, removing more than 550,000 lower-quality listings while promoting “Guest Favorites,” its top 10% of listings by rating. Those listings now account for roughly half of bookings.

The result has been fewer support interactions and better booking conversion. Together with a broader set of product changes designed to reduce friction, these improvements have helped push Airbnb’s NPS to its highest level since the pandemic—“by far,” according to Chesky.

This selectivity advantage, combined with Airbnb’s disproportionate share of non-professional hosts, has lifted the average listing rating to roughly 4.8 out of 5 stars. Competitors do not disclose directly comparable figures, but the gap is visible. In my own cross-platform analysis across major tourist cities, weighted-average ratings for first-page OTA results tend to fall around 4.2 to 4.3—and slightly lower when hotels are included.

Summing it up: better supply → better experience → more demand → greater selectivity → even better supply. That, in a nutshell, is the Airbnb flywheel.

There are, however, two further advantages that help keep it spinning.

1. User Interface (UI)

Spend a few minutes on each of the major platforms and the difference is hard to miss: Airbnb’s product is, by some distance, the most pleasant to use. That has been true for years, and it is likely to remain true. There are several reasons.

First, two of Airbnb’s three co-founders, Chesky and Gebbia, came from design backgrounds. But this is more than a matter of education or training: they live and breathe design. It is part of their DNA and, by extension, the company’s. As Chesky put it:

“I obsess over our app: its design, its interface … Everything must be perfect.”

Second, Airbnb is, unlike its competitors, a technology company operating in travel—not the other way around. The distinction can sound cliché, but it explains a great deal. By all accounts, Airbnb has some of the strongest technical and design talent in the world, helped by its San Francisco base and reflected, not coincidentally, in its materially higher SBC.

Finally, design quality is simply something Airbnb has chosen to optimize for—even at the expense of near-term conversion. That stands in contrast to its OTA rivals, where conversion is the organizing principle.

Below are comparisons of Booking’s and Airbnb’s webpages, both at 100% zoom, followed by a few brief observations.

It is not a knock on Booking to say that it feels, by comparison, like a high-performance marketplace funnel—because that is exactly what it is. The classic e-commerce persuasion cues are all there: urgency prompts, competing colors and font sizes, icons, badges, and so on. Meanwhile, the prominently featured filters nudge the user to narrow the search if the “top picks” do not appeal.

Airbnb does the opposite. The filter button is discreetly tucked away. The wider layout gives the page more breathing room. There are fewer visual interruptions, stronger aesthetic cohesion, and a general sense of openness. The experience feels less like shopping and more like browsing. The page seems to have one job: make the listings look appealing.

The contrast is just as clear at the listing level.

This “beautiful merchandising,” as Chesky describes it, directly reinforces Airbnb’s core advantage: having the largest catalogue of unique (i.e. non-commoditized) supply.

The second advantage is less tangible, but no less important.

2. Community and Trust

Reviews are a form of currency—proprietary reviews, that is—and Airbnb has more than 500 million of them. Engagement is unusually high, with roughly two-thirds of guests leaving a review after a stay, and those reviews tend to be longer and more detailed.

That is not accidental. It is a direct consequence of Airbnb’s deliberate effort to foster a sense of “community”—a word that can sound soft until one considers what the product actually requires: persuading people to stay in, or hand over, private homes.

Unlike other platforms—where, as Chesky has noted, “you often don’t even know who people are”—Airbnb profiles are rich and consistent. Users have profile photos, visible review histories, trip counts, and tenure on the platform. None of these are shown in quite the same way on Vrbo or Expedia.

Booking does somewhat better, displaying country of residence and profile photos. But unlike Airbnb, where profile photos generally show the actual user, adoption remains limited.

At first, these differences may sound minor. Then you see them.

Compared to Booking, top, and Vrbo, bottom:

These small touches combine to produce a not-so-small effect: Airbnb feels warmer, more personal, and less transactional.

More recently, the company has even started allowing users to see who else has signed up for a group Experience—and even to message those users afterward. As a result, Airbnb is beginning to look less like a traditional travel marketplace and more like a social platform layered onto one.

That helps reduce uncertainty on both sides of the transaction—something especially important when one of the top concerns for would-be hosts is the idea of a stranger staying in their home.

On Airbnb, guests feel less like strangers.

Closing Note

And that’s a wrap for Part I. Hopefully, readers now have a better understanding of how Airbnb’s marketplace works—and why its competitive position is stronger than the market often appreciates.

Part II, however, is where things get more interesting. There, we’ll examine the reacceleration case in both the core business and new verticals, address the key bear arguments—my favorite part—and walk through valuation, catalysts, and related considerations.

The valuation discussion is especially relevant now, given Airbnb’s rally and Booking’s sell-off. As many readers will know, Booking has become an increasingly obvious candidate for value-oriented investors, given its stronger recent growth—off a larger base—and materially higher profitability.

Thanks for reading, and stay tuned.

In this post, I’ll provide a brief update on the Aquitaine Model Portfolio’s (AMP) performance to date. I’ll also clarify a few points from my introductory post and provide additional context on my approach to stock selection and portfolio construction.

Revisiting that first piece, it’s clear I could have done a better job explaining what AMP is—and, just as importantly, what it isn’t—and why I’ve chosen to run it this way. I’ll try to remedy that here.

Performance

For those new here, AMP is fully invested and consists of just four stocks. The largest position accounts for roughly 40% of the portfolio; the smallest is about 10%.

As of December 31, AMP was up approximately 14%, compared with roughly 1% for the S&P 500 over the same period. While I don’t optimize for—or pay attention to—monthly or quarterly results, it’s not a bad start (or “restart”) for the portfolio.

Notably, that year-end gain added roughly 29% to total returns since inception—a neat illustration of the power of compounding. All told, cumulative returns now stand at approximately 140%, versus about 52% for the S&P 500 over the same time frame.

What is AMP?

AMP is a concentrated, long-only portfolio consisting of companies I’ve covered (or intend to cover) in deep dives. My own equity portfolio consists of the same companies, held at broadly similar weights.

The objective is straightforward: to significantly outperform the market over the long run without the use of margin, options, or leverage.

Note: Going forward, AMP’s positions and sizing will be available only to paid subscribers.

On Concentration

The more diversified you are, the harder it is to beat the market. As Bill Ruane—Warren Buffett’s longtime friend—once noted:

I’m not Peter Lynch. But I do believe I can do a really good job investing in a small number of deeply researched businesses. While the number of holdings may grow over time, I don’t expect AMP to ever exceed 8–10 stocks.

That level of concentration naturally comes with heightened volatility, making it unsuitable for most investors. I’ve found, however, that I have an unusually high tolerance for price swings. Some of that is innate; most of it comes from (a) knowing exactly what I own, and (b) staying alert to information that might contradict my views.

That second point is crucial. According to Michael Platt—one of the world’s most successful hedge fund managers—it’s also the primary reason investors blow up. “Ego gets in the way,” he said. “They just don’t want to be wrong.”

Time Horizon

Bargains can exist even in expensive markets, but whether a stock is truly a “bargain” also depends on your time horizon. A compelling risk-reward profile over two to five years can look like a terrible proposition over the next twelve months—which increasingly appears to be the market’s default horizon.

When I buy a stock, it’s because I believe it will be worth substantially more a few years from now—preferably sooner—regardless of what the broader market does in the interim.

Portfolio Criteria

As noted in the first post, AMP holdings are not selected solely on expected IRRs. While return potential clearly matters, each position must meet three high-level criteria:

• It offers a robust asymmetric risk-reward profile

• I have a differentiated angle or analytical lens

• It’s likely to be of interest to a reasonably broad audience

That final point generally excludes most small-cap stocks and also explains why AMP remains long-only; short ideas tend to appeal to a much narrower subset of readers.

Beyond that, several additional boxes must be checked.

Catalysts

Undervalued stocks can remain undervalued for a very long time—and the same is true in reverse. While I’m willing to wait years if necessary, I generally avoid situations where I don’t see a plausible reason for the valuation gap to narrow/close within the next 12–18 months.

Beta

(Some of you may have seen my recent note on beta. I’m reiterating it here for ease of reference.)

Beta was never something I paid much attention to—until I read Hedge Fund Market Wizards by Jack D. Schwager.

“Buying low-beta stocks is a common mistake investors make,” said Martin Taylor, a fund manager Schwager credits with “possibly the best performance record in emerging markets.” Between 1995 and 2011, Taylor generated compounded net returns of 27%—more than double the 12% delivered by the emerging-markets index.

Taylor continued:

“Why would you ever want to own boring stocks? If the market falls 40% for macro reasons, they fall 20%. Wouldn’t you rather just hold cash? And if the market rises 50%, the boring stocks go up only 25%. You get negative asymmetric returns. It’s what I call a pigeon-and-elephant trade—you eat like a pigeon and s**t like an elephant. To get equity-like returns from a portfolio of boring stocks, you have to leverage it up. If things go wrong, the loss will be massively asymmetric because of the leverage.”

His reasoning is hard to argue with: when you’re right, you want the stock to be able to really move. Of course, beta cuts both ways, underscoring the importance of buying with a margin of safety.

AMP’s weighted beta is currently around 1.25 on a one-year basis and 1.36 over five years.

Financials

With a few exceptions, the businesses I want to own share the following characteristics—or have a clear path toward them:

• High normalized operating margins. There are excellent structurally low-margin businesses—Costco is the canonical example—but they rarely make attractive long-term investments. High margins provide room for error, reinvestment, and compounding.

• Strong balance sheets. Minimal leverage, ideally none. If a business can’t weather a downturn without access to credit or additional equity—often available only on punitive terms—it’s unlikely to make it into the portfolio.

Management

Few CFOs are as highly regarded as Barry McCarthy, formerly the long-time CFO of Netflix and Spotify. In an interview a few years ago, he made an observation that stuck with me:

“In my experience sitting on boards—and I’ve sat on many—you never really know what’s going on inside a company, no matter how hard you try. Board members are almost entirely dependent on the management team and the CEO to assist them in their understanding of the business.

It’s a sobering reality. While this doesn't suggest we should forgo deep research—on the contrary, it is essential—it does argue for a measure of humility regarding what we can truly know as outsiders. For McCarthy, this means successful investing often boils down to one thing: backing the right leadership.

“As an outsider, you can’t ever know enough to really make an informed decision. So you need to have the pattern-matching skills to know the difference between great and not.”

This is why I spend so much time listening to—and studying—management teams. Whether it’s earnings calls, investor days, conference appearances, or media interviews, I want to understand who these people are, how they think, and what drives their decisions. When they exist, podcasts and other informal settings are particularly revealing.

Yes, it’s far more efficient to read transcripts—and I do that too. But a tremendous amount of information never makes it onto the printed page: tone, cadence, hesitation, confidence, deflection. Over time, those subtleties accumulate. This qualitative work helps me get closer to answering the questions that matter most:

• Competitive Edge: If I were a competitor, would I rather square off against this team—or someone else?

• Talent Magnetism: If I were a professional in this industry, would I want to work here? (Top talent, after all, wants to be around top talent.)

• Capital Allocation: Do they deploy capital in a scrupulous, shareholder-friendly manner?

• Engagement: Do they sound genuinely energetic and passionate about the business?

• Healthy Paranoia: Perhaps most importantly, do they seem sufficiently paranoid? (Because the competition never stands still.)

Final Thoughts

As should be clear by now, AMP is shaped by a variety of idiosyncratic factors and is therefore a “model” portfolio in name only. As I emphasized in my first post, it’s not intended to be copied—even loosely—but rather to serve as a public scorecard and a mechanism for accountability around my research.

I’m excited about the work I currently have in progress, including Aquitaine’s next deep dive, which is set to be published later this week.

Until next time.

A brief recap of the quarter:

• Financials: Revenue and operating income (OI) exceeded consensus estimates by 3.5% and 18%, respectively.

• Americas: Same-store sales (SSS) in the Americas declined 5% y/y—largely in line with expectations.

• International: Results were a meaningful upside surprise, driven by 25% comps growth in China (FX-neutral), accelerating from 16% in Q2.

The most consequential development, however, came on the management front. CEO Calvin McDonald announced he will step down after seven years in the role, effective January 31, 2026. While revenue tripled under his tenure, the stock’s immediate reaction suggests the market views his departure as a positive.

But is that reading correct?

In my deep dive, I argued that many of the criticisms directed at McDonald were—when properly contextualized—broadly unfair, and that I remained confident in his leadership. I did, however, flag an important caveat: if the Spring 2026 assortment failed to reignite growth, all else equal on the macro front, a leadership change would likely follow.

Against that backdrop—and with the stock now up roughly 20% since publication—this feels like an appropriate moment to take a temperature check and reassess the portfolio implications.

Read more

• Toast ($TOST): $34/share • $20bn mcap • ~23× EV/EBIT (NTM)

• Airbnb ($ABNB): $117/share • $71bn mcap • ~21× EV/EBIT (NTM)

• Match Group ($MTCH): $33/share • $7.8bn mcap • ~11× EV/EBIT (NTM)

Let’s get into it.

Toast

I spend a lot of time listening to CEOs, and because I’m usually hunting for exceptional businesses at reasonable prices—the leaders I study tend to be of a high-caliber. Even so, it’s rare for one to make me stop and think, “Wow—this CEO is exceptional.”

That’s exactly what happened when I encountered Toast’s co-founder and CEO, Aman Narang. There are many reasons, but they boil down neatly into two points:

1. If I were a competitor, he’d be the CEO I’d least want to go up against.

2. If I worked in this space, he’d be the CEO I’d most want to work for.

(I suspect most of you would reach the same conclusion.)

Founded in 2013, Toast runs a cloud-based technology platform for the restaurant industry—an all-in-one POS. Growth has been exceptional: from under $1B pre-pandemic to more than $6B today, with plenty of room ahead. The company turned profitable last year and should ultimately reach adj. EBITDA margins of 40% or more. It’s also extremely well capitalized, on track to end 2025 with roughly $1.8 billion in net cash.

Most people (myself included, until recently) think of a POS as a fancy cash register. That may have been true once, but today—as Aman notes—it’s better understood as a restaurant’s central nervous system, coordinating orders, payments, menus, staffing, inventory, online channels, customer data, and more.

Toast now powers more than 150,000 restaurants across the U.S.—primarily SMB and mid-market operators—including more than half of those with Michelin stars.

And there are still several hundred thousand restaurants in the U.S. alone that don’t (yet) use Toast—not to mention international markets, where the company has barely begun to scale.

Toast’s U.S. market share has more than doubled since 2021—from roughly 7% to more than 15% today—an especially impressive climb given how sticky POS systems tend to be. (Replacing one has been compared to removing a root canal.)

But the national figure understates Toast’s strength in its most developed cities. In many of these “flywheel markets”—dense pockets of customers that amplify word-of-mouth—Toast often holds 30%+ share. In Cambridge, MA (where Toast is based), for example, Aman observed: “you’d think Toast is the only point of sale.” Win rates and conversion are meaningfully higher in these areas.

And its share of new restaurant openings is higher still.

There’s also substantial room to grow monetization per location.

And Toast’s TAM isn’t limited to restaurants. Management sees several adjacent retail categories where they believe they have a “right to win.” Retail food and beverage is a notable example—still early, but traction so far has been strong.

According to management, these adjacencies—along with Toast’s budding enterprise business—could eventually rival or even exceed the size of its core restaurant business.

To sum it up: Toast has its competitors playing catch-up, and I expect that dynamic to persist. I also share management’s confidence in the company’s growth potential, both within the core and across these emerging categories. (In my experience, when you back exceptional CEOs at sensible valuations, the surprises tend to break your way.)

Valuation: Toast currently trades at 23× and 17× EV/EBIT for 2026/27 based on consensus estimates (OI margins of ~10.5% and 12%) and an implied 19% revenue CAGR. Assuming a stable macro backdrop, I expect the company to outperform on both growth and profitability (15–20% OI margins by 2027). If so, I’d expect at least 40–50% upside from current levels—and potentially much more.

The balance sheet also supports continued share repurchases, providing a measure of downside protection if the stock were to pull back. And unless the thesis changes, I’d view that as an opportunity. For long-term investors, the risk/reward remains compelling.

Subscribe

Airbnb

“I keep telling people that one day this company is going to be huge—it’s going to have thousands of users.” — Brian Chesky, CEO and co-founder, recalling Airbnb’s early days.

Airbnb hardly needs an introduction. Its name is now shorthand for the entire vacation-rental category—so much so that it’s become a verb. (Their recent marketing campaign plays on this: “Now you can Airbnb more than an Airbnb.”)

The company has delivered impressive growth post-COVID, doubling revenue from ~$6B in 2021 to about $12B today, with ample runway ahead. It boasts strong economics—OI margins around 22% and rising—and a fortress balance sheet with over $10B in net cash.

The platform now has over 8 million active listings—roughly 4× Vrbo’s ~2 million—even after pruning ~450,000 lower-quality listings in recent years. Airbnb’s verb status, massive installed base, and billions of reviews—roughly two-thirds of guests leave one—combine to form one of the strongest moats in consumer tech.

This is evident in the fact that ~90% of its traffic comes through direct or unpaid channels, far higher than Vrbo, Booking, or Expedia. (For the latter two, alternative accommodations remain add-ons rather than core offerings.)

And while most engagement has shifted to the app—used on more than 1.6 billion devices each year—the chart of Google Search interest from 2011–2025 offers a crisp view of how decisively Airbnb won the market-share war.

Airbnb now operates in more than 220 countries and over 100,000 cities. Yet its top five markets—the U.S., Canada, Australia, France, and the U.K.—account for 70% of revenue, underscoring substantial international headroom.

In the U.S., its most mature geography, Airbnb now accounts for roughly one in ten nights away from home. That’s impressive—though not the existential threat to hotels many once feared. Hotels still dominate short, highly urban, or one-night stays. But for longer trips (about 70% of nights booked are for month-long stays), and/or for families and groups, Airbnb’s value proposition is hard to beat: more shared space, kitchens, outdoor areas, and privacy.

So what’s interesting here, and why now?

The international runway is clear. Less obvious, though, is that Airbnb has a plausible path to doubling its share in mature markets (which is, in my view, not priced in). Several recent developments point in that direction:

1. Improved UI and reduced user friction
Airbnb has addressed many of its most persistent pain points—quality consistency, affordability, and pricing opacity. For instance, new pricing tools help hosts price more competitively, while the shift to a single-fee model (guests no longer see separate service or cleaning fees) makes the experience cleaner and easier to compare.

2. Co-hosting unlocking supply
Many people with homes would like to host but don’t have the time; many hosts (or would-be hosts) would like to host more but don’t have the capital for another home. As Chesky put it, “We’re basically existing in a very narrow Venn diagram of people that have time to host and have a home.” Co-hosting widens that Venn diagram. Launched late last year, it has already generated more than 10 million booked nights.

3. Adding hotels to fill inventory gaps
A large share of travelers begin their search on Airbnb but defect when they can’t find a suitable home—particularly for short, urban stays. To capture that spillover demand, Airbnb has begun listing boutique and independent hotels directly on the platform. During the latest earnings call, Chesky described his pitch to New York hotel operators—where Airbnb sees millions of searches:

“We believe that the majority of people who come to New York on Airbnb would be open to booking a hotel if there wasn’t a home available. Many of these people are subsequently opening other apps and booking hotels elsewhere. So, we said: if we added hotels, gave you a best-in-class commission, beautiful custom-built product pages, and brought you a lot of demand—often high-income, young American travelers, which are some of the most appealing consumer sets—would you be interested?”

The answer, unsurprisingly, was an enthusiastic yes.

4. Building the “Amazon for services”
The most consequential opportunity may sit outside homes entirely. Suppose you needed to book a photographer, private chef or caterer, makeup artist, or on-site massage therapist—where would you start? There is no dominant discovery platform, and certainly no unified marketplace that combines:

• identity verification (Airbnb has over 200 million verified IDs—equal to or more than any other U.S. company)

• integrated payments and scheduling

• pre-screened, quality providers

• standardized profiles with searchable supply

• and, critically, trusted, platform-wide reviews

Airbnb launched Services earlier this year, alongside a relaunch of Experiences. Below are a few examples from Toronto (where I live): the tasting menu is a Service; the other two are Experiences.

Experiences were relaunched because several frictions needed to be addressed: they were hard to find, poorly merchandised, and not integrated with social media. (As you can see below, discovery will no longer be an issue.)

This placement, however, is not risk-free. Putting Services and Experiences on an equal footing with Homes—Airbnb’s highest-margin and highest-volume category—may hit conversions and could frustrate hosts. Yet management is clearly confident the long-term upside outweighs the potential trade-offs. (For what it’s worth, so am I.)

Many guests wouldn’t think to book a chef, photographer, or massage therapist when planning a trip. But when those options surface naturally in the booking flow, you’re prompting consideration when the guest already has intent. Experiences, conversely, are often explicitly searched for—so consolidating them within the same platform makes good sense.

Early results are encouraging. Roughly 10% of bookings are coming from entirely new customers, and locals are heavily engaging with Experiences. For example, in Paris—one of Airbnb’s largest markets—locals account for about 70% of Experience bookings.

Taken together, these developments are apt to both expand Airbnb’s audience and make the platform even stickier. As Chesky noted during the last earnings call:

“We believe that services, experiences, and hotels could each be multi-billion-dollar businesses.”

If that’s even directionally right, then the incremental lift to the core Homes business should likewise be measured in the billions.

Valuation: At around $115/share as of this writing, Airbnb sits well below its all-time high of ~$220/share and meaningfully under its five-year average of roughly $140/share. The stock now trades at about 21× EV/EBIT (NTM). Given the company’s growth trajectory and expected margin expansion, that strikes me as very reasonable.

Based on my work so far, I expect Airbnb to exceed earnings expectations in 2026/27 (assuming a stable macro backdrop). Consensus implies a 2027 EV/EBIT multiple of ~17×; my base case is closer to 15×. And unless I’ve misread/overlooked something material—which, as always, is possible—I’d expect the stock to work its way into the $150–$200 range.

Match Group

While it’s far more efficient to read transcripts, I try to listen to as many earnings calls as I can—because the text alone often leaves out a surprising amount of signal. Match is one of the clearest examples of why that habit matters. It also makes this pitch the hardest to capture persuasively in writing.

“We need to improve the perception of the [online dating] category. And the way to do that is to prioritize user outcomes over short-term revenue and profit. That has not been the way this company has operated historically, and that’s been to our detriment.” – Spencer Rascoff, CEO, Match Group

With more than 20 brands and roughly $3.5B in revenue, Match is the clear category leader in online dating. The business has solid economics—operating margins around 25%, with room to expand. Unlike the previous ideas, however, it does carry leverage: just under $3B in net debt, or roughly 2.3× adj. EBITDA.

Match owns two of the three major Western apps—Tinder and Hinge (with Bumble as the third). Hinge is the growth engine; Tinder, the #1 dating app by downloads in more than 100 countries, is the cash cow with some brand baggage.

As many of you know, the online-dating category has been shrinking—or at best stagnating—in recent years, though it remains the primary way people meet. The open question is whether the industry is structurally broken or simply overdue for a reset.

Only time will answer that. But one person who clearly believes it’s fixable (yes, you guessed it) is Match’s new CEO, Spencer Rascoff—who wouldn’t have taken the role if he thought otherwise.

Spencer is the co-founder and longtime CEO of Zillow Group, which, for those unfamiliar, dominates the U.S. online real-estate marketplace. Put differently: he doesn’t need the paycheque, and he wouldn’t have stepped in unless he believed he could engineer a turnaround.

I cannot overstate the contrast between Spencer and his predecessors—three CEOs since 2017. All were capable operators and, by all appearances, good people. But that doesn’t necessarily make someone a great leader, capable of inspiring thousands of employees to do their best work.

The contrast with his immediate predecessor is especially stark—so stark, in fact, that the difference calls to mind Churchill and Chamberlain in 1940. Only one could rally a nation when it mattered. (If you’ve listened to their respective earnings calls, you know that’s not hyperbole.)

Spencer joined the board in 2024 and became CEO in February 2025. His first priority was rebooting the company culture—though he described it more forcefully: “shaking the company from a slumber.”

Match, he noted, is a category roll-up—spun out of IAC—that never fully extracted the benefits of its combined scale. Zillow, by contrast, is successful in part because it did exactly that. As Spencer explained:

“When I got here 100 days ago, it was really run as 20 different companies. Each app basically as their own company with its own Head of Marketing, its own Head of Technology, its own Head of Engineering. And I’ve changed a lot of that, and I’ll continue to, because there are enormous synergies to be had by recognizing the combined power of Match Group.”

Somewhat ironically, that independence is what allowed Hinge to thrive and maintain a “very impressive and distinctive company culture, with highly engaged employees,” as Spencer noted—in sharp contrast to Tinder.

During his first quarter, Spencer reduced the workforce by 13%, including one in five managers. But he was quick to emphasize that’s not what he came to do:

“I know how to cut costs, but that’s not what gets me up in the morning. What gets me up in the morning is being a product innovator, and building cool stuff that drives user outcomes, that grows audience, that grows revenue. I’m a growth CEO.”

Recent data might make that growth comment sound out of sync with reality. (The same goes for his remark: “I’m optimistic that one day we’ll look back and see the TAM in online dating was much larger than any of us expected.”)

Yet as you can see below, that optimism isn’t baseless.

Conventional wisdom frames the online-dating model as a paradox: the better the product is at making the right matches, the faster it loses its customers—and the worse that is for revenue. Hinge, which dominates the “intentioned” dating category, is a useful counterexample. Its success comes from a precise understanding of the job it exists to solve: to be the last dating app you’ll ever need. It is, literally, “designed to be deleted.”

That mindset—prioritizing user outcomes—is now being pushed across Tinder and the rest of the portfolio. I won’t unpack all the recent developments or new features here—several of which I’m optimistic about—but it’s clear the company is operating with renewed urgency and focus. As Spencer put it:

“By many measures, such as code commits or experiments that we have in-flight, we’re operating at about twice the pace as we were just a couple of quarters ago.”

I’m optimistic the products will continue to improve and resonate more with users, and that the perception of the category will gradually recover. As the category leader, Match has outsized influence in making that happen.

From a competitive standpoint, Match also holds the best hand. Its primary challenger, Bumble—also in turnaround mode—is roughly a quarter of the size and, in my view, far less likely to successfully land the plane. The more interesting threat comes from social-media giants like Facebook, which launched its dating product in 2019.

Its main advantage is simple: it’s free. It now has 21 million daily users, the vast majority over 30. (For reference, Hinge has 15 million.) It’s a trend worth monitoring, but it may sound more ominous than it is—especially given that dating is, and always has been, a multi-app category, as shown below.

Taken together, I believe extrapolating recent trends here would be unwise—especially given the scale of the internal changes Match has undergone. Turnarounds, especially at scale, don’t happen overnight. The chart below, borrowed from Rebound Capital, is a useful reminder of why projecting the recent past forward without accounting for shifting dynamics can lead you astray:

I’m not suggesting the same outcome awaits Match—only that a meaningful improvement from here seems more likely than what the current valuation implies.

Valuation. Match currently trades at roughly 11× EV/EBIT (NTM). If the company can maintain momentum at Hinge—which still has substantial international runway—and simply stop the bleeding at Tinder, I’d expect the multiple to re-rate closer to 13–14× (roughly the 3-year average). If Tinder can return to even modest growth, I think 15× or higher is plausible.

In a nutshell, the potential upside significantly outweighs the downside over the next 12–24 months, assuming no fundamental changes to the thesis.

As always, thanks for reading. If you enjoyed this post, a quick thumbs-up—or a share with your network—goes a long way.

Share

Disclosure: I or accounts under my control hold positions in the stocks discussed. This is not investment advice. Readers are encouraged to conduct their own due diligence. Some quotes were lightly edited for clarity and relevance.

“People are crazy and emotional. They buy and sell things in an emotional way, not in a logical way, and that’s the only reason why we have any opportunity … So, if you have a way to value businesses that’s disciplined and makes sense, you should be able to take advantage of other people’s emotions.”
- Joel Greenblatt, in Richer, Wiser, Happier by William Green

It’s remarkable how quickly sentiment can flip.

Less than a year ago, in December 2024, Jim Cramer told Lululemon CEO Calvin McDonald that he thought the company was “the best story out there in retail right now.” He went on to elaborate:

“Everyone I talk to says China is a disaster; there’s no selling. Everybody tells me there’s no such thing as new; it’s all old. Everybody tells me it’s the most promotional Christmas in history; you have no promotions, you have newness, and you’re doing well in China. What are you doing right that the others are doing wrong?”

If Cramer’s starry-eyed tone didn’t already give it away, Lululemon was—until very recently—the crown jewel of the apparel category. Over the past decade, the company compounded revenue at roughly ~19% annually, despite already being a multibillion-dollar business at the outset. More impressive still, it did so while sustaining ~20.5% operating margins—more than twice the peer-group average of ~8% (Nike, Adidas, Under Armour, Puma). Businesses don’t produce numbers like that at scale without a moat of real depth.

The prevailing view today, however, is that the moat has drained. Even after a roughly 65% drawdown from its peak less than two years ago, short interest remains at a multi-year high of 6.2%. That might make sense if Lululemon were truly “running out of gas and it’s not fixable,” as one analyst claimed. The evidence, though, points elsewhere.

The bear case centers on falling U.S. comps, heavier discounting, and “hot” challenger brands nibbling at the edges. These are fair concerns—but context matters. The recent declines (–2% and –4% in Q1 and Q2, versus 0% and –3% a year earlier) come on the heels of a decade of near-uninterrupted hypergrowth. They also appear more symptomatic of internal complacency than external defeat—a lapse management has acknowledged.

“I now believe we have let our product life cycles run too long,” Calvin McDonald admitted. “We have become too predictable within our casual offerings and missed opportunities to create new trends.”

Given the company’s extraordinary run, it’s hard to fault them for sticking with a winning formula a little too long. (Most of us would have done the same.)

These declines also occurred against a softer category backdrop. “Consumers are spending less on apparel overall, spending less in performance activewear, and are being more selective in their purchases,” McDonald explained. Yet even then, Lululemon continued to gain share within premium activewear in both Q1 and Q2—just as it did in Q2 and Q3 of 2024, when comps were also slightly negative.

So while there’s clearly work to do, the notion that Lululemon is “broken” doesn’t hold up. Even a director at athleisure “disruptor” Vuori dismissed the claim outright: “I think the argument that they are getting stale is not a valid one.”1 An executive at competitor Fabletics agreed: “As a millennial myself, I don’t think Lululemon is getting stale.”2

Timing matters, too. Much of the current assortment still reflects the tail end of the former Chief Product Officer’s work, given an 18-month design cycle. The first full season shaped by the new creative director—and by Lululemon’s revamped design and merchandising model—won’t arrive until Spring 2026.

And by McDonald’s account, this design team—comprised of both seasoned and newly hired design leads—is “the best that we’ve ever had in the history of this organization.”3 If that’s even directionally true, and my research strongly suggests it is, then declaring the brand finished seems premature at best.

The question, then, isn’t whether Lululemon can course-correct, but whether it will. And if you’ve read this far, you can probably guess where I land on that question. Even if the recovery takes longer than investors might like, today’s valuation offers ample margin of safety for those willing to wait.

That margin already prices in the usual anxieties: fiercer competition, “dupes,” shifting style trends, Gen-Z fickleness, tariffs, and macro uncertainty. We’ll return to each in due course.

But before we dive in, a little historical perspective:

• Jim Cramer, September 2016: “Should we be worried about Lululemon here, given the recent heinous action in the stock and the constant worry of the analyst community that athleisure has peaked?”

• Professor Aswath Damodaran, April 2019: “I’d bet on Levi’s over something like a Lululemon, which will be a distant memory. The problem with brands today is that they build up really fast and don’t have roots. And millennial consumers don’t have loyalty.” (Notably, Damodaran said this despite Lululemon already being 20 years old—with a customer base among the most loyal in the industry.)

Investor anxieties, in other words, are rarely a reliable predictor of what happens next.

Thesis Overview

1. Highly profitable. Lululemon’s operating margins remain exceptional—roughly 16 points higher than peers—reflecting a business model that converts growth into profit with unusual consistency. (The company is also strikingly frugal for one so profitable—a rare and underappreciated trait, and not a coincidence.)

2. Store productivity. Despite industry-leading digital penetration among scaled players (~44%), stores generate nearly $1,600 per square foot—about four times the mall average.

3. Loyal, resilient customer base. Lululemon’s customers are exceptionally sticky. “Ninety-two percent of our loyal guests continue to shop with us year in and year out,” McDonald said in 2019. “I’ve never seen another [apparel] brand with that type of engagement.” They also skew more affluent and are therefore less exposed to economic shocks. Even in the depths of the 2009 financial crisis, operating margins held at 19%.

4. Marketing leverage. Lululemon’s stores double as brand billboards. Combined with strong brand equity, this allows the company to maintain marketing spend well below peers—a structural advantage that compounds returns and frees up capital for reinvestment elsewhere.

5. R&D edge and the flywheel. Product innovation is one of those reinvestment areas, where Lululemon invests more heavily than most. Continuous improvements sustain premium pricing and repeat purchases, which in turn fund further innovation—a self-reinforcing loop that strengthens the brand over time.

6. Extraordinary economics. These forces together produce exceptional returns on tangible capital, far above industry norms.

7. Growth runway. Only about 25% of sales come from outside North America—versus roughly 55% for comps—leaving significant international whitespace. Lululemon also remains underpenetrated in men’s, which represents just 24% of sales, well below the 50%+ typical for activewear brands (Nike is closer to 70%). Further gains are likely from co-located store expansions, which have historically driven productivity increases well in excess of the associated growth in square footage.

8. Financial strength and downside support. The company holds roughly $2 billion in cash, carries no debt, and continues to repurchase shares aggressively—roughly $4.5 billion since 2019—providing considerable downside protection. (The scale and prices of those buybacks have raised concerns, which we’ll address later.)

9. Valuation gap. Despite superior economics, vast international whitespace, and robust balance sheet, Lululemon trades at a steep discount to peers—a disconnect unlikely to last.

Lululemon also trades at a steep discount to its closest private-market competitors, Alo Yoga and Vuori, each valued in the $5–10 billion range, or roughly 5.5× sales. By comparison, Lululemon trades at just ~1.8× forward sales.

Taken together, Lululemon’s historically washed-out valuation, coupled with ample capacity for additional repurchases, provides significant downside support at current prices.

Disclosure: I and/or accounts under my control hold a position in Lululemon’s stock. This is not investment advice. Readers are encouraged to conduct their own due diligence. All figures are in USD unless otherwise noted. Some quotes may be paraphrased or lightly edited for clarity and relevance.

Brief History

“Demand and growth have been steadily up and to the right for really two decades now, and so not to say Lulu is invincible, but I have been impressed with how they’ve been able to prove their staying power over time. And I think the moment I realized that Lulu was here to stay was probably when I found myself happily spending a couple hundred bucks on their men’s clothing with many of my friends doing the same, after thinking it was a brand only for women for about a decade.”
- Shawn O’Malley, The Intrinsic Value Podcast

Founded in Vancouver in 1998 by apparel entrepreneur Chip Wilson—an eccentric and occasionally controversial figure who left the company in 2015 but remains a major shareholder (~8%)—Lululemon began as a women’s yoga line, a notable inversion in an industry long geared toward men.

Until then, leggings were simply something you wore to work out. Wilson’s billion-dollar insight was to make them look good, too. By merging form and function, Lululemon’s leggings were technical enough for the studio yet stylish enough for the street—helping ignite the trend of what would soon be called athleisure.

The company broke convention in other ways as well. It sold exclusively through its own stores, allowing it to “own” the customer relationship and cultivate an aura of exclusivity. Its refusal to discount—almost heretical in retail—only reinforced that mystique, aided by an unusually generous return policy.

Then there were the bags. The brand’s red-and-white reusable shopping totes—“the first grocery bag, at least in New York, for sure,” as one former executive recalled—became ubiquitous, doubling as both status symbols and walking advertisements.

Together, these elements gave Lululemon something rare in fashion: genuine cross-generational appeal. Ordinarily, being cool to soccer moms and cool for their daughters is mutually exclusive—but Lululemon managed both.

Over time, the company expanded well beyond women’s yoga wear—broadening its athletic range (tennis, golf), adding lifestyle apparel, and moving into men’s clothing (2014), accessories, and footwear. It would, of course, also make its mark internationally.

Company Overview and Business Model

Fast forward to today: Lululemon operates roughly 800 stores—about two-thirds in North America—generates around $11 billion in annual sales, and counts 30 million members. Its product mix skews about 60% performance and 40% casual, anchored by durable core franchises; around 40% of styles sell year-round.

About 90% of revenue comes from company-owned stores and e-commerce. The remainder stems from temporary pop-ups, select wholesale accounts (including university retailers), and outlets for slower-moving or prior-season inventory—52 globally, mostly in the Americas—plus a handful of international licensees.

Over the five years ending 2024, Lululemon added 276 net new stores, roughly two-thirds of them outside North America.

Stores are concentrated in premium retail corridors—street-front locations, tier-one malls, and other high-traffic zones. Designs can vary significantly by region, but the brand identity is unmistakable.

Average store size has expanded about 50% over the past decade—from 3,000 to 4,500 square feet—through a mix of new builds and co-located remodels. While sales per square foot have remained roughly flat in nominal terms, e-commerce penetration has more than doubled—from ~21% of sales in 2016 to ~44% today (two-year average). That places Lululemon among the highest digital-mix apparel brands at scale.

It’s also one of the most profitable, with operating margins in the low 40s. That profitability reflects three structural advantages:

1. Unusually high customer retention

2. long-lived core franchises (Align, ABC, Scuba)

3. Highly consistent fits—once a customer knows they’re a size 30 in ABC, they can reorder online with confidence

Fit consistency, in turn, drives cross-category purchasing, while in-store returns help minimize reverse-logistics drag on margins. Lululemon’s near-pure direct-to-consumer model carries a structurally lighter net-working-capital load than wholesale-heavy peers, offset by a heavier fixed-asset base tied to its owned-store footprint. Taken together, the differences tend to net out.

Style

Lulu describes its product as “high performance, high style”—always in that order. “It all starts at the fabric level for us,” McDonald explained. “We obsess over the performance and feel of fabrics.” We’ll return to performance shortly; first, the aesthetic.

You might think of “high style” as a kind of refined minimalism—tailored fits and a quietly polished look. While more adventurous competitors might lean into bright neons or bold tri-color patterns, Lululemon’s palettes stay simple: cool or warm (not hot or cold), and rarely both. The payoff is versatility—pieces that move easily from one activity to another.

Internationally, Lululemon may adjust its assortment to local tastes, tweaking color and fit on a portion of its global line to better resonate with regional guests. In China, for instance—its most localized market—these adaptations account for roughly a quarter of sales, with another 10–15 percent designed and developed locally.

Over the past year, new styles—“newness” in industry parlance—have made up about 23% of the assortment globally. (Newness can also refer to new colors or prints applied to existing silhouettes.)

“The Bread and Butter”

“I think the fabric; that’s the bread and butter. Working at Under Armour, we had a lot of consumer insights that our fabric wasn’t where it needed to be. We used to look at Lulu and Nike and say like, “How can we replicate this?” With Lulu’s fabric, it was just impossible.”4 - Former Senior Manager at Under Armour

Lululemon sits at the frontier of technical apparel innovation—and for good reason. The company employs hundreds of specialists devoted to product development through its Whitespace Lab and dedicated R&D facilities. To put that in perspective, a director at Gap-owned Athleta explained:

“Athleta had a couple of raw materials people, but they were servicing the entire business, whereas Lulu has PhD-level scientists working on fabric innovation. That’s going to make a difference if you’re bringing your own research to the table with the mills and actually developing proprietary fabrics.”5

As implied above, most of the scientific heavy lifting in apparel happens at the mill level, not within the brands themselves. Brands typically request adjustments to an existing mill construction—tweaks to fiber ratios, yarn counts, gauges, or finishes—rather than commissioning a fabric from scratch. Any exclusivity they receive is usually brief, often limited to a single season.

Lululemon’s scale, deep supplier relationships, and collaborative development model extend that advantage dramatically—often securing exclusivity windows of 18 to 24 months or more. In many cases, it goes further still, owning the patents outright.6

Read more