Walmart

A month ago we wrote about Costco and concluded it was a wonderful business at an unwonderful price. Walmart is the natural companion, because it sits in the same aisle running the opposite model — and arrives at a strikingly similar problem.

Walmart is the largest retailer in the history of the world. It did $713 billion of revenue in the year to January 2026, serves something like 72% of American households in groceries alone, and operates a logistics network so effective that it has been studied for forty years as the definitive example of scale advantage. Its method is simple enough to state in one line: buy in colossal volume, run costs lower than anyone can match, pass most of the saving to the customer, and make a thin profit on an enormous number of transactions. How thin? Walmart keeps about four cents of operating profit on every dollar it takes. That is the whole model — a business of astonishing size and almost no margin for error.

And here is the fact that shapes this entire report, which we found in our own decade of data rather than in any research note. Over the ten years to January 2026, Walmart's revenue grew 47% — from $486 billion to $713 billion. Its operating income over the same period grew just 31%, from $22.8 billion to $29.8 billion. Which means the operating margin actually fell, from 4.68% to 4.18%. In that same window, the stock re-rated from roughly fifteen times earnings to forty. Something has to explain that gap, and the explanation the market has settled on — that Walmart is becoming a high-margin advertising and membership business — is real, important, and, so far, not visible in the operating margin at all.

"This is my 33rd year at the company, and I can't remember a year where competitive pricing isn't always top of mind." — John Furner, who became Chief Executive on 1 February 2026

What has genuinely changed is the shape of Walmart's profit. Alongside the merchandise business sits a fast-growing collection of high-margin income: advertising (roughly $6.4 billion globally in FY2026, growing 46%), membership fees from Walmart+ and Sam's Club, and marketplace and fulfilment fees from third-party sellers. Chief Financial Officer John David Rainey disclosed that advertising and membership income together now account for about one-third of Walmart's operating income — a threshold that simply did not exist a year earlier. Since core US retail earns about four cents on the dollar, and an advertising dollar carries perhaps seventy or eighty cents of incremental margin, one dollar of advertising is worth roughly fifteen to twenty dollars of extra merchandise sales. That arithmetic is the entire bull case, and it is a good one. The question this report has to answer is why, if it is working, ten years of it have not yet lifted the margin.

Walmart's history matters here for an unusual reason: it contains one of the best-documented mistakes Warren Buffett ever admitted to, and then a second one that he has never really discussed.

The two Buffett errors are worth holding side by side, because they teach opposite lessons and this report needs both. The first — refusing to pay up a few cents for a wonderful business, and forgoing billions — is the classic warning against excessive price-sensitivity, and it is the argument every Walmart bull will make to you today. The second — selling a wonderful business because a new competitor frightened him — is the warning against abandoning a durable franchise on a narrative. Both cut in Walmart's favour. But notice what neither error involved: paying forty times earnings. Buffett's thumb-suck happened at a price in the low twenties as a multiple; his exit at a mid-teens one. Applying "don't be too cheap about great businesses" to a stock at 40× is a misuse of the lesson, and Part X is where we test whether the price can be defended on its own terms rather than by analogy.

The machine, in five steps:

Walmart is exactly the kind of business Buffett spent a career describing: a simple, essential, unglamorous operation with an economic advantage anyone can explain in a sentence. There is no technological risk to speak of, no patent cliff, no regulatory body setting its prices, no scientific outcome to underwrite. People will buy groceries in ten years, and Walmart will very probably sell them more of those groceries than anyone else in America. On the business test, this passes as comfortably as anything we cover. Which is precisely why the analysis has to be conducted almost entirely on price and on the quality of the margin — because when a business is this knowable, the only two ways to lose money are to overpay for it or to misjudge whether its economics are improving or decaying. Both of those questions are live here.

Why the Sam's Club discovery matters — and why it should not be oversold. When we analyzed Costco we established the central mechanic: the membership fee is essentially pure profit, collected upfront, renewed by ~92% of members, and it is what permits merchandise to be sold at a deliberately capped markup. Sam's Club runs that same machine, and on the FY2026 numbers it runs it in an even purer form — the fee is larger than the whole segment's operating profit. On roughly $93 billion of revenue, Sam's Club is a genuinely large club business that would be a major public company if it were spun out, and the market does not price it separately at all.

Two honest caveats before anyone gets carried away. First, that $2.525 billion is a blended "membership and other income" line which includes breakage on unredeemed Sam's Cash rewards; the underlying fee growth is mid-single-digit, not the headline 8.7%. Second, and more tellingly, Walmart does not disclose a renewal rate, a Walmart+ member count, or a clean membership-fee figure — where Costco reports paid households, cardholders, executive-tier members and renewal rates every quarter. That disclosure gap is not an accident; it is a signal about which company thinks its membership economics are the point of the business. If Walmart+ disappeared tomorrow, Walmart would lose perhaps 1–2% of operating income. If Costco's fee disappeared, more than half its profit would go with it and the entire pricing model would have to be rebuilt. The models differ in kind, not degree — and that difference is a large part of why Costco has historically carried the richer multiple.

Walmart's moat is the textbook case of scale as a competitive advantage, and it is genuine. But this is the section where an honest analysis has to report data the bulls tend not to mention.

What is strong. Purchasing power at $713 billion of volume is unmatchable — Walmart buys cheaper than anyone and can therefore sell cheaper than anyone, a self-reinforcing loop Sam Walton set spinning sixty years ago. The distribution network, now paired with 4,600+ US stores acting as forward-deployed depots, lets Walmart reach ~60% of the US population within 30 minutes; fast-delivery sales grew 50% last quarter. Groceries — 59% of Walmart US — make the whole thing defensive, because people buy food in recessions. And household reach hit a record 72%. There is no plausible world in which this business is disrupted quickly.

What is eroding, and it is not trivial. Walmart's share of US groceries has fallen three consecutive years — 20.4% to 20.0% to 19.9% — while Costco (7.6% → 8.2%) and Amazon with Whole Foods (6.4% → 8.0%) both gained; Amazon has now essentially drawn level with Costco in consumer packaged goods, having added 1.6 points in two years. Walmart's share of US physical retail visits has slipped from 10.2% in 2019 to 9.7%. And across a full decade, the clearest possible evidence sits in our own table: gross margin drifted DOWN from 25.6% to 24.9%, and operating margin from 4.68% to 4.18%. A widening moat normally shows up as pricing power and expanding margins. Walmart's shows up as volume and reach without either.

The new moat, and the honest verdict. The genuinely additive development is the data flywheel: Walmart Connect advertising, the VIZIO acquisition (which turned Walmart into an owner of connected-TV supply with closed-loop attribution against real purchase data), the marketplace, and Walmart+. This is a real, high-margin, defensible asset built on something only Walmart has — knowledge of what 250 million weekly customers actually buy. It is the reason the stock re-rated and it deserves respect. But scale it honestly: Amazon's advertising business is roughly eleven times larger, and Amazon's single year of incremental ad revenue was about twice the size of Walmart's entire ad business. Rainey himself concedes Walmart has "long ways to go to get in the neighborhood of some of the best-in-class competitors." I score the moat an 8 — one of the widest in existence on reach and cost, marked down because the metrics that measure a widening moat (share, margin, pricing power) have all been going the wrong way.

Everything about owning Walmart at forty times earnings rests on a single proposition: that this is no longer merely a retailer, but a retailer bolted to a high-margin advertising, membership and marketplace business that will drag the whole company's profitability upward. The evidence for that proposition is genuinely strong. The evidence against it is that it has not happened yet.

Our read is that both columns are true, and reconciling them is the key to the whole company. The most likely explanation is this: the high-margin income is real and growing fast, but it is being consumed — deliberately — by price investment and cost inflation rather than allowed to fall to the bottom line. Rainey said it almost explicitly when asked where tariff-refund money would go: "the single best return that we can have on a dollar of capital right now is to invest in the customer and invest in price." Walmart is running ~7,200 active rollbacks against a historical baseline of 5,000–5,500, and its own like-for-like inflation ran around 1% while food-at-home CPI ran 2.7%. In other words, management is taking the advertising windfall and handing it to shoppers to defend share and traffic.

That is a perfectly rational strategy — arguably the correct one for a business whose entire moat is being the cheapest — but it has a sharp implication for an investor. If the ad profits are structurally recycled into lower prices, they are not margin expansion; they are the cost of standing still. They show up as comparable sales, household penetration and traffic — all of which are indeed healthy — rather than as earnings power. And the market has priced them as though they were earnings power. That is the gap. There is a genuine bull case that the recycling ends once the price war cools and the automation capex peaks (Furner says supply-chain investment "probably peaks this year and next year"), at which point a decade of accumulated high-margin income arrives at once. We simply note that this has been the promise for several years, the operating margin has not moved, and the shares now cost forty times earnings on the strength of it.

Set the three side by side and something uncomfortable emerges for the defensive-retail investor. Amazon does slightly more revenue than Walmart, grows two and a half times faster, earns nearly three times the operating margin — and trades roughly ten points of P/E cheaper. Amazon's worst segment (International, at 2.9%) is about equal to Costco's entire company; its North American retail segment (6.9%) comfortably beats Walmart's whole company. We are not making an Amazon recommendation here — our Amazon report did that, and its ~$200 billion capex commitment is a real and unproven bet. The point is narrower and it is about Walmart: defensive retail has re-rated to growth multiples while the actual growth company in the sector is the cheapest of the three. If you are paying up for quality and durability, you should at least know that the market is charging you a premium for the slowest-growing, thinnest-margin business of the trio.

Against Costco specifically, the honest scorecard is mixed and the two are less substitutable than people assume — a membership-gated bulk run and a weekly stock-up are different errands, and affluent households increasingly do both. Costco wins decisively on returns (a 38.4% ROIC against Walmart's 11.9% is not a close contest), on growth, and on share. Walmart wins on breadth, on the sheer indispensability of 72% household reach, and on the optionality of an advertising business Costco has explicitly declined to build — CFO Gary Millerchip has said Costco routes 80–90% of retail-media value back into member pricing rather than booking it as revenue. The genuine competitive threat to Walmart's grocery share, meanwhile, is not Costco at all; it is Amazon, which has added 1.6 points of CPG share in two years, and Aldi, which is committing $9 billion and 180+ new stores a year through 2028.

Two things deserve flagging that a promotional write-up would skip. First, the automation targets were missed. At its 2023 investor day Walmart guided that by the end of FY2026 roughly 65% of stores would be serviced by automation and about 55% of fulfilment-centre volume automated. The actual figures, given by Rainey on the very call marking that target date, were approximately 60% and 50%. Furner's own characterisation a quarter later was blunter still: "we're about halfway there. So, we have more to do." This is not a scandal — these are enormously complex programmes — but it matters because the automation payoff is a load-bearing part of the future-margin story, and it is arriving later than promised.

Second, the capital intensity is real and it is why free cash flow looks so poor. Walmart's free cash flow per share is just $1.58 against earnings per share of $2.89 — cash is barely half of reported profit, which is why the stock trades at an eye-watering 72× free cash flow versus 40× earnings. As with Eli Lilly and the AI hyperscalers, this is a capex distortion rather than an earnings-quality problem, and Furner says supply-chain investment "probably peaks this year and next year." If that holds, free cash flow converges toward earnings and the picture improves markedly. It is, however, another instance of the same pattern running through this whole report: the good news is always in the future. I score management a 7 — a genuinely strong operating record and an orderly succession, marked down for missed automation targets and a margin that a decade of strategy has not moved.

This table is the least flattering thing in the report and it is drawn entirely from Walmart's own audited numbers. Over ten years, a company universally described as executing a brilliant transformation grew its revenue by 47% and its operating profit by only 31%. Both its gross margin and its operating margin went down. Earnings per share did better than either — up 87% — but a meaningful slice of that came from retiring 14% of the shares rather than from the business earning more on each dollar of sales. This is the profile of a company winning volume and reach while its unit economics quietly deteriorate.

To be scrupulously fair, the recent trend is better than the decade: FY2026 operating income of $29.8 billion on 4.7% revenue growth, net income up to $21.9 billion, EPS up 13%, e-commerce finally at scale and profitable, and a genuinely strong balance sheet (Altman-Z of 6.15, net debt around 1.3× EBITDA). The last three years are the best stretch of this whole run, and the mix shift is plainly contributing. But an investor paying forty times earnings is not buying the last three years — they are buying the next ten, and the ten-year record in this table is the single most relevant evidence available about what this business does with time. It grows. It does not get more profitable. Whether that changes is the question, and the price assumes the answer is yes.

Here is where Walmart becomes genuinely difficult, because — unusually for an expensive stock on our board — the models actually like it. The discounted-cash-flow model says $145.70, some 27.5% above the price. Analysts average $140.26, some 22.8% above, and remarkably even the lowest target on the Street ($120) sits above today's price. Nobody covering this stock thinks it goes down. That is worth taking seriously, and it is the opposite of what we found at Lilly and UnitedHealth.

Now set it against what you are actually paying.

We take the bullish models seriously, and we should say why they get there: a DCF rewards Walmart's extraordinary predictability, and if you assume capex normalises (freeing that suppressed free cash flow) and the advertising mix finally lifts margins, $145 is a perfectly reasonable output. That is a coherent case and it may well prove right.

But weigh it against the single most damning arithmetic in this report. Consensus has Walmart earning $4.05 per share in FY2031 — five years from now, after five more years of e-commerce growth, advertising growth, automation and share buybacks. At today's price you would be paying 28.2 times those 2031 earnings, which is still above the company's own ten-year median multiple of ~30.5× — barely below it — and roughly double what Walmart traded at for the whole of the 2010s. Put plainly: on the market's own forecasts, five years of growth barely gets you back to a normal multiple. Every dollar of value creation between now and 2031 has already been paid for, in advance, at today's price. Compare that to Eli Lilly at ~18× its 2030 estimate, or Amazon at ~29.6× trailing with three times the margin and 12% growth.

Add the context that the market has already shown you how this unwinds. On 21 May 2026 Walmart beat on both revenue and earnings — and the stock fell 7.1%, its worst day of the year, purely on cautious guidance; it dropped about 12% over that month from a high of $135.16. That is the signature of a stock priced for perfection: good results are not enough, because good results are already in the price. I score valuation a 3 — the same mark we gave Costco, for the same reason. The buy zone opens toward $95, near the 52-week low and around 29× the FY2028 estimate, and we will say honestly that even that is not cheap by this company's own history. Walmart has not been genuinely inexpensive since 2020, and an investor waiting for 20× may wait forever. But paying 40× for 4% margins and 4.7% growth is a decision that requires the future to be flawless.