Why it's the best retailer alive · and why it's never cheap
Charlie Munger, who sat on this company's board for a quarter-century and owned it in his personal account until the day he died, said of it: "I love everything about Costco. I'm a total addict, and I'm never going to sell a share." He also said, in the same breath, across many years: "The trouble with Costco is it's 40 times earnings." Both statements are true, and the space between them is the entire investment case. This report is not about whether Costco is a wonderful business — it obviously, almost boringly, is. It is about the older and harder question Munger spent his life on: what do you do when a wonderful business is never offered at a wonderful price?
Here is the trick that makes Costco unlike any other retailer. It sells ~$270 billion of merchandise a year at a deliberately capped markup — a gross margin of just ~12.8%, roughly half of Walmart's — running the actual retailing close to breakeven on purpose. Its profit comes from somewhere else entirely: the annual membership fee. Members paid $5.3 billion in fees last year, and that single line is ~51% of all operating income and the majority of net profit. In other words, Costco is not a store that also sells memberships; it is a subscription business that runs a giant store at cost to justify the subscription. The merchandise is the bait; the annual fee, renewed by 92% of members like clockwork, is the catch. Understand that inversion and you understand why this is the most durable retail model ever built — and why the market has never, not once in decades, let you buy it cheaply.
"I'm not saying I'm buying Costco at this price. But I'm certainly not selling any." — Charlie Munger, Daily Journal meeting, 2022
That sentence is the most useful thing ever said about this stock, and it is, in the end, our verdict too. Munger — the man who taught Buffett that a great business at a fair price beats a fair business at a great price — looked at Costco at forty times earnings and could neither bring himself to sell it nor to call it a bargain. He held it because the compounding was real and the quality was permanent; he wouldn't add because the price offered no margin of safety. Today, at ~48 times earnings, Costco is priced above the level Munger flagged as "the trouble." An honest analysis has to hold both truths at once, and resist the temptation — strong on a business this admirable — to let love for the company become an argument about the price.
| Founded | 1983 (Seattle) by Jim Sinegal & Jeff Brotman · descended from Sol Price's Price Club · IPO 1985 |
| Sector / Industry | Consumer Defensive · Warehouse Club (economically: a subscription business) |
| CEO | Ron Vachris (since Jan 2024; started as a forklift driver in 1982) · CFO Gary Millerchip |
| Makes money from | Membership fees (~51% of op income) + a near-breakeven merchandise machine |
| Revenue (FY2025, ended Aug) | $275.2 B · net income $8.1 B · membership fees $5.3 B |
| Market capitalisation | ~$423 B · 82.9M paid households · ~148M cardholders |
Sol Price's idea, and the culture that became the moat
Costco's history matters because its culture is its moat, and the culture was hand-built by men who are named, quotable, and gone — which raises the one real question about its future.
| Year | Milestone |
|---|---|
| 1954–76 | Sol Price — a San Diego lawyer, not a retailer — invents the warehouse club (FedMart 1954, Price Club 1976). His insight: charge a membership fee, which locks in loyalty AND eliminates the advertising budget, a permanent structural cost edge. He calls himself a 'low-margin retailer, not a discounter.' |
| 1983 | Jim Sinegal — who unloaded mattresses for Sol Price at $1.25/hour in 1954 and 'learned everything, everything I know' from him — co-founds Costco in Seattle with Jeff Brotman, at an 8–9% markup. |
| 1993 | The student acquires the teacher: Costco merges with Sol Price's ailing Price Club. By 1997 every Price Club is rebranded Costco — the man who unloaded Price's mattresses had erased Price's brand from Price's own stores. |
| 1997 | Buffett declines Sinegal's offer of a board seat and sends Charlie Munger instead. Munger stays a director for 26 years, until his death, and never sells a share. |
| 2012 | Sinegal (famous $350k salary: 'if I make 12 times the floor worker, that's fair') retires; Craig Jelinek takes over and holds the line — including refusing to raise the $1.50 hot dog. |
| 2023 | Munger dies at 99, weeks short of 100, having eaten $1.50 hot dogs to the end. Costco announces his passing inside its monthly sales release — the most Costco thing imaginable. |
| 2024–26 | Ron Vachris — who started as a 17-year-old forklift driver in 1982 and has no college degree — becomes CEO. The bloodline that built the culture still runs it; the first fee increase since 2017 (Sept 2024) lifts membership income. |
Two lessons for an owner. First — the reassuring one — the culture is deep, self-replicating, and internally promoted: a forklift driver runs the company, warehouse managers are grown not hired, employees are paid the highest wages in retail (~$32/hour average) and stay (turnover ~6% versus 60–70% industry-wide). That culture is the source of the moat, not a decoration on it. Second — the sobering one — within roughly two years Costco lost Munger, its 40-year CFO, and a former CEO from its board, and Munger's board seat was not backfilled with another value-investor conscience. The institution now rests unusually heavily on one man, Vachris, and on the durability of habits set by people who are no longer in the room. The business is built to outlast them. That is the bet.
As knowable as a business gets
The machine, in five steps:
This is the rare business where the ten-year future is almost embarrassingly predictable. People will still need groceries, bulk staples and cheap gasoline; they will still pay a small fee to buy them at the lowest price; the fee will still renew at ~90%; the assortment will still be too small to shop anywhere else for the same trust. The circle of competence here is not the question — a ten-year-old grasps a warehouse club that charges to get in. The only genuine uncertainty is what a share of it is worth, and that uncertainty is not about the company at all. It is about the price the market insists on charging for certainty.
The fee is the profit · the store is the moat
The economics are the most elegant on our entire board, and worth stating precisely (FY2025):
The structural advantage that never shows up in the P&L headline. Because Costco stocks fewer than 4,000 items — against 120,000+ at a supercenter — it buys each one in colossal volume, which wins it the best price from suppliers, which becomes the lowest price on the shelf, which renews the membership, which funds the profit. It is a flywheel, and it spins tighter every year. And there is a beautiful piece of financial engineering hiding in the balance sheet: Costco's accounts payable ($19.8B) exceed its entire inventory ($18.1B) — it sells the goods, on average, before it has paid its suppliers for them. Its members' cash and its suppliers' patience finance the whole operation, which is why Costco runs on negative net debt — it holds more cash than it owes — while still opening warehouses and paying special dividends. This is what a moat looks like in the cash-conversion cycle: the suppliers stock the shelves for free.
The widest in retail — and, unusually, widening
Costco's moat is not one wall but a self-reinforcing loop, and — rarest of all in retail, the industry Buffett most distrusts — it is widening, not eroding. Four springs feed it:
1 · The membership flywheel. 82.9 million paid households renewing at ~92% is a synthetic bond — recurring, high-margin, and re-priceable every few years. Once a household is a member, every purchase defaults to Costco, and the switching cost is the sunk annual fee plus the habit. This is the core.
2 · Scale and SKU discipline. Fewer than 4,000 items, each bought in enormous volume, gives Costco supplier terms and per-unit costs no broadline retailer can match. Walmart, five times Costco's SKU count, cannot concentrate its buying the same way. The gap compounds; it cannot be closed by spending.
3 · The price-trust brand. The capped markup is a promise, kept so consistently that members stop comparison-shopping. The proof is the gold bars: Costco began selling one-ounce bullion — a perfectly fungible commodity with zero merchandising edge — and it sells out instantly, rationed to two per member per day, because people trust Costco's price on gold more than a bullion dealer's. That trust, applied to a commodity, is the moat made visible.
4 · The culture. Highest wages in retail, ~6% turnover, internal promotion, and a fanatical cost discipline that treats every penny of markup as sacred. A competitor can copy the format; it cannot copy 40 years of institutional habit.
Where could it crack? Not at the business level — the live bear case is entirely about price, not quality (Part X). The only operational watch-items are the slow deceleration of paid-member growth (5.2% → 4.1% over the past year) and an international build-out that remains rhetoric ahead of footprint. Neither threatens the flywheel. I score the moat a 9 — among the two or three widest we have examined, and one of the very few still getting wider.
The only debate that matters here is the price
For almost every company on our shelf, the central question is about the business — will the moat hold, will the turnaround work, will the drug survive the patent cliff. For Costco, uniquely, the business question is settled, and the central question is the one thing value investors are supposed to be best at and most often get wrong: is a wonderful business worth any price? Munger's answer was subtle, and it is worth sitting with, because it is the whole verdict.
| The bear (it's the price) | The bull (it's the quality) |
|---|---|
| ~48× earnings for ~10% growth — a PEG near 3.8; the multiple, not the moat, sets your return from here | A wonderful business compounds you out of a rich price — hold long enough and the earnings grow into the multiple |
| A 21% crash didn't make it cheap — it bottomed at ~45× last December; the stock has never been a bargain in living memory | It has 'always been expensive' and always won — buyers who waited for a cheap Costco have waited 20 years and missed the compounding |
| The DCF says $319 (−67%); a fund (Bell Global) exited in 2026 purely on 'the risk-reward at ~45× is no longer compelling' | Analysts see $1,102 (+15%), and even their lowest target ($1,000) is above today's price — the Street won't call it dear |
| Reverting to its own 10-yr median (~38×) is a ~18% haircut — 'growing into a valuation' means flat for two years, which is exactly what 2025 delivered | Munger, who saw the 40× 'trouble,' still refused to sell — the compounding + the special dividends reward patience through the flat stretches |
Our read is Munger's, and it is not a dodge — it is the disciplined answer. The business deserves a premium; this premium deserves caution. The honest history is that buying Costco at ~48× has, in the recent past, returned you almost nothing for a year or more even as the business performed beautifully — the stock is roughly flat over the trailing twelve months while earnings grew ~15%, precisely the "growing into a valuation" that a rich multiple forces on you. The bull retort — that Costco has "always been expensive" and always rewarded holders — is true and important, and it is why the answer for an existing owner differs from the answer for a new buyer. There is no analytical trick that makes 48× a bargain; there is only a judgment about whether certainty is worth paying up for. Munger paid up, held, and never added at the top. We think that is exactly right.
Nobody out-Costcos Costco
| Arena | Who | Where Costco stands |
|---|---|---|
| Warehouse clubs (US) | Sam's Club (Walmart) · BJ's | Higher renewal, ~2–3× the sales per club — the category leader |
| Membership as a rival | Amazon Prime | Munger: 'Amazon may have more to fear from Costco' |
| Broadline retail | Walmart · Target | Can't match the ~12.8% capped margin without breaking their model |
| China | Sam's Club (~60 clubs) | Behind — ~8 clubs; the growth runway, not yet the win |
| Grocery / low-income squeeze | Aldi · Trader Joe's · dollar stores | Costco's affluent base insulates it from the trade-down |
The competitive picture is the simplest on our board: nobody out-Costcos Costco. Sam's Club, backed by all of Walmart, has competed for forty years and still trails on renewal and on sales-per-club; Munger's blunt verdict was that "Walmart just let them do it… a terrible mistake." BJ's is regional. Amazon is the one rival that matters at the membership level, and even there Munger argued the fear ran the other way — "Amazon may have more to fear from Costco than the reverse," because Costco's price trust and purchasing power are things a marketplace can't easily replicate. The one place Costco is genuinely behind is China, where Sam's Club has ~60 clubs to Costco's ~8 — but that is a statement about untapped runway, not about a contested moat. In its own arena, Costco has no equal, and hasn't for decades. That, precisely, is why it is never cheap.
The forklift driver, and a discipline that refuses to overpay for itself
Management is a genuine strength — and its capital allocation contains a lesson most companies never learn.
Here is the capital-allocation point that separates Costco from most of corporate America, and it is a virtue disguised as a limitation. Many companies buy back their own stock indiscriminately, at any price — we flagged Nike doing exactly that at ~$97, then pausing at $45. Costco does the opposite: it refuses to buy back its own shares aggressively when they are expensive, keeping repurchases to a token amount that merely offsets compensation, and returning excess cash through special dividends instead — the CFO all but says out loud that the stock is too dear to repurchase. That is precisely the discipline Buffett preaches and few practise: what is smart at one price is stupid at another. A company that won't overpay for its own stock is telling you something honest about the price — and, incidentally, is the kind of steward you want holding your capital. I score management an 8: an elite, self-replicating culture and genuinely disciplined capital allocation, dinged only by the post-Munger thinning of the board's institutional conscience.
Modest margins, a fortress balance sheet, relentless compounding
| Metric | Value | Read |
|---|---|---|
| Revenue (FY2020 → FY2025) | $166.8B → $275.2B | ▲ ~10.5%/yr — relentless, low-drama compounding |
| Membership fees (FY2025) | $5.3B | ▲ ~51% of operating income — the annuity |
| Gross margin · operating margin | 12.8% · 3.8% | ◆ thin BY DESIGN — the capped markup is the strategy |
| Return on equity · invested capital | 28.3% · 19.0% | ▲ elite returns on a low-margin model |
| Membership renewal (US/Canada) | ~92.2% | ▲ the most important number in the company |
| Balance sheet | NET CASH · Altman-Z 8.9 | ▲ a fortress — holds more cash than debt |
| Free cash flow (FY2025) | $7.8B | ◆ real, but ~48× it — the whole debate |
| Cash conversion cycle | negative | ▲ payables ($19.8B) exceed inventory ($18.1B) |
| Shares outstanding (5 years) | ~flat at 444M | ◆ no dilution, but no meaningful buyback either |
Read these numbers and the paradox sharpens. The margins look pedestrian — a 3.8% operating margin, a 3% net margin, the thinnest on our entire board except the insurers — because that thinness is the strategy: Costco chooses to hand its scale advantage to members as lower prices rather than book it as profit, trusting the membership fee to do the earning. Yet on that razor-thin margin it generates a 28% return on equity and 19% on invested capital, because it turns its inventory ~13 times a year and its suppliers finance the shelves. The balance sheet is pristine — net cash, an Altman-Z of 8.9, among the safest we have scored. And the top line has compounded at ~10.5% a year without drama for a decade. There is genuinely nothing wrong with this business; the entire argument, from here, is about the one number in the middle of the table — $7.8 billion of free cash flow, which the market is pricing at forty-eight times.
The definitional wonderful-business-at-an-unwonderful-price
| Yardstick | Today | Context | Read |
|---|---|---|---|
| P/E — trailing earnings | ~48x | 10-yr median ~38× · 2022 trough ~34× | above its own rich history |
| P/E — forward (FY2026) | ~46x | ~34× the FY2030 estimate | still rich even on the out-years |
| PEG (forward) | ~3.8 | for ~10–12% EPS growth | you pay ~4× the growth rate |
| P / Free cash flow | ~48x | — | no cash-flow support at this price |
| Dividend yield | ~0.6% | + periodic specials | not an income story — a compounding one |
Costco produces the strangest valuation page we have drawn. The automated DCF reads $319 — sixty-seven percent below the price, the most extreme overvaluation on our board, because no discounted-cash-flow model can rationalise ~48× earnings for ~10% growth. And yet Wall Street's analysts sit at $1,102 (+15%), and — remarkably — even their single most bearish published target, $1,000, is above today's price. Both cannot be right, and the resolution is the whole point: Costco is not mispriced, it is fully priced, and it has been fully priced for two decades. The model says it should be cheaper; the market says it never will be; history sides with the market. The most useful anchor is neither extreme but the middle: reverting to Costco's own ten-year median multiple of ~38× would put the stock near $760 — roughly a 20% haircut, which is what a market panic or a single soft quarter could deliver, and which is the level at which a long-horizon buyer finally gets paid for the quality rather than paying ahead for it. At $956 there is no margin of safety — you are buying certainty at retail, with the return over the next few years resting almost entirely on whether the multiple holds, not on whether the business delivers (it will). That is a bet on the market's continued affection, not on Costco. The business is a 9; the price is a 3; and honest arithmetic will not let the first excuse the second.
The whole risk is the multiple · verified July 2026
Verified the week of publication — and the headline is that the ruby risk is not a lawsuit or a competitor, it is the price. On litigation, Costco's docket is remarkably light for a company its size and, notably, contains no securities fraud, no short-seller campaign, and no regulatory investigation — the bear case is purely valuation, never integrity. The live suits are ordinary consumer-class matters: a January 2026 action alleging the rotisserie-chicken 'no preservatives' claim is false; a March 2026 customer class action alleging Costco cited tariffs to raise prices then sought refunds without reimbursing shoppers (Costco won its Supreme Court challenge to the tariffs in February 2026 and is processing refunds); a membership auto-renewal suit; and routine wage/hour settlements. None is material to the thesis. The amber risks are the real ones an owner monitors: the de-rating that happens through time (the stock was flat for a year while earnings grew — the tax a rich multiple always collects), the international build-out that trails Sam's Club badly in China, wage inflation from the Teamsters ramp, and the post-Munger thinning of the board's value-investor conscience. What is absent is any threat to the moat or the balance sheet: this is a net-cash fortress with a 92% renewal rate. The only way to lose money in Costco is to overpay for it — which, at ~48×, is the one risk that is fully present.
Charlie Munger ate a one-dollar-and-fifty-cent Costco hot dog in the last days of his life, at ninety-nine years old, having sat on that company's board for a quarter of a century and having owned its shares in his personal account the entire time — refusing, through every high and every panic, to sell a single one. He is the reason I have to write carefully about Costco, because I loved Charlie, and because his verdict on this company is the most honest thing anyone has ever said about it, and it contains both halves of the truth. "I love everything about Costco," he said. "I'm never going to sell a share." And then, without contradiction: "The trouble with Costco is it's forty times earnings." He held it and would not add to it. That is not indecision. That is the discipline of a man who understood, better than anyone alive, the difference between a wonderful business and a wonderful investment — and knew they are not always the same stock.
Let me give the business its full and honest due, because it is extraordinary. Costco is very possibly the finest retail business ever constructed. It does not really sell groceries; it sells a membership — eighty-three million households paying a yearly fee, renewed ninety-two percent of the time, and that single fee is roughly half of all the profit. The merchandise is run at a whisker above cost on purpose, a capped markup treated as a sacred promise, so that the price is always the lowest and the member never leaves. Its suppliers finance its shelves; it holds more cash than debt; it turns its inventory thirteen times a year and earns twenty-eight percent on its equity while charging its customers as little as it possibly can. It pays the highest wages in retail and keeps its people for decades and grows its own chief executive from a seventeen-year-old forklift driver. And it will not buy back its own stock at these prices — the surest sign, from a management that knows the value of a dollar better than any I follow, that it considers its own shares dear. On the quality of the enterprise I have not one reservation. It is a perfect damn company, as Charlie said.
And it costs forty-eight times its earnings. That is not a typo and it is not, at these growth rates, a bargain by any arithmetic I know how to do. A discounted-cash-flow model, the honest machine that valued Amazon and Coca-Cola for us, says this business is worth three hundred and nineteen dollars a share; it trades at nine hundred and fifty-six. The model is not wrong about the math — it simply cannot believe anyone would pay forty-eight times for ten-percent growth. The market has believed exactly that for twenty years, and the buyers who waited for a cheap Costco are still waiting, and have missed one of the great compounding stories of our lifetimes. So I will not insult you by pretending the model settles it. But I also will not do the thing that love of a business tempts every analyst to do, which is to let the quality of the company become an argument about the price. The plain truth is that at these levels your return over the next several years rests almost entirely on whether the market keeps paying this premium — not on whether Costco performs, because it assuredly will. That is a bet on affection, not on earnings, and a margin of safety is the one thing this price does not contain. When it fell twenty-one percent last winter, it merely went from absurdly priced to very expensively priced. It has never, in living memory, been cheap.
So what would Charlie have me tell you? He told us himself, and I can improve on it not at all: "I'm not saying I'm buying Costco at this price. But I'm certainly not selling any." If you own this company, hold it, and hold it the way Charlie did — through the flat years when the earnings grow and the stock sits still, collecting the occasional fat special dividend, letting a wonderful business slowly grow into even a rich price, and never, ever mistaking a full valuation for a reason to sell one of the best businesses on Earth. The tax on owning greatness at a premium is patience, and patience in Costco has always, eventually, been paid. But if you do not own it, I cannot in conscience tell you to chase it here, at a price above the one Charlie himself flagged as "the trouble," with no cushion beneath you and your whole return riding on the multiple holding. Wait. This is a company that occasionally hands you a gift — a market panic, a soft quarter, a rotation out of quality — and drops it toward its own ten-year-average multiple, somewhere near seven hundred and sixty dollars, where the arithmetic finally works in your favour and you are paid for the quality rather than paying ahead of it.
My verdict, then, is the most Munger-ish one I have ever issued, and I make no apology for it: hold it if you have it; do not chase it if you don't. Keep this magnificent company on your watchlist, keep the price trigger set toward the mid-seven-hundreds, and understand that you are waiting not for the business to get better — it cannot get much better — but for the market, just once, to offer it to you at a price a value investor can defend. Charlie waited his whole life for wonderful businesses and then, when he found this one, held it forever and refused to overpay to add. I can think of no better instruction. The hot dog is still a dollar-fifty. The business is still perfect. The price is still the only problem — and, for the patient, the only opportunity.