Costco Sells Groceries at Cost and Trades Like a Tech Stock

---

Let us begin by celebrating Costco properly, because the case here is not that it is a bad company — it is one
of the best companies in the world — but that being one of the best companies in the world and being a good
investment at any given price are two entirely different questions. In its third quarter of fiscal 2026, Costco
grew net sales 11.6% to about $69 billion, with comparable sales up nearly 10% (and a healthy 6.6% excluding
volatile gasoline and currency effects), digital sales up more than 21%, and net income rising to $2.19 billion.
Its membership renewal rate sits around 90% worldwide and an extraordinary 92-93% in the United States and
Canada, across 148.5 million cardholders. These are the numbers of a business operating at the peak of its
powers, with a customer loyalty most companies can only fantasize about.

The genius of Costco is well understood and genuinely remarkable. But the financial structure underneath it is
less widely appreciated, and it is the key to understanding both why the business is so good and why the stock
is so strange. So this essay does two things: it explains what Costco actually earns its money from — which is
not what most people think — and then it asks the only question that matters for someone considering the stock
at today's price, which is whether even a perfect business can be a poor investment when you pay fifty times
earnings for it.

Costco is a subscription company wearing a retailer's apron

Here is the fact that reorganizes everything once you see it. Costco's merchandise gross margin runs around 11%
— meaning that for every $100 of goods it sells, it keeps about $11 before any of its own costs. Out of that $11
it must pay for the warehouses, the famously well-compensated employees, the utilities, the logistics, and all
the overhead of running one of the largest retail operations on Earth. By the time those costs are paid, almost
nothing is left. The actual selling of merchandise — the rotisserie chickens, the flat-screens, the
fifty-pound bags of rice — produces, on its own, close to zero operating profit.

So where does Costco's $2-plus billion of quarterly profit come from? Overwhelmingly, from membership fees. In
the quarter, membership fee income was about $1.37 billion, and because that income carries margins approaching
75-80% — there is almost no incremental cost to collecting an annual fee — it flows almost entirely to the
bottom line. Costco, in other words, is not really a retailer that happens to charge a membership fee. It is a
subscription business that happens to run a chain of near-breakeven warehouses as the thing its subscription
gives you access to. The merchandise operation exists to make the membership worth buying and renewing; the
membership is where the profit lives.

This is, to be clear, a brilliant design — it is the reason Costco deserves a premium to ordinary retailers,
because recurring, high-margin, 90%-renewing subscription income is worth far more than the thin, cyclical
margins of selling goods. The membership model is the closest thing in physical retail to the recurring-revenue
annuity that makes software companies so valuable. The bulls are absolutely right about this. But recognizing
that the profit engine is the membership fee also tells you exactly where the vulnerability is — because a
business whose entire profit rests on one recurring revenue stream is a business whose entire profit rests on
that stream continuing to renew at the rate it always has.

The mix shift the renewal rate is hiding

Which brings us to a subtle but important wrinkle that the headline numbers gloss over: the renewal rate, the
single most important metric for a subscription business, remains near record highs — 89.7% worldwide and around
92-93% in the United States and Canada — and that strength is genuine and central to the bull case. But beneath
the headline figure sits a compositional concern that analysts have begun to flag. A growing share of Costco's
newer members are signing up digitally or through promotions, and these members have historically renewed at
slightly lower rates than the traditional, deeply-committed in-warehouse member. As that newer cohort grows as a
share of the base, it exerts a quiet downward pull on the blended renewal rate over time — a headwind the
record-high current number masks rather than disproves.

Now, this is a subtle effect, and it would be alarmist to treat it as a crisis — the renewal rate is, today,
about as strong as it has ever been. But it matters disproportionately because of the valuation. When a stock
trades at a normal retail multiple, the future composition of its membership renewals is a footnote. When a stock
trades at fifty times earnings on the premise that its membership model is a perpetual, ever-strengthening
annuity, any hint that the mix is shifting toward lower-renewing members is more significant than its size
suggests — because the premium multiple is, in effect, a bet that the renewal rate and the membership growth
never disappoint, in either level or trajectory. The risk does not have to be large to matter when the price
assumes there will be no risk at all. The market has priced perfection into a metric whose underlying mix is
quietly becoming a little less perfect, even as the top-line number still gleams.

Fifty times earnings for high-single-digit growth

Now the central question, stated plainly: what should you pay for this? Costco trades at roughly 50 times
earnings — with various measures putting it anywhere from the high 40s to the mid-50s — against an S&P 500 median
closer to 23 or 24 times. The market is valuing a mature, brick-and-mortar warehouse retailer at more than twice
the multiple of the average company, and at a level historically associated with fast-growing technology firms.

Set that multiple against the growth it is buying. Costco grows its earnings at perhaps 10-15% a year in a good
year — genuinely excellent for a retailer, but a fraction of the growth rates that justify a 50x multiple in
technology. When you pay 50 times earnings for a company growing earnings around 12%, you are making a very
specific and very demanding bet: that the company will keep compounding at that rate for a very long time and
that the market will keep paying a premium multiple the entire way, so that the multiple never compresses against
you. If either assumption fails — if growth slows even modestly, or if the multiple simply reverts toward
something more normal for a retailer — the stock can deliver years of poor returns even as the underlying
business keeps performing well. This is the central, counterintuitive risk of a wonderful company at a rich
price: the business can do everything right and the stock can still go nowhere, because all the good news was
already in the price when you paid it.

Consider the arithmetic of the multiple itself. A 50x earnings multiple is an earnings yield of about 2% — below
what you can earn on a risk-free Treasury bill. You are accepting a 2% earnings yield from a retailer on the faith
that its earnings will grow enough, for long enough, to make that worthwhile, while a government bond pays you
more to take no business risk at all. That is not an impossible bet — Costco's quality may well justify it over a
long horizon — but it is a bet with no margin of safety, and "no margin of safety on a 2% starting yield" is a
precarious place to stand, however magnificent the company.

A long-duration asset in disguise

There is a technical but crucial point about what a 50x multiple actually makes Costco's stock, and it explains
why the valuation is more fragile than its rock-solid business implies. When you pay a very high multiple for a
stock, the overwhelming majority of the value you are buying sits far in the future — most of the cash flows that
justify the price arrive years and decades from now, not soon. In the language of finance, a high-multiple stock
is a long-duration asset, mathematically similar to a very long-dated bond: its present value is dominated by
distant cash flows, which makes it acutely sensitive to the interest rate used to discount them.

This is why richly valued stocks fall hardest when interest rates rise — and it applies to Costco despite the
homely, recession-proof nature of its underlying business. The 50x multiple means Costco's stock price is heavily
weighted toward earnings far in the future, so even a modest rise in long-term interest rates, or a modest
increase in the return investors demand, mathematically compresses its present value more than it would a
cheaper stock whose value is anchored in near-term cash. A bag of groceries sold today is about as un-speculative
a thing as exists in business — but a claim on Costco's grocery profits forty years from now, bought at fifty
times current earnings, is a long-duration financial instrument whose price can swing sharply on
macroeconomic forces that have nothing to do with how many memberships Costco sells. Investors who think they own
a safe, defensive consumer staple may actually own one of the more interest-rate-sensitive positions in their
portfolio, precisely because of the price they paid for the safety.

The thin margin cuts both ways

It is worth dwelling on the flip side of Costco's celebrated low-margin model, because the same thinness that is
a competitive weapon is also a source of fragility that a 50x multiple does not price. Costco's roughly 11%
merchandise gross margin is its great strength — it is how the company delivers unbeatable value and drives the
membership flywheel. But operating on such thin merchandise margins means Costco has very little cushion to
absorb cost shocks on the goods themselves. In an environment of tariffs on imported merchandise, supply-chain
disruptions, or sustained input-cost inflation, a retailer keeping only eleven cents on the dollar has almost no
room to absorb higher costs without either raising prices — which risks the value proposition that is its entire
identity — or accepting even thinner margins, which on a near-breakeven merchandise operation is perilous.

Costco manages this masterfully today, and its scale gives it enormous buying power to push back on suppliers.
But the structural point stands: a business that makes its merchandise profit on a sliver is more exposed to
merchandise-cost shocks than its membership-fee profit engine suggests, and the macro environment of trade
frictions and cost volatility is exactly the kind that tests thin-margin operators. Add to this that a meaningful
chunk of Costco's recent headline comparable-sales growth was flattered by gasoline prices and currency effects —
strip those out and the underlying merchandise comps, while still good, are more pedestrian than the
double-digit headline — and the picture is of a great operator working harder than the top-line numbers suggest
to sustain the growth that the premium multiple demands. None of this threatens the business. All of it argues
that the flawless, friction-free future the 50x price assumes is less certain than the price implies.

What the bulls genuinely get right

In fairness, the bull case for Costco's premium is serious and partly correct, and the skeptic must grant it real
weight. The membership model genuinely is exceptional: recurring, high-margin, 90%-renewing revenue is worth a
premium multiple, and a simplistic comparison to a thin-margin retailer like Target understates how different and
how superior Costco's economics are. The flywheel is real and self-reinforcing: Costco's scale and razor-thin
markups create unbeatable value, which drives membership and renewal, which funds the scale, in a loop
competitors cannot easily break. The Kirkland Signature private-label brand is a genuine moat and a growing
profit source. The international runway is real — Costco is underpenetrated outside North America and has decades
of warehouse-opening growth ahead. And the company has compounded shareholder wealth magnificently for decades,
rewarding exactly the investors who ignored valuation worries and simply held. "Costco is too expensive" has been
wrong, expensively wrong, for a very long time, and a skeptic must respect that track record.

The honest synthesis is that Costco unambiguously deserves a premium multiple — the debate is entirely about how
large a premium is reasonable. There is a vast difference between paying 30 times earnings for Costco, which its
quality arguably justifies, and paying 50 times, which requires not just that Costco be excellent but that it
keep growing briskly for decades and that the market never reprice it lower. The bulls are right that Costco is
worth more than an ordinary retailer. The question the 50x multiple poses is whether it is worth this much more
— whether a price that yields 2% and assumes uninterrupted compounding has confused a great company for a great
investment.

The Nifty Fifty lesson

There is a historical precedent that every Costco bull should know, not because it predicts doom but because it
illuminates the exact risk. In the early 1970s, a group of blue-chip American growth companies — Coca-Cola, IBM,
Disney, Xerox, and others — became known as the "Nifty Fifty," the one-decision stocks you could supposedly buy
at any price and hold forever because their quality was so unassailable. They were, genuinely, wonderful
companies, the best of their era. And they were bid up to extraordinary multiples — many to fifty times earnings
and beyond — on precisely the logic applied to Costco today: that for a truly great compounding business, price
does not matter, because the growth will eventually justify any entry point.

It did not work out that way. Many of those companies were, in fact, excellent for decades afterward — and yet
investors who bought them at their peak multiples endured a decade or more of poor returns, because the
businesses grew into their valuations slowly while the multiples compressed. The lesson was not that the Nifty
Fifty were bad companies; most were superb. The lesson was that the price you pay determines your return, and
that even the best company on Earth becomes a mediocre investment if you pay enough for it. Costco is a Nifty
Fifty company — magnificent, durable, beloved — trading at a Nifty Fifty multiple, and the history of that exact
combination is that the company keeps winning while the stock, for a long while, does not.

The kicker

Costco has earned every bit of its reputation. It is a masterpiece of retailing, a fortress of recurring
membership revenue, a company that has made patient shareholders rich for forty years and may well continue to.
None of this essay disputes that, and anyone betting against Costco's business is betting against one of the
surest things in American commerce. But the stock is not the business, and at fifty times earnings — a 2%
earnings yield, more than double the market's multiple, a valuation built for hypergrowth attached to a retailer
growing at a steady high-single-digit clip — the price has already paid the company the compliment of assuming
decades of flawless compounding, with no discount for the possibility that growth slows, the renewal rate keeps
softening, or the multiple simply reverts to earth. You can admire a company without limit and still pay too much
for it. Costco's quality is not in question. Its price is the entire question.

The greatest retailer in America sells you a forty-dollar bag of groceries for thirty-nine and makes its money
on the card in your wallet, and the market has decided that card is worth fifty years of earnings paid today.
Somewhere in the mix of all those renewals a quietly larger share of members signs up on a promotion they may not
renew, and the question is not whether Costco is wonderful — it is — but whether wonderful, at fifty times, was
ever the same thing as cheap.
The market has a long and expensive history of forgetting that the two are different, and of relearning it most
painfully in exactly the companies it admired most. Costco may be the rare exception that grows fast enough, long
enough, to make even this price look reasonable in hindsight. But that is the bet, and it is worth naming as a
bet rather than mistaking it, as the multiple invites you to, for a sure thing — because the surest things in the
market are, again and again, precisely the ones that get priced so high they quietly stop being sure at all.