Costco's Strangest Business: How a Retailer That Barely Profits on What It Sells Built One of the Widest Moats in America

The Model That Nobody Fully Gets

Some scale first as that sets up everything else. Costco operates 924 warehouses worldwide has 81 million paid member accounts and generated $275 billion in revenue in fiscal 2025. Now let's start with the income statement because it explains where all that revenue actually turns into profits. Costco sells you products with a gross margin of between 12 and 13 percent low enough to barely cover the cost of runningThe warehouses. It then charges you an annual fee just to get in the door and that fee is where almost all of the profit comes from. Membership fee revenue in the first quarter of fiscal 2026 was $1.329 million a year-over-year increase of 14 percent. One line item and it represents the vast majority of Costco's operating income

Everything else the rotisserie chicken the Kirkland cashmere sweaters the $1.50 hot dog is the delivery mechanism. You're not actually a customer buying paper towels. You're a subscriber and paper towels are Costco's way of keeping you subscribed

Most retailers try to make money on what they sell. Costco built a model where it makes almost nothing on what it sells on purpose because the implicit promise of "we'll always be the cheapest" is what keeps you renewing. And the renewal rate is what makes that membership fee extraordinary: 92.3 percent of members in the United States and Canada renew every year. Software companies would kill for that curve

A Worked Example: What the Membership Fee Is Actually Funding

The claim that membership dues fund nearly all of Costco's profits is easy to assert and difficult to imagine so let me build an illustrative and simplified warehouse from scratch. These are not Costco's actual per-warehouse numbers;The company does not disclose financial data in this level of detail. Call it a clean independent example created to show the mechanism not a reported figure

Suppose a warehouse generates $250 million in annual merchandise sales. With a gross margin of 13 percent consistent with the range Costco actually discloses that produces a gross profit of 250,000,000 times 0.13 or $32.5 million. Now let's assume that the operating expenses of that single warehouse labor rent utilities everything necessary to maintain the floor amount to 12.6 percentof sales or $31.5 million. Merchandise operating income is $32.5 million minus $31.5 million which is $1 million. On $250 million of sales this is an operating margin of 0.4 percent. Call it rounding error. Here's what "barely breaking even by design" looks like in numbers

Now add the membership side. Let's say the warehouse serves 100,000 households that pay at an average rate combining the standard and executive levels of $70 a year. That's 100,000 times 70 or $7 million in membership fee revenue. Serving that member base the call centers cards and renewal processing costs relatively little. Call it 5 percent of fee revenue or$350,000.Members' net operating income is 7 million minus 350,000 or $6.65 million

Add the two together. The warehouse's total operating income is 1 million plus 6.65 million or $7.65 million. The membership dues alone are responsible for 6.65 divided by 7.65 or about 87 percent of that total. The merchandise operation is not really a profit center. It's more like a very expensive and very effective advertisement for membership one that also happens to be a warehouse full of products you really want to buy

There's a second layer here that many people overlook which is how maximum profit margin affects the return on equity not just the margin. The cost of goods sold in this example is 250 million minus the 32.5 million gross profit or $217.5 million. Let's assume the warehouse has an average of $12.5 million in inventory at any given time. Inventory turnover is the cost of goods sold divided by the average inventory217.5 divided by 12.5 which is 17.4 times a year. Divide 365 days by 17.4 and a unit of inventory will remain on the shelf for about 21 days before being converted to cash. The gross profit for the year $32.5 million is 2.6 times the capital invested in inventory at any one time. A small margin executed through capital that turns over every three weeks still produces a really strong return on capital.employee. That's the part that a quick look at the 13 percent gross margin doesn't reveal

Why Costco Stocks Fewer Than 4,000 Things

Here's a fact that explains much of the purchasing power behind those numbers. A conventional supermarket sells tens of thousands of different products sometimes reportedly more than 30,000 once all the sizes and flavors are counted. Costco by most accounts has between 3,700 and 4,000 SKUs short for stock keeping units at any given time. That's not a rounding difference. It's entirely a philosophy.different

Fewer SKUs means much higher volume per item. If a supermarket divides demand for laundry detergent into fifteen brands and thirty sizes Costco could offer two options one of them Kirkland. All the volume that would have been spread among fifteen suppliers is concentrated in one or two and Costco can enter that negotiation asking for a price that a fragmented supermarket buyer could never obtain. Fewer options for the buyer but much more leverage for the buyer. That leverage is what makes the gross margin of 12at 13 percent is sustainable rather than suicidal. A supermarket has a broad assortment at a higher margin in part because it needs the margin to cover the cost of carrying all that redundancy. Costco has a narrow assortment at a lower margin because first volume concentration allows it to extract a better cost base from suppliers

It also explains the reputation of the treasure hunt the feeling that you never know what will be on the runway this month. A reduced rotating assortment is not just a marketing gimmick. This is what a purchasing strategy based on volume concentration looks like from the buyer's side of the counter

The Flywheel Most People Miss

Put the pieces together and you have a flywheel that reinforces itself in a way that is really hard to alter. High membership volume gives Costco enormous purchasing power. That power funds lower prices. Lower prices generate higher renewal rates. Higher renewal rates fund greater warehouse expansion. More warehouses attract more members. It goes around

The binding constraint on the flywheel is real estate not capital. Costco needs large parcels of land with strong nearby demographics and there is a limited amount of that in any given market. Still the company will open 25 to 30 new warehouses a year through 2026 including recent forays into Sweden South Korea and China. Compare that to the roughly 924 warehouses already open around the world. Let's call the new class warehouses27.5 and 27.5 divided by 924 equals about 3 percent unit growth per year. Slow and steady and deliberately

The newest layer is digital and is growing faster than the physical footprint. Comparable e-commerce sales increased 22.6 percent in the second quarter of fiscal 2026 about triple the comparable growth rate in stores. The old assumption about Costco was that the treasure hunt and in-person store experience is what creates loyalty and that it couldn't be replicated online. This is probably still true for most companies. But the growth of e-commerce over20 percent suggests that members are increasingly willing to also shop Costco's carefully curated assortment digitally expanding what Costco can sell without threatening what built loyalty in the first place

Renewal Rates Are the Number That Actually Matters

Of all the metrics Costco reports the renewal rate is the one I actually look at first more than revenue growth more than the number of memberships. Here's why the arithmetic convinces me. If 92.3 percent of members renew each year about 7.7 percent don't. Under a simple model in which the churn rate remains constant year after year the expected length of an average membership is 1 divided by the churn rate or1 divided by 0.077 which is about 13 years. This is a rough approximation actual turnover is never perfectly constant and this is a simplified model but it tells you the shape of the object. A membership base that renews at more than 92 percent doesn't actually behave like a list of retail customers. It behaves like a long-term annuity

That's also why a small change in the renewal rate matters so much more than a change of the same size in say quarterly comparable sales. Comparable sales bounce around with the economy the weather gas prices and all sorts of noise.13 years to 10 years a significantly shorter annuity. A small change in the holder's number translates into a big change in what that number is supposed to represent

Case Study: The Dollar Fifty Hot Dog

If you want to see the flyer applied in public look at the hot dog and soda combo in the food court which has been priced at $1.50 since Costco introduced it in 1985 and has remained there through four decades of inflation wars a financial crisis and a pandemic.direct statements that Jelinek has repeated in interviews since then to leave him alone. The price never moved

From a purely accounting standpoint this is an odd decision. Input costs for meat and bread have risen enormously since 1985 and Costco absorbs that cost every time someone buys the combination. Viewed as a stand-alone product line it's not obviously rational. Viewed as marketing to members it's nearly perfect. A $1.50 hot dog held flat for decades is an expensive visible and easy-to-verify signal that the company is serious when it says prices don't.They will go up. It costs real money to send that signal and that's exactly why it works. It's easy to not believe a promise that costs nothing. A promise that the company has been paying to keep since 1985 is not

Relate that to the renewal math above. The hot dog isn't really about hot dogs. It's a public demonstration of the same commitment that is displayed less visibly throughout the assortment and that commitment is what keeps 92.3 percent of members renewing each year

The Valuation Problem

None of this is cheap. Costco trades at about 56 times earnings versus about 35 times Walmart and about 21 times the S&P 500 overall. That premium is explicitly for the quality and durability of the membership model and the market pays for future visibility rather than just this year's earnings. The bear case is simple: At 56 times earnings there's almost no room for error. Any slowdown in membership growthany surprise drop in the renewal rate any real margin compression and the stock can return to trading hard and fast

The good thing is that Costco has earned a trust premium that is truly rare in retail. In late 2025 the company sued the U.S. government for tariff refunds an unusually aggressive stance that signals the administration won't just quietly absorb policy costs. CEO Ron Vachris reportedly personally approves every Kirkland Signature product. That level of operational obsession sustained over decades and multiple CEO transitionsfrom Sinegal to Jelinek and Vachris it's the real moat. Whether that culture is worth paying 56 times revenue for is a judgment call. In my opinion it's not unreasonable

The Counterargument: Hard to Copy, Also Hard to Grow

I want to reinforce the skeptical case here because it's real and not just a token objection. Start with something everyone already agrees on: the model is hard to copy. Walmart with more scale than Costco in almost every dimension has run Sam's Club for decades and has never matched Costco's turnover rate or its reputation for value at least according to the general perception among retail analysts and shoppers I've seen discuss. If the largest retailer on the planet can't fully replicate this with a nearlyidentical that means the moat is real. It's culture and execution more than a strategy that anyone can photocopy

But hard to copy isn't the same as easy to grow. The old unit growth math about 3 percent more warehouses a year is slow by the standards of a company trading at a growth multiple. Costco can't just open 200 warehouses next year to speed up the process because the limitation is finding the right parcels of land in the right places not raising capital. That's a real limit to how fast this business can grow even though thebusiness itself is excellent

The model also relies on a lever that Costco rarely pulls: the membership fee increases on its own. Historically these increases have come only every five years or so sometimes more because raising the fee too frequently would undermine the confidence on which the flywheel depends. A significant part of the bullish case is effectively waiting for a catalyst that by design is rare and not something management can time to Wall Street's preferences

Finally I think much of the real debate about Costco has never really been about whether business is good. Hardly anyone seriously argues that business is bad. The discussion has been almost entirely about the multiple whether 56 times earnings is already too much good news not about whether the flywheel is real. This is a different narrower disagreement than it is sometimes given credit for and it's worth being precise about what argument you're really having

How I'd Actually Use This

My reading is that the most useful thing Costco teaches is not really about Costco. It's a model for spotting a specific type of business: one in which a recurring fee funds a deliberately balanced core operation and the balancing operation is what makes it worth paying the fee in the first place. Once you've seen it once you start to notice the way in other places in warehouse clubs obviously but also in jack-knife-style models and in certain marketplace businesses that subsidize on the side.to keep the other loyal

The way I would really use this is as a checklist when looking at a new company. First find the part of the business that looks unprofitable and ask if it's not profitable by accident or by design. Second find what funds it and check to see if that funding source has a renewal or retention metric that you can actually track over time the same way Costco discloses the renewal rate every quarter. Third do the capital efficiency calculations not just the margin calculationsbecause a small margin based on capital that is constantly renewed may still be an excellent business and a large margin based on capital that is barely renewed may not be

I will say clearly that I find the valuation decision really difficult. I don't think 56 times earnings is obviously wrong and I don't think it's safe either. This is not stock picking and I'm not telling anyone to buy or sell anything. It's a way of reading the business so that if you form an opinion on price at least you're arguing about the right the multiple rather than confusing a debate about price with a debate about quality

What to Watch

Two numbers matter more than the rest for this thesis. The first is the renewal rate itself. The second is executive membership penetration. Executive members pay about twice the standard rate and currently generate 75.8 percent of all sales despite being a minority of total members. An increase in executive penetration from around 40 million to 50 million would be a real profit catalyst because that mix change alone increases average revenue per member without Costco changing a singleprice. A drop in the renewal rate below 90 percent would be the first serious sign that the moat is eroding. According to the figures in this article none of that has happened. It is worth checking both every quarter not just once a year

The Bottom Line

Costco makes almost nothing by selling you things and that's the point not a flaw. The 12 to 13 percent gross margin on merchandise is limited on purpose spread across a deliberately narrow assortment of about 4,000 SKUs that concentrates volume and purchasing power and runs through inventory that turns over every few weeks so a thin margin still produces a strong return on the capital involved. The membership fee not the merchandise provides theThe vast majority of operating income and dues only continue to flow because renewal rates are above 92 percent which itself follows prices that members are confident will not increase all the way to a $1.50 hot dog that has been stable since 1985. Low prices drive renewals and renewals finance low prices. It's hard to copy for exactly that reason and its growth is slow for a related one: the limitations ofTerrain and infrequent rate increases limit how quickly the wheel can turn faster. The real market disagreement over Costco has focused primarily on whether 56 times earnings is the right price for all this not on whether the underlying business is solid. My view is that the business is as good as advertised. What you pay for it is a separate and more difficult question