Shopify

IThe business

Shopify rents a shop to anyone who wants to sell something. A merchant pays a monthly fee, gets a store, a checkout, an inventory system and a point-of-sale terminal, and then pays Shopify again each time a customer pays, because most of those payments run through Shopify Payments. The merchant keeps the brand and the customer. Shopify runs what sits behind them. The company was incorporated in 2004, is based in Ottawa, and describes itself as internet infrastructure for commerce.² It serves millions of merchants in more than 175 countries, from a kitchen table to brands such as SKIMS and Supreme, and no merchant has ever been more than 5% of revenue.²³

The number that matters most is GMV, gross merchandise volume: the dollar value of orders that pass through the platform, including shipping, duty and sales tax, net of refunds.² GMV is what merchants sell. It is not what Shopify earns. In the twelve months to 30 June 2026 it was $432.2B C (calculated).³²¹⁰ In Q2 it was $115.6B, up 32%, or 30% at constant currency.⁵ That was the fifth straight quarter of constant-currency growth at 29% to 30%, on the CFO’s count.⁷

Where the money comes from

There are two lines. Subscription solutions were $802M in Q2, up 22%: plan fees, variable platform fees, apps, themes and domains. Merchant solutions were $2,781M, up 37%, and are 78% of revenue.² Most of that is the processing and currency fees from Shopify Payments. The rest is lending, referral fees from partners, shipping labels, hardware and advertising.² Interest and fees on Shopify’s loans alone were $91M in the quarter.²

MRR, monthly recurring revenue, is the plan fees in force at the end of a period. It was $221M, up 19%, and the 10-Q says the rate is helped by the comparison with a year when paid trials were lengthened.² So subscription growth runs at roughly 20%. The 30% and more comes from payments, which grow with volume and with penetration.

Penetration is the share of GMV that runs through Shopify Payments. It has risen every year I can see, from 41.5% in 2018 to 68% in Q2 2026, when $78.1B passed through.⁴² Each point moves revenue, because Shopify books the payment fee gross and pays the card networks and processors out of it.

Revenue is 63% American and 21% from Europe, the Middle East and Africa. The second is the faster one: 42% growth in 2025 against 28% in the United States C (calculated).³ In Q2, North American GMV grew 28% and Europe 34% at constant currency, and management says European penetration rose more than 3.5 points in a year.⁷ The business peaks in Q4. Merchants sold $14.6B over the 2025 Black Friday weekend alone, up 27%.¹²

What Shopify keeps

Here is the fact that shapes this page. Take $100 of merchant sales in the last twelve months. About $3.07 became Shopify revenue. About $1.47 became gross profit. About $0.43 became operating income C (calculated).³²

The revenue figure has climbed for years. The gross profit figure has not. It was 1.42% of GMV in 2019, 1.41% in 2021, 1.53% in 2024 and 1.47% now C (calculated).⁴³² Every extra point of revenue take has come with a matching point of cost. Gross margin went from 54.8% to 47.8% as payments grew.

This is not a flaw. Payments are low margin but need little selling or engineering, and the 10-Q says so.² It does mean that “take rate up” is a weaker claim than it sounds. Management’s own Q3 outlook has revenue up in the low thirties and gross profit up in the mid-to-high twenties.⁵ When I think about what the owner of this business gets, I start from gross profit, and I value the company on GMV and that 1.5 cents. Part of the gap in 2025 was one-off: the CFO named PayPal and paid-trial comparisons as temporary, and said they have largely passed.¹³

What it stopped doing

Shopify is also smaller than it looks. It tried to run warehouses and robots, buying 6 River Systems and then Deliverr, and in 2023 it sold the logistics businesses, took a net impairment of $1.3B, kept a stake and a commercial deal with Flexport, and cut about 23% of its people.⁴¹¹³ Headcount was about 7,600 at the end of 2025, down from 8,100 a year before, while revenue rose 30%.³¹¹ TTM revenue is about $1.75M an employee on that year-end count C (calculated). Capital spending in the last twelve months was $25M on $13.3B of revenue.³² Almost nothing here needs a building.

It also holds a set of stakes that came from partnerships: $1,655M of Affirm, $768M of Global-E, $280M of Klaviyo and a large private portfolio, with $5.4B of equity and other investments in all, counting Flexport.² These swing reported net income so much that GAAP profit tells you nothing. Net income was $1,502M in Q2 and a loss of $581M in Q1, on the same business.²

IIThe moat

The question is simple. If Shopify raised prices 10%, how many merchants would leave? I think few, and the reasons have changed over time.

Switching costs. A store is not only a website. It is the product catalog, order history, tax settings, payment set-up, staff habits, the apps wired into all of them and now the point-of-sale system in the shop. Moving is a project, and the merchant’s revenue is at risk during it. Most plans are monthly, so the contract holds nobody.³ The work holds them. Management says merchants who reach $1M of annual GMV retained at 92% over five years, and 97% at $10M.⁷ Those are management’s numbers, from a call, without a definition I can check. The 10-K is vaguer but points the same way: revenue from older cohorts has grown by more than the revenue lost from merchants who leave.³ The CFO said the Q1 2015 cohort now does five times its first quarterly GMV.⁷

The small get big, and stay. Shopify takes every entrant and most fail. The few who win need more plans, more payments and more loans, and pay more each year. A rival would have to win them young, or pay to move them later. This is the part I find hardest to attack, because it doesn’t depend on any single product.

The ecosystem. More than 21,000 apps sit in the Shopify App Store.³ Developers build where merchants are, and merchants come where the apps are. This compounds slowly, and it also makes Shopify cheap to run: it builds what most merchants need and lets partners build the rest.

The checkout. With 68% of GMV in Shopify Payments, the company sees the payment, the buyer and the merchant’s sales history at once. That lets it make loan offers others can’t. Shopify purchased $2.8B of loans and advances in the first half, up from $1.8B.²

Where it is thin

Two things worry me. First, Shopify Payments sits on other people’s rails. The agreements with Stripe and PayPal renew every twelve months, and the 10-K says any disruption could hurt.³ Stripe processed $1.9T in 2025, up 34%.¹⁴ It is a supplier, and it is also building a rival route to the customer.

Second, and bigger, is the agent. A store’s value rests on the idea that a shopper arrives at it. If a shopper tells an AI assistant to find a standing desk and the assistant does the buying, then the assistant holds the customer, and the merchant’s site may be reduced to a data feed. Stripe built an Agentic Commerce Protocol with OpenAI, and says it powers the first shopping experiences inside ChatGPT.¹⁴ Shopify co-built a rival standard, the Universal Commerce Protocol, and says every merchant on it is ready to sell through agents.⁷ Management named OpenAI, Google, Meta and Microsoft as partners, and said on the Q1 call that Amazon, Meta, Microsoft, Salesforce and Stripe joined the protocol’s council.⁷¹⁵ So far this looks like a good defense. Shopify’s Catalog holds more than a billion products, and management says AI searches using it convert at twice the rate of scraped data.⁷

It is too early to say how this ends. The honest position is that the protection against other e-commerce software is wide, and the protection against an agent that owns the shopper is untested. Management says agentic volume is small and carries the same economics as a storefront sale, with no new fees.⁷ That is the right answer today. It also means Shopify has not yet shown it can charge for that position.

IIIThe competitive advantage

Shopify’s advantage is being the default for anyone who wants their own store, at a scale no independent rival is near. The figures below measure different things, so read the growth column. Shopify’s GMV grew 32%. US e-commerce grew 12.2%, Amazon’s third-party seller revenue 9% and eBay’s GMV 15%, with marketplaces and processors slower still.⁵⁶¹⁶¹⁷

GMV growing about 2.6 times as fast as the market is the main evidence of an edge C (calculated).⁵⁶ Stripe’s 34% is a useful check on it: being the rails for online sellers has been a good place to be, whoever you are.¹⁴ North American GMV growth of 28% against Census growth of 12.2% is the cleaner comparison, though not a clean one.⁷⁶

On share, the CFO said Shopify represents over 14% of US e-commerce, and that, on eMarketer’s data, its merchants have taken nearly half of the incremental US e-commerce dollars since the start of 2025.⁷ I can’t reproduce either. Census e-commerce was $340.2B in Q2, about $1.36T a year.⁶ If US GMV matched the 63% US share of revenue, that would be about $272B in the TTM, or 20% of the market, but GMV includes tax, shipping and shops, so the two aren’t the same thing C (calculated).³⁶ Treat 14% as management’s figure. Management also says Shopify has under 1% of global retail sales, which is a true-sounding number that doesn’t tell you much about what can be won.⁷

Scale shows up in costs. Research and development was $445M in Q2, up 13%, against revenue up 34%; sales and marketing rose 20%.² Each dollar of engineering is spread over more merchants than any independent rival has. Wix, the closest listed software rival, had $563M of revenue in Q2, up 15%, about a sixth of Shopify’s.¹⁹² For large merchants, the comparison is custom builds and enterprise suites, where management says migrations now take weeks; it named Balmain, Burton and General Motors as recent wins.⁷

The advantage is real and also narrower than the headline. It doesn’t produce pricing power that reaches gross profit per dollar sold, as the 1.5 cents shows. What it produces is volume, and low costs on that volume. That is still a very good thing to own. It is a different thing from a toll that gets richer each year, and I’d keep that distinction in mind when comparing it with Visa or Mastercard.

IVManagement

Tobias Lütke, 45, founded the company and has been on the board since 2004. He is chief executive and head of research and development.⁸ Harley Finkelstein is president and Jeff Hoffmeister is CFO. Jessica Hertz, formerly general counsel, became COO in October 2025 after Kasra Nejatian left that September.²¹²²

How they are really paid

Lütke’s salary is C$1, at his own request. His 2025 pay was $35.0M of options, all of it. In 2024, the year he moved to a new structure, it was $150.0M: $112.5M of options and $37.5M of RSUs.⁸ Shopify calls this a boxcar model. Awards are made years ahead, and each one starts vesting only when the last has finished. His 2025 grant has a four-year cliff, then monthly vesting over one year, and it doesn’t begin until the 2024 award has vested. Others have a two-year cliff.⁸

There are no performance conditions. The circular says plainly that the company doesn’t believe in metric-based incentive designs.⁸ So what does this pay them to do? It pays them to stay, and to get the share price up over about five years. There is no return-on-capital test, no cash flow test and no annual bonus. For a founder with this much stock, I have no objection. For the others it’s simple time-and-price pay. At $157.49, Lütke’s 2.67M options at $80.20 are about $206M in the money, and the 2025 options at $127.05 about $16M C (calculated).⁸¹ Options gain from a higher price and from nothing else, and the company pays no dividend, so the design leans toward growth and away from returning cash. The buyback, below, is the first sign of a shift.

Below the top, employees choose how much of a dollar-denominated wallet to take as cash, RSUs or options, and can change that each quarter.⁸³ This is the most interesting pay system I’ve seen on these pages. It explains why stock pay was only $482M in the TTM, 3.6% of revenue C (calculated).³² On my Snowflake page stock pay takes 30% of revenue. It does mean a falling share price can raise cash pay, as people switch.

How much of their own money is in

Lütke and his holding companies control 77.75M Class B shares and 1.55M Class A, which includes 258,000 held by his foundation.⁸ That is 79.3M shares, about $12.5B at the price and about 6.1% of the shares C (calculated).⁸² He has 40.02% of the votes, and the circular names no other holder above 5% of a class; BlackRock has since reported 5.1% of Class A.²³ The Founder Share gives him however many votes he needs to reach 40%, capped at 49.9% with his other shares, and the 10-K says this holds even if the Class B shares all convert.⁸³ So outside holders would struggle to remove him, and new share issues do not dilute his vote. The vote can end. It falls to one vote after a sunset date, set by the board between 9 and 18 months after the earliest of three events: he stops being both an executive and a director, he dies or is disabled, or he and his family hold fewer shares than 30% of the Class B he held in April 2022, as adjusted.⁴⁵ The 2022 filing sets these terms, and I read them there; the articles themselves are scanned images that I could not search.

The advisory vote on pay in June passed with 81.24%.²⁴ If Lütke’s 40% of about 2.04B votes went in favour, the rest of the shareholders voted about 64% for and 36% against E (estimate). Hoffmeister adopted a plan to sell up to 20,000 shares by June 2027, which is small.² Directors and staff can’t hedge the stock.⁸ One governance point: when the audit chair left for the US Commerce Department in July, chairman Joe Natale was expected to take that job too, so one person holds both roles for now.²⁵

What they did with capital over a full cycle

The record has three acts, and I’d grade them differently.

Raising when the stock was high. From 2016 to February 2021 Shopify sold about $6.6B of stock in public offerings, including $2.6B in 2020 and $1.5B in February 2021 at $131.50 a share, split-adjusted, plus $920M of convertible notes in 2020.²⁶⁴ Selling at those prices was smart. The stock closed at $169.06 that November and at $25.67 eleven months later, a fall of 85%.¹

Spending it badly. Deliverr cost $1,962M in cash and $10M in stock in 2022.¹¹ A year later the whole logistics business was gone, with a net $1.3B impairment.¹¹ That is a large, plain mistake on capital. To their credit, they reversed it fast, took the loss and got out of a business that would have needed warehouses for ever. Fewer managements would have done that within a year.

Paying it back. The $920M notes matured in 2025 and took $1,043M of cash to settle, with no new debt.³ In February 2026 the board approved a $2B buyback, then raised it to $5B in June.²²⁷ In the first half Shopify bought 16.86M shares at an average of $114.63, for $1.93B.² They bought below the $131.50 they sold at in 2021, and $42.86 below today’s price. That’s good execution, and the pace was brisk: 5.0M shares in May at an average of $105.70, just after the shares fell 15.6% on the Q1 report C (calculated).²¹ The buyback used more cash than free cash flow produced, and cash and securities fell by $831M.²

They admit errors quickly and bought back well. They have also made a $2B bet that failed inside a year, and with 40% of the vote nobody can stop the next one. The stakes taken through partnerships have done better: Shopify’s group reports 33.4M Klaviyo shares and warrants, 20.83% of that class.²⁸²

VFinancials

Shopify has grown revenue from $389M in 2016 to $13.3B in the TTM, about 45% a year C (calculated). GMV rose about 28 times in the same period C (calculated).⁴² Growth went 86% in 2020, 57% in 2021, then 21% in 2022, and has run at 26% to 30% a year since 2023, and 34% in the first half of 2026 C (calculated).⁴³²

Margins, after a detour

Operating income reached 14% of revenue in the TTM, the best in its history.³² The path ran through a hole: an operating loss of $822M in 2022 and $1,418M in 2023, the latter with the logistics impairment.²⁶ In Q2 operating income was $488M, 13.6% of revenue, up 68%.² Operating expenses were 34% of revenue, against 38% a year earlier C (calculated). In the first half gross profit rose $783M and operating income rose $376M, so 48 cents of each new gross-profit dollar reached operating income C (calculated).² That figure matters for my base case, which needs more than that for a long time.

One line runs the other way. Transaction and loan losses were $141M in Q2, up 76%, and 3.9% of revenue.² Of the $61M rise, $41M came from lending.² Sales and marketing also rose $85M on marketing programs in the quarter.²

Three warnings about free cash flow

Shopify’s reported free cash flow was $2,352M in the TTM, about 18% of revenue, against $2,007M in 2025, and capital spending is close to nothing.²⁹⁵ That looks like a clean number, and it isn’t. Three things sit underneath it.

The rule changed. Since April 2026, merchant cash advances have moved from operating to investing cash flow, in line with loans. It was prompted by a Canadian regulatory change.² The effect was a $37M lift in Q2, about a point of margin, and the CFO guided to under a point in Q3.⁵⁷ It’s small, disclosed and reasonable. It still means all of Shopify’s lending now sits below the line that investors watch.

Lending is real cash. Net of repayments, Shopify put $707M into loans and advances in the TTM, about 30% of free cash flow C (calculated).³²¹⁰ The book is now $2.18B. Credit losses add back as a non-cash provision, while the actual write-offs show up only in investing.

The book looks fine so far. The allowance is 10.0% of gross, and 5.7% of balances are 180 days or more late, the same as in December.² But it has not been through a hard recession at this size, and it grew 22% in six months C (calculated).

Interest counts twice. Interest income was $301M in the TTM, about 13% of free cash flow, earned on cash that any valuation adds back on its own C (calculated).³²¹⁰ If I add the cash, I can’t also capitalize its interest.

What an owner can really take out

So here is the number I’d use. Capital needed to stand still is almost zero. The real deductions are stock pay, the interest on cash I count separately, and the loans that go bad. That leaves about $1,451M of owner cash. If the loan book keeps growing as it has, the figure is about $922M.

Against reported earnings, that is close. Shopify’s own measure, net income excluding equity investments, was $1,760M in the TTM C (calculated).⁵ Owner cash is 82% of that and 62% of free cash flow. At the price, the market value is 141 times owner cash, a yield of 0.7%, against 4.94% on the ten-year Treasury C (calculated).³⁰

What to read on 3 November

Q3 results come before the open on 3 November.³¹ The outlook implies revenue of about $3.73B, gross profit of about $1.77B and operating income of about $0.52B, if each lands mid-range E (estimate).⁵ If GMV grows 29%, it comes to about $119B, which would put gross profit at 1.49% of GMV E (estimate). Q3 2025 is also an easier comparison on losses, because transaction and loan losses that quarter were an unusually high $148M.² I would look past revenue and go to four lines: GMV at constant currency, gross profit dollars, net lending in the cash flow statement, and the 180-day column in Note 6.

The balance sheet

Cash and securities were $5.5B at 30 June, including $525M of longer-dated securities. There is no debt.² Equity and other investments carry at $5.4B, of which private holdings rose by $522M in Q2 on a single observed price.² Together that’s about $8 a share C (calculated). Shares outstanding were 1,286.6M at 31 July, down from 1,304.1M in early February.²³ Buybacks have more than covered stock pay.

VIValuation

How this goes badly wrong

Before any numbers I ask what would most likely wreck this, and work back. I see three ways, and only two of them involve the business.

The agent takes the customer. Buying moves into assistants, and the assistant owner sets the terms. Shopify still processes the order, but as a back end. Merchants find the store worth less and the app ecosystem thins. The 1.5 cents starts to fall, and volume follows more slowly. This is what my bear case describes: gross profit falling to 1.25% of GMV, with growth slowing to about 10% a year.

A hard recession hits the merchants and the loan book at the same moment. Shopify’s customers are small, consumer-facing and thinly funded. A buying slump cuts GMV, raises merchant failures and raises credit losses on a book that has never been tested at $2B. GMV growth went from 46% in 2021 to 12% in 2022, and that was with consumers in decent shape.⁴ If 2027 and 2028 came in at 5% and 10%, with the rest of my base unchanged, the base value falls to about $75 E (estimate).

The price itself. This one needs no failure at all. From $169.06 in November 2021 the shares fell 85% to $25.67 in October 2022, while GMV still grew 12%.¹⁴ They did not close above that 2021 high again until October 2025, and today’s price is still below it C (calculated).¹ The way people lose most in a stock like this is to pay for a decade of perfection, get a good business, and still lose money.

The first two are slow. Neither would show up in this year’s accounts. So I look at what the price pays for, and at how it would look if one of these came true.

What I can’t judge

I can’t tell you how much commerce moves to agents, or who ends up holding the customer. Nobody can yet. It’s the largest input to the value, and the range below is wide because of it. I can’t tell you how long 30% volume growth lasts: management won’t give a number past next quarter, and there is no long-range target to test. I can’t see the economics of the Catalog or of agent-placed orders, beyond management saying they match the storefront. I can’t verify merchant retention, because the only figures come from a call. And results arrive on 3 November, four days after this price, so the Q2 numbers here will be stale soon.³¹ The shares fell 5.6% on 30 October on heavier volume; I found no source for why, so I won’t guess.¹

What the price implies

At $157.49 the market value is $204.3B on 1,297.0M diluted shares: the 1,286.6M outstanding on 31 July, plus options by the treasury method and RSUs C (calculated).¹² The close is confirmed on a second feed.¹³² Take out $5.5B of cash and securities, $5.4B of stakes and about $1.4B of second-half cash, and the operating business is priced at $192.0B C (calculated). That is 30 times TTM gross profit on enterprise value, and the stock is 108 times Nasdaq’s 2026 consensus of $1.46 and 86 times 2027’s $1.84, with the basis unstated.⁹

The market has paid more than this and much less. At the end of 2021 the market value was 71 times gross profit. A year later it was 16 times. Today the market value is about 32 times C (calculated). GMV grew 12% in those twelve months, so the fall in the multiple measures mood as much as merit.

A reverse DCF asks what growth the price needs. Starting from free cash flow, cash must compound about 27% a year for ten years. From owner cash, about 34%. Revenue has grown about that fast lately, and with margin gains cash can outrun volume, so these are not absurd. They are also numbers very few companies have held for ten years from this size.

The discount rate

I use 10.5%, the rate on my Snowflake page and a little above the 10% on NVIDIA and Palantir, so that the pages can be compared. I want to be plain about what that hides. The capital asset pricing model, using the 4.94% Treasury and a 4.20% equity premium, gives 16.2% on Shopify’s five-year beta of 2.68, and 13.7% on its ten-year beta of 2.09 C (calculated).^3033341 I don’t think a company with no debt, $5.5B of cash and this much cash generation deserves 16%. That beta was earned in the 2021 to 2022 collapse. But the stock has behaved that way, and a reader who takes the model at its word gets a base value near $47.

Three cases

I build each year from GMV, gross profit as a share of GMV, and the costs below gross profit. I model gross profit dollars because revenue flatters growth when payments rise. My 2026 estimate takes the first half, then GMV up 29% in the second half, and the Q3 outlook, giving about $495B of GMV, $7.1B of gross profit and $2.2B of operating income E (estimate).⁵ I run fifteen years because a business growing like this needs a longer fade than ten. My operating income, after 20% tax and with interest added, comes to about $1.51 a share for 2026 and $2.05 for 2027, a few percent above Nasdaq’s consensus of $1.46 and about a tenth above its $1.84 E (estimate).⁹ So I am not low in the near years. The difference is how long the growth runs.

Base, $94. GMV grows 24% in 2027 and fades to 9% by 2036, about 15% a year, to $2.0T. That’s about 1.5 times the whole US e-commerce market today C (calculated). The 1.5 cents holds. Costs fall from about 70% of gross profit to 48.5% in 2036 and 45% by 2040. That takes operating income to $15.5B in 2036. To reach it, about 58 cents of each new gross-profit dollar must reach operating income for ten years, against 48 cents in the first half C (calculated). So my base is not timid on margins.

Bull, $188. Growth stays at 20% or more through 2032, averaging 21% over ten years, to $3.3T. Advertising, lending and agent fees lift gross profit to 1.70% of GMV. Costs fall to 42% of gross profit by 2036. Operating income reaches $32.5B. Everything has to work.

Bear, $44. Agents and competition squeeze gross profit to 1.25% of GMV, growth averages 10%, and costs stop at 60% of gross profit. Shopify still grows faster than retail and earns $6.2B of operating income in 2036. This is a decent business at the wrong price, with the shares down 72%.

The price is 68% above my base and 16% below my bull. Weighting bear, base and bull at 25%, 50% and 25% gives about $105 E (estimate). That weighting is a judgement, and the shape is what matters: terminal value is 52% of the base, and a tenth of a point on gross profit per dollar sold moves it about $6 a share C (calculated).

What closes the gap

My base cash flows are worth the price at a discount rate of 8.0%, not 10.5% E (estimate). With everything else unchanged, the price needs GMV growth of about 23% a year for ten years, or $3.8T in 2036, against my 15%. Or, with volume as in my base, it needs gross profit of about 2.56% of GMV, which Shopify has never come near.

The simple way to see this: three generous changes together, two points of faster growth, a tenth of a point more gross profit and 10% lower costs, reach $125. It takes a lower required return as well to get close, and a 9.5% rate still leaves $150.

Now multiply. The price rests on three things happening together. GMV must grow about 23% a year for a decade: I’d put that at about 30%, since few companies at this scale have done it. Gross profit per dollar must hold while agents reshape the checkout: about 70%. And costs must fall to under half of gross profit: about 55%. Multiply them and you get about 12% E (estimate). They aren’t independent, since strong growth makes cost leverage easier, so the real figure may be higher. It’s still a long way from likely. I am not saying the price is wrong. I am saying it asks for a lot at once.

Against the next best use of the same dollar

A ten-year Treasury pays 4.94% with no effort.³⁰ Shopify’s owner cash yields 0.7% today, so nearly all the return must come from growth. On my base, the price works out at an implied rate of about 8.0% a year.

The best comparisons on this site are the other tolls on the same dollars. Visa’s page gives an implied rate of 8.2% in its base case, and Mastercard’s gives 8.4%, both measured against a 9.0% hurdle.³⁵³⁶ So on the base case Shopify offers about what those do, with far more spread: my bear value is 72% below the price and my bull 19% above. Uber is the other useful one. It also takes a slice of a large gross volume, and its page has the price 24% below the base value, at 10.6%.³⁷ Amazon’s page has its price 47% above base, which is closer to this page.³⁸ The pages use different rates, so these are rough. The plain point stands: this price pays up for a longer runway, and several pages on this site ask for less.

The margin of safety

Two questions are getting mixed together in most discussion of this stock. How good is the business? Very good: it has no debt, needs almost no capital, grows at 30% with a founder who owns 6% and a system of compounding merchant cohorts. How good is the price? That is a separate matter, and today it looks like a great business at a full price.

My margin-of-safety price is $66, my base value less 30%. I use 30% because the range is wide and more than half of the value sits past year fifteen. The shares last closed there in August 2024, and the lowest close of the past year was $95.40, in May.¹ At $66 the bear value would be 33% lower and the implied rate on my base about 13.0% E (estimate).

Where this could be in 5 and 10 years

Each figure is my model’s value at that date: the present value then, in that year’s dollars, of the owner cash after it, at 10.5% with 3.5% growth after 2041. Cash and stakes stay at today’s level, and I divide by today’s diluted share count, about 1.3B.C (calculated) Owner cash is after 20% tax and after funding the loan book.

All per share. Owner cash is the sum from 2027 to 2036, not discounted. Return a year is the year-10 value plus that cash, against the $157.49 price, over ten years.C (calculated)

For the base to hold at five years, GMV must reach $1.19T in 2031 from about $495B this year, 19% a year, with gross profit at 1.5% of GMV and costs down to 58% of gross profit. At ten years GMV is $2.0T, costs are 48.5% and operating income is $15.5B. Only $13 a share of cash arrives in the first five years, so 74% of the year-10 value is terminal value after 2041.E (estimate)

The bear has GMV growing 12% a year to 2031 and 7% after, gross profit falling to 1.25% of GMV and costs stuck at 60%. The bull has 26% and then 16%, gross profit at 1.70% and costs at 42%. The rate matters as much: the base year-10 value is $219 at 9% and $144 at 12%.E (estimate)

The most likely way the base is wrong is growth. If GMV grows two points a year slower than the base, with everything else the same, the year-10 value is $109, not $174.C (calculated)

These are estimates of value, not forecasts of the share price. The market price can sit above or below value for years.

VIIThe case for the price

What most people miss

Staff choose how much of their pay comes as stock, and few take much. Stock pay was 3.6% of TTM revenue and the share count is falling. Software-style growth with almost no dilution is rare, and it means reported cash is closer to owner cash than at most peers C (calculated).³²

Headcount fell while revenue grew 30%. Research and development fell to 12% of revenue from 15% in a year, even with AI usage costs rising.² If that continues, my cost path is too cautious.

Partners pay partly in equity. Shopify has booked revenue from non-cash consideration, $94M in 2024 and $49M in 2025, and took stakes such as Klaviyo through partnerships. The $5.4B of investments on the balance sheet is about $4 a share, and it costs nothing to run.³²²⁸

The bull case

The best version goes like this. E-commerce is still only 17.1% of US retail.⁶ Shopify is taking a large share of what’s new. Merchants now span from the kitchen table to General Motors, international is earlier in the same curve, and B2B GMV grew 76% in Q2.⁷ The CFO says merchants took nearly half of new US e-commerce dollars since early 2025, and cohorts keep compounding.⁷

Then comes the agent, which the bulls read as an opening. ARK’s fintech analyst Varshika Prasanna wrote on 11 May 2026 that Shopify is “positioning itself at the center of agentic commerce”, citing AI-driven traffic up 8 times and orders from AI-powered search up about 13 times; her note gives no price target.³⁹ Finkelstein’s argument is that agents help the long tail: on the Q2 call he said 75% of AI-attributed purchases came from outside the top 100 categories.⁷ If agents find the odd, specific thing a shopper wants, the independent merchant wins, and those merchants are on Shopify. ARK held 4,155,925 shares at 30 June.⁴⁰ Baillie Gifford held 28.4M and Durable Capital 2.3M, up from 1.0M in March; the filings don’t say why.⁴¹⁴² Altimeter held 570,210 at the end of 2025 and no line in its March or June filings.⁴³ Those are the only named views I could verify on this company.

Against reality

Today’s numbers support the first half of this. GMV is growing 30% at constant currency, costs are falling as a share of gross profit, and first-half operating income rose 76% C (calculated).⁵² If Shopify grows GMV about 32% next year and 23% a year across the decade, keeps 1.5 cents, and takes costs to under half of gross profit, the price is possible. That’s $3.8T of GMV in 2036, nearly three times the US e-commerce market today C (calculated).

The gap is in the second half. Agent volume is still small, by management’s account, and earns nothing extra.⁷ Gross profit per dollar sold has not risen in seven years. For the bull value of $188 it must reach 1.70%. Growth has been steady for five quarters, but 23% for ten years starts from about $495B, not from $15B.

What a 10x would take. Ten times $157.49 is $1,575, or about $2.0T of equity. By 2036 that’s about 25% a year C (calculated). At 30 times earnings it needs $68B of net income. At a 35% operating margin and 20% tax, that’s about $243B of revenue, more than twice my bull case’s 2036 figure. At today’s 3.07% of GMV it means about $7.9T of GMV, some 5.8 times annualized US e-commerce C (calculated).⁶ It has been done from this size: NVIDIA went from about $218B in May 2020 to ten times that by March 2024, in under four years C (calculated).⁴⁴ That came from a new product in shortage at huge margins. Shopify keeps 1.5 cents of a dollar sold in a competitive market. On that model a 10x by 2036 is close to impossible. It would need a different business, one that charges for the agent layer at a rate it doesn’t charge today.

What they have to hit

Six numbers, each on a regular schedule. The thresholds are what the price needs, and they agree with my signposts below.

Where I come down

The cost discipline moved me: I gave margins more credit than the first-half run rate supports. The agent story didn’t, because it adds no earnings yet and management says it carries no extra fee. A gross profit share above 1.55% of GMV for four quarters, or any paid agent product with a disclosed rate, would raise my values.

VIIIRisks, and what would change my view

Agents. Covered above. Stripe and OpenAI hold one protocol, and Shopify co-built another. If shoppers settle inside one or two assistants, the assistant can ask for a cut, and that cut comes from somewhere.¹⁴⁷

Payments dependency. Stripe and PayPal support Shopify Payments on agreements that renew yearly.³ Stripe is both a supplier and a rival for the same merchants. Regulation of interchange or of payment facilitators would hit the largest revenue line directly.

Credit. The book is $2.4B gross and grew 23% in six months, and transaction and loan losses rose 76% in Q2.² Canadian rules have already changed the product once.² The buyer here is a small merchant in a downturn.

Consumers and trade. Merchants sell mostly to consumers, and 63% of revenue is American. Shopify’s own forward-looking statements name tariffs and consumer spending as risks.²⁷³

Capital allocation and control. A founder with 40% of the votes can make another Deliverr. He also made this company. Key-person risk is real: the 10-K names him, and he also runs research and development.³⁸

Portfolio noise. $2.7B of the stakes move with public share prices and $1.6B are private holdings carried at observed prices.² A fall takes a few dollars a share off value, and much more off reported net income.

AI costs. Cloud costs including AI usage rose $37M in subscription cost of sales in Q2, and computer costs in R&D rose $34M.² These are still small. Free AI tools for merchants could become a real line.

What would raise my base value: GMV growth above 25% through 2027, gross profit above 1.55% of GMV for four quarters, or a disclosed agent fee. What would lower it: GMV growth under 20% in the next two years, gross profit under 1.40% of GMV, or the share of loans 180 days late above 7%. The first test comes on 3 November.³¹