Shopify Inc.

01 — Executive summaryA profitable compounder trading like a broken story

Shopify reported revenue of $3.58 billion for the quarter ended 30 June 2026, an increase of 34% over the prior year and 33% in constant currency. Gross merchandise volume was $115.6 billion, up 32%. Gross profit rose 31% to $1.71 billion. Operating income rose 68% to $488 million. Free cash flow was $654 million, an 18% margin, up roughly 55% year over year. Growth of 30% or more was recorded on every one of those lines simultaneously — GMV, revenue, gross profit, operating income and free cash flow — for a second consecutive quarter. By any ordinary standard this is a business performing at the top of its peer group.

The stock is down approximately 19% year to date while the S&P 500 has risen more than 10%. It trades at $128.50 against an analyst consensus target of $168. The gap between the operating record and the share price is the subject of this report, and we do not think it is explained by a deterioration in the business. It is explained by a single question that the market cannot yet answer: when an AI agent does the shopping, whose storefront does the buyer visit?

Shopify's answer is that the storefront becomes the agent's data source rather than its victim. The company has spent two years building an infrastructure layer for this — a structured product catalog of more than a billion items, a shared protocol for agent-to-merchant transactions, and an AI assistant embedded in the merchant's administration console — and it reports that AI-attributed traffic and orders both tripled year over year in the June quarter. It also reports that those transactions carry exactly the same economics as an ordinary online order: no new fee, no separate take rate.

That last disclosure is the crux. If agentic commerce monetises at the same rate as everything else, then the entire thesis of Shopify's AI investment is defensive. It is a large, expensed cost incurred to prevent a channel that currently produces a trivial share of GMV from becoming a channel that produces a material share of it on someone else's terms. Management frames this as offensive — "Catalog will be one of Shopify's most important assets for years to come" — but from an owner's perspective, the honest description is that Shopify is paying to remain the layer that agents route through, and cannot yet charge more for the privilege.

Our rating is Neutral with a twelve-month target of $132, essentially the current price. That is an uncomfortable place to land on a company growing revenue in the mid-thirties with an improving cash conversion and a fortress balance sheet. We land there for three reasons.

First, the valuation already discounts a great deal. At 9.4 times forward sales Shopify trades at a 57% premium to the technology sector and roughly three times Amazon's multiple. Roughly 70% of the enterprise value sits beyond the next five years of modelled free cash flow. A company can be excellent and still be priced for more excellence than it can plausibly deliver.

Second, the AI investment is real cost with no visible price. Agentic transactions monetise at the same rate as ordinary ones. Sidekick's model costs sit inside subscription gross margin, which the company has held flat only through offsetting efficiencies. The accounting is honest, but it means the market cannot observe whether the AI strategy is working in the only terms that eventually matter — incremental revenue per unit of incremental cost.

Third, the credit book is growing faster than the disclosure that surrounds it. Loans and merchant cash advances have grown from $0.69 billion at the end of 2023 to $2.18 billion at 30 June 2026, and transaction loans and losses have risen from 0.9% to 3.9% of revenue over a similar period. Management attributes the increase to volume rather than deterioration and says both payment and credit loss rates are at normalised levels. That is plausible. It is also unverifiable from the outside, in a business that is lending against the cash flows of small merchants, at a moment when the US consumer is carrying more tariff-driven cost than a year ago.

None of those three observations is a prediction of failure. Each is a reason why the risk-adjusted return at $128.50 does not compensate for the range of outcomes. Our bull case — agentic commerce becomes a genuine incremental demand channel and the multiple holds — produces $198. Our bear case — agents commoditise the storefront and the multiple compresses — produces $82. With the base case at $132, the distribution is roughly symmetric around a price that already reflects a lot of good news.

The remainder of this report works through the evidence. Section 2 describes what the business actually is, which is more subtle than the "e-commerce platform" label suggests. Sections 3 through 8 work the financial record, including a section on the accounting items that make reported earnings noisier than they look. Sections 9 and 10 cover payments and the lending book. Sections 11 through 15 work the agentic strategy and the geographic expansion in detail. Sections 16 through 19 cover competition, regulation, and the reason the shares have fallen. Sections 20 through 25 contain the valuation, the cases, and the conclusion.

02 — What Shopify actually isA toll road, not a marketplace

The most common misconception about Shopify is that it is an e-commerce company. It is not. It does not own inventory, it does not set prices, it does not hold title to the goods it helps sell, and it takes no position in the demand it mediates. It is a software and payments business that charges a fee for routing a merchant's commerce through its rails. The distinction matters because it determines what can go wrong.

A marketplace — Amazon, for instance — is a principal. It buys attention, aggregates demand, owns the customer relationship and bears the inventory and fulfilment risk. Its economics improve with scale because scale in a marketplace is self-reinforcing. Its exposure is concentrated: if consumers change how they shop, the marketplace loses the demand outright.

A toll road is an agent. Shopify provides the storefront software, the payment processing, the point-of-sale terminal, the lending, the tax calculation, the shipping labels and the analytics. The merchant owns the customer, the brand and the inventory. Shopify earns a cut of the transaction. Its exposure is not to any single merchant but to the aggregate volume of independent commerce — and, critically, to whether independent commerce continues to be a distinct thing.

Two numbers establish the shape of the business. The first is the take rate, the proportion of GMV that Shopify retains as revenue: 3.10% in the June quarter, up from 3.05% a year earlier. The second is payments penetration, the proportion of that GMV processed through Shopify Payments rather than a third-party gateway: 68%, up from 64%. Those two figures move the model more than any product launch, because they determine how much of each incremental dollar of merchant volume converts into Shopify revenue.

It is worth being precise about the difference between Shopify and its most obvious point of comparison, because the confusion is pervasive. Amazon's third-party seller services business and Shopify's merchant solutions business both reported roughly $150–170 billion of annualised GMV at various points. But Amazon's business is a closed marketplace where the seller's customer belongs to Amazon, and Amazon's fee is closer to 15% of the transaction. Shopify's business is an open layer where the customer belongs to the merchant, and its fee is roughly 3%. Same volume, radically different economics, and radically different strategic positions.

That composition is why the company has survived three distinct shocks that killed weaker platforms — the end of the pandemic demand surge, the 2022 rate shock that de-rated the entire software complex, and the current de-rating across application-layer software. In each case the merchant base absorbed the shock and Shopify collected a smaller toll on a still-growing volume of commerce. The model is resilient in a way that a marketplace is not, because it is diversified across the survival of millions of merchants rather than dependent on its own demand capture.

It is also why the company has a structural weakness that section 11 examines at length. A toll road only earns a toll if the traffic passes over it. If buyers begin their shopping inside a general-purpose AI assistant that answers a question and completes a purchase without ever visiting a storefront, the traffic still exists — but the road it uses may not be Shopify's.

The counter-argument, and the one the company makes most forcefully, is that the data required to complete such a transaction — accurate price, availability, dimensions, shipping options, return policy, tax treatment — must come from somewhere authoritative, and in independent commerce that somewhere is Shopify. We find that argument materially stronger than the market currently does. Whether it is sufficiently stronger is a different question, and it is the question valuation ultimately has to answer.

03 — The financial recordFive years of compounding, and the first two years of profit

Shopify's financial history divides cleanly into two phases. From 2021 to 2023 the company grew revenue rapidly while losing money on a GAAP basis, and the losses were large enough to raise the question of whether the model could ever produce durable profit. From 2024 onward the company has grown at a similar rate while producing expanding operating income and free cash flow. The transition happened faster than almost anyone modelled, and it is the single most important fact in the company's investment history.

Several things about that table deserve comment, because the headline numbers conceal more than they reveal.

The 2021 and 2022 net income lines are not meaningful. The $2.9 billion profit in 2021 and the $3.5 billion loss in 2022 are dominated by the mark-to-market treatment of Shopify's equity investment in Affirm, an investment that was revalued violently as rates moved. Neither year's GAAP earnings described the operating business. This is the first of several places where Shopify's reported results require adjustment, and it recurs every quarter — see section 7.

The margin decline is monotonic and structural. Gross margin has fallen from 53.8% in 2021 to 47.8% on a trailing basis. The company has been consistent and transparent about why: the fastest-growing parts of the business are payments and capital, both of which carry third-party cost. In 2021, subscription solutions — the high-margin software line — was 29% of revenue. It is now 23%. The mix has shifted toward the lower-margin line every year, and management has explicitly said it expects this to continue.

The operating leverage is real and large. Operating margin has gone from 3.7% to 17.8% in three years. This is not an artefact of a single year's cost control; it is the result of a deliberate decision, taken in 2023, to stop growing headcount faster than revenue. The company cut roughly 10% of its staff in 2022 and again in 2023 and has held the line since, with 7,600 employees supporting a $13.3 billion revenue run-rate. The revenue per employee figure, approximately $1.75 million, is among the highest in enterprise software.

What the table does not show is the shape of the quarterly data, which is more informative than the annual aggregates. Revenue has grown at a remarkably stable rate for six consecutive quarters — between 27% and 34% — and constant-currency GMV growth has sat in a 29-to-30% band for five quarters running. Management calls this "a tight band" and is proud of it, for good reason: it means the business is not one quarter away from a growth break, and it means the deceleration that investors fear is not visible in the actual data.

04 — Revenue is not GMVThirty times the volume, and the gap is the story

Shopify's most-quoted metric is gross merchandise volume, and its least-understood relationship is the one between that metric and revenue. In the June 2026 quarter, merchants sold $115.6 billion of goods through Shopify and Shopify recognised $3.58 billion of revenue. The ratio is roughly thirty-two to one.

Traders who follow GMV and investors who follow revenue are watching two different businesses, and the difference is not academic. GMV grew 32% while revenue grew 34%, which means the take rate rose slightly. But the composition of that rise matters: effectively all of the increase came from payments penetration and capital, both of which carry cost of goods, rather than from increases in subscription pricing, which does not.

This is why it is possible for Shopify to report a quarter in which volume grew 32%, revenue grew 34%, gross profit grew 31%, and gross margin fell — all simultaneously and all consistent with one another. The company is monetising a larger share of a larger volume, but the incremental monetisation is lower-quality revenue than the average of what it already earns. In a business growing at 30%, the incremental mix dominates the reported margin, which is why the reported margin falls even as the absolute profit rises.

There is a second observation buried in the GMV data that gets less attention than it deserves. Management disclosed that GMV per merchant varies by cohort in a specific and favourable way: the $25-million-plus cohort is the fastest-growing by size, and merchants in the $2 million to $25 million band contribute the most incremental GMV. The company's own framing is that these are overwhelmingly existing shops graduating upward rather than new large merchants being acquired. That is a much better growth engine than new logo acquisition, because it requires no incremental sales cost.

The concern with a GMV-centric model is that GMV is not audited the way revenue is, and it can be inflated by definitions. Shopify defines it as the total value of merchandise sold through its platform, including through third-party payment gateways, but the precise boundary — whether it includes cancelled orders, whether it includes taxes and shipping, whether it includes the B2B and offline channels — has varied. We treat GMV as directionally reliable and precisely unreliable, which is the appropriate posture for any non-GAAP metric that a company chooses to feature.

The practical implication for the model is this: the take rate is the single most important conversion factor in the business, and it has been range-bound between 2.9% and 3.1% for four years. If agentic commerce turns out to route more volume through Shopify, the take rate rises and the model compounds faster than revenue. If it routes volume around Shopify, the take rate falls and the entire GMV line becomes progressively less valuable. Everything else is detail.

05 — The margin mix problemWhy gross margin will keep falling and why that is not necessarily bad

Shopify's gross margin has fallen roughly 600 basis points over five years, from 53.8% to 47.8%. If the trend continues at the rate of the last two years, it reaches the mid-forties by 2028. Investors who treat gross margin as a proxy for pricing power read this as deterioration. We think that reading is wrong, but the reason it is wrong is subtler than the company's explanation suggests.

The company's explanation is mix: merchant solutions is growing faster than subscription solutions, and merchant solutions has a lower gross margin because payment processing requires paying interchange to card networks and issuers. That is accurate as far as it goes. Payment processing is a pass-through business with a thin spread — Shopify collects a percentage of the transaction and pays most of it away — and the more payments grow relative to software, the more the blended margin falls.

But there is a more useful way to think about it. Gross margin percentage is the wrong metric for a business whose revenue is growing 30% and whose absolute gross profit is growing 31%. What matters is gross profit dollars per unit of operating cost, because operating costs are largely fixed and largely paid in people rather than in cost of goods. Consider the two lines separately over the trailing twelve months:

Gross profit grew 80% over three years while operating expenses grew 22%. That is the entire operating leverage story, and it is why the falling gross margin percentage is a distraction: the company is converting a growing share of a growing revenue base into gross profit dollars faster than it is adding the cost required to service it. The percentage compresses because the denominator includes a pass-through cost line; the absolute profit expands because the numerator is real.

There is nonetheless a genuine risk in the mix shift, and it is not about the margin line. It is about capital intensity. Subscription solutions requires essentially no capital to scale — the marginal merchant costs a few dollars of hosting. Merchant solutions, and particularly Shopify Capital, requires balance sheet. Every dollar of lending requires a dollar of funding, and every dollar of lending carries a risk of loss. A business that grows by shifting toward lending is a business that grows by taking more balance-sheet risk, and that risk does not appear on the income statement until it materialises.

06 — Operating leverageThe single strongest thing in the file

If there is one number in Shopify's disclosure that justifies the premium multiple, it is not the growth rate. It is the trajectory of operating income. In the first quarter of 2025 the company earned $203 million of operating income. In the second quarter of 2026 it earned $488 million. Over five quarters that is a 140% increase, on a revenue base that grew roughly 35%.

The mechanism is headcount discipline. Shopify employs approximately 7,600 people. That is fewer than it employed in 2022, when revenue was less than half its current run-rate. The company cut headcount sharply in 2022 and again in 2023, and — unusually — has not rebuilt. Most software companies that cut during the 2022–2023 period resumed hiring as conditions normalised. Shopify did not, and the decision to hold the line is the reason operating margin has expanded from 3.7% to 17.8% in three years.

Whether that discipline persists is one of the least-discussed and most consequential open questions. The company is making large investments in AI infrastructure — model inference for Sidekick, the catalog, the agentic integrations — and it is doing so with a headcount that has not grown. Management's framing is that AI leverage substitutes for hiring: "we've moved from a place of just reflexive use of AI to a place of AI leverage," as the CFO put it, describing an internal policy of using frontier models where they matter and cheaper distilled models where they do not.

That framing is plausible, and the evidence so far supports it — operating expenses fell from 37% of revenue in Q3 2025 to a guided 33–34% in Q3 2026, even as AI investment scaled. But it also creates a structural exposure: if the company has genuinely substituted AI for headcount, then its cost base now has a direct, variable link to AI inference pricing. Software businesses with fixed headcount cost have unusually predictable margins. A software business that has swapped headcount for inference has reintroduced a variable cost that it does not control.

The company has quantified this only once, and obliquely. The majority of AI costs relating to Sidekick sit inside subscription solutions gross profit, and subscription gross margin was held at just under 80% in the quarter despite usage scaling. The implication is that either inference costs are smaller than feared, or offsetting efficiencies are masking them, or both. Section 13 examines the mechanics.

07 — Quality of earningsReported net income swings by a billion dollars on an item unrelated to commerce

Shopify's GAAP net income is not a reliable measure of what the business earned in any given quarter, and the reason is a single line item. The company holds equity investments — most prominently its stake in Affirm Holdings — and marks them to market through the income statement. When those stakes rise, reported profit balloons. When they fall, reported profit collapses.

The recent record is instructive. In the first quarter of 2026 Shopify reported a net loss of $581 million. In the second quarter it reported net income of $1.50 billion. The swing is $2.1 billion across two consecutive quarters in which the underlying commerce business grew at a consistent 30–34% and produced consistent operating profit. The company disclosed that mark-to-market gains on equity investments, net of tax, contributed $1.06 billion to the second quarter, against $568 million in the prior-year period.

Strip those gains out and the picture is far more stable. On a Farstar construction — reported net income less the disclosed equity investment gains, net of tax — Shopify has earned approximately $430 million in each of the last three quarters, growing at a rate in the high twenties. That series is the one that describes the business. The reported series describes the market's appetite for consumer credit, which is a different thing entirely.

One further accounting observation deserves attention, because it is unusual. Shopify's free cash flow exceeds its operating income, and the gap is large. On a trailing basis operating income was $2.37 billion and free cash flow was $2.35 billion — approximately equal. But free cash flow is struck after tax and capital expenditure, which means the operating cash flow line contains a substantial working-capital benefit that offsets both. Section 8 works through where that benefit comes from.

08 — Free cash flow and its texture$2.35 billion a year, and part of it is a float

Free cash flow is the strongest number Shopify reports and the one we weight most heavily. On a trailing basis the company generated $2.35 billion, a 17.7% margin, up from $2.01 billion and a 17.4% margin the prior year and from $905 million in 2023. In the June quarter alone it produced $654 million at an 18.3% margin, up roughly 55% year over year.

The guidance for the third quarter is a high-teens-to-low-twenties percent margin, inclusive of a sub-one-point tailwind from the cash-flow reclassification. That is a wide range, and the width is telling: it implies management does not have precise visibility into the quarter's cash conversion, which in turn implies the working-capital line is volatile.

The bridge above shows where the money comes from and where it goes. Of the 47.8 points of gross margin available on a trailing basis, roughly 25.5 points are consumed by research and development, sales and marketing, and general and administrative expense, leaving 17.8 points of operating margin. Taxes and other items consume about 6.9 points. Capital expenditure is negligible at 0.2 points — Shopify is an almost pure software business on the asset side — and working capital and other items contribute about 7 points. The result is a 17.7 point free cash flow margin.

That 7-point working-capital contribution is the item to understand. Shopify processes payments on behalf of merchants and settles with them on a lag. In a business whose GMV is growing 32% a year, the pool of money held between receipt from the buyer and payment to the merchant grows mechanically — and that growth appears as operating cash flow. It is not fictitious cash; it is genuinely available for a period. But it is a function of growth rather than of profitability, and it reverses if growth stalls. A useful mental model: roughly 40% of Shopify's free cash flow margin in a strong quarter is a consequence of the business getting bigger, not of the business being more efficient.

This is not a criticism unique to Shopify — every payments business has this characteristic, and Stripe, Adyen and PayPal all exhibit it. It is simply the reason we would not extrapolate a 20%+ free cash flow margin as a permanent state. Management declined to do so when asked directly, saying only that it continues "to drive more and more leverage through the system." That was the right answer, and we treat the high teens as the more likely steady state.

09 — Payments and the take rateThe most reliable engine, and the one that costs the most to run

Shopify Payments is the largest and most important component of merchant solutions, and the most reliable growth engine in the company. Penetration reached 68% of GMV in the June quarter, up from 64% a year earlier, and gross payments volume was $78.1 billion. Over two years penetration has risen one percentage point per quarter, almost mechanically.

The mechanical quality of that progression is unusual and reassuring. Shopify does not need to win a competitive battle to grow payments penetration; it needs to keep signing merchants and keep improving the geographic coverage. Each new country where Shopify Payments goes live converts a portion of that country's GMV from a third-party gateway to Shopify's own rail, and Shopify keeps the spread rather than paying it away. Payments are now live in 40 countries, with European penetration up more than 350 basis points year over year off a much lower base and the United Arab Emirates added during the quarter.

There is a second-order argument here that matters more than the penetration number. Payments gross margin is higher internationally than in the United States, because international mix skews toward debit transactions and carries lower interchange. So the international expansion is not merely additive to revenue; it is accretive to payments margin at the same time. That is a favourable interaction that the market has not, in our view, priced.

The countervailing consideration is the one that drives the margin discussion in section 5. Payments revenue is the lowest-quality revenue Shopify earns: it is high-volume, high-cost-of-goods, thin-spread, and structurally exposed to interchange regulation. Every incremental point of payments penetration raises revenue and lowers the blended gross margin. The company has traded margin percentage for margin dollars, which is the correct trade at this stage of the business, but it means the reported margin line will keep looking worse for as long as the strategy continues to work.

Two adjacent payments products deserve mention because they change the character of the franchise. Shop Pay grew GMV 53% year over year and passed $400 billion in cumulative accelerated GMV. Shop, the consumer application, grew native GMV 70% year over year, with the cart-sync feature accounting for more than 30% of Shop application GMV. Both are strategically significant for a reason that is not obvious from the growth rates: they are the only places where Shopify has any relationship with a buyer rather than a merchant. In a world where agents mediate demand, owning any part of the buyer-side identity is worth a great deal more than the revenue either product currently produces.

10 — Shopify Capital and creditThe fastest-growing line in the file, and the least disclosed

Shopify Capital lends money to merchants, either as a term loan or as a merchant cash advance repaid as a percentage of daily sales. The product is strategically elegant: it uses the payment data Shopify already has to underwrite merchants that traditional lenders cannot assess, and it increases the merchant's dependence on the platform. It is also, from a risk standpoint, the least transparent part of the company's disclosure.

The balance sheet line has grown from $690 million at the end of 2023 to $1.78 billion at the end of 2025 to $2.18 billion at 30 June 2026 — an increase of roughly $400 million in six months, against a book that was $1.06 billion at the end of 2024. The growth rate is approximately 50% per year, materially faster than either revenue or GMV.

Meanwhile the loss line has grown faster still. Transaction loans and losses reached 3.9% of revenue in the June quarter, up $80 million year over year to $141 million, from roughly 0.9% of revenue three years ago. Management's explanation is that the increase reflects the growth of the capital book and the payments business rather than deterioration in credit quality, and that both payment and credit loss rates are at normalised levels. The company also said capital, not credit, was the larger driver of the increase this quarter, and that it is improving its measurement and forecasting to keep loss rates low as adoption grows.

We take that explanation at face value while noting two things that make it difficult to verify. First, Shopify does not disclose the vintage-level performance of its loan book — no cohort-by-cohort loss curves, no delinquency buckets, no allowance coverage ratio presented in a form that allows trend analysis. Second, the entire book is lending to small and medium merchants, a category that historically experiences elevated losses in the late stage of a consumer cycle, and the current environment includes tariff-driven input cost inflation for exactly this cohort.

For modelling purposes we assume the loss rate stabilises in the 4% to 5% range rather than continuing to rise, and we do not assume the lending book becomes a source of incremental profit. At $2.2 billion it is 1.3% of the market capitalisation. It is not large enough to sink the equity, but it is large enough to matter to a multiple that already assumes clean execution everywhere.

11 — Agentic commerceThe thesis, the evidence, and what is actually disclosed

This is the section that matters most. Every other line in the financial statements is performing well and is reasonably well understood. The reason Shopify's multiple has compressed from 18 times sales at the end of 2025 to 12.5 times today is that the market has re-priced the probability that AI agents will insert themselves between the merchant and the buyer — and that the insertion will happen on the agent's terms rather than Shopify's.

The bull argument in its strongest form goes like this. An AI agent completing a purchase requires structured, accurate, real-time product data: price, availability, variants, shipping cost and time, tax, return policy, and increasingly richer attributes such as fit, material and suitability for a use case. Scraping a website produces unreliable data. The authoritative source of that data for independent merchants is the commerce platform. Shopify has built the structured catalog and the transaction protocol, and it has done so with a two-year head start and a billion-product corpus. If agents become the interface, Shopify becomes the data layer behind the interface, and the take rate survives because the transaction still has to be processed.

The bear argument is equally coherent. If buyers transact inside a general-purpose assistant, the merchant's brand, storefront and customer relationship all degrade to a supplier record. The assistant becomes the demand aggregator, exactly as Amazon did to retail two decades ago, and extracts the economics accordingly. In that world Shopify's catalog becomes a commodity input — necessary but not valuable — and the platform is squeezed toward a lower take rate, the way payment processors were squeezed toward a smaller spread.

What does the evidence actually show? The company disclosed the following for the June quarter:

Several of those disclosures are genuinely favourable, and one is the most important number in this report. The 75% figure is the strongest single argument for the bull case. If three-quarters of agent-mediated orders are landing in product categories outside the top 100, then agents are not simply redirecting demand that already existed — they are surfacing long-tail inventory that conventional search would never have ranked. That is incremental demand, and incremental demand is the only kind that grows a platform without cannibalising it.

The 2.5× product-page landing rate is also meaningful. Agent-referred buyers arrive at the specific product rather than at a homepage or a category page. That is higher-intent traffic, and it explains the elevated new-buyer order rate — an agent that has already resolved the comparison sends a buyer who is ready to transact rather than one who is still exploring.

But the disclosure that constrains the whole thesis is the one the company states most plainly and inverts most carefully: agentic transactions carry identical economics to ordinary transactions. No new fee. No separate pricing. This means that in the current construct, agentic commerce is revenue-neutral per unit of GMV. If it grows, Shopify's revenue grows proportionally with GMV. If it does not grow, nothing is lost. What it does not do is expand the take rate — and expanding the take rate is the only mechanism by which the AI investment produces a return above the return on ordinary volume.

There is a defensible strategic logic to monetising agentic commerce at par. Charging a premium for agent-mediated orders would push merchants and agents toward alternative rails, and Shopify's priority is adoption — becoming the default. Standard commercial practice: subsidise the new channel until it is structurally necessary, then re-price. The question is whether the re-pricing is possible once the channel is established, or whether by then the buyer relationship has moved somewhere Shopify cannot reach.

12 — Catalog and the discovery layerThe asset the company says it owns, and why that claim is testable

Shopify Catalog is a structured index of products drawn from the merchants on the platform — more than a billion items, per the company — normalised into a schema that AI agents can query. Management describes it in unusually absolute language: "Catalog will be one of Shopify's most important assets for years to come… it is the discovery engine for the future and Shopify built it and owns it."

The claim deserves scrutiny because it is the load-bearing element of the entire bull case. If Catalog is a durable, defensible asset, Shopify sits at the centre of agentic commerce and the multiple is too low. If Catalog is a commodity data feed that any sufficiently motivated competitor can replicate, the position is much weaker. Three things make the claim testable.

The conversion differential. Shopify claims Catalog-powered AI searches convert at twice the rate of scraped data, and roughly 80% higher than traditional organic search. If that differential is real and persistent, it means agents have an economic incentive to route through Catalog rather than scraping — the same logic that made Google's index more valuable than the open web it indexed. If the differential decays as scraping improves, the advantage erodes.

The attribute depth. The company is investing in what it calls taste-driven attributes — whether a garment is formal enough for a wedding, whether a fabric is breathable, whether it wrinkles, whether a shoe suits an endurance run. These attributes cannot be derived from a product description page; they must be either supplied by the merchant or inferred from aggregated transaction and return data. Shopify has the second. This is a genuine moat if the inference is good, because it is the kind of data that improves with volume and cannot be bought.

The identity layer. Catalog integrates with Shop sign-in, which means an agent querying Shopify can recognise a returning buyer and personalise on that basis. Shopify states that no other catalog API can do this. That is a strong and specific claim — it points to the Shop consumer application and the payment credentials attached to it as the mechanism, and it is the reason the Shop and Shop Pay growth numbers in section 9 matter more than they appear to.

Against those, the commodity risk is real. Structured product feeds are not technically difficult to build, and the major agent platforms have strong incentives to build or buy their own. Google has a product catalog and a shopping graph. OpenAI has Instant Checkout and the Agentic Commerce Protocol, developed with Stripe. Amazon has the largest product catalog in existence. Each of those parties would prefer not to depend on Shopify for the data their agents need. The question is whether they can assemble an equivalent corpus with equivalent attribute depth, and the honest answer is that they can build something — but not with Shopify's merchant-supplied attribute data, and not with Shopify's return-rate history.

We would characterise Catalog as a probable but unproven moat. It is the strongest strategic asset in the company and the one whose value is hardest to quantify. The Universal Commerce Protocol, which Shopify co-developed and which every Shopify merchant is ready to support, is the distribution mechanism for Catalog — and the fact that it is a shared protocol rather than a proprietary one is both the reason it may achieve adoption and the reason it may not capture the value.

13 — Sidekick and switching costsThe metric that matters is not the one being reported

Sidekick is an AI assistant embedded in the Shopify administration console. It answers questions about the merchant's business, builds themes, writes product descriptions, analyses performance and creates custom applications. The reported usage figures are striking: 34 million conversations in the June quarter, daily active merchants up 3.6 times year over year, daily sessions up 4.8 times, and 36,000 custom apps created in the quarter against 12,000 in the prior quarter — a tripling in three months.

The reason we care about the custom-app figure more than the conversation count is what it implies about switching costs. A merchant that has built internal workflows, data connections and storefront features by describing them conversationally inside Shopify's admin has accumulated a set of dependencies that do not travel to another platform. Every custom app is a small increase in the cost of leaving. Aggregate 36,000 of them in a quarter across a merchant base, and the platform is quietly converting a software subscription into infrastructure.

The usage-pattern data reinforces this. Shopify disclosed that in a merchant's first thirty days, roughly half of Sidekick conversations concern setup, design and theme configuration. By year five, setup falls to about 8% while analytics and reporting exceed 40%. The same product is being used for a completely different job as the merchant matures — which means Sidekick's addressable utility grows with the merchant rather than being consumed at onboarding. That is an unusually favourable retention dynamic, and it is not something a competitor can easily replicate without the merchant's operational history.

14 — Plus, enterprise and offlineThe migration argument, and why it is credible

Shopify's historical weakness was that it served small merchants well and large ones poorly. That has changed materially over the last two years, and the change is visible in the numbers. Shopify Plus — the enterprise tier — represented 34% of monthly recurring revenue in the quarter and grew 19%. B2B GMV grew 76%. Offline GMV, meaning point-of-sale transactions, grew 32%.

The named-logo list is more informative than the aggregates. Brands reported as newly live or migrating include e.l.f. Cosmetics, Claire's, Burton, Suitsupply, Barnes & Noble, General Motors, Amer Sports — the parent of Wilson, Salomon and Peak Performance — Alo, Vuori and FIGS. Guess, Fred Segal and Avon were cited as new or in migration. Balmain migrated its storefront in what management described as a matter of weeks. In retail infrastructure, a fast migration is not a convenience feature; it is the primary purchase criterion, because the cost of a botched platform migration is measured in lost holiday revenue.

Management's framing of the enterprise pitch is worth quoting because it is the sharpest articulation of the strategy in the transcript: the old model of enterprise commerce required a migration every few years, and the pitch is that this is "their final migration." That is a claim about architectural completeness — that Shopify has enough depth in B2B, POS, payments, multi-currency, tax and fulfilment that a large retailer never needs to leave.

The cross-sell motion is the mechanism. A brand typically enters through one channel — often the online storefront — and then expands into unified commerce: POS in physical locations, B2B for wholesale, payments, and agentic selling. Shopify cited the retailer Arhaus as the pattern: online first, then offline. For business models that are not a clean fit for full migration, such as certain marketplaces, the company is willing to sell Shop Pay as a standalone commerce component and expand from there.

15 — InternationalThe second engine, and the one management talks about most

Shopify's management discusses international expansion with a frankness that is unusual in investor communications. The President has said on multiple calls that he "gripes about this all the time" and that the company has captured less than 1% of global retail sales. The frustration is a statement of opportunity rather than of failure, and the recent data supports the optimistic reading.

In the June quarter, international GMV grew 37% and European GMV grew 34% in constant currency — both ahead of the 28% growth in North America, and both against tougher prior-year comparisons. Payments penetration in Europe rose more than 350 basis points year over year even though many of the served countries launched only in the prior year. Shopify Payments launched in the United Arab Emirates, bringing the total to forty countries, and the company added its first local payment method in Mexico.

The structural change announced in the quarter is more consequential than any single market launch. Managed markets — the product that handles cross-border tax, duties, currency and localisation for merchants selling internationally — was expanded beyond United States merchants for the first time, to Canada and the United Kingdom. Historically, international expansion meant selling to merchants located outside the United States. Managed markets means enabling United States and other merchants to sell into foreign markets, which is a much larger addressable opportunity because it monetises existing merchants rather than requiring new ones.

There is also an interaction between international expansion and the agentic thesis that is worth making explicit. Shopify disclosed that 75% of AI-attributed orders came from outside the top 100 categories, which is a long-tail phenomenon. Long-tail inventory matters most in markets where conventional retail distribution is thin — precisely the international markets Shopify has not yet penetrated. If agents are genuinely better than keyword search at surfacing obscure products, the value of that capability is greatest where the product is hardest to find through conventional channels. That is an argument for international and agentic investment being complementary rather than competing uses of capital.

The scale of the opportunity is best expressed through the share-of-market numbers, which the company itself cites. Shopify represents just over 14% of the United States e-commerce market, and management cites data indicating that since the start of 2025 Shopify merchants have captured nearly half of all incremental United States e-commerce dollars. Those two figures describe very different things, and the gap between them is the point: a platform with 14% of a market can be winning 48% of the market's growth if the incumbents are standing still. Management says international results are on a similar early trajectory.

We would treat the incremental-dollar figure with more caution than the company does. It is sourced from eMarketer rather than from audited data, the definition of "incremental" is not disclosed, and the comparison implicitly counts a market — United States e-commerce — in which Amazon holds roughly 40% of the units and against which Shopify mostly does not compete directly. A platform can capture half of the growth in the segment it serves without capturing half of the growth in the market.

16 — CompetitionThe peer set is not growing, and that is the point

Shopify's direct competitors in commerce software are Wix.com, Commerce.com (formerly BigCommerce and Web.com), and a long tail of regional and open-source platforms. The comparison is uncomfortable for the competitors rather than for Shopify.

Shopify grows roughly three times faster than Wix and four times faster than Commerce.com at a comparable or superior operating margin. Over the period surveyed, Commerce.com returned 73.8% and Wix 12.1% while Shopify fell — which is a striking illustration of how the market has priced this sector. The smaller competitors have been re-rated upward on the theory that they are cheap and could be acquired, while the market leader has been de-rated on the theory that its category is structurally threatened. Both cannot be right.

Commercially, the dynamic favours Shopify. Wix competes primarily on price and ease of use for very small businesses, a segment where Shopify's entry price of $39 a month is not disadvantaged. Commerce.com competes on enterprise flexibility for mid-market brands, a segment Shopify has been winning with Plus, B2B and unified commerce. Neither has the payments infrastructure, the capital book, the catalog, or the merchant data that Shopify has. Both are materially smaller and both are growing slower.

The real competitive question is not about commerce software at all. It is about whether the next generation of commerce happens on a platform at all, or whether it happens inside assistant interfaces and social applications where Shopify is a backend rather than a surface. That is section 17.

17 — Amazon, Temu and the demand shockThree different threats wearing one coat

Three distinct competitive pressures are frequently bundled together in discussion of Shopify. They have different mechanisms, different time horizons and different implications.

Amazon: the channel competitor

Amazon operates services explicitly designed to serve merchants off its own marketplace: Buy with Prime, which lets a merchant offer Prime shipping on its own site; Multi-Channel Fulfilment, which uses Amazon's logistics network for orders from other channels; and Amazon Pay. Collectively these let a merchant sell outside Amazon while using Amazon's infrastructure — and therefore let Amazon capture a share of commerce it does not host.

The strategic intent is clear and the competitive overlap with Shopify is partial rather than total. Amazon's offer is a bundle of logistics and payment capability; Shopify's is a complete commerce operating system. A merchant choosing Buy with Prime still needs a storefront, an inventory system, tax handling and analytics — and the most likely place for those to come from is Shopify. The realistic outcome is coexistence: Amazon takes the fulfilment and payment spread, Shopify keeps the platform, and the merchant pays both. That is not a disaster for Shopify, but it is a leak from its take rate.

Temu and Shein: the demand competitor

Temu and Shein compete with Shopify's merchants for consumer spending rather than for platform share. Their model depends on direct-from-factory sourcing and cross-border parcel shipping, and it was built in substantial part on the United States de minimis exemption, which exempted shipments below $800 from duty.

That exemption has now been curtailed. As of 2026, low-value shipments from all origins are subject to duty under a phased regime, following the earlier restriction applied to Chinese origin goods and the subsequent extension to all countries. The effect on Temu and Shein's unit economics is significant, and both have been adjusting — shifting toward local fulfilment, raising prices, and in Temu's case moving toward a marketplace model that recruits local sellers.

The implication for Shopify is favourable at the margin, and it operates through two channels. First, a competitor whose cost advantage was regulatory has lost part of that advantage. Second, and more subtly, the tariff regime raises landed cost for all import-dependent small merchants, which is a headwind to the long tail of Shopify's merchant base. The net effect is ambiguous: Shopify's competitive position improves relative to cross-border discounters while its merchants' margins compress. On balance we read it as roughly neutral for GMV and modestly negative for merchant quality.

The agent platforms: the structural competitor

The genuinely novel competitive threat is not a retailer at all. It is the general-purpose assistant that mediates the purchase. OpenAI's Instant Checkout, built on an Agentic Commerce Protocol developed with Stripe, allows ChatGPT users to complete transactions without leaving the conversation. Google operates a shopping graph and a universal commerce initiative of its own. Both are attempting to become the place where commerce is decided, leaving the platform as the place where it is merely processed.

This is the threat the market is pricing, and it is the reason Shopify's multiple has compressed even as its results improved. We address the evidence in section 11, but the structural point belongs here: an intermediary that controls discovery controls the economics of the value chain, and Shopify has never controlled discovery. It has always depended on merchants to generate their own demand through paid advertising, search, social and email. If discovery migrates from those channels into assistants, Shopify's merchants lose the ability to own their demand, and Shopify's take rate becomes renegotiable at exactly the moment its position is weakest.

18 — Regulation and tariffsA light regulatory load and a heavy macro one

Shopify's regulatory exposure is materially lighter than that of the large technology platforms. The company is not designated a gatekeeper under the European Digital Markets Act, is not the subject of a United States antitrust action, and does not operate a consumer-facing network at the scale that attracts content regulation. Its principal regulatory surfaces are payments and lending, both of which are supervised and both of which constrain the business proportionately rather than existentially.

Payments. Interchange fees are regulated in several jurisdictions, most significantly in the European Union and in the United States for debit. Lower interchange compresses the spread Shopify earns on processing, though the company's payments economics are less dependent on interchange than a pure acquirer's because it captures value across the platform. The more relevant regulatory trend is the expansion of instant payment rails and account-to-account payments, which bypass card networks entirely and could in time reduce the cost base of processing — a tailwind rather than a headwind, if it arrives.

Lending. Shopify Capital is subject to consumer and commercial lending law in each market and to disclosure requirements around merchant cash advances. The product's structure — repayment as a percentage of sales rather than a fixed schedule — has attracted regulatory attention in some jurisdictions on the question of whether it constitutes a loan subject to usury limits. Shopify has managed this without material incident, but a book growing at 50% a year invites scrutiny.

Tariffs. The most consequential policy development for Shopify is not regulation of the company at all; it is the collapse of the de minimis exemption and the broader rise in tariff barriers. Shopify's merchants are disproportionately small, import-dependent businesses selling consumer goods. The tariff regime raises their input costs, compresses their margins and, for the subset that imports directly, raises the cost of the goods they sell. The company does not bear this cost, but its merchants do, and merchant health is the ultimate determinant of GMV.

We would note one offsetting consideration. Tariff complexity is itself a reason for small merchants to consolidate onto a platform that handles duty calculation, landed cost, and cross-border compliance natively. Shopify has invested in exactly those capabilities through managed markets. A more complicated trade environment raises the value of the software that navigates it. That argues for tariff risk being partly self-hedged at the platform level while remaining a genuine headwind at the merchant level.

19 — Why the stock fellNineteen percent down in a year the market rose ten

Shopify's share price has fallen approximately 19% since the start of 2026 while the S&P 500 has gained more than 10%. The stock peaked at $158.70 in August and closed at $128.50 on 18 September. There is no operational event in the period that explains the decline; the June quarter beat expectations on every line.

The decline decomposes into three components, in descending order of importance.

First, a sector-wide de-rating of application-layer software. The market has spent 2026 gradually re-pricing the entire software complex on the thesis that AI agents will compress the economics of applications that sit between a user and a task. This has been called the "SaaSpocalypse," and it has been indiscriminate. Companies with no meaningful AI exposure have fallen alongside companies whose entire business is AI. Shopify is not an application in the sense the thesis describes — it is infrastructure — but it is classified as software, it has a high multiple, and it has been sold accordingly.

Second, specific concern about the agentic position. Shopify's own disclosures invite the question. If agentic transactions monetise at par, and if the absolute size of the channel is undisclosed, then the market cannot verify that the AI investment is producing a return. A company that tells investors it is making a large strategic bet, that the bet is working, and that the bet cannot be measured in revenue, is asking for trust at the exact moment trust is scarce. We think the disclosure is honest; we also think it is the reason the multiple has compressed.

Third, growth deceleration optics. The guidance for the third quarter is low-thirties revenue growth against 34% reported — a deceleration, even if a small one — plus gross profit growth in the mid-to-high twenties against revenue growth in the thirties. Investors who model on gross profit rather than revenue see a company decelerating on the metric that matters more. And the fact that this guidance was issued alongside a beat made it read as conservatism about the future rather than precision about the present.

What has not happened is worth stating clearly. Revenue estimates for 2026 and 2027 have been stable for thirty days. The consensus third-quarter earnings estimate of $0.44 has not been revised. Full-year 2026 consensus earnings of $1.89, implying 61.5% growth over 2025, remain unchanged. Analysts have not cut numbers; they have cut multiples. In September, Bernstein raised its price target to $160, Piper Sandler to $180, Rosenblatt to $175 and RBC to $180. The consensus target is $168 against a price of $128.50.

20 — ValuationNine point four times forward sales, and what has to be true

Shopify's valuation is high on every conventional measure, and the honest analytical task is not to argue about whether it is high but to specify what it requires. At $128.50 the company has a market capitalisation of $165.3 billion, a trailing price-to-sales ratio of 12.5 times, a forward price-to-sales ratio of 9.4 times, a trailing price-to-earnings ratio of 86.7 times, a forward price-to-earnings ratio of 61.5 times, and a trailing price-to-free-cash-flow ratio of 70.3 times. It trades at a 57% premium to the computer and technology sector's forward sales multiple.

The peer comparison is instructive but requires care. Wix.com trades at 1.98 times sales and Commerce.com at 0.88 times — but Shopify grows three to four times faster than either, at a comparable or better margin. Amazon trades at 2.93 times sales on a business whose margins are structurally lower. Sector medians blend companies with very different growth profiles. The appropriate conclusion is not that Shopify is expensive relative to peers, because it is a better business than most of them. It is that Shopify is expensive in absolute terms, and that the absolute level of the multiple is what creates the risk.

The structure of that table tells you what the market is doing. Roughly 70% of the enterprise value is attributable to cash flows beyond the next five years. For the current multiple to be justified, the business must compound free cash flow at approximately 25% annually for a decade. That is not an unreasonable assumption for a company growing revenue in the low thirties with 17–18% free cash flow margins and improving operating leverage — but it is an assumption that leaves no room for the growth rate to settle in the high teens, which is what a successful agentic disintermediation would produce.

Two more precise observations. First, the forward multiple of 9.4 times sales is the lowest Shopify has traded on that basis in six years, and it is the source of the bull argument that the stock is cheap relative to its own history. That argument is weak on its own terms: multiples mean-revert to what growth and rates support, not to what they were when rates were zero.

Second, the free cash flow multiple is the most defensible way to look at the company, because free cash flow is the metric least distorted by the equity investment accounting and the mix shift. At 48.6 times our 2027 estimate, Shopify is expensive against the broad market but not against a cohort of high-growth compounders, and it is materially cheaper on this basis than its P/E suggests. An investor who believes the growth rate persists should be looking at this line, not the earnings line.

Our scenario model produces a $116 spread between the bear and bull cases. We have set the scenarios on a single dominant variable — GMV growth and the multiple that growth supports — because in our judgement that is where the variance genuinely lies. To be explicit about the assumptions:

The base case of $132 assumes GMV growth settles in the high twenties, free cash flow margin expands modestly toward 19%, the take rate holds at approximately 3.1%, and the multiple settles near 8 times forward sales. On that path the stock is worth very close to what it costs today, which is why we rate it Neutral rather than either Buy or Sell. The interesting quality of the distribution is that it is roughly symmetric: the bull case is as large as the bear case, and both are material. That is a genuinely two-sided situation, and two-sided situations do not warrant conviction in either direction.

21 — The bull caseSix arguments, ordered by strength

One. The growth is real, broad and has not decelerated. Revenue has grown between 27% and 34% for six consecutive quarters. Constant-currency GMV growth has sat in a 29–30% band for five. Growth above 30% was recorded simultaneously on GMV, revenue, gross profit, operating income and free cash flow. This is not a story about a business losing momentum; the momentum is intact and management's guidance suggests it continues.

Two. The operating leverage is exceptional and largely untapped. Operating income quadrupled in five quarters on a revenue base that grew by roughly a third. Operating margin has gone from 3.7% to 17.8% in three years. There is no obvious reason the leverage stops here, given that headcount has been held flat while the business doubled, and AI tooling is substituting for hiring at the margin.

Three. Retention among the merchants who matter is near-total. Ninety-two percent five-year retention above $1 million of GMV; ninety-seven percent above $10 million. The 2015 cohort has grown fivefold. A platform with that retention profile has a compounding base that does not need to be re-won every year.

Four. Payments penetration is a mechanically reliable growth engine. One point of penetration per quarter, forty countries live, Europe up 350 basis points year over year off a low base, and international payments carrying a higher margin than domestic. The company does not need to win a competitive battle to grow this line; it needs to keep executing the country rollout.

Five. The balance sheet provides optionality that peers do not have. $5.29 billion of net cash, $178 million of debt, negligible capital expenditure, and a lending book funded from operations. Shopify can fund its own expansion, buy back stock, or acquire capability without touching the capital markets. In the current rate environment that is worth more than it was two years ago.

Six. Catalog is a genuinely differentiated asset with hard evidence behind the claim. A billion products, two-year head start, merchant-supplied attribute data, return-rate history, Shop identity integration, and a measurable conversion advantage over scraped data. If agentic commerce becomes the next interface for shopping, this is the most valuable asset in independent commerce, and Shopify owns it.

22 — The bear caseSix arguments, ordered by strength

One. The agentic bet has no visible price and no disclosed size. Agentic transactions monetise at identical economics to ordinary transactions, so the AI strategy cannot be observed as incremental revenue per incremental cost. The absolute size of the channel is undisclosed; the company says only that it remains small relative to total GMV. A large, growing expense with no measurable return, in a business being asked to justify a 9.4-times-sales multiple, is the core of the bear case.

Two. The multiple leaves no room for error. Nine point four times forward sales against a sector at 6.0, and a value score of F. Roughly 70% of enterprise value depends on cash flows beyond five years. If growth settles in the high teens, the compression is severe irrespective of how good the business is.

Three. Gross margin is structurally declining and management has confirmed it will continue. Six hundred basis points over five years, and guidance that gross profit growth will trail revenue growth by several points. The mix shift toward payments and capital is deliberate and correct at this stage, but it means the company is becoming a lower-margin business by design.

Four. The credit book grows at 50% a year with limited disclosure. Two point two billion of loans and advances, transaction losses up from 0.9% to 3.9% of revenue, no vintage-level loss curves and no delinquency disclosure in a form that permits trend analysis. The company's explanation is plausible and unverifiable.

Five. Discovery is the value chain's controlling layer, and Shopify does not control it. Every intermediary in commerce history has eventually been disintermediated by whoever controlled demand. Shopify sits beneath the demand layer, which protects it from displacement and exposes it to commoditisation.

Six. Retail conditions are deteriorating at the merchant level. Tariffs have raised landed cost for import-dependent small merchants; the de minimis exemption that enabled a generation of direct-import sellers is gone; and consumer spending at the low end of retail is under pressure. Shopify's merchants bear this cost before Shopify does, but GMV is ultimately a function of merchant health.

23 — Risk matrixOne axis dominates

The risk map below expresses our judgement about probability and impact over a twenty-four-month horizon. We have included two upside risks alongside the five downside risks, because the distribution is genuinely two-sided and presenting only the downside would misrepresent our view.

The two uppermost risks are the same risk seen from opposite sides, and that is the defining feature of this investment. If agentic channels do not disintermediate the storefront, Shopify's multiple is too low. If they do, Shopify's business model is impaired. There is no scenario in which the agentic question is immaterial, which is why the stock behaves like a binary and why the volatility is high.

24 — CatalystsWhat will move the shares

Q3 2026 results, 3 November 2026. The nearest and most informative event. Watch three things in order of importance: the actual revenue growth rate against the "low-thirties" guidance, whether gross profit growth trails revenue growth by more or less than the 300 basis points implied by guidance, and whether the company begins to quantify agentic commerce in absolute terms. Any of the three could move the stock several percent.

Disclosure of agentic commerce magnitude. This is the single largest potential catalyst and it is entirely within the company's control. If Shopify begins reporting the absolute share of GMV or orders attributable to AI channels, the market can price the thesis rather than speculate about it. We regard the absence of this disclosure as the most important information gap in the file, and its arrival as the most likely trigger for a re-rating in either direction.

Third-party data on AI-referred commerce. Independent measurement of the share of e-commerce traffic and orders arriving from assistant interfaces would let investors triangulate the company's claims. Several data providers are building this capability. Its emergence would reduce the information asymmetry that currently drives the volatility.

Competitive announcements from the agent platforms. New commerce capabilities from OpenAI, Google or Amazon that bypass platform data — particularly any move that establishes a rival structured catalog at scale — would be read as a direct threat. Conversely, an expansion of the Universal Commerce Protocol's adoption beyond Shopify's own merchant base would strengthen the position materially.

Rate policy and the software multiple. Shopify's multiple is more sensitive to the discount rate than most, because 70% of its value sits beyond five years. A sustained decline in long rates would lift the entire complex including Shopify, irrespective of company specifics.

Shopify Capital disclosure. Any deterioration in the lending book, or alternatively any decision to disclose vintage-level performance, would resolve a question that currently attracts a discount. We regard deterioration as the more likely of the two and the more consequential.

25 — ConclusionNeutral at $132, and why the answer is genuinely close

What would change our mind

We would move to a constructive rating on any two of the following three developments: the company beginning to disclose agentic commerce in absolute terms with a conversion advantage that persists over multiple quarters; gross profit dollars resuming growth at or above revenue growth, indicating the mix shift has stabilised; and free cash flow margin holding above 20% for two consecutive quarters without a working-capital tailwind.

We would move to a negative rating if the take rate began to fall, if transaction loans and losses exceeded 5% of revenue for two consecutive quarters, or if management disclosed a material slowdown in the growth of large-merchant GMV — which would indicate the enterprise migration strategy had stalled at the same moment the long tail came under pressure.

26 — AppendixFinancial tables

Appendix A — Quarterly income statement, $ millions

Appendix B — Segment revenue, $ millions

Appendix C — Key operating metrics

Appendix D — Scenario assumptions

Appendix E — Consensus and target dispersion

27 — SourcesPrimary documents and third-party data

This report is constructed from primary documents wherever possible. The following sources were used. Company figures are drawn from filings and earnings materials; where a number is a Farstar estimate or construction it is labelled as such at the point of use.

• Shopify Inc., Quarterly Report on Form 10-Q for the quarter ended 30 June 2026, filed with the US Securities and Exchange Commission.
• Shopify Inc., Annual Report on Form 40-F for the fiscal year ended 31 December 2025.
• Shopify Inc., Q2 2026 earnings press release and accompanying financial statements, 5 August 2026.
• Shopify Inc., Q2 2026 earnings conference call transcript, 5 August 2026, including prepared remarks by Harley Finkelstein, President, and Jeff Hoffmeister, Chief Financial Officer, and the analyst question-and-answer session.
• Shopify Inc., Q1 2026 earnings press release and conference call, for the comparative quarterly data used throughout.
• Shopify Inc., Q3 2022 through Q4 2025 quarterly earnings releases, for the historical series in Appendix A and Appendix B.
• Shopify Editions, Spring 2026 product announcements, covering the agentic commerce release set.
• stockanalysis.com — Shopify (NASDAQ: SHOP) financial statements, statistics and valuation ratios. Accessed 19–20 September 2026.
• Zacks Investment Research — "Shopify Plunges 13% in a Month: Buy, Sell or Hold the Stock?", 16 September 2026, for the forward P/S comparison, Value Score, consensus earnings estimates and competitive overview.
• Benzinga — "Shopify Stock Lags the Market, But Top Analyst Sees a Buying Opportunity", 13 September 2026, for analyst price targets and the Bernstein commentary.
• CNBC — "Shopify has been under pressure this year. Bernstein says the decline is overblown", 11 September 2026.
• eMarketer data on US e-commerce market share and incremental dollar capture, as cited by Shopify management on the Q2 2026 earnings call. Not independently verified.
• Market price and index data as at the close on 18 September 2026.
• United States executive actions and Customs and Border Protection guidance regarding Section 321 de minimis treatment, 2025–2026, for the tariff analysis in section 18.

Where a third-party figure and a company figure conflict, this report uses the company figure and states the discrepancy. Where a metric is company-reported but not disclosed in filed financial statements — including GMV, payments penetration, cohort retention, and the agentic commerce growth rates — it is attributed to management and identified as unverified. Three data providers report Shopify's analyst price target at different values ($167.72, $171.15, $172.77 and $168 depending on the sample); this report uses $168, which is the figure corroborated by the largest recent sample.

28 — DisclosureConflicts, limitations and revision policy