Growing faster than it was, priced as though it has stopped

Overall Investment CaseiOur conclusion in one of three words. Qualifies means the business and the price both work. Watchlist means the business may work but the price or the evidence does not yet. Does Not Qualify means we would not buy it.

Adobe sells the software that professional creative work is made with, and on the evidence it qualifies. The research timeframe is 5 years and beyond. The disposition is Qualifies. This is the first pick on this page to reach that conclusion, so it is worth saying plainly why, against the two tests that stopped the others. The business grows, and it is growing faster than it was. Revenue rose 12.9% in the quarter to 28 August 2026, to $6,760 million from $5,988 million. Over the nine months it rose 12.5%. Annual recurring revenue, which is the subscription base that renews, reached $27.50 billion at the end of August, up 11.2%. Contracted revenue not yet recognised stood at $22.16 billion. None of those is a currency effect. Adobe reports in dollars and earns mostly in them. The accounts are clean and the balance sheet is close to empty of risk. Gross margin was 88.7% in the quarter and operating margin 34.8%. Debt of $6,363 million against cash and short-term investments of $5,639 million leaves net debt of $724 million, against free cash flow of $9,852 million in the last full year. There is no investigation, no accounting dispute and no controlling shareholder. And the price has collapsed anyway. At $242.27 the shares are 33.4% below the 52-week high of $363.70 set on 28 October 2025, and down 30.5% over twelve months while a US market tracker rose 15.7%. They trade at 13.4 times the midpoint of this year reported earnings guidance and 9.9 times the adjusted figure. The market is pricing the risk that generative artificial intelligence makes these tools matter less. That risk is real and it is the subject of the bear case below. What the numbers do not yet show is any sign of it happening.

Why It Could GrowiThe products, plans or market changes that could raise what this company earns over the next several years. Management saying so is not enough on its own.

Artificial intelligence is adding revenue rather than taking it, so far. AI-first annual recurring revenue passed $650 million in the quarter, growing more than 150% on the year. That is small against $27.50 billion of total recurring revenue, and it is the fastest-growing thing the company has. Price and tier increases. A subscription business with 88.7% gross margins and a product customers use daily can raise prices, and the gap between 12.9% revenue growth and 11.2% recurring revenue growth is partly that. The enterprise side. Adobe sells marketing and analytics software to large companies alongside the creative tools, which is a different buyer with a different budget. Buybacks. This one grows earnings per share without growing the business at all. The company spent $6,820 million buying 26.1 million of its own shares in nine months, and has $24.55 billion still authorised. Diluted share count fell from about 427 million for the last full year to 395 million in the latest quarter. At that pace the same profit is divided among roughly 7% fewer shares each year.

What Makes It StrongiWhat stops a competitor taking the business, and whether that protection is getting stronger or weaker.

The file formats are the industry standard. A profession work, its archives and its handoffs between firms all run through them, and that is a far deeper hold than a feature advantage. Workflow lock-in. The cost of moving a studio off these tools is retraining every person in it and re-opening every past project. Few switch. 97% of revenue is subscription. Of $19,776 million over the nine months, $19,196 million was subscription, $246 million product licences and $334 million services. That is about as recurring as a software business gets. 88.7% gross margin. Each additional dollar of revenue costs under twelve cents to deliver, which is what funds both the research and the buyback.

Financial HealthiWhether the company makes money, turns it into cash, and carries a sensible amount of debt. All three, because a company can pass one and fail another.

Profitability is strong. Gross margin 88.7% and operating margin 34.8% in the quarter to 28 August 2026. Net income was $1,827 million on $6,760 million of revenue. One figure needs explaining rather than quoting. Return on equity looks extraordinary, near 60%, but that is an artefact: years of buybacks have shrunk the equity base to $11,764 million, so the ratio flatters. Return on invested capital lands in a similar place for the same reason. The honest quality signals here are the gross margin and the cash conversion, not the return ratios. Cash generation is strong. Free cash flow was $9,852 million in the last full year against net income of $7,130 million, which is 1.38 times. Operating cash flow over the five full years went $7,230 million, $7,838 million, $7,302 million, $8,056 million and $10,031 million, so about 8.5% a year with one down year in the middle. Operating cash flow in the nine months to 28 August 2026 was $7,646 million. Debt is low. Net debt of $724 million is less than four weeks of free cash flow. Of the $6,363 million gross, $1,597 million is classified as current, which the cash balance covers more than three times over.

Is Today's Price Attractive?iWhat we think the business is worth now against what the share costs today. A great company can still be a bad purchase at the wrong price.

Yes. On our estimate the shares are undervalued. We value this on a multiple of reported earnings, and the choice of reported over adjusted is deliberate. The gap between the two is mostly stock compensation, which is a real cost to shareholders even though no cash leaves the company. Using the adjusted figure would make this look cheaper and would be the less honest basis. Normalised earnings per share are taken at $18.145, the midpoint of this year reported guidance of $18.12 to $18.17. The multiple range is 16 to 20 times. That is below where a business with 88.7% gross margins, 97% recurring revenue and 11% growth would normally be valued, and the discount is for the artificial intelligence question, which is not settled and will not be settled soon. 16 times $18.145 is $290. 20 times $18.145 is $363. Rounded, our estimated fair value today is $290 to $360. The close on 9 October 2026 was $242.27, which is 16.5% below the bottom of that range. A cross-check, because one method is not evidence. Free cash flow of $9,852 million across 395 million shares is $24.94 a share. At 14 to 17 times that gives $349 to $424, which is higher than the range above. We publish the lower of the two. Worth noticing in passing: free cash flow per share of $24.94 is almost exactly the adjusted earnings guidance of $24.475. The adjusted figure is not fiction, it is roughly the cash. What it leaves out is that stock compensation hands shares to employees, and the answer to that is the buyback, which the company is running hard.

Potential OutcomesiThree estimates of where the share could be at the end of our timeframe. Bear is the bad case, Base is the likely one, Bull is the good one. They are estimates, not forecasts.

These are estimates, not forecasts, and all three start from the same $242.27. Adobe pays no dividend, so for this company the price return and the total return are the same number. Bear case, $171 a share by 2031. That is 29.4% below today, or 6.7% a year. Generative artificial intelligence does erode what these tools are worth. Revenue growth falls toward 2%, pricing power goes, operating margin compresses from 35% toward 28%, reported earnings per share stalls near $19 even with the buyback helping, and the multiple falls to 9 times. This is the case the market is currently leaning toward. Base case, $435 a share by 2031. That is 79.6% above today, or 12.4% a year. Revenue growth moderates to about 8%, margins hold, the buyback keeps retiring shares, reported earnings per share compounds at about 10% a year to roughly $29, and the multiple settles at 15 times. Note that 10% is below the 13.6% a year the company actually delivered over the last four full years, and 15 times is still below the bottom of our fair value range. Bull case, $700 a share by 2031. That is 188.9% above today, or 23.6% a year. Artificial intelligence turns out to expand creative work rather than replace the tools, AI recurring revenue scales from $650 million into the billions, growth re-accelerates to the mid teens, earnings per share reaches about $35 and the multiple returns to 20 times. The spread between those three is unusually wide, and that is the honest shape of this investment. A business this profitable bought at 13 times earnings produces large numbers if it is fine and loses a third if it is not. We are not putting a probability on any of these. We do not have a method that would make one honest.

Main RisksiThe short version of the threats. The write-up explains each one.

Generative artificial intelligence eroding what the tools are worth. The central risk and the reason the shares are where they are. If a model can produce professional output from a written instruction, the value of mastering a complex tool falls, and so does what anyone will pay for it. This is the risk that would invalidate the case rather than reduce it. Pricing pressure. Related but not the same. Adobe could keep every customer and still earn less per customer if free or cheap alternatives reset what the market expects to pay. The chief executive transition. Anil Chakravarthy takes over on 1 December 2026 after Shantanu Narayen eighteen years. The handover is orderly and the successor is internal, which is the good version of this event. It is still an unproven chief executive arriving at the moment the company faces the hardest strategic question in its history. Enterprise budgets. The marketing and analytics software is sold to corporate buyers who cut software spending in a downturn. Stock compensation dilution. Real, ongoing, and currently more than offset by the buyback. If the cash flow that funds the buyback ever falters, the dilution stops being offset.

What Would Change Our View?iThe specific developments that would make us reconsider. If none of them happens, nothing has changed.

Four things. Two would confirm the case and two would break it. Annual recurring revenue growth falling below 8% for two consecutive quarters. This is the number that would show the disruption arriving, and it would move us out. It was 11.2% in the quarter to 28 August 2026. Gross margin falling below 85%. That would distinguish the two bad outcomes from each other. Losing customers shows up in revenue. Being forced to charge less shows up here first. AI-first recurring revenue passing $2 billion while total recurring revenue growth stays above 10%. The second half of that is the test. AI revenue growing while the total stalls would mean it is replacing the old business rather than adding to it. The new chief executive first full year delivering growth and margins in line with the record above. Not a number we can specify in advance, and the thing we will actually be judging through 2027.

What We're Watching NextiThe results and milestones ahead that could strengthen or weaken the case.

Anil Chakravarthy becoming chief executive on 1 December 2026, with Shantanu Narayen moving to executive chair. Then the fourth quarter and full year results for the year ending in late November 2026. The company has not published a date for them, so none is given here. Those will confirm or revise the guidance this valuation rests on, and will carry the first recurring revenue reading under the new leadership.

Year EstablishediThe year the company was founded. Not the year it listed its shares, which is usually later.

1982.

Company ProfileiWhat the company does, what it sells, and which countries it operates in.

Adobe makes software for creating and managing digital content. The creative tools are the ones professionals use for images, illustration, layout, video and documents, and the document format it created is the standard way a fixed page is sent anywhere in the world. Alongside that it sells marketing, analytics and content management software to large companies. It sells almost entirely by subscription, worldwide, and reports in US dollars. Revenue was $23,769 million in the last full year and $19,776 million over the nine months to 28 August 2026, with guidance of $26.576 billion to $26.626 billion for the year now ending.

Business ModeliHow it makes money, where the sales come from, and how much it leans on any one customer or product.

Customers pay monthly or annually and keep paying. Of $19,776 million of revenue over the nine months to 28 August 2026, $19,196 million was subscription, which is 97.1%. Product licences were $246 million and services $334 million. Two figures describe the forward book. Annual recurring revenue, the annualised value of the subscriptions in place, was $27.50 billion at the end of August and grew 11.2%. Remaining performance obligations, which is contracted revenue not yet recognised, were $22.16 billion. A business can be read some way into the future from those two. There is no customer concentration worth naming. Revenue comes from millions of individual subscribers and from enterprises, and no single one of them matters to the total. The split between the creative business and the enterprise business is not presented in the quarterly filing, which disaggregates revenue by type rather than by segment, so no percentages for it are given here.

Business StrengthiIts position against competitors, how loyal its customers are, and who it is actually up against.

Adobe is the standard toolset for professional creative work. The competitive advantage is wide and it rests on the file formats, on the cost of retraining a workforce off the tools, and on the archives of past work that have to stay openable. The competitors are not short of money. Canva has taken the lower end of the market, Figma took interface design, and the large artificial intelligence laboratories now produce images and video directly from written instructions. The first two have been visible for years and the recurring revenue above grew through them. The third is new and is what this whole case turns on. A retention figure is not disclosed. Recurring revenue growth of 11.2% and $22.16 billion of contracted revenue not yet recognised are the closest published proxies.

Growth Track Record & PlansiHow fast it has grown in the past, what it plans next, and what evidence exists that the demand is there.

Revenue went $15,785 million, $17,606 million, $19,409 million, $21,505 million and $23,769 million across the five full fiscal years to late 2025. That is about 10.8% a year. Reported earnings per share went $10.02, $10.10, $11.82, $12.36 and $16.70 over the same years, about 13.6% a year, ahead of revenue because of margins and the shrinking share count. The pattern inside those numbers matters. Revenue growth was 22.7% in the first of those years and has sat between 10.2% and 11.5% in every year since. So growth stepped down once, four years ago, and has been stable since rather than continuing to decay. The latest quarter at 12.9% is the fastest in that period. We rate expansion progress as on track. Guidance for the current year was raised alongside the September results, recurring revenue is compounding at 11.2%, and the artificial intelligence product line went from nothing to over $650 million of recurring revenue.

Financial PerformanceiSales, earnings, margins and cash flow, and which way each has been moving. Anything unusual or one-off is called out.

Revenue was $23,769 million for the last full year and $19,776 million for the nine months to 28 August 2026, against $17,575 million in the same nine months a year earlier. Reported earnings per share were $16.70 for the last full year and $13.47 for the nine months. Guidance for the year now ending is $18.12 to $18.17 reported and $24.45 to $24.50 adjusted. The difference between those two is mostly stock compensation and acquisition-related amortisation. Margins. Gross margin 88.7%, operating margin 34.8% and net income of $1,827 million in the quarter to 28 August 2026. Operating income was $7,010 million over the nine months on $19,776 million of revenue, which is 35.4%. Cash flow. Operating cash flow across the five full years went $7,230 million, $7,838 million, $7,302 million, $8,056 million and $10,031 million. Free cash flow went $6,882 million, $7,396 million, $6,942 million, $7,873 million and $9,852 million. Capital spending is tiny for a company this size: the gap between the last year operating and free cash flow is under 2%. Operating cash flow was $7,646 million in the nine months to 28 August 2026. Two things in those series deserve naming rather than smoothing over. Cash flow fell in the middle year of the five, so this is not a company that has never had a soft year. And net income jumped from $5,560 million to $7,130 million in the final year, a 28% rise against 10.5% revenue growth, which is margin and tax rather than trading.

Balance Sheet StrengthiCash, debt and what has to be repaid when. This is what decides whether a bad year is survivable.

Cash was $4,359 million and short-term investments $1,280 million at 28 August 2026, so $5,639 million together. Total debt was $6,363 million, being $1,597 million current and $4,766 million long term. Net debt is therefore $724 million. Total equity was $11,764 million, and that figure is small relative to the earnings for a reason worth understanding: sustained buybacks reduce equity, so a long history of returning cash leaves a thin looking balance sheet. It is a sign of capital discipline here rather than of weakness. Could it withstand a difficult period? Net debt of $724 million against free cash flow of $9,852 million a year is not a solvency question. The current portion of debt, $1,597 million, is covered more than three times by cash alone. The risk to this company is competitive, not financial.

Management & GovernanceiWho runs the company, whether they deliver what they promise, how much they own, and any accounting or board concerns.

Shantanu Narayen has been chief executive for eighteen years and announced on 12 March 2026 that he would hand over once a successor was named. Anil Chakravarthy, who ran the customer experience and worldwide field operations side of the business, was named president and chief executive in September 2026 and takes over on 1 December 2026. Narayen becomes executive chair and stays involved through the transition. Our assessment is strong, and it is an assessment of the record rather than a prediction about the new appointment. Revenue went from $15,785 million to a guided $26.6 billion across the period covered here, the subscription transition was completed, margins rose, and $6,820 million of stock was retired in nine months. The succession was announced six months before it was filled and the successor came from inside. That is the orderly version of a difficult event. Governance concerns are none. There is no regulatory investigation, no accounting dispute, no controlling shareholder and no related-party pattern that we found. The company has no dual-class share structure, so one share is one vote. The precise insider holding was not established in this pass. On capital allocation judgement there is one episode worth remembering. Adobe agreed in 2022 to buy Figma for about twenty billion dollars and abandoned the deal in 2023 after competition regulators objected, paying a break fee for the privilege. Reasonable people read that two ways: as a willingness to overpay for a threat, or as a discipline to walk away. Nothing of that scale has been attempted since.

Use of Company MoneyiWhere the profits go. Back into the business, into buying other companies, out as dividends, or into buying back its own shares.

Research and development and the sales force are funded out of operating cash flow, and capital spending is immaterial: the gap between operating and free cash flow in the last full year was under 2%. Dividends. None. Adobe has never paid one and returns cash entirely by buying its own shares. Buybacks. $6,820 million spent on 26.1 million shares in the nine months to 28 August 2026, with $24.55 billion still authorised under a programme approved in April 2026. At the current price that authorisation is about a quarter of the company. Acquisitions. Not a meaningful part of the recent story, and the one large attempt was abandoned, as covered above. Share dilution is net negative, which is the good direction. Stock compensation issues shares and the buyback retires more of them: diluted share count went from about 427 million for the last full year to 395 million in the quarter to 28 August 2026.

ValuationiThe ratios, the comparisons with similar companies and with its own history, and the working behind our estimate of what it is worth.

Market capitalisation is about $95,697 million, being 395 million diluted shares at the $242.27 close. On that close the shares trade at 13.4 times the $18.145 midpoint of reported earnings guidance, 9.9 times the $24.475 adjusted midpoint, and 14.5 times the $16.70 actually reported last year. Market capitalisation is 9.7 times last year free cash flow of $9,852 million and about 3.6 times guided revenue. Against its own recent history the de-rating is stark and can be shown in one comparison. At the 52-week high of $363.70, set on 28 October 2025, the shares were on 21.8 times the $16.70 reported for the last full year. At $242.27 they are on 14.5 times the same figure. The earnings did not fall. The multiple did. The shares sit 30.0% of the way up their 52-week range, above the $190.12 low of 18 June 2026, and below both the $257.88 fifty-day and the $256.98 two-hundred-day average. Twelve-month return was minus 30.5% against plus 15.7% for a US market tracker. A peer multiple comparison was not established in this pass, so none is quoted. Adobe does appear on published screens of the cheapest wide-moat US companies, which is how it reached this research in the first place. How we got to the range. Reported earnings per share of $18.145, the guidance midpoint, on the reported basis rather than the adjusted one because the difference is mostly stock compensation and that is a real cost. A multiple of 16 to 20 times, below what this margin and recurring revenue profile would normally carry, with the discount taken for the unsettled artificial intelligence question. 16 times gives $290 and 20 times gives $363, rounded to $290 to $360. The free cash flow cross-check above gives a higher range and we publish the lower one. The metrics this sector is actually judged on are annual recurring revenue and its growth rate, remaining performance obligations, gross margin and the subscription share of revenue, and all four are given above with the periods they cover. A price to earnings ratio on a subscription software business tells you less than the recurring revenue growth rate does, because the multiple is almost entirely a judgement about whether that growth continues.