Adobe Inc

Adobe passes our Buffett-style tests, with a wide base-case margin and a thin bear-case margin.

Adobe sells software mainly on subscription, through Creative Cloud, Acrobat, Express, Firefly and its marketing platform, Adobe Experience Platform. Revenue was $23.8B and net income was $7.1B in the latest year, and our tests found no loss year across fifteen years of filings. Return on equity averaged 34%, and free cash flow has averaged 142% of reported profit, so the earnings look real in cash terms. On value, the base case is $466.77 per share against a price of $254.80, a 45% margin, but the bear case of $266.97 clears the price by only 4.6%. The pass status means the tests pass and the base case clears the required 25% margin; it does not mean the price is safe under every assumption.

Adobe sells software mainly as subscriptions. Its main products are Creative Cloud, Acrobat, Express and Firefly for individuals and creative professionals, plus Adobe Experience Platform and related marketing tools for enterprises. Subscription revenue is recognised ratably over the contract term, so income arrives steadily over time. Revenue was $23.8B and net income was $7.1B in the latest year.

• Creative Cloud and Acrobat subscriptions
• Enterprise marketing and Experience Platform contracts
• Ratable recognition of subscription revenue
• Generative AI credits bundled into paid plans
• Price and plan changes across a large user base

Inside the circle? The subscription model and a long run of consistent earnings give the business a steady revenue base that a careful owner can model. The filing also warns that competition, AI change, regulation and rising AI costs could shift pricing and margins within a ten-year window, and the excerpts do not show how the AI products will earn over that period. The cash flows look strong, but the moat depends on product choices that could change quickly.

Adobe's protection comes mainly from switching costs, because its products are built to work together and customers gain little by moving between tools. Its Firefly models are described as commercially safe, and eligible enterprise customers can obtain an intellectual property indemnity, which adds a second point of difference. Returns on capital are high: return on invested capital averaged 41% and the lowest year was 26%, while gross margin averaged 87%. The filing itself describes limited barriers to entry and customer price sensitivity, which argues against a wide moat, and the excerpts do not show how the AI products will earn over a ten-year period. Our base-case value is $466.77 per share against a price of $254.80, but the bear case of $266.97 shows how much this depends on growth assumptions.

Adobe returns surplus cash mainly through share repurchases, and it pays no dividend. The filing says the repurchase programme is designed to return value to stockholders and to limit dilution from stock issued to staff, and the share count has fallen by -1.8% a year. Management has also agreed a large cash acquisition and spends heavily on AI development, but the excerpts give no return target for either. Buffett's $1 retained test is supported by 1.4×, yet that is a backward-looking measure and the value per share figures are estimates. The filing discloses a terminated deal and a termination fee plainly, which counts in management's favour on candour.

• Share repurchases are the main way cash reaches owners, and the filing describes the programme as designed to limit dilution from stock issuances.
• No cash dividend is planned, and the filing says management does not anticipate paying one in the foreseeable future.
• A large cash acquisition, Semrush, was agreed in November 2025, and the excerpts do not give a return target for it.
• Total debt is $6.2B and net cash is $385.0M. The senior notes carry no financial covenants, and the filing shows a termination fee from the Figma deal in fiscal 2024.

• PassesDemonstrated, consistent earning powerPasses

  No loss years across 15 years of filings, with operating margin varying by 3.5% around its average.

  Why this test
• PassesHigh returns on equity without leveragePasses

  Return on equity averaged 34%, reaching 15% in 10 of the years measured; the weakest year was 16%.

  Why this test

Several principles support the case. Consistent earning power and high returns on equity and capital are the strongest evidence, with return on invested capital averaging 41% and the lowest year at 26%. Little debt also applies: total debt is $6.2B, payback is 0.9 yrs of normalised owner earnings, and interest cover is 33×. The one dollar retained test is met on the numbers (1.4×), but that is a look backwards. The moat is weaker than the returns suggest. Our moat read points to switching costs, yet the filing itself says there are 'limited barriers to entry' and cites 'customer price sensitivity', which are the conditions that test pricing power, the single most important question in Buffett's framework. Owner earnings are $7.9B reported and $7.1B normalised, and the reported maintenance capex estimate of $0.0 looks more like a gap in the data than a fact, so it should be checked before relying on it. On rates, the owner earnings yield of 6.6% sits close to the ten-year Treasury yield of 5.2%, so the cushion over a risk-free return is thin. On management, the filing says the repurchase programme is designed to 'return value to our stockholders and minimize dilution from stock issuances', and it discloses a terminated deal and its fee plainly, which fits Buffett's test of candour. The filing gives no return target for the large AI spending or the Semrush acquisition, so the circle of competence test cannot be fully met yet.

• Growth in Total Adobe ARR and whether AI products are named as a source of it

  The filing says it is uncertain whether new AI solutions will be monetisable, so the revenue line is the clearest test.

• Cost of revenue and gross margin as AI inferencing costs grow

  The filing says AI costs 'could adversely impact our margins', and the gross margin average is a core part of the moat case.

• Any change in pricing, bundling or plan tiers for Creative Cloud and Acrobat

  This is the direct test of pricing power that Buffett treats as the most important question.

• Repurchases compared with free cash flow and any new borrowing

  The filing notes new senior notes were issued in January 2025, and buybacks funded by debt would weaken the little debt test.

• Completion and early results of the Semrush acquisition, expected in the first half of fiscal 2026

  The filing gives no return target for the deal, so the outcome is the only evidence of its value to owners.

Questions the filings couldn't answer

• Does the spending on AI products earn a return above its cost? The filing gives no target and the excerpts do not show the economics.
• Is the recent rise in return on equity partly a recovery from a weak prior year, given the termination fee and the effect of buybacks on book equity?
• Is the maintenance capex estimate of zero a real figure or a gap in the reported data?