Franchise and De-Rating

Franchise and De-Rating

Adobe is the incumbent of professional content creation — Photoshop, Illustrator, Premiere, Acrobat — sold almost entirely by subscription. In fiscal 2025 it earned $23.8 billion in revenue at a 37% operating margin and threw off $9.9 billion of free cash flow [1]. Yet the stock has fallen roughly 70% from its 2021 peak while those fundamentals kept compounding. That gap between the business and the price is what this report exists to examine.

What Adobe sells

Adobe makes software that professionals use to create and manage digital content, and it collects for that software as a recurring subscription rather than a one-time licence. Its mission, in the company's own words, is "to empower everyone to create," a franchise built "for over four decades" across imaging, design, video and documents [2]. Through fiscal 2025 the business reported in three segments — Digital Media, Digital Experience, and a small legacy Publishing and Advertising line — which management is collapsing into a single reportable segment beginning in the first quarter of fiscal 2026 [3].

Digital Media is the core. It houses Creative Cloud — the flagship creative apps — and Document Cloud (Acrobat), and it generated $17.65 billion in fiscal 2025, up 11% year on year [4]. Digital Experience, the marketing-software business for enterprises, added $5.86 billion [5]. What matters for the investment case is how durable those dollars are: subscription revenue was $22.9 billion of the $23.8 billion total, so roughly 96% of the company recurs [6]. Management tracks the subscription base as Annualized Recurring Revenue; Digital Media ARR reached $19.20 billion at year-end, growing 11.5% [7], and the deferred backlog — remaining performance obligations — stood at about $22.52 billion, visibility that a transactional software vendor does not have [8].

Sources: FY2025 Form 10-K — revenue and operating margin from the Consolidated Statements of Income [9] and free cash flow from the Consolidated Statements of Cash Flows [10]; Total Adobe ending ARR from Q4 FY2025 call [11].

Source: FY2025 Form 10-K, MD&A segment results [12].

The shape of the business is unusual: a near-90% gross margin, three-quarters of revenue from one segment, and a customer base that renews. This is the asset the market is repricing.

The compounding machine

For most of the last decade Adobe was a textbook compounder. Revenue rose from $5.9 billion in fiscal 2016 to $23.8 billion in fiscal 2025 — a 16.8% annual growth rate sustained across nine years [13]. Earnings grew faster still, because margins widened and the share count shrank.

Source: Adobe Form 10-K filings, FY2016–FY2025, as reported; FY2025 anchor [14].

Diluted GAAP earnings per share climbed from $2.32 to $16.70 over the same span — about 24% a year — reaching the FY2025 figure on net income of $7.13 billion [15]. Part of that per-share growth is buybacks: Adobe repurchased nearly $12 billion of stock in fiscal 2025, cutting shares outstanding by more than 6% in a single year [16], part of a steady reduction from roughly 504 million shares in fiscal 2016 to 427 million in fiscal 2025 [17].

Source: Adobe Form 10-K filings, FY2016–FY2025, diluted EPS and share count, as reported; FY2025 anchor [18].

Cash generation kept pace with earnings rather than lagging them: fiscal 2025 operating cash flow exceeded $10 billion, and with negligible capital intensity, free cash flow was $9.85 billion — a 41% margin [19]. The quality of that cash — how much survives stock-based compensation, and how it is deployed — is a question later chapters can test. What is not in dispute is that the operating record through fiscal 2025 shows no break: revenue, ARR and cash all grew double digits into the most recent year.

The de-rating

The share price tells the opposite story. Adobe traded near $95 in mid-2016, peaked at an intraday high of $699.54 in November 2021, and closed at $211 on July 1, 2026 — roughly 70% below the peak. The decline was not a single event: the stock reached $597 at the end of 2023, then fell to $350 by the end of 2025 and lower still through the first half of 2026.

Source: market price data, as reported (year-end closes; 2026 as of July 1). Intraday peak of $699.54 recorded November 22, 2021.

Because earnings rose while the price fell, the valuation multiple compressed sharply. On year-end price against fiscal-year GAAP earnings, Adobe traded near 57 times earnings at its 2021 peak; at $211 against fiscal 2025 EPS of $16.70, it trades near 13 times [20]. Against management's fiscal 2026 GAAP EPS guidance of $17.90 to $18.10, the forward multiple is closer to 12 times [21]. A franchise software business with 37% operating margins now carries a market multiple usually reserved for slow growers.

Source: derived from year-end price data and fiscal-year GAAP diluted EPS; EPS anchor [22]. 2026 uses the July 1 price against trailing FY2025 EPS.

The AI fork

The reason for the divergence is a single question the market cannot yet settle: what generative AI does to Adobe. The case rests on two readings of the same technology, and Adobe's own filings supply both.

The bear reading is disruption. Tools that generate images, video and documents from a text prompt let "users of all skill levels" produce content that once required Adobe's software — and the company states plainly in its risk factors that this "could significantly disrupt industries in which we operate and our existing products, services and solutions" [23]. Adobe describes its own markets as having "limited barriers to entry, short product lifecycles, customer price sensitivity" and "the frequent entry of new solutions or competitors," now explicitly including "AI or cloud-native companies" [24]. If a prompt replaces the tool, the recurring revenue and the 90% gross margin are exposed.

The bull reading is monetization. Adobe is embedding its own commercially-safe Firefly models and partner models across Creative Cloud, Acrobat and its enterprise platform, and it has begun to size the result: by the end of fiscal 2025, "new AI-influenced ARR" exceeded one-third of the total book of business, and the narrower set of AI-first products was still small but growing fast [25]. By the second quarter of fiscal 2026, management put AI-first ARR at more than $500 million, roughly triple a year earlier [26]. In this reading, AI widens the moat by giving Adobe more to sell to a larger audience.

The most recent evidence complicates both. In June 2026, management said AI "is accelerating customer behavior at an unprecedented speed" [27] and announced a strategic pivot toward free, "friction-free" onboarding — a shift it acknowledged "lowers our second half ARR growth expectations from individual subscribers" and "will come at the cost of short-term ARR" [28]. At the same time, the company is mid-transition at the top: CEO Shantanu Narayen is moving to Board Chair with a search underway, and long-time CFO Daniel Durn departed [29]. A company defending a franchise against a technological shift is changing both its business model and its leadership at once.

The evidence available today points to a business that is still compounding — revenue, ARR and cash all grew double digits into fiscal 2025, and management guides to another year of it in fiscal 2026 [30]. The strongest fact against reading that as safety is that the deceleration risk is now self-inflicted as well as competitive: Adobe is choosing to trade near-term recurring revenue for user reach, precisely because AI is changing how customers arrive. What would change the read is direct evidence on either side — durable acceleration in AI-first monetization that outruns the freemium give-up, or the reverse: net-new ARR that stalls as free tiers cannibalize paid ones.

The question this report is built to answer: does generative AI widen Adobe's dominant, cash-rich creative-software franchise or dissolve it — and is the market right to have cut the multiple from roughly 57 times earnings to about 13 while revenue, recurring revenue and free cash flow keep compounding at double digits? Everything that follows examines one piece of that question.