Adobe Earnings Preview: The Valuation Deadlock Under the Shadow of AI Self-Cannibalization
N.R. Finch
Adobe reports Q2 results after Thursday's close. Analysts expect EPS of $5.81 and revenue of $6.45 billion, up roughly 15% and 10% year-over-year. Yet the stock has fallen 43% over the past year — the real question is not whether Adobe can beat estimates, but whether its AI business can grow fast enough to offset the erosion of its legacy products.
Earnings keep beating — so why is the stock down 43%?
Over the past year Adobe has posted a 100% beat rate on both EPS and revenue, yet the stock has dropped 43% in the same period.
This means → the market is no longer pricing "how much Adobe earned this quarter" — it is pricing "can Adobe still earn this much three years from now."
In plain terms = investors are not worried about today's report card; they are worried that AI will eat Adobe's legacy business faster than new AI revenue can replace it.
This "earnings up, valuation down" divergence points to one concept: self-cannibalization — new products stealing customers from the company's own older products.
Firefly is selling well — so why is that bad news?
Adobe's AI image-generation tool Firefly has genuinely driven new users and higher product adoption.
But management has already acknowledged that its legacy cash cow Adobe Stock — the licensed-image library — is declining faster than expected.
In plain terms = designers can now generate custom images at low cost with Firefly, so they stop paying for stock photos. Adobe's left hand is taking business from its right hand.
To fend off lightweight AI art competitors, Adobe has adopted a freemium acquisition strategy — it locks in users but suppresses annual recurring revenue (ARR) in the short term, leaving the market seeing volume without profit.
AI spending is surging — can margins hold?
Generative AI demands sustained investment in compute and model R&D, while competitive pressure is pushing AI tool pricing down across the industry.
This means → Adobe is squeezed on both ends: costs rise as compute spending grows, while revenue from AI subscriptions is hard to reprice upward.
In plain terms = if AI tools cannot command premium prices and compute costs do not fall fast enough, Adobe's traditionally high gross margins in its software business will erode over time, lowering the long-term profit ceiling.
After a 43% drop, is it cheap now?
After the selloff, Adobe shows some relative-value characteristics versus other high-multiple tech peers and has broken below its 200-day moving average — a key long-term technical support level.
Wall Street remains cautious: the consensus view is that Adobe's profit growth has not yet caught up with the pace of legacy-business erosion, which is insufficient to justify multiple expansion.
This means → "cheap" is only relative. The valuation deadlock loosens only if management can demonstrate on tonight's earnings call that AI revenue can offset legacy pressure at a faster pace and with stronger profitability.
Content is for reference only, not financial advice.