Cash machine prints records while the multiple prices a melting moat

• Records on every line, a 70% drawdown on the tape. FY25 revenue grew 10.5% to $23.8B, free cash flow stepped up 25% to $9.85B, and ROIC hit a decade-high 26.7%. The stock peaked at $688 in November 2021 and trades at $210 today — roughly $240B of market cap erased while revenue grew over 50%.
• Q2 FY26 cracked the operating-margin pillar. GAAP operating margin printed 33.8% on June 11, 2026 — down 210 bps year-on-year and 400 bps from Q1, the lowest in eight quarters — and management deferred the planned H2 FY26 Creative Cloud price increases, pivoting to freemium acquisition. That coincidence is the whole debate.
• The valuation prices an impairment. Adobe trades at 8.8× EV/FCF and 12.6× trailing P/E — the lowest large-cap software multiple in over a decade, against an eight-year EV/EBITDA average of 27×. Bears read it as correct pricing of a melting moat; bulls read it as a multiple that funds the long thesis on flat fundamentals.

Top-decile margins, decade-high ROIC, and a buyback running at 13% of market cap

Subscription mix runs 96.4% — customers pay up front, revenue recognizes ratably, and deferred-revenue growth funds working capital. Capex is a trivial $179M (0.8% of revenue), so FCF prints $1.15B ahead of GAAP operating income and 138% of net income. The buyback retired 5.1% of shares in FY25, the largest single-year shrink of the modern era, and the April 2026 $25B authorization equals ~29% of today's market cap.

A beat-and-raise, then a 6.8% drop, seven targets cut, and a CFO walking out for Marvell

• June 11 — beat-and-raise, but margin gave back 400 bps. Q2 FY26 revenue grew 12.7% to $6.62B, non-GAAP EPS beat by $0.15, total ARR rose 12.5% to $27.1B, and the FY26 guide was lifted to $26.5–26.6B. GAAP operating margin printed 33.8% on heavier AI infrastructure and go-to-market spend with no matching monetized AI uplift yet.
• June 12 — seven sell-side shops cut targets the next morning. Stifel 350→200 with a downgrade, Evercore 325→225 with a downgrade, KeyBanc 235→195 with a downgrade, Wells 330→250, BMO 285→230, Barclays 275→250, DA Davidson 300→250. Consensus rating moved to Hold; the new $190–$510 target range is the widest dispersion in the post-Figma era.
• June 15 — CFO Durn departs for Marvell. Outgoing CFO Dan Durn publicly cited a 'once-in-a-generation AI expansion' at his next stop. SVP Steve Day takes the seat on an interim basis, meaning the first FY27 guide on December 10 will likely be set without a permanent CFO — and three months after CEO Narayen's March 12 succession announcement, still no successor named.

Wide on Pro creative and PDF; eroded at the entry tier where Canva already wins

• Wide on the franchise (~75% of revenue). Photoshop, Illustrator, Premiere, and Acrobat are the file-format standard and the hiring spec — 85% of 2026 Sundance films were cut in Creative Cloud, 99 of the Fortune 100 use Adobe Experience Platform, and no Fortune-500 account has been reported lost. Pro Creative ARR grew 10.5% in FY25 and Business & Consumer grew 14.7%.
• Eroded at the entry tier (~4–5% of revenue). Canva claims 260M monthly users and 95% of the Fortune 500 use it in some seat; Figma IPO'd in 2025 at a ~$19B target valuation — three years after Adobe's $20B bid was blocked and a $1B break fee paid. Adobe Express is the only surface where the moat is observably absent today, and the Q2 freemium pivot is the defensive recognition.
• Unproven on the AI commercial-safety wedge. Firefly trained only on licensed Adobe Stock and public-domain content with full enterprise IP indemnification — a real legal wedge versus frontier image and video tools. AI-first ARR tripled to $500M+ and 24B+ Firefly assets have been generated, but Foundry's standalone ARR has not been disclosed eight months after launch. The AI optionality is invisible in the multiple.

8.8× EV/FCF prices 0–2% long-term growth on a franchise that compounded 17% for a decade

Across large-cap software peers, Adobe holds the highest ROIC (26.7%), the second-highest operating margin behind Microsoft, and the lowest leverage — and trades on the cheapest P/E, P/FCF, and EV/EBITDA of any profitable name except Salesforce. Scenario math: a halfway re-rate to the 8-year norm on flat fundamentals would imply ~$325; the bear's $145–$200 path requires a 200-bps FCF-margin reset and operating margin settling below 35%.

Both CEO and CFO seats open while a $25B buyback runs through them

• Narayen sold $18.3M six weeks after announcing his own exit. The 75,000-share sale on April 28, 2026 at $243.54 was filed via Form 144 with no 10b5-1 plan disclosed, removed ~17% of his direct holdings, and landed one week after the board authorized $25B of buybacks. His Retention Agreement also accelerates 100% of outstanding equity on any change of control — a single-trigger payout no other named executive carries.
• The interim CFO will likely set the first FY27 guide. Dan Durn left for Marvell citing a 'once-in-a-generation AI expansion'; the permanent CFO search is open. FY25 buybacks ran at 115% of FCF and 158% of net income — funded with $1.99B of fresh debt that grew long-term debt from $4.13B to $6.21B — and whether the next CFO sustains that pace is the swing variable behind the per-share-FCF math.
• Strong plumbing, weak optics. 10 of 11 directors are independent, anti-pledge and anti-hedge are real, and the say-on-pay vote cleared at ~80% — but Lead Director Frank Calderoni running the CEO search has served alongside Narayen for 14 years, Snowflake, Uber and Expedia were added to the FY26 peer group to drag the comp median down, and the March 13, 2026 DOJ/FTC $150M settlement is forcing Adobe to dismantle the cancellation flow that suppressed individual and SMB churn.