ADBE — Investment Thesis

Date: 2026-04-26 Price: $245.44 | Forward P/E: 9.3x | Trailing P/E: 14.3x | Graham Score: 4-5/7 | Decision: WAIT (do not deploy T1 yet)


TL;DR

Adobe at $245 is structurally cheap on Graham's Growth Stock Addendum — the market is pricing 0.4% growth into a software franchise that has compounded GAAP EPS at 13%/year over 4 years and is about to retire ~25% of its share count via a $25B buyback. Forward P/E of 9.3x is the cheapest the stock has been since 2013. SBC-adjusted P/FCF is 11.9x.

But we are not buying today. A fresh competitive RED flag — Anthropic Claude Design and OpenAI GPT Images 2.0, both launched in mid-April — directly targets Creative Cloud's user base. Per shr-034, we don't deploy T1 with an unresolved RED and no positive catalyst within 30 days. Q2 FY26 earnings on June 11 is the resolution event: it will reveal whether Creative Cloud ARR is decelerating. Until then, ADBE is a watch-list candidate, not a buy candidate.


Hypothesis

The market is mispricing Adobe by treating two distinct risks as one:

  1. AI competitive disruption is real but partially priced. Generative AI from OpenAI, Anthropic, Midjourney, and Canva genuinely threatens Adobe's lower-end design tools and image-generation use cases. The 42% drawdown from the $423 high (~$180B in market cap erased) reflects this fear.

  2. The cash-generative core franchise is intact. Creative Cloud + Document Cloud + Experience Cloud is a $24B revenue platform serving enterprise and pro creators with multi-year switching costs. Revenue still grew 12% in Q1 FY26 (Feb 2026). Operating margin expanded to 37.8%. FCF $10.3B TTM. None of this looks like a business in disruption — yet.

The thesis is that risk #1 is overpriced relative to risk #2's resilience. If Q2 FY26 earnings shows ARR holding at 10%+, the buyback alone (~5.5%/yr EPS accretion at current pace) plus modest organic growth justifies a re-rating from 9.3x forward to 13-15x — implying $300-$370 over 18-24 months.

If Q2 shows ARR decelerating to 7-8%, the thesis is wrong and we should not own this.

This is why the buy decision is conditional on June 11.


What Changed (Why Now)

Three things happened in the last 60 days that changed Adobe's setup:

  1. 30-day -9% drawdown (Mar–Apr 2026) brought the forward P/E from ~12x to 9.3x — the deepest discount in a decade for ADBE.
  2. $25B buyback authorization (8-K filed Apr 21, 2026) — represents ~25% of the current $99B market cap. At the current $2.5B/quarter pace, this retires ~10% of the share count per year.
  3. Anthropic Claude Design + OpenAI GPT Images 2.0 launched Apr 17, 2026, in the same 24-hour window. Figma dropped 10% on the news; Adobe fell ~2%. This is the fresh competitive shock that's still unresolved.

The Adobe Summit (Apr 21-22) showed Adobe responding aggressively — AI monetization push, Firefly integration deepened, Nvidia CEO Jensen Huang publicly endorsed Adobe as "the future of marketing." But Summit was talk; June 11 earnings is data.


Why 9.3x Forward P/E Is Cheap (The Growth Stock Addendum Test)

Per shr-004, Graham's Growth Stock Addendum says: buy a non-Defensive-passing stock when the implied growth from current P/E is less than the growth rate you believe is achievable, with margin.

Implied growth from forward P/E 9.3x: (9.3 - 8.5) / 2 = 0.4%

Achievable growth rate? Adobe's track record:

To believe the market's 0.4% implied growth, you have to believe:

That's a triple-failure scenario. The probability isn't zero — generative AI could plausibly cause one of those — but pricing all three with high confidence is what the current 9.3x multiple does.

Margin of safety table (Graham V = EPS × (8.5 + 2g), trailing GAAP EPS $17.16):

Growth V MoS at $245
0% $146 -40% (price above)
3% $249 1.5% (~at price)
5% $317 29%
7.5% $401 64%
10% $489 99%

Even if you cut Adobe's actual 4yr GAAP EPS CAGR (13%) in half to 6.5% and discount further, you land at ~$370 IV — 51% margin of safety from $245.


SBC Adjustment (per shr-001)

Tech-company FCF must always be adjusted for stock-based compensation, which is a real cost via dilution.

Metric Value
TTM Reported FCF $10.32B
TTM SBC $1.98B
TTM SBC-Adjusted FCF $8.34B
P/FCF reported 9.6x
P/FCF SBC-adjusted 11.9x

The adjustment adds 2.3 turns. This is meaningful but not the doubling seen in genuinely high-SBC names like RDDT (where the adjustment took P/FCF from 42x to 84x). At 11.9x SBC-adjusted, ADBE remains cheap for a 38% op-margin software platform.

SBC has grown from $1.44B (FY22) to $1.98B (TTM) — roughly stable as a % of revenue (~8%). Not accelerating. Not a structural concern, just a structural cost.


AI-Washing Test (per shr-032)

Graham Ch. 18 says narrative without financial confirmation is exit liquidity. The shr-032 filter for AI-era names: does the AI claim show up in segment financials within 2+ quarters?

Adobe's AI claim:

Verdict: mixed. ADBE is not a "fog machine" (CSCO pattern) — margins are genuinely expanding and revenue isn't masking decline. But it is also not showing AI as a discrete growth driver visible in financials. Adobe is embedding AI rather than monetizing it as a new revenue stream. The bull case requires Adobe to eventually disclose AI-attributable revenue (Firefly, Acrobat AI Assistant, etc.) — and the longer that disclosure is delayed, the more the shr-032 concern strengthens.

Operational rule: if Adobe reaches Apr 2027 (2 years post-Firefly launch) without disclosing AI-attributable revenue, downgrade the thesis.


Insider Activity (per shr-002)

Last 6 months:

This is a neutral-to-weak signal. Not the systematic C-suite bailout shr-002 warns about (no CEO/COO/CTO selling, modest dollars), but no bullish CFO open-market buy of the kind that anchored the AIG thesis. A 42% drawdown from the high without any insider stepping in is a small AMBER flag — they have the conviction to dictate $25B in buybacks but not to put their personal money in.


Framework Reasoning

Graham (Quantitative)

Dorsey (Moat)

Klarman (Downside)

Lynch (Classification)


Key Numbers

Metric Value Why It Matters
Forward P/E 9.3x Deepest forward multiple in a decade
Trailing P/E 14.3x Passes Graham's <15 filter
P/FCF SBC-adjusted 11.9x Cheap for 38% op-margin software (per shr-001)
Buyback as % mktcap 25% ($25B / 4yr) ~5.5%/yr EPS accretion mechanically
Net cash $0.23B Fortress; minimal leverage
Operating margin 37.8% (latest qtr) Expanding from 34.6% (FY22)
Revenue growth (Q1 FY26) +12% YoY Stable, not accelerating, not decelerating
GAAP EPS 4yr CAGR 13% $10.10 (FY22) → $16.70 (FY25)
Insider open-market buys (6mo) 0 Neutral-to-weak signal
Insider C-suite open-market sells (6mo) $816K (CFO) Modest, not systematic bailout
Mean analyst PT $329 +34% upside; recent PT cuts post-Summit
Analyst Buy/Hold/Sell 22/19/4 Net positive but distribution shifted Hold
Short interest 4.07% Benign; DTC 2.67 days
IV rank ~61% Moderately elevated; options expensive

Red Flags and What Could Kill the Thesis

  1. Generative AI competitive shock (the dominant risk) — Anthropic Claude Design + OpenAI GPT Images 2.0 launched Apr 17, 2026 directly targeting Creative Cloud. Figma dropped 10% on the same day; Adobe -2%. The damage to ARR (if any) won't show until Q2 FY26 earnings June 11. This is the unresolved RED that drives the WAIT decision.

  2. AI-washing risk (shr-032) — Adobe doesn't disclose AI-attributable revenue. The AI narrative has been running 2+ years (Firefly launched Mar 2023). If 2027 still has no AI revenue line, the thesis weakens to "buyback compounding only."

  3. No insider buying — Despite a 42% drawdown from the high and a $25B buyback authorization, no officer or director put personal money in via open market in the last 6 months. The strongest bullish signal is absent.

  4. Sector/macro pressure — ADBE is underperforming NDX (-9% over 30 days vs NDX +). ServiceNow Apr 23 warning about Iran/geopolitical slowing enterprise software deals is a real headwind. Software-sector P/E compression could continue regardless of ADBE-specific fundamentals.

  5. No dividend — Capital return is buyback-only. If buyback pace slows for any reason (M&A spending, regulatory pressure, mgmt change), the EPS-accretion thesis weakens immediately.

  6. Semrush integration risk — $1.3B acquisition closing Q2 FY26. Small relative to ADBE's scale, but failed integrations have killed prior software M&A theses.


How This Fits the Portfolio


Decision Framework

WAIT (do not deploy T1 today)

Reason: Per shr-034, RED competitive flag (Claude Design / GPT Images 2.0 launched Apr 17) is unresolved AND the next positive catalyst (Q2 FY26 earnings) is 46 days out — outside the 30-day window. Cash is a position.

Triggers to deploy T1

Deploy a $300 / 1-share T1 in the $230-270 zone if at least one of these flips:

  1. Q2 FY26 earnings (Jun 11, 2026) confirms Creative Cloud ARR growth ≥ 10% AND management reaffirms full-year guide. (Primary trigger.)
  2. Adobe pre-announces Q2 ARR strength before the print (rare, but watch for 8-Ks).
  3. Insider open-market buy by CEO, CFO, or board members > $250K aggregate — would be a strong shr-002 GREEN flip.
  4. Software-sector re-rating — ADBE re-couples with NDX over a sustained 3-week window AND no further generative-AI competitive shocks.

Triggers to abandon

Move ADBE off the watch list and write a retro if:

  1. Q2 FY26 earnings shows ARR < 9% AND management cuts FY26 guide.
  2. Major C-suite open-market sells > $5M before Jun 11.
  3. Buyback paused in favor of a large speculative AI acquisition (>$10B).
  4. Operating margin compression below 34% in Q2 FY26.

If executed (contingent plan)

Watchlist alerts to set (only after positive Q2 trigger)


Bottom Line

ADBE at $245 passes the Graham value test on the framework that matters for tech (Growth Stock Addendum + SBC-adjusted P/FCF), but fails the same-day shr-034 readiness check because of fresh, unresolved competitive disruption.

The disciplined move is to wait for Q2 FY26 earnings on June 11, 2026 — a 46-day hold with no capital deployed. If ARR holds, deploy T1; if ARR cracks, this stays a no-go and a useful negative case study for the playbook.

This is the same lesson shr-034 was written from: passing Graham filters at trough valuations is necessary but not sufficient — you also need evidence that the headwinds are stabilizing, not intensifying. Today, the headwinds are intensifying.