Adobe (NASDAQ: ADBE)

The call

We are initiating on Adobe with a HOLD and a $280 fair value, +15.6% against the $242.27 close on Friday, October 9, 2026.

Adobe trades at 9.9 times the non-GAAP earnings it has guided for this fiscal year. For a company growing revenue 13% at a 44.0% non-GAAP operating margin, a multiple that low is not a valuation. It is an accusation. The market is saying that the reported growth is borrowed from a future that will not arrive, and that generative AI is going to take the creative software franchise apart.

So the only question worth modeling is whether Adobe's own disclosure supports the accusation. Our finding is that it partly does, which is not what we expected to write.

The headline numbers are excellent and the numbers underneath them are not. Revenue grew 13% last quarter. Total annual recurring revenue reached $27.50bn and grew 11.2%. Remaining performance obligations, which is the contracted backlog and the closest thing to a floor under future revenue, grew 8%. GAAP net income, from fiscal 2025 to the company's own guidance for fiscal 2026, grew 1.8%. And net new annual recurring revenue, the figure an analyst put to management on the third-quarter call and which management did not dispute, fell by about 36%.

Every one of those is a growth rate for the same company in the same quarter. They run from 13% to minus 36%, and they sort almost perfectly by how far each one sits from cash actually arriving. That ordering is the report.

Revenue grew 13 percent. Profit did not.

Here is every growth rate Adobe disclosed for the quarter ended August 28, 2026, sorted from worst to best.

Read it from the bottom. Revenue is the fastest-growing number Adobe reports and it is also the most lagging, because subscription revenue recognized this quarter was contracted in previous ones. Then ARR, then the backlog, then profit. The further you move toward money the company gets to keep, the slower it goes.

The two lines that matter most are the first and the last. GAAP net income grew 1.8% on revenue growth of 13%. GAAP earnings per share, meanwhile, grew 8.7%. The entire difference is the share count, which fell 6.8% year over year as Adobe retired 9.5 million shares for $2,232m in the quarter alone.

Why did the profit not follow the revenue? Because Adobe chose that. GAAP operating margin fell from 36.6% in fiscal 2025 to 34.8% in the third quarter. Management is spending on a freemium acquisition strategy, and it explicitly deferred Creative Cloud pricing actions. On the earnings call the outgoing chief executive said he was "really happy that we didn't focus on the pricing actions," arguing that user acquisition matters more than short-term pricing relief.

That is a defensible strategy and it may well be the right one. It is also, precisely, a decision to trade near-term profit for future optionality, and it means the current profit line understates the earning power of the installed base. Bulls and bears agree on the mechanism and disagree on whether the conversion ever happens.

The evidence Adobe offers for the conversion is real but small. AI-first ending ARR passed $650m and grew more than 150% year over year. Firefly ARR grew 40% quarter on quarter. Monthly active users among business professionals and consumers passed 900 million, up more than 25%, and creative freemium users passed 100 million. Set $650m of AI ARR against $27.50bn of total ARR, though, and it is 2.4% of the company. It is growing fast from approximately nothing, which is what things do.

What the price assumes

This is the method this publication uses on every company: take the price as given, because it is the one number nobody has to estimate, and solve for the growth it requires. See the reverse DCF for the mechanics and ask what the price assumes for why we prefer it.

Discounting at 10% with a 2.0% terminal rate over 10 years, Adobe's $242.27 price implies that cash earnings grow +1.75% a year for a decade and then 2.0% forever. Against 2.5% inflation, that implied figure is -0.73% in real terms. The market is paying for a business whose profits shrink slightly, in purchasing power, every year from here.

Put that next to what Adobe actually reports.

The contracted backlog is growing 8%. That is the most conservative forward number Adobe publishes, it is a contractual obligation rather than a forecast, and it alone would be worth $380 a share on this model. The price is $242.27.

The discount rate is the one input with no observable anchor, so here is the same solve across a range of them. A capital asset pricing calculation on Adobe's 1.42 beta would give 11.1%, and we think that is too high for reasons set out in the methodology.

Across every hurdle we would defend, the price requires cash earnings growth between roughly minus one half of one percent and plus three and a half percent, nominal, forever. For a business with $27.50bn of recurring revenue growing 11.2%, that is a strong claim about the future. It is not, however, an absurd one, and the next section explains why we cannot dismiss it.

Which earnings are real

Everything above depends on a number we have not yet justified: what Adobe's cash earnings actually are. There are three candidates and the choice moves the answer by more than any other judgment in this report, so we are making it in the open.

Reported free cash flow of $10,591m is the wrong number and we will not use it. It adds back $2,071m of stock-based compensation, which is a real cost: the company either issues shares or spends cash preventing the issue, and Adobe is visibly doing the second, repurchasing $9,695m over the trailing twelve months against that $2,071m charge. That is 4.7 times cover. The cash is going out of the door. See non-cash is not free.

GAAP net income of $7,258m is too low. Adobe spends $214m a year on property and equipment against a $484m depreciation and amortization charge. A company whose depreciation runs at more than twice its capital spending is not consuming its asset base, and deducting the full charge understates what an owner gets.

Owner earnings of $8,520m is too high, for a reason that is easy to miss. Roughly $992m of the operating cash flow, which is 12% of owner earnings, is cash collected ahead of revenue recognition. Deferred revenue stood at $7,204m at the quarter end. That tailwind exists only while the subscription base keeps expanding. Pairing it with a growth path that fades toward 2% books the same optimism twice: it assumes the business both stops growing and keeps collecting the cash that growth produces.

So we use the middle: $7,528m, which is guided GAAP net income plus depreciation and amortization less capital expenditure, excluding the growth-dependent working capital. The shares trade at 12.5 times that figure, a 8.0% cash earnings yield.

If you disagree, the model is explicit about what your answer would be. On owner earnings the price implies growth of +0.01% a year and this is a clear BUY. On GAAP net income it implies +2.26% and the stock is close to fairly valued. On reported free cash flow it implies -3.10% and the stock is conspicuously cheap. We have picked the conservative middle, and we want readers to notice that the rating would change if we had picked differently.

The bear case, taken seriously

The bear case is not that Adobe is a bad business. It is that generative image and video models collapse the value of the tool layer, that Photoshop and Illustrator become one option among many for work that increasingly starts from a prompt, and that a franchise built on thirty years of switching costs and trained muscle memory stops being defensible when the thing being switched away from is the act of manual creation itself.

Three pieces of Adobe's own disclosure support it.

Net new ARR fell about 36%. This is the single most forward-looking number in the whole report. ARR is a stock and net new ARR is the flow that feeds it. A stock growing 11.2% on a flow that is down by more than a third is a stock whose growth rate is going to fall, arithmetically, whatever anybody believes about AI. Management attributed the decline to the freemium strategy and to seasonality and did not dispute the figure.

The backlog is growing more slowly than revenue. Remaining performance obligations grew 8% against revenue growth of 13%, and the current portion, the contracted revenue due over the next twelve months, grew 9%. The backlog is still being added to, but more slowly than revenue is growing. Unless bookings accelerate, revenue growth has to converge down toward the growth of the backlog it is drawn from, and the current portion puts that near 9%.

Pricing power is being deferred rather than demonstrated. A company confident in its moat raises prices. Adobe chose not to, and said so. The charitable reading is a land-grab ahead of monetization. The uncharitable one is that the pricing actions were withheld because the market would not have borne them. We cannot distinguish between those two from the outside, and neither can anyone else, which is the central problem with this stock.

And two further risks that are not about AI at all. Adobe changes chief executive on December 1, 2026, when Anil Chakravarthy succeeds Shantanu Narayen after eighteen years and Narayen becomes executive chair. The chief financial officer's seat is also open: Dan Durn left earlier in the year and Steven Day is serving on an interim basis. A company facing a genuine question about the durability of its franchise is about to face it with two new people in the two seats that matter most.

Why this is a HOLD and not a BUY

Our central case is that cash earnings grow 5% and fade to 2% over ten years, then grow at 2% in perpetuity. That path sits between what the GAAP line is doing now, about 2%, and what the contracted backlog is doing, 8%. It is below every growth rate Adobe currently discloses except the GAAP net income line itself. It is not a bullish path.

It produces $278.41 a share, which we round to $280. And we are still not calling it a BUY. Here is the grid that explains why.

Every cell in that grid rests on assumptions we would defend in public. The cells run from $225 to $333. The price is $242.27, which sits inside the range, in the lower third of it.

A fair value 16% above the price, drawn from a distribution that wide, is not a margin of safety. It is the midpoint of an argument we cannot settle. We hold the view that Adobe is more likely undervalued than overvalued, and the honest expression of a view that weak is HOLD rather than a BUY we would have to caveat into meaninglessness.

There is one more reason, and it is the asymmetry running the other way. The bear case here is not a multiple de-rating that mean-reverts. It is a permanent impairment of a franchise, and if it is right, the terminal value in every cell of that grid is wrong rather than merely optimistic. Our bear path, cash earnings flat and then declining 2% a year forever, gives $171 a share, -30%. Our bull path, 8% fading to 2.5%, gives $330. We do not have a confident basis for weighting those two, and anyone who tells you they do is describing a conviction rather than an analysis.

For context rather than for guidance: the consensus of 39 analysts is an average target of $275.06 and a median of $265.50, with a range from $195 to $373, and 23 of the 39 rate the shares Hold. Our $280 is near that average. We arrived at it independently and mention the coincidence only because a reader is entitled to know when an answer is unremarkable.

Risks, and what would change our mind

Upgrade to BUY if net new ARR stabilizes. This is the cleanest single trigger in the report. The flow feeding the recurring base is down by more than a third. Two consecutive quarters of it flattening or turning up would remove the arithmetic case for a decelerating ARR line and would move us.

Upgrade if the deferred pricing actions resume and stick. Pricing power is the whole question. A price increase that the installed base absorbs without elevated churn would answer it directly, in a way that no amount of AI-first ARR growth from a small base can.

Downgrade to SELL if ARR growth breaks below the high single digits while margins keep compressing. That combination, a decelerating recurring base and a profit line that does not respond, is what the early innings of a structural impairment would look like, and it would mean the price was right and we were slow.

Watch the operating margin, not the non-GAAP one. GAAP operating margin has fallen from 36.6% to 34.8%. Guidance assumes a 45.0% non-GAAP margin for the year, which excludes $2,071m of stock compensation. The gap between the two is $6.33 a share of guided earnings, and it is where the cost of defending this franchise is going to show up first.

And the risk to our own reasoning. We use a 10% cost of equity against a reported beta of 1.42, which a capital asset pricing calculation would put at 11.1%. We think the beta is high because the share price has been falling on disruption fear, and using a hurdle derived from that fear to discount cash flows we have already haircut for the same fear would charge the stock twice for one risk. A reader who disagrees should read the 11% row of both grids, where the price looks roughly fair. We have also valued the business on a cash earnings definition of our own construction, and as we argued in free cash flow is not a fact, every such definition is a choice that someone could have made differently.

Related research

Reports and notes that bear on this one. Every call on this site is dated, public and scored the same way, whether it worked or not.

• PayPal, the closest analogue on the board: another franchise where volume grows, profit does not, and the price underwrites a decline that has not started. We reached the opposite rating there, and the difference is instructive.
• McDonald's, the other HOLD, where a fine business was fairly priced and we said so rather than reaching for a call.
• Palantir, the same software sector at the opposite valuation extreme, which is a useful calibration for what the market will pay when it believes a growth story.

Also non-cash is not free on the stock compensation adjustment that drives our earnings base, and growth is not a virtue on why a growing top line and a flat profit line is the pattern worth noticing. The full list of calls, with the price each was made at, is on the scoreboard.

Methodology, sources & disclosures

Methodology

Every figure in this report is computed by a model script rather than typed, and the script asserts that the reverse solve reconciles to the observed market capitalization, that the three earnings bases nest in the correct order, and that the trailing twelve month cash flow figures reconcile from the primary filings. Valuation is a reverse discounted cash flow: we take the price as given, because it is the one number we can observe, and solve for the growth it requires. The central case uses a linear fade rather than a constant growth rate, because a company decelerating this visibly will not grow at one rate for a decade.

The earnings base

Trailing twelve month operating cash flow is computed from primary filings as fiscal 2025 ($10,030m) less the nine months to August 29, 2025 ($6,871m) plus the nine months to August 28, 2026 ($7,646m), giving $10,805m. Capital expenditure and stock-based compensation are reconciled the same way and agree with an independent compilation to within rounding on all three lines. The valuation base of $7,528m is guided fiscal 2026 GAAP net income plus depreciation and amortization less capital expenditure, excluding the $992m working capital tailwind, for the reasons set out above.

The reference price

The $242.27 reference price is the close on Friday, October 9, 2026, verified before any modeling against five independent sources. stockanalysis, Yahoo Finance and MarketBeat each reported the identical close, date and change of $1.22, or 0.5%; TradingView reported the same $242.27 with the market shown closed; and Robinhood showed $242.40 intraday in the same session. Four further sources were examined and rejected as stale rather than contradictory: Google Finance was serving a September 25 close, Morningstar a snapshot carrying a 404.2 million share count, Investing.com an October 7 page, and Macrotrends a 2024 price. We note the rejections because a reader checking our work against one of those four would find a different number and should know why.

The net new ARR figure

The decline of about 36% in net new annual recurring revenue is not a company disclosure. It was put to management by an analyst on the third-quarter earnings call of September 10, 2026, and management did not dispute it, attributing the slowdown to the freemium strategy and to normal seasonality. We treat it as reliable enough to reason from and flag it as second-hand wherever it appears, which is the same standard we would want applied to us.

Primary sources

Quarterly results, guidance and cash flow detail read directly from the company's own filings rather than through an aggregator.

• Third Quarter Fiscal 2026 Earnings Release, exhibit 99.1 to the Form 8-K filed September 10, 2026, for the quarter ended August 28, 2026. Source for revenue, segment subscription revenue, operating income, earnings per share, ARR, remaining performance obligations, repurchases and all guidance figures.
• Fourth Quarter and Fiscal 2025 Earnings Release, for the fiscal 2025 comparatives, full-year operating cash flow and the prior-year ARR base.
• Form 10-Q for the quarterly period ended August 28, 2026, condensed consolidated statements of cash flows, both nine-month columns, for operating cash flow, capital expenditure, stock-based compensation and depreciation and amortization, and the condensed balance sheet for deferred revenue, cash, investments, debt and equity.
• Third quarter fiscal 2026 earnings call transcript, September 10, 2026, for the AI-first ARR and monthly active user figures, the deferred pricing commentary, and the net new ARR exchange described above.
• Adobe leadership update, March 12, 2026, and subsequent reporting confirming Anil Chakravarthy succeeds Shantanu Narayen as chief executive on December 1, 2026.

Correction, October 13, 2026

The version of this report published on October 10, 2026 said that revenue was being recognized faster than new contracted revenue was being added. In dollar terms that was wrong: remaining performance obligations grew, so new bookings exceeded the revenue recognized out of the backlog. The correct statement, now in the bear case section, is that the backlog is growing more slowly than revenue, which pulls revenue growth down toward the backlog's rate over time. The argument it supported is unchanged, and no figure, the fair value or the rating changed.

Disclosure

The author holds no position in ADBE and has no plans to initiate one within 72 hours of publication. This report is opinion published for educational purposes and is not investment advice. 2 Comma Investor is not a registered investment adviser or broker-dealer. Analyst consensus figures are cited for context only and no part of the valuation is derived from them. Corrections, if any are needed, will be published on this page with a date attached.