The Legendary Debate

Warren Buffett: adjusts his glasses and leans forward "Software businesses fascinate me because the best ones print money — and Adobe prints more money than almost any of them. But here's the question that's kept me up this week: Shantanu Narayen just announced he's leaving after eighteen years and a hundred earnings calls. He took this company from selling shrink-wrapped boxes to running a $24 billion subscription engine. Now he's handing the keys to someone who hasn't been named yet, right in the middle of the biggest technology shift since the internet. So let me ask it plainly — is Adobe a business that succeeds because of its CEO, or because of its architecture? Because the answer to that question determines whether this transition is a speedbump or a cliff."

Charlie Munger: "Warren, I'd answer your question by inverting it — what would have to go wrong for Adobe to actually die? Think about it concretely. Photoshop has to become irrelevant, which means AI-generated images have to replace precision editing so completely that the 85% of Sundance films made with Adobe tools — as they just reported — suddenly switch to prompting Midjourney for their post-production work. The PDF standard, which is literally ISO 32000 and embedded in every government filing system and legal workflow on Earth, has to be replaced by some format that nobody has invented yet. And the Experience Platform that processes 35 trillion segment evaluations daily for 99 of the Fortune 100 has to be displaced by — what exactly? Salesforce doesn't have the creative workflow. Microsoft doesn't have the marketing orchestration. You'd need all three legs to fail simultaneously, and I can construct scenarios for each individually, but the probability of all three collapsing at once approaches the kind of number that doesn't move my pencil. The Narayen departure worries me — I'll grant you that — but the business architecture is more durable than any one person."

Dev Kantesaria: "Charlie, I want to push back on your confidence in the enterprise tier because I think you're conflating switching costs with inevitability, and those are fundamentally different things. When I look at my holdings — FICO, Moody's, Visa — the toll is mandatory. You literally cannot originate a mortgage without a FICO score. You cannot issue a bond without a Moody's or S&P rating. The activity cannot occur without paying the toll. Adobe's enterprise position is strong — I grant that — but it is earned through competitive superiority, not structural mandate. A Fortune 500 CMO uses Experience Manager because it is currently the best option, not because some regulation requires it. The moment a sufficiently compelling AI-native alternative emerges for content management or marketing orchestration, that CMO can switch. The PDF standard is closer to a genuine toll booth — I'll acknowledge that — but Photoshop and Illustrator are competitive advantages with different durability profiles. And here is what genuinely concerns me: the company itself acknowledged that its stock photography business is declining 'steeper than expected.' When management's own internal models cannot predict the pace of AI substitution in one product category, I have to ask whether the models for the adjacent categories are similarly optimistic."

David Tepper: "Dev, I hear your framework, but you're describing a theoretical risk while ignoring the empirical reality sitting right in front of us. This business just posted its best financial year in four decades — record revenue, record margins, record cash flow — and the stock is trading at a multiple that implies it's about to permanently break. I've been in this game long enough to recognize what's happening here: growth-oriented funds can't hold a stock below 20x earnings, GARP investors are style-drifting out, and the AI narrative has created a consensus fear that hasn't shown up in the actual numbers. The stock photography business that everyone keeps citing as the canary in the coal mine represents maybe three to four percent of revenue. Meanwhile, the enterprise CXO segment is growing at 30-plus percent. The Business Professionals segment is accelerating at 15%. Firefly ARR just exceeded $250 million growing 75% quarter-over-quarter. The market is looking at the stock photography canary and ignoring the enterprise eagle that's gaining altitude. That's not analytical rigor — that's narrative bias."

Pulak Prasad: "David, I respect your ability to find asymmetric setups in beaten-down situations, but I think you're underweighting something that my evolutionary framework is specifically designed to detect: the rate of environmental change. Adobe has survived technology transitions before — desktop to web, web to cloud, perpetual licenses to subscriptions — and I acknowledge that track record. But each of those transitions happened over five-to-seven year cycles with relatively predictable competitive dynamics. What I'm observing now is qualitatively different. Canva went from zero to over 200 million users in roughly a decade. Midjourney went from nothing to producing images indistinguishable from professional photography in approximately three years. CapCut became the dominant mobile video editor in perhaps two years. And Microsoft embedded AI-powered design into the 365 suite that 400 million people already use — at zero incremental cost. That pace of competitive entry is the fastest I have observed in enterprise software. My framework asks a simple question: is the habitat getting more or less hostile for the incumbent? And the answer, for Adobe's creative tools tier specifically, is unambiguously more hostile. The enterprise CXO tier may be different — I genuinely believe it would pass my evolutionary filter independently. But I cannot own that tier without also accepting exposure to creative tools where the competitive landscape is mutating faster than management can forecast."

Robert Vinall: "Pulak raises the right question, but I think the answer is more nuanced than his framework allows. What I see in Adobe is a bifurcated business — and the market is applying a blended discount that dramatically undervalues the widening-moat enterprise tier while arguably being fair-to-generous about the pressured consumer tier. AEP and apps growing 30-plus percent, GenStudio ARR growing 30-plus, Firefly Enterprise customer acquisition up 50% — those are not the metrics of a business being disrupted. Those are the metrics of a business whose deepest switching costs are getting deeper. The 35 trillion daily segment evaluations on AEP represent data gravity that compounds with each year of accumulated customer data, making switching progressively more irrational. That is the widening moat I look for. But — and this is where I apply my Myth Number Five — I worry that Adobe's 40-year dominance in creative tools has created institutional processes optimized for a per-seat subscription world that may not adapt quickly enough to consumption-based pricing. The AI transition requires fundamentally different capabilities than what built the subscription business. And the CEO who understood both the old and new worlds is leaving. So my conviction is real, but it is tempered by uncertainty about the transition's execution."

Mohnish Pabrai: "I'm going to be direct, because my framework makes this straightforward. Adobe is an exceptional business — I'm not disputing that. The ROIC trajectory, the gross margins, the cash conversion — it is genuinely world-class. If this business had a $30 billion market cap, I would be writing my largest check. But it has a $102 billion market cap, and my framework requires 3:1 asymmetric upside-to-downside. For me to achieve that, Adobe needs a realistic path to a $300 billion enterprise value. Run the math: even if you assume 15% annual FCF per share growth for five years and slap a 20x terminal multiple on that future cash flow, you get roughly 2.7x from today's price — below my 3:1 threshold once you risk-adjust for the CEO succession, the AI transition, and the balance sheet dynamics that Robert mentioned. I also want to flag something nobody else has addressed: management spent roughly $20 billion on buybacks over the last two fiscal years at average prices of $330 to $380 per share, funded in part by $4 billion in new debt. The stock now sits at $248. That is a capital allocation error of extraordinary magnitude — billions of dollars of shareholder value destroyed by buying shares at prices 30 to 50 percent above where we are today. It does not kill the business, but it tells me that management's price discipline is suspect, and that matters when the balance sheet is getting levered during a period of strategic uncertainty."

Warren Buffett: "Mohnish raises a fair point about the buyback timing, and I want to address it honestly because I've made the same mistake at Berkshire — we bought IBM at prices that proved too high, and I've learned that even wonderful businesses can be over-bought. But I draw a different conclusion from the buyback data than Mohnish does. Yes, Adobe repurchased shares at $330-380 that are now worth $248. But the reason they're worth $248 is not that the business deteriorated — it's that Mr. Market changed his mind about AI risk. The business itself just posted its best year ever. If you believe — as I do — that the enterprise moat is intact and the AI transition is navigable, then those $330-380 buybacks will look reasonable in hindsight because the stock won't stay at $248. The more important signal to me is that at current prices, each buyback dollar retires 30 to 50 percent more shares than it did last year. The compounding math improves dramatically from here."

PHASE 2: THE FINANCIAL EVIDENCE

Charlie Munger: "Let's put numbers to the qualitative story, because the financial trajectory here tells you something that sentiment cannot override. ROIC went from 3.9% in 2014 — the absolute trough of the subscription transition when they were deliberately cannibalizing their own perpetual license revenue — to 36.3% in fiscal 2025. That is a nearly ten-fold improvement in capital efficiency achieved while revenue quadrupled from $5.9 billion to $23.8 billion. In my experience, a business that can simultaneously grow and become more capital-efficient is one of the rarest things in capitalism. It means the moat is actually widening — each incremental dollar of revenue requires less capital to produce than the last. Gross margins have held at 89% for a decade, which is the financial expression of pricing power that customers are not pushing back on. Operating margins expanded from 25.5% to 36.6%. Free cash flow per share went from $4 to $23 in nine years. These are not the numbers of a business being disrupted. These are the numbers of a business at the peak of its competitive powers. Now — is that peak also the beginning of decline? That is the real debate, and I concede that neither I nor anyone in this room can answer it with certainty."

Dev Kantesaria: "Charlie, I want to acknowledge that the financial trajectory is extraordinary — and I say that as someone who has spent 18 years in venture capital studying what real compounders look like from the inside. The ROIC curve is genuinely comparable to what I see in FICO and Moody's, which is the highest compliment I can pay a business. But I need to address the capital allocation honestly, because it cuts both ways. SBC was $1.9 billion in fiscal 2025, roughly 8% of revenue. Net buybacks were $10.9 billion, exceeding free cash flow by over a billion dollars — the gap filled by new debt. Equity has declined from $16.5 billion to $11.6 billion in just two years. On a per-share basis, the share count has gone from 498 million in 2016 to roughly 413 million — which is genuine accretion, but the rate is closer to 1.7% annually over the full period, not the 4% that the recent debt-fueled acceleration suggests is sustainable. The question I ask is: what happens if AI creates a 12-to-18-month margin trough similar to what the subscription transition created in 2013-2015? During that earlier transition, ROIC fell to 3.9% and the business burned through its flexibility. Today, the balance sheet has less cushion."

David Tepper: "Dev, I think you're building a bear case that the financial data directly contradicts. Let me point to the most recent quarter — Q1 FY2026. Revenue was $6.4 billion, growing 11%. GAAP EPS was $4.60, up 11%. Non-GAAP EPS was $6.06, up 19%. Total ending ARR hit $26 billion. The Business Professionals segment grew 15%. Enterprise CXO grew 30-plus. New AI-first ARR tripled year-over-year. These are not the numbers of a business preparing for a margin trough. The 850 million MAU — while I grant that management warned MAU growth 'dampens ARR in the short term' — represents the single largest creative software distribution platform on Earth. Even if only 5% of those users convert to paid, that is 42 million subscribers, each paying enough to generate the $24 billion revenue base we see today. The freemium funnel is a strategic investment, not a dilution risk."

PHASE 3: VALUATION AND VERDICTS

Robert Vinall: "I want to lay my cards on the table, because I think the valuation debate exposes a genuine philosophical split in this room. At $248.15, using the conservative end of free cash flow — let's call it $8.5 billion after reconciling the different definitions across data sources — you get roughly a 12x FCF multiple on a business with 36% ROIC and 89% gross margins. Using FY2025 EPS of $16.74, the P/E is about 14.8x. Both are historically anomalous for a business of this quality — Adobe has typically traded at 25-35x earnings over the past decade. The market is embedding significant structural disruption risk. My calculation of fair value, blending EPS-based and FCF-based approaches, centers around $330-370. At $248, that is 33-48% upside — meaningful. But I want the margin of safety that accounts for the CEO transition and the Myth Number Five risk I described earlier. I would build a full position at $225, where the margin from my conservative $340 fair value exceeds 33%. At $248, I am interested but want to see one more data point — either the CEO successor announcement or Firefly ARR crossing $500 million — before committing significant capital."

Mohnish Pabrai: "The business is exceptional — I have said that clearly and I mean it. But at $102 billion market cap, even Robert's $370 fair value — which I consider reasonable — offers only 50% upside. With the CEO transition, the AI uncertainty, and the debt-funded buyback overhang, the risk-adjusted ratio falls below my 3:1 threshold. I would re-engage below $195 per share, which implies roughly $80 billion market cap — and at that level, owner earnings of $15-17 per share produce an 8-9% yield with genuine tripling potential over five years."

Dev Kantesaria: "I want to be precise about my position, because I respect this business enormously. The enterprise tier genuinely passes my quality filters. AEP is approaching toll-booth characteristics — the data gravity, the integration depth, the 99 of Fortune 100 customer base. But my framework requires that the toll be structurally inevitable, not merely competitively superior. And when I look at the creative tools tier — which is still the majority of revenue — the toll is earned through competitive dominance that AI is actively challenging. The stock photography decline is not a small thing to me. It is the first empirical evidence that AI substitution is moving faster than management predicted. I need more evidence before I can say with conviction that this is a toll booth I'd own for a decade."

Pulak Prasad: "My position is clear: I avoid industries where the competitive landscape is changing faster than management can forecast. Adobe's own stock photography miss is my evidence. I would revisit if Creative Cloud subscription growth sustained above 10% for four or more quarters in the face of AI competition, Firefly ARR crossed $1 billion demonstrating the consumption model works at scale, and a CEO successor articulated a clear and testable AI strategy. Those are specific, measurable conditions — not a vague 'wait and see.'"

Warren Buffett: "At $248.15, I see a business generating roughly $10 billion in annual operating cash flow with 89% gross margins, 95% recurring revenue, and switching costs embedded in institutional reality. I would buy a starter position here at 2-3% and build to 5% below $225. The risk-reward from $248 to a conservative $330-340 fair value is attractive, and the downside is anchored by a franchise that even the bears in this room acknowledge is genuinely world-class."

PHASE 4: SYNTHESIS

Warren Buffett: surveys the room "Let me try to pull the threads together, because this was one of the more genuinely divided debates we've had. On the qualitative side, there's broad agreement that Adobe has built something extraordinary — you don't get to 850 million monthly active users, 89% gross margins, and 99 of the Fortune 100 as customers without a deep structural moat. Charlie's inversion test is powerful: you have to construct a triple simultaneous failure across creative tools, document management, and enterprise marketing to kill this business, and that probability is vanishingly small. Even Pulak and Mohnish, who are avoiding the stock, do so because of framework constraints — market cap, competitive pace — not because they dispute the franchise quality.

Where we split is on the AI transition and its second-order consequences. Dev and Pulak see the stock photography decline as a leading indicator of broader creative tool substitution — a pattern that's changing the habitat faster than the organism can adapt. And they have a point: management itself admitted the decline was steeper than their models predicted, which means the bears are not imagining things. David, Robert, and I see the same data but draw a different conclusion — the enterprise tier where the deepest switching costs live is accelerating at 30%, Firefly is monetizing at $250 million and tripling, and the 850 million MAU base provides a distribution advantage that no AI-native startup can replicate. We believe the market is pricing the stock photography canary while ignoring the enterprise eagle.

The financial evidence resolves the debate in favor of the business, if not the stock. ROIC climbing from 13% to 36% while revenue quadrupled — that is a widening moat expressed in dollars. EPS growing from $2.35 to $16.74 in nine years — that is a compounding engine. Free cash flow per share going from $4 to $23 — that is real cash, not accounting fiction. The balance sheet bears watching — equity declining from $16.5 billion to $11.6 billion, debt rising from $3.6 to $6.2 billion, buybacks at $330-380 now underwater — but the operating cash flow of $10 billion provides resilience that most businesses would envy.

So five of us — Charlie, David, Robert, Dev on the enterprise merits, and myself — believe this business warrants ownership at or near current prices, though Robert wants a slightly lower entry. Mohnish and Pulak avoid it on framework grounds that I respect even where I disagree. The consensus: Adobe is a franchise-quality compounder being priced by Mr. Market as though its best days are permanently behind it. We think Mr. Market is wrong — but we're sizing our positions to acknowledge that he might not be."