“For the investor, a too-high purchase price for the stock of an excellent company can undo the effects of a subsequent decade of favorable business developments.”
— Warren Buffett
Listen to the deep-dive discussion – Why Alphabet dilutes shareholders for AI
Download the Slide Deck (PDF) The_Growth_Ledger
I. The Headline
On June 1st, 2026, Alphabet announced it would raise $80 billion by selling its own stock. The capital will fund what management calls “world-class AI compute infrastructure.” The company’s full-year capital expenditure budget now stands at $190 billion. The CEO explained that demand exceeds the capacity the company can deliver.
On the same day, Berkshire Hathaway agreed to buy $10 billion of that stock at approximately $350 per share—a modest discount to the market price.
Two signals in one headline. Alphabet is selling its own shares to finance growth. Buffett’s successor is buying. One reveals the cost of the wave. The other recognizes the architecture beneath it. And neither offers the one thing the framework demands at the moment of purchase—a gap.
II. The Cost of the B
In the Averaging Up framework, every business has an A—the toll it collects today—and a B—the wave that compounds the toll over time. The critical question is never whether the B exists. It is what the B costs.
When the B is free—when growth arrives inside the moat, at zero marginal cost, without capital deployment—the business is a Freesurfer. Revenue grows. Costs do not move. Margins expand. Shareholders are concentrated through buybacks funded by excess cash.
When the B is mechanical—when growth requires capital expenditure, human effort, and competitive spending—the business must pay for its future. Sometimes from cash flow. Sometimes from debt. And now, for the first time among the technology giants, from the pockets of its own shareholders.
Alphabet is the first mega-cap technology company to fund the AI race by diluting its equity holders. Microsoft and Amazon have financed through cash flow and debt. Google’s $80 billion equity raise reveals what the balance sheet alone does not: even the cash flow of the most profitable company on earth is not enough to pay for the mechanical B. When the cost of growth exceeds the cash a business generates, the growth is not free. It is the opposite of free.
III. Growth Inside the Moat, Growth Outside It
Here is the principle that separates the Freesurfer from the Imperfect B, and it is the heart of this entire discussion.
Growth captured inside the moat is the B. Growth that requires leaving the moat is a cost.
Google Search grows inside its moat. More queries, more advertisers, more intent expressed through the dominant platform—all collected automatically by a toll that no competitor can replicate. The moat that protects Search also captures its growth. This growth costs nothing because the infrastructure already exists and no rival can enter. This is the true B.
AI compute and Cloud infrastructure is growth outside the Search moat. The dominance of Search does nothing to protect Google in cloud computing. There, Google faces Microsoft Azure and Amazon Web Services on a level playing field—three giants of comparable resources, each spending $100 to $200 billion to capture the same demand. The Search moat collects none of this growth automatically. Every dollar of AI revenue must be fought for, built for, and paid for.
This is the distinction the market misses. When a company grows by expanding inside its moat, the growth is the B—free and compounding. When a company grows by entering a new market where its moat has no power, the growth is not the B at all. It is the cost of building a second business on contested ground. The $190 billion is not the expansion of Google’s franchise. It is the construction of a new one, in mid-flight, against entrenched competitors.
IV. The Magnificent A
This is not a post against Alphabet. Or Microsoft. Or Amazon.
Google Search is a behavioral monopoly with over 90% market share—arguably the most powerful toll in technology. Microsoft Windows and Office are embedded in the workflow of every enterprise on earth, with immense switching costs. Amazon Web Services runs a third of the world’s cloud, with prohibitive migration costs and real data gravity.
All three possess a magnificent A. A dominant toll. A deep moat. Cash generation that dwarfs most industries. The A is not in question. It never was.
The B is the question. And the AI race answers it in real time.
V. The Three Layers Test
The Freesurfer captures three layers of growth at zero cost: real volume, pricing power, and inflation pass-through. The Imperfect B captures some layers naturally and pays for others.
Google: Volume grows naturally inside the Search moat. Pricing power is partial—ad pricing is auction-based and competitive, cloud pricing is deflationary. Inflation pass-through is partial. Roughly two of three layers are captured inside the moat. The third must be purchased outside it.
Microsoft: Volume grows naturally. Pricing power is strong—Office and Windows subscriptions increase regularly. Inflation pass-through is moderate. Roughly two and a half of three layers—but Azure still demands the capex arms race.
Visa: Volume grows naturally. Pricing power is structural. Inflation pass-through is automatic—the percentage toll captures every price increase without changing a fee. Three of three layers. Zero cost. No equity raise. No $190 billion.
VI. Why Berkshire Buys
Greg Abel is not making a mistake. Berkshire’s $10 billion in Alphabet recognizes what the framework recognizes: in the absence of a pure Freesurfer in technology, the magnificent A with an imperfect B is the next best thing.
Search is so dominant, so deeply embedded in human behavior, so impossible to replicate, that even $190 billion of mechanical capex cannot destroy the economics. The natural B—Search—generates enough cash to subsidize the mechanical B—Cloud and AI—while still returning value to shareholders. Berkshire buys the A and accepts the impurity of the B.
But notice what Berkshire does not do. It does not sell its Freesurfers to buy Google. It does not abandon the payment networks or the rating agencies for the AI narrative. It adds Google with new capital, in a portfolio that already holds the purer architectures. The hierarchy is preserved. The Imperfect B is an addition, never a replacement.
VII. The Missing Gap
And here the epigraph does its work. Google at $367 offers no gap. The stock trades above any reasonable zero-growth floor. The investor who buys today pays full price for the magnificent A and pays again, through dilution, for the mechanical B.
Buffett’s warning is precise: a too-high purchase price for an excellent company can undo a decade of favorable business developments. Google may win the AI race. The $190 billion may compound into something extraordinary. But if the entry price already reflects that optimism, the investor receives no margin of safety—no gap to protect against disappointment, no free gift of growth that was excluded from the price.
Contrast this with the Freesurfer sitting 22% below its zero-growth floor while the market watches AI headlines. One business is priced for a decade of perfect execution. The other is priced as if it will never grow again—and grows anyway, for free.
VIII. The Architecture Difference
On June 1st, 2026, Alphabet sold $80 billion of its own stock to fund growth. On the same day, Visa continued buying back its own shares—funded entirely by free cash flow, with no capital expenditure required.
One business dilutes its shareholders to purchase the wave. The other concentrates its shareholders while the wave arrives for free. One needs $190 billion to grow. The other needs zero.
The A can be magnificent in both. The B is where the architecture diverges. Google is an extraordinary business—but it is not a Freesurfer, and the $80 billion equity raise is the proof. When the cost of growth exceeds the cash flow of the most profitable company on earth, the B is not free. And when the B is not free, the compounding is not geometrical.
$190 billion to grow. Zero dollars to grow.
The price of the B is the architecture made visible. And the price of the stock is whether you receive a gift—or pay for one that may never arrive.
———
Related posts on averagingup.com:
Visa vs Mastercard · The Imperfect B · Cheap Money Hides Impurity · The Architecture Beneath the Margin · Three Layers, Zero Cost · The Freesurfer · Toll Position
Based on the framework from The Infinite Investor, available at averagingup.com.