Home Case-Studies Content Strategy Portfolio Financial Modelling Articles Contact Why Pay 2trn for Anthropic if you can own Alphabet?

Why Pay $2 Trillion for Anthropic When You Can Own Alphabet?

gcwagner.com | October 2026 | G.C. Wagner

TL;DR: Alphabet fell 12% in a day after a messy quarter. Anthropic is about to list at a target value of $2 trillion. Alphabet owns around 14% of it, runs a rival model, and makes serious profit. For roughly 1.7 times the price, you get the whole Alphabet machine. I think that is the better deal.

Alphabet had a bad week. Anthropic is having a very good one, at least on paper. I think the market has got the two the wrong way round.

First, what went wrong. The quarter was a miss on both lines. Revenue was $127.6 billion against $128.2 billion expected, and adjusted earnings per share were $2.67 against $3.02. It was the first double miss in nearly four years. The shares closed down 12.3% at $282.32. About $480 billion of market value went with them.

The cash problem

Investors were upset about cash, not profit.

Alphabet spent $51.4 billion on capital expenditure in the quarter. Free cash flow was negative $9.1 billion. The quarter before it was negative $5.9 billion. Over the last twelve months it is still positive, but only by $19.7 billion.

The spending is set to climb. Alphabet expects capex of $195 to $205 billion this year and says it will rise significantly in 2027, with annual capex expected to top $300 billion. Operating cash flow over the last twelve months was about $180 billion. So the company is funding the gap. In the first nine months it bought back no shares, raised about $43 billion by selling new ones, and took long-term debt from $46.5 billion to $125.1 billion.

That's the price of the AI race. It is also the reason this is not a classic value stock. A classic value stock hands cash back. Alphabet is spending it, and the question is what it buys.

What the money buys

Mostly, it buys cloud. Google Cloud revenue doubled to $30.7 billion, up 103%. Cloud operating income rose to $11.1 billion from $3.6 billion a year ago. That's a margin of about 36%. One analyst told CNBC that cloud was the only business that beat, and everything else missed.

The Gemini app also passed one billion users. Search query growth was the best in five years. The demand is real.

The Anthropic question

Now the exciting bit.

Anthropic has filed to list on Nasdaq. It plans to raise $75 billion to $100 billion, with a debut set for 20 November. The talk was a $2 trillion valuation. Some investors now say $1.5 trillion is more likely. I'll use $2 trillion, because that's the target.

Anthropic is an impressive business. It told investors that third-quarter revenue was $16.8 billion and that it reached profit on an adjusted basis. But adjusted profit is not the same as profit. Earlier documents showed an operating loss of more than $8 billion and a net loss of nearly $42 billion. And its prospectus shows $431 billion of compute contracts that can't be cancelled.

Here are the two side by side. I annualised each company's latest quarter.

Alphabet Anthropic (at $2tn target)
Value about $3.47tn $2.0tn
Annualised revenue about $510bn about $67bn
Value divided by revenue about 6.8 times about 30 times
Latest quarter operating profit $40.6bn (32% margin) Adjusted profit only
Free cash flow, last 12 months +$19.7bn Not reported
Cancellable? Capex is a choice $431bn of compute contracts are not

My calculation from Alphabet's results and the reports above. Alphabet's market value is 12.29 billion shares at $282.32. Anthropic's revenue is its third-quarter figure multiplied by four.

Anthropic is growing faster, and that's what the extra price pays for. But look at what you get for the money. Alphabet charges about 7 dollars for each dollar of revenue. Anthropic charges about 30.

Now the key point. Alphabet owns roughly 14% of Anthropic, and its contract caps the stake at 15%. At $2 trillion that stake is worth $280 billion. Strip it out and the rest of Alphabet is priced at about $3.2 trillion. So buying all of Anthropic at $2 trillion costs about 63% of what the entire remaining Alphabet empire costs.

That remaining empire includes:

  • Search and YouTube, which make most of the profit
  • Google Cloud, doubling in a year
  • Gemini, a direct rival to Claude
  • Google's own AI chips and data centres
  • Waymo, and stakes in other companies

Alphabet is even a supplier to Anthropic. It can lend the company up to $20 billion through convertible notes. If Anthropic wins, Alphabet gets a share. If Gemini wins, Alphabet gets that instead. That's two bets on one outcome.

So my question is simple. Why pay $2 trillion for one AI lab with non-cancellable commitments and an adjusted profit, when $3.5 trillion buys the same AI exposure and a profit machine on top?

What else is hiding inside Alphabet

Here's a rough list of what's inside the share price.

Item Value Share of market value
14% of Anthropic (at the $2tn target) $280bn 8.1%
SpaceX shares (sale restricted for now) $75.2bn 2.2%
Cash and securities minus long-term debt $140bn 4.0%
Total about $495bn about 14%

My calculation. I left out Waymo and other holdings because Alphabet doesn't give its exact share. New shares sold in the Anthropic IPO will dilute the 14% a little.

On the quote page I used, Alphabet trades on a forward price to earnings ratio of about 17.3. The S&P 500 is on 18.7, according to FactSet. Take $495 billion of assets off Alphabet's value and the multiple on the rest drops to about 14.8. That's well below the market, for a business growing revenue about 25%.

The Siemens lesson

Look at what happened to Siemens shareholders. Over the years the group has spun off and floated businesses, and owners got shares in each one. The latest plan is to pass 30% of Siemens Healthineers directly to Siemens shareholders, with a vote at the February 2027 annual meeting. Investors had pushed for a simpler structure because the market valued the holding company below the sum of its parts.

Alphabet has the same set-up. Waymo was valued at $126 billion in February. Yet the whole Other Bets division made $372 million of revenue and lost $2.05 billion last quarter. One analyst puts the parts at about $337 a share against today's $282. Alphabet has not said it plans a spin-off. But the Anthropic IPO, on 20 November, will put a public price on the largest piece, and that alone could change how the market sees the rest.

The advertising worry

The bear case is that chatbots steal searches and ad money follows.

The quarter gave both sides something. Search revenue grew 12% to $63.4 billion, which missed expectations even with record query growth. Google says AI Mode queries grew faster than the money made from them. Its business chief said AI Mode monetised less well, and that tests of ads in it looked promising. OpenAI is adding new ad formats to ChatGPT, and Google admits lower monetisation in markets where ChatGPT ads have launched.

But think about where advertisers can actually go. Big brands need scale. Google has two huge places, Search and YouTube, where YouTube ads alone made $11.2 billion. The main alternative is Meta, and Meta grew revenue 26% to $64.6 billion. So money does move. But the pool is small, and Google owns two of its biggest corners.

What I'm watching

Three things:

  1. Cloud keeps growing, with the margin holding as the buildout continues.
  2. AI Mode ads close the gap, so Search growth turns into revenue growth.
  3. Capex levels off, and free cash flow turns positive again.

The thesis breaks if Search revenue slips, or if the 2027 capex plan arrives with no growth to match.

My takeaway

When a company is cheap, it's usually because the market sees a problem. Here the problem is cash burn, and that is real. But it's the same cash burn Anthropic has, only Alphabet funds it from a business that earns $160 billion a year in operating profit.

If you believe in AI, you have two ways to own it. One costs $2 trillion for a lab with an adjusted profit and $431 billion of fixed compute bills. The other costs $3.5 trillion, and comes with the profits, the cloud, the chips, the data and a 14% slice of the first one. I know which I'd rather own.

Disclaimer: This content is intended for general informational and educational purposes only and does not constitute financial, investment, legal, tax or accounting advice. I make no representations or warranties about the accuracy, reliability, or completeness of the information provided. The author has a position in Alphabet. Please conduct your own research and/or consult with a qualified professional for advice tailored to your individual circumstances before making any investment decisions.

G.C. Wagner writes on finance, culture and the occasional inconvenient philosophy, and helps businesses with content, SEO and GEO. gcwagner.com

Bibliography