The IPO headline everyone is chasing may not be the real investment opportunity.

Imagine a 261-page IPO document lands on your desk.

Everyone immediately screams:

$2 TRILLION VALUATION! 🤯

Then:

$42 BILLION LOSS! 😱

Meanwhile, buried underneath all that drama is a much more interesting question:

Who gets paid every time Anthropic spends another billion dollars building AI?

That, to me, is the real story.

Anthropic — the company behind Claude — has reportedly generated $4.59 billion of 2025 revenue, roughly 12 times its 2024 figure. Impressive.

But it also reported a nearly $42 billion net loss, including roughly $34 billion tied to accounting charges involving financial instruments. The operating loss was about $8.06 billion.

So yes, the $42 billion headline deserves context.

But let's not turn accounting into magic.

Losing $8 billion operationally is still losing $8 billion.

The really interesting number comes next.

💰 Follow the invoice

Anthropic reportedly has around $518 billion of future cloud, computing and infrastructure commitments.

That's more than 100 times its 2025 revenue.

And Reuters reports that a large majority of these commitments are binding regardless of actual usage.

Suddenly, the Anthropic IPO looks less like:

and more like:

That's where the investment map gets interesting.

🛣️ The AI money machine

Think of the cycle like this:

Investors fund AI labs

↓

AI labs buy computing capacity

↓

Cloud companies, chip companies and infrastructure suppliers get revenue

↓

Those companies invest more in AI infrastructure

↓

AI labs grow and attract higher valuations

↓

More capital becomes available

↓

Repeat.

This doesn't automatically mean “AI bubble.”

It may simply be what building a new technological platform looks like at enormous scale.

But it creates feedback-loop risk.

If AI demand keeps accelerating, the machine compounds.

If customers slow spending while infrastructure commitments remain enormous, the same machine can work in reverse.

And that distinction matters enormously to investors.

🟢 5 stocks sitting around the money flow

🥇 1. Alphabet — GOOGL

Google may be one of the most interesting beneficiaries because it potentially gets paid in multiple ways.

Anthropic reportedly has more than $111 billion of commitments involving Google, while Google also has a strategic relationship and investment exposure.

That gives Alphabet exposure to:

• Google Cloud revenue
• Tensor Processing Units (TPUs), Google's AI chips
• Anthropic investment upside
• Enterprise AI distribution

But here's the catch:

Revenue isn't profit.

Google still has to build data centers, networking and computing capacity.

So watch Google Cloud growth, margins, capital expenditure and whether Anthropic's commitments translate into profitable utilization.

Verdict: 🟢 High conviction

🥈 2. Amazon — AMZN

Amazon is another fascinating “sell the infrastructure” play.

Anthropic reportedly has around $110 billion of infrastructure commitments involving Amazon, while Amazon also owns a major strategic stake and supplies computing through Amazon Web Services (AWS).

Amazon also has its own Trainium AI chips.

That creates an attractive combination:

Cloud + custom silicon + Anthropic relationship.

Amazon doesn't necessarily need Claude to become the world's only AI winner.

It needs companies to keep needing compute.

Verdict: 🟢 High conviction

🥉 3. Broadcom — AVGO

This is the less obvious one.

The AI infrastructure war isn't simply:

Nvidia vs everyone.

Custom chips are becoming increasingly important.

Broadcom sits behind a growing ecosystem of custom AI accelerators, networking and infrastructure.

Why do I like this angle?

Because if hyperscalers decide they want more specialized computing rather than relying exclusively on one merchant GPU supplier, Broadcom can potentially benefit from the diversification itself.

That's a different thesis from simply betting on which AI model wins.

Verdict: 🟢 High conviction infrastructure thesis

4. Microsoft — MSFT

Microsoft reportedly has at least $31.4 billion of Anthropic-related infrastructure commitments.

But Microsoft's bigger advantage may be distribution.

Enterprises don't just ask:

They ask:

Microsoft already owns enormous enterprise distribution through Azure and its software ecosystem.

That's valuable real estate.

Verdict: 🟢/🟡 Strong business; Anthropic is an additional catalyst, not the whole thesis.

5. Salesforce — CRM

This one is spicy. 🌶️

Salesforce reportedly invested about $50 million in Anthropic, and its stake was subsequently valued at roughly $5 billion after later funding rounds.

But remember:

$5 billion valuation ≠ $5 billion cash profit.

It is an unrealized gain until Salesforce actually monetizes the investment.

The interesting part is that Salesforce is also integrating Claude into its enterprise AI strategy.

So CRM potentially gets two bites:

investment upside + operating benefits.

But investors should separate those two.

Otherwise, you may accidentally mistake a rising private-company valuation for improving Salesforce operating performance.

Verdict: 🟢/🟡 Interesting optionality

⚠️ What about Nvidia?

This is where I would be careful.

Nvidia is obviously a major AI infrastructure winner.

But don't make the lazy equation:

Anthropic grows = Nvidia wins automatically.

Anthropic is building a multi-platform compute strategy involving Nvidia GPUs, Google's TPUs and Amazon's Trainium.

That is strategically smart for Anthropic.

It also means investors shouldn't assume Nvidia owns the entire Anthropic wallet.

And as AI models become more efficient, the industry may eventually need more intelligence per chip rather than simply more chips.

Nvidia remains a phenomenal AI company.

But the Anthropic thesis alone isn't enough to justify buying it at any price.

🚨 The trap I would avoid

Buying Anthropic on launch day simply because everyone else is buying it.

A reported $2 trillion target is an aspiration — not automatically the IPO valuation investors will receive.

The final economics depend on:

• IPO price
• share count
• dilution
• voting structure
• lock-up provisions
• customer concentration
• cash burn
• future commitments
• revenue quality
• free cash flow

Anthropic itself confidentially submitted an S-1 — the registration document for a U.S. IPO — but explicitly said the number of shares and offering price had not yet been set.

So don't buy the headline.

Buy the economics — if the price makes sense.

📉 The canary I would watch

Forget the daily AI stock noise.

Watch hyperscaler capital expenditure (CapEx) — spending on data centers, chips, networking and other long-lived infrastructure.

If Amazon, Google and Microsoft keep increasing AI infrastructure spending while customer demand and cloud utilization remain healthy, the AI investment cycle is probably still functioning.

But if:

CapEx slows + AI customer growth slows + chip orders get cancelled + software renewals weaken

then we have a much more interesting warning.

The first-order story says:

“AI spending is enormous.”

The second-order question is:

“Who still wants all that capacity six months later?”

That is the signal I'd rather trade.

🧠 My 7-point Anthropic investor checklist

Before buying any AI stock, ask:

☐ Who gets paid? Follow the money.

☐ Is the revenue recurring? Big revenue isn't necessarily durable revenue.

☐ Who are the customers? Two customers producing ~25% of sales deserves attention.

☐ Are contracts binding? A commitment is more valuable than a handshake — but still examine the economics.

☐ Who bears the CapEx? Revenue can look fantastic while capital intensity quietly eats returns.

☐ What happens if AI becomes cheaper? Great for users; potentially complicated for suppliers.

☐ What would prove my thesis wrong? Decide before buying, not after the stock falls 30%.

🎯 The three scenarios

🟢 Bull: AI demand keeps exploding → cloud utilization rises → infrastructure suppliers benefit.

🟡 Base: AI revenue grows rapidly but spending remains extreme → selective stock-picking becomes more important.

🔴 Bear: customer spending slows → infrastructure commitments become burdensome → AI valuations compress.

Notice something?

AI doesn't have to “die” for AI stocks to fall.

Sometimes the technology wins while the stock loses.

You can be completely right about the future and still overpay for it.

That is one of investing's cruelest jokes. 😂

💡 Why Wealth Builder matters

The biggest problem for retail investors isn't a lack of information. It's too much information arriving too quickly — IPO leaks, AI headlines, valuation targets, earnings calls and social-media rocket emojis all competing for attention.

Wealth Builder, passive-income ideas and structured investing newsletters can help turn that noise into a repeatable process: identify who benefits, challenge the obvious thesis, check valuation, examine cash flow, test downside scenarios and create rules before emotions take over. You don't need to predict every AI winner. You need a framework that helps you recognize opportunities and avoid expensive traps.

If you want more practical investing research from like-minded newsletters, explore the collection here. Your portfolio doesn't need more noise. It needs better filters.

Follow The Invoices. 💰

📚 Notes, Sources & Fine Print

Primary source

1. Reuters — “Anthropic’s IPO prospectus shows sweeping AI vision, surging costs,” September 28, 2026. This is the principal source for the reported 2025 revenue of approximately $4.6 billion, nearly $42 billion net loss, more than $8 billion operating loss, approximately $7.33 billion spent on computing and infrastructure, the potential $2 trillion valuation, and the reported $518 billion of future cloud, computing and infrastructure obligations. Reuters​
2. Reuters — “Anthropic’s $518 billion AI buildout hinges largely on deals that cannot be canceled,” September 29, 2026. This provides the detailed breakdown of the infrastructure commitments, including approximately $111.1 billion with Google, $110 billion with Amazon and $31.4 billion with Microsoft, plus approximately $161.2 billion of Broadcom-related equipment lease obligations. Reuters reported that roughly 80% of the $518 billion total is non-cancelable or requires payment regardless of usage. Investing.com
3. Reuters — “Anthropic IPO prospectus lays bare deep dependence on Big Tech partners,” September 29, 2026. Important for the less-obvious risk: approximately 47% of Anthropic's 2025 sales flowed through Amazon and Google's cloud marketplaces, while two unnamed customers each represented approximately 12% of revenue. Reuters also reported that many major customers were not bound by long-term contracts. Investing.com

Company sources

1. Anthropic — “Anthropic confidentially submits draft S-1 to the SEC,” June 1, 2026. Anthropic confirmed that it had confidentially submitted a draft S-1 (Registration Statement) to the U.S. Securities and Exchange Commission. It also stated that the number of shares and offering price had not yet been determined and that any IPO would depend on market conditions and other factors.
2. Anthropic — “Anthropic and Amazon expand collaboration for up to 5 gigawatts of new compute,” April 20, 2026. Anthropic said it was committing more than $100 billion over ten years to AWS technologies, with up to 5 gigawatts (GW) of new capacity, while Amazon announced an additional $5 billion investment with the possibility of another $20 billion.

Additional reporting / context

1. Reuters — Big Tech dependency and revenue concentration. The Reuters reporting is particularly useful because it reveals that Amazon and Google are simultaneously investors, cloud providers, distributors, customers and competitors to Anthropic. That creates both an enormous distribution advantage and potential conflicts of interest. Investing.com
2. Salesforce / Anthropic investment: The newsletter's Salesforce discussion is based on reporting that Salesforce invested approximately $50 million in Anthropic and subsequently held a stake valued at roughly $5 billion following later financing rounds. Importantly, this is an unrealized valuation, not $5 billion of realized cash profit.

🧠 Important interpretation notes

• IPO = Initial Public Offering, the first sale of a company's shares to public investors.
• S-1 = the registration statement a company files with the U.S. Securities and Exchange Commission (SEC) when preparing for a U.S. IPO.
• SEC = U.S. Securities and Exchange Commission.
• AI = Artificial Intelligence.
• AWS = Amazon Web Services, Amazon's cloud-computing business.
• TPU = Tensor Processing Unit, Google's specialized AI processor.
• GPU = Graphics Processing Unit, a processor widely used for AI computation.
• CapEx = Capital Expenditure, money spent on long-lived assets such as data centers, servers and networking equipment.
• Operating loss = the loss generated by the company's underlying operations before certain non-operating items.
• Free Cash Flow (FCF) = cash generated after operating needs and capital expenditures.
• Unrealized gain = an increase in the estimated value of an investment that has not yet been sold for cash.
• Non-cash accounting charge = an accounting adjustment that affects reported profit or loss without representing an equivalent current-period cash payment.

⚠️ One crucial disclaimer for the article

The figures discussed above are based substantially on Reuters' reporting of Anthropic's confidential IPO materials and company disclosures. The final publicly available registration statement, IPO valuation, share count, offering price, terms and financial figures may differ. Anthropic itself said its confidential S-1 submission did not establish the number of shares to be offered or the offering price. Anthropic

This newsletter is for educational and informational purposes only and does not constitute financial, investment or trading advice. Individual stocks discussed are examples for research, not recommendations to buy or sell securities. Investors should conduct their own due diligence and consider their own risk tolerance, valuation assumptions, portfolio concentration and investment horizon.