A company can publish more than 300 pages about its business, results, and risks, and investors can still be left with basic questions unanswered.

That’s what happened after Anthropic published its registration statement for an initial public offering.

The document reveals exceptional revenue growth, along with extensive details about competition, risks, and computing. But it also leaves gaps in some important figures, particularly third-quarter results and certain future infrastructure commitments.

This highlights an important rule for analyzing IPOs:

Missing information can sometimes matter just as much as the information that’s available.

What is an S-1 filing?

When a company wants to go public in the United States, it files a Form S-1 with the Securities and Exchange Commission.

The document gives investors a picture of the company: its business, revenue, losses, shareholders, risks, and plans for using the proceeds from the offering.

But it doesn’t guarantee that every question will be answered right away.

Anthropic’s preliminary prospectus, for example, reported revenue of $16.3 billion in the first half of 2026, up from $1.16 billion in the same period a year earlier, while recording an operating loss of about $693 million.

Huge, clear numbers.

But investors need more than that.

First question: What happened after the latest reported figures?

Fast-growing companies can change more in three months than a traditional company does in a year.

That makes the latest financial period especially important.

In the preliminary version of Anthropic’s prospectus, some figures for the quarter ended in September appeared as blanks, even though the company indicated that Claude continued to grow.

This illustrates the concept of information freshness.

First-half results may be excellent, but if growth started to slow in the third quarter, the company’s valuation should reflect that.

Markets don’t buy the past; they try to price the future.

Second question: Who is paying the company?

High revenue isn’t enough; investors need to know where it comes from.

According to Anthropic’s disclosures, about 48% of first-half 2026 revenue came through Amazon, Google, Microsoft, and their cloud platforms, up from 11% in 2023.

These partnerships are valuable because they give Anthropic fast access to business customers.

But they raise another question:

What happens if distribution terms change, or if these companies—which are simultaneously partners, competitors, and computing providers—become less reliant on Claude?

This is where the concept of customer and channel concentration comes in.

The more a company depends on a limited number of parties, the more of its revenue is exposed to a single relationship or a handful of key relationships.

Third question: How much has the company committed to spend?

This is especially important for AI companies.

Anthropic says computing is its largest operating expense, and it has signed massive contracts to secure chips, data centers, and power for years to come.

But the preliminary version of the prospectus left some totals for future commitments blank, even though the details and footnotes reveal contracts worth hundreds of billions of dollars.

There’s a big difference between:

A company that can cut costs when demand falls

and a company committed to paying huge sums whether or not it uses all of its computing capacity.

That’s why investors should look not only at revenue growth, but also at how flexible the cost structure is.

Fourth question: What price will we pay for all this?

Even the best company in the world can be a bad investment if its price is too high.

This is the part that can’t be fully assessed until the offering price and final share count are known.

Market expectations pointed to a potential valuation of around $1.5 trillion to $2 trillion, but the final offering figures were not yet available in the preliminary version.

This brings us back to a basic investment principle:

A good company and a good stock aren’t the same thing.

A company may be growing quickly, but investors need to know how much they’ll pay for each dollar of future revenue or earnings.

Reading between the blanks

A good prospectus doesn’t eliminate uncertainty.

Its purpose is to give investors enough information to understand where the risks lie.

That’s why reading an IPO filing shouldn’t be limited to searching for revenue figures or growth rates.

The more important question may be:

What do I still not know after reading the prospectus?

In an IPO, the published figures help us understand the company.

But sometimes the figures that haven’t been published yet determine whether the price we’ll pay makes sense.