Anthropic and one of the greatest venture returns in history

Hello friends, and welcome to The Unsophisticated Investor! Brought to you by Scott & Rob, the founders of Shuttle!

If you want to invest alongside the VC funds who've backed breakout companies like Revolut, Asana, JustEat, Bolt, Lets Get Checked, Loom, Runna, Charlotte Tilbury, Deel, Aircall, AngelList, Carta, TransferWise and many more, regardless of your knowledge, network or net worth, then sign-up using the link below.

Now, let’s get into it 👇

In the summer of 1602, a housemaid in Amsterdam named Neeltgen Cornelis walked into a counting house and handed over 100 Dutch guilders. In return she received something no servant anywhere had ever been able to own: a tradable share in a company.

The company was the Dutch East India Company (VOC), and what it did that summer was new in the world’s history. Article 10 of its charter declared that "all the residents of these lands may buy shares in this Company," with no minimum and no maximum. Around 1,143 people did, from grandees to bakers to Neeltgen, raising roughly 6.4 million guilders. The shares could be traded, and an exchange opened up to trade them. An ordinary person could buy a piece of the venture at the very beginning, before a single ship had returned, and share in whatever came after.

The VOC paid its shareholders nothing for the better part of a decade, which provoked enough discontent that in 1608 a disgruntled former director named Isaac Le Maire mounted what is now remembered as the world's first short-selling raid against it. But those who held on owned a slice of what became the most valuable corporation of its age. The apparatus Neeltgen used, the public share, the secondary market, limited liability, is the direct ancestor of the one a modern saver uses to buy a fund in a pension today.

On the 1st of June 2026, Anthropic filed to use the very same machinery Neeltgen would recognise: the public share, the exchange, the prospectus. But over the four centuries since, the thing that machinery does has quietly inverted. What was built to let outsiders in at the beginning, before the ships had sailed, now serves mostly to let them in at the end, once the voyage is over and the cargo sold. The modern public does not buy into the venture. It buys out the people who did.

What the filing unlocks

Anthropic, the maker of the Claude AI assistant, confidentially submitted a draft registration statement to the SEC on the 1st of June, days after closing a $65 billion round that valued it at $965 billion. The filing is the formal start of a path to a listing reported to be targeted for the autumn, though a confidential filing carries real caveats: no price has been set, the financials remain private until shortly before any roadshow, and such filings can be amended or withdrawn entirely.

If it does list near that valuation, the event will convert into cash one of the largest privately built fortunes ever assembled. The single largest venture return in recorded history is about to be claimed the same season, by Founders Fund, whose roughly $20 million investment into SpaceX in 2008 is reported to be worth more than $60 billion at that company's listing, assuming SpaceX lists anywhere near the $1.75 trillion that has been floated for it. Anthropic's backers are not in line for the single largest return. They are in line for one of the largest, and they got there in a quarter of the time.

The compounding, and who captured it

The shape of Anthropic's valuation is the story here. Its first priced round, a Series A in May 2021, valued the company at around $623 million, according to Forge Data, with the earliest cheques coming from individuals: the Skype founder Jaan Tallinn, the former Google chief Eric Schmidt, the Asana and Facebook co-founder Dustin Moskovitz. A year later a Series B led by the now-convicted, and imprisoned, Sam Bankman-Fried put the figure at $4 billion. By May 2023 it was $4.1 billion, the round in which Spark Capital led and Google took its first stake. Then the line bends almost vertical: $18.4 billion in early 2024, $61.5 billion in March 2025, $183 billion that autumn, $380 billion in February 2026, and $965 billion in May. From the 2022 round to today is roughly a 240-fold increase in four years.

Because the S-1 is confidential, the exact ownership split is not yet public. What can be assembled from round announcements and reporting is pretty stark though. The earliest backers at $623 million are on a paper return of something like 1,500-fold. Spark Capital, leading at roughly $4 billion, is on the order of 240 times its entry. Salesforce Ventures, which invested in 2023, has seen its stake reported at around $5 billion, having appreciated by more, the firm has noted, than the combined market-value gains of many of its acquisitions; that single passive position is now worth more than half of Salesforce's entire operating profit for the year... Cathie Wood's ARK, which entered in mid-2023, is reported to have multiplied its position roughly 95-fold. Amazon, with up to $8 billion committed, and Google, with several billion, hold strategic positions whose paper gains run to many tens of billions.

The mechanism underneath those numbers is simple and worth stating plainly, because it explains everything that follows. Every financing round does two things at once. It typically raises the price and it removes risk. The 2021 investors bought into a research lab with no revenue and an unproven thesis, and were compensated with a price that assumed it might fail. The 2023 investors bought a company with a working product and real customers, and paid $4 billion for the privilege. By the time the public buys at the listing, the existential risk has been engineered out, and the price reflects its absence. The early cheque is paid to bear the chance of zero. The late cheque pays a premium for its removal.

The public investor inherits a different arithmetic

This is where the analysis has to become unsentimental, because the gap between the early backer and the public buyer is not a matter of unfairness or sentiment. It is a matter of numbers, and the numbers are unforgiving.

Consider what it would take for someone buying a share at the listing to earn what Spark Capital is on course to earn. Spark's entry implies roughly a 240-fold return at a $965 billion listing. For a public buyer to make the same multiple, Anthropic would have to reach a valuation of around $230 trillion. That is close to twice the annual output of the entire world economy, and more than the combined value of every public company currently listed anywhere on earth. The figure is not ambitious. It is arithmetically impossible. Even a tenth of that multiple, a 24-fold return, would require Anthropic to be worth more than $20 trillion, comfortably larger than any company that has ever existed, the Dutch East India Company included, even on the most generous inflation-adjusted estimates sometimes made for it.

There is a second reason the public buyer stands in a different place, and it has nothing to do with arithmetic. By the time the shares trade freely, the people who built the compounding are on their way out. An IPO is, mechanically, a liquidity event for existing holders. Lock-up agreements typically stop insiders and early investors from selling for 90 to 180 days, and when those expire a wave of supply arrives from the investors whose conviction carried the company through its risky years. The marginal holder of the stock shifts from a venture partner who bought at $4 billion and knows precisely what they own, to a public buyer who arrived at the listing and is pricing the future from a prospectus. 

At $965 billion against a reported revenue run-rate of around $47 billion, Anthropic would list at roughly twenty times its revenue, a multiple that only holds up if growth stays aggressive for years. Revenue obeys the same law of large numbers as valuation: a company tripling from 10 to 47 billion is a different animal from one tripling from 47 to 140, and harder again beyond that. When the public disclosures arrive, they will be read for precisely this. Reporting suggests Anthropic's strength is heavily enterprise rather than consumer. Its losses and compute obligations are large, and it is carrying a public dispute with the Pentagon that it has said could threaten billions in revenue. None of this makes Anthropic a weak company. It makes the listing price a demanding one, which is the fact that matters for whoever buys at it.

What the listing actually is

Put the pieces together and a structural shift comes into focus, one that runs well beyond Anthropic. The function of an IPO has changed. A generation ago a company went public because it needed capital to grow, and the public markets were where that capital lived. Today the largest private companies raise tens of billions from crossover funds and sovereign wealth without filing a prospectus, and they use tender offers and secondary sales to hand early shareholders and staff their liquidity along the way. That leaves the listing doing a narrower job: less fundraising for the next phase of growth, more an exit for the holders of the last one. The median US company now lists at twelve to fourteen years old rather than the six to eight of the 1980s and 1990s, and at a median value many times higher in real terms. By the time the public is admitted, most of the steep part of the curve has already been climbed in private.

The politics are sharpening around this point also. On the same day Anthropic filed, Senator Bernie Sanders used a New York Times op-ed to propose what he calls the American AI Sovereign Wealth Fund Act, a one-time levy that would transfer half the equity of the largest AI companies, Anthropic and OpenAI among them, paid in stock rather than cash, into a public fund granting ordinary Americans a direct ownership stake, voting rights and seats on the board. The bill faces long odds, and there are numerous valid reasons why this might not work in practice, from valuation to precedent. But the idea is less fringe than it first sounds: OpenAI has floated a public wealth fund of its own, and Anthropic's chief executive has spoken approvingly of sovereign funds holding AI equity. Strip away the politics and the proposal is simply the bluntest possible answer to the asymmetry described here. If the public cannot buy in early, the argument runs, then hand it a stake by statute.

There is a quieter, market-based answer to the same problem, and it is the one we work on at Shuttle. The case against letting ordinary investors into private markets was never really that they cannot handle the risk. It is that the infrastructure connecting them to private opportunity has been missing, clumsy or expensive, so that access, where it exists at all, arrives late and on poor terms. Shuttle is built to supply that infrastructure properly: a regulated platform, through which a vetted community of investors can co-invest in private companies, with the diligence, structure and disclosure that separate genuine access from being sold the riskiest assets at the top of the cycle.

The bottom line

The Anthropic listing, if it happens, will be a genuine landmark, one of the largest realisations of privately created wealth on record. It is worth seeing it clearly for what it is. It is not the public being invited to share in the next phase of growth, in the way Neeltgen Cornelis was invited to share in the voyages of 1602. It is a settling-up of value already created, in rounds the public could not join, on terms it could not see. The early backers were paid for taking a risk that no longer exists. The public buyer is asked to pay for its absence, at a size that caps the upside by simple arithmetic, while the investors who made the return become the sellers on the other side of the trade.

None of this is an argument against public markets, or against owning a piece of an extraordinary company. It is an argument for being honest about where in the company's life you are being allowed to buy, and what that position is actually worth. That distinction, between the start of the voyage and the end of it, is the one the next decade of reform will be judged on.

What we’ve working on at Shuttle

• Building a job-board of VC-backed companies for our readers 👷

• Planning a new content series for founders and investors 🎥

• Heads-down in product, with more to share shortly 🧑‍💻

The Unsophisticated Investor is brought to you by Scott & Rob, the founders of Shuttle. We’re both sick of private markets being a playground exclusive to the ultra-wealthy so we started a company to challenge the status-quo. Shuttle’s singular focus is to unlock private markets for Millennial and Gen Z tech professionals and help them build wealth through the highest performing private market opportunities.

Scott & Rob
Shuttle Co-Founders