Commentary

    ANT-Hangover: why didn't I invest in Anthropic earlier?

    Anthropic's float has made its founders, staff and early backers extraordinarily rich. Was there ever a way in for an ordinary UK investor before the listing, and what would we need to do differently to catch the next one?

    Portrait of Raj PatelRaj PatelHead of Research14 min20 November 2026

    Looking at the Anthropic share price this morning is not a comfortable experience.

    The company priced its IPO at $83 a share on Wednesday 18 November, above the $70 to $78 range it had marketed. When trading began on Nasdaq yesterday, under the ticker ANT, the shares opened at $111.50, which briefly valued the business at more than $2 trillion. They closed at $97.17, still 17% above the IPO price. The base offering totalled $81.3 billion, the largest on record, with about $79.4 billion going to Anthropic and the balance to selling shareholders.

    For founders, early employees and venture capital funds, the sums are hard to take in. For the rest of us there is one nagging question: why didn't I invest in Anthropic?

    Perhaps I should have bought in 2021. Or in 2023, when Claude first appeared. Or in 2025, when institutions were already putting in billions. In hindsight all three look obvious.

    But hindsight is a wonderful investment adviser. It forgets the companies that failed and makes yesterday's winners look as if they were always going to win.

    So before deciding what to buy next, I went back over how Anthropic was financed, who was allowed to invest at each stage, and whether ordinary investors missed an opportunity that was ever open to them.

    From a $124m start-up to a $1.5tn IPO

    Dario and Daniela Amodei, who are siblings, founded Anthropic in 2021 with a group of former OpenAI colleagues. Their aim was to build advanced AI with safety, interpretability and reliability as the priority.

    It is easy to forget how early that was. ChatGPT was still eighteen months away and Claude nearly two years, and nobody knew whether large language models would ever make serious money.

    Investors backed the company anyway, in rounds that kept getting bigger.

    DateRoundRaisedValuation
    May 2021Series A$124mNot disclosed
    April 2022Series B$580mNot disclosed
    May 2023Series C$450mNot disclosed; reported at $4.1bn to $5bn
    Dec 2023 to early 2024Series DAbout $750m (reported)About $18.4bn (reported)
    March 2025Series E$3.5bn$61.5bn post-money
    September 2025Series F$13bn$183bn post-money
    February 2026Series G$30bn$380bn post-money
    May 2026Series H$65bn$965bn post-money
    November 2026IPO$81.3bnAbout $1.5tn

    Series A to C and E to H are the figures Anthropic announced. Series D was never formally announced, and press reports disagreed on whether $18.4bn was its value before or after the new money. The Series H total includes $15bn that the big cloud providers had already committed, $5bn of it from Amazon. The IPO valuation is on a fully diluted basis. The table leaves out the separate strategic investments by Amazon and Google described below.

    The Series A in May 2021 was led by Jaan Tallinn, the Skype co-founder, with Eric Schmidt and Dustin Moskovitz among the other backers. These were established technology investors with existing ties to the field, and the round was never offered to the public.

    A year later Sam Bankman-Fried, then chief executive of FTX, led the $580 million Series B, and Caroline Ellison was among the other investors. FTX collapsed that November, so even Anthropic's early backers do not all look like obvious partners in hindsight. Anthropic itself was still a young research company, and nothing guaranteed it would succeed commercially.

    Claude launched in March 2023, and a different kind of money followed. That September Amazon agreed to invest up to $4 billion, which it completed in two payments, the second in March 2024. It committed a further $4 billion in November 2024, and Google put in billions of its own. These deals came with cloud contracts, chips and access to the computing power that frontier models need, so they were partnerships as much as investments.

    They were also very good investments. Amazon had put in about $18 billion by the time of the IPO and came out of it owning 14.7% of Anthropic, with Google on 13.7%. Much of the AI opportunity was captured by a small group of venture funds and the largest technology companies in the world.

    The valuation took off in 2025. The Series E in March raised $3.5 billion at $61.5 billion. Six months later the Series F raised $13 billion at $183 billion, almost exactly three times the price.

    Two more private rounds followed in 2026. February's Series G raised $30 billion at $380 billion, and May's Series H raised $65 billion at $965 billion, which made Anthropic one of the most valuable private companies in the world. Less than six months later the IPO valued it at about $1.5 trillion.

    Going from $124 million in 2021 to a $1.5 trillion listing took under six years. Every round along the way was private, open to institutions, venture firms and strategic partners and to almost nobody else.

    Could an ordinary UK investor have got in before the IPO?

    The usual assumption is that retail investors had no way in at all. That is not quite right. There were four possible routes, and each came with serious limits.

    Investing in the funding rounds

    The direct way to share in Anthropic's growth would have been to invest alongside the venture firms in the Series A, B or C. In practice that was close to impossible.

    These were private placements. The company and its existing shareholders chose who took part, and nothing was offered through a stock exchange or a retail platform. Money alone would not have been enough either: you also needed the relationships, the legal eligibility and a company willing to have you on its register. For nearly all UK retail investors this route did not exist.

    Buying existing shares on the secondary market

    In a secondary sale an existing shareholder sells to a new investor, and the company raises nothing. It is how employees at successful private companies turn some of their equity into cash, and how early backers trim a position.

    Anthropic's prospectus describes two such events. In June 2025 the company bought back 27.9 million shares from current and former staff at $5.61 each, about $157 million in total. Anthropic was the buyer and it retired the shares, so no outside investor could take part.

    The second was a tender offer that closed in April 2026. A group of outside buyers approved by the board paid $25.9136 a share for 179.9 million shares, about $4.66 billion. Both prices are adjusted for the ten-for-one share split in September 2026, so they compare directly with the $83 IPO price. Whoever bought in April paid less than a third of what the shares fetched seven months later.

    So Anthropic shares did change hands before the IPO, but only once with outside buyers, and only with buyers the board had approved.

    Anthropic also policed this market. In February 2026 it published a warning about unauthorised share sales: any transfer its board has not approved is void, and special purpose vehicles (SPVs) are not permitted to acquire its shares at all. In May it went further and named platforms it said were not authorised to sell its stock.

    Anyone browsing pre-IPO platforms should take note. A platform that advertises access to a private company has not shown that the company will recognise the deal, and exposure to a company is a different thing from shares it accepts on its register.

    Buying a listed investment trust

    This is the route that was open to ordinary investors.

    Some investment trusts are allowed to hold private companies alongside listed ones. Buy the trust's shares on the London Stock Exchange and you own a slice of its private holdings, with no need to qualify for a funding round yourself.

    Seven UK-listed trusts held Anthropic before the IPO, according to the Association of Investment Companies (AIC):

    Investment trustAnthropic as % of fundGrowth since first investment
    Baillie Gifford US Growth6.8%About 2.5x
    Schiehallion Fund6.1%About 3x
    Scottish Mortgage2.8%About 2x
    RIT Capital Partners1.7%Not disclosed
    Monks1.0%About 1x
    HarbourVest Global Private Equity0.7%Not disclosed
    Pantheon International0.2%Not disclosed

    Source: AIC, 29 October 2026. Weightings are the latest each trust had reported, at dates between June and September 2026, so all of them predate the IPO.

    Scottish Mortgage, Baillie Gifford US Growth and Schiehallion all made their first investment in August 2025. Scottish Mortgage put in $229 million between then and May 2026, and by 31 July 2026 the holding was valued at $460 million. Its money had doubled before Anthropic listed. At the $83 IPO price, QuotedData puts the stake at about £664 million, 3.9 times the £170 million invested.

    Anyone with an ordinary investment account could have bought Scottish Mortgage shares in August 2025. The same was true of most of the other trusts. Schiehallion was the exception: until it moved to the Main Market in December 2025 it traded on the London Stock Exchange's Specialist Fund Segment and was aimed at professional investors.

    Owning a trust is not the same as owning Anthropic. You are buying a diversified portfolio in which Anthropic is one holding, 6.8% at most in the table above, and the trust's own shares can trade above or below the value of its assets. It was still a real and easy way to own part of a private company before it listed.

    For me this is the main lesson of the Anthropic story. The opportunity was open to retail investors, but only to those prepared to look beyond buying individual listed shares.

    Applying in the IPO itself

    Finally there was the IPO, which UK retail investors could apply for directly. RetailBook ran the UK offer and seven platforms took orders, among them AJ Bell, Hargreaves Lansdown and interactive investor.

    The legal basis was the UK's public offer platform regime, in force since 19 January 2026, which lets a company offer shares to the UK public without a London listing or a UK prospectus. SpaceX had used the same route earlier in the year.

    About 110,000 UK investors bought $300 million of stock between them, according to the Financial Times. We cover the allocation, the platforms and the scaling back in a separate piece: Anthropic's $81bn IPO: retail investors get access, but how much of the deal did they really get?

    But applying in an IPO is not getting in early. By the time ordinary investors could buy, Anthropic was valued at about $1.5 trillion, and most of the returns its early investors enjoyed had already been made.

    So did we actually miss an opportunity?

    My answer is partly, though not in the way it feels this morning.

    We did not miss the chance to buy Anthropic at its Series A valuation. For most retail investors that chance was never on offer. What we may have missed is exposure to a new generation of private technology companies through listed investment trusts. That is a different kind of miss, and one we can do something about.

    It also says something about how capital markets have changed. Some of the most valuable companies of the next decade may stay private for much longer than earlier technology businesses did. By the time they list, venture funds, private equity firms, strategic investors and employees may already have taken a large part of the growth. Waiting for the IPO can mean arriving after much of the opportunity has gone.

    We should be careful, though, about treating Anthropic as a blueprint. It is an extraordinary case: a company at the centre of a potentially transformative technology, enormous infrastructure spending and a commercial market that is growing fast.

    NVIDIA is another. Look at its share price history today and it seems absurd that anyone failed to buy it a decade ago, yet at the time it was far from obvious that graphics chips would become basic hardware for generative AI.

    It is tempting to study the winners and forget every company, technology and investment thesis that failed. Convincing ourselves that we could have predicted Anthropic gets us nowhere. Building a better process for finding and reaching opportunities earlier might.

    What a strategy for the next Anthropic could look like

    A better question than "should I have bought Anthropic in 2021?" is how to be positioned for the next generation of private companies. There are a few ways to go about it.

    Earlier-stage venture capital

    One option is exposure to Series A, B and C companies through venture capital funds or co-investment structures. The logic is attractive. A company at Series A or B is valued at a fraction of what it might become, so investors at that stage can capture far more of the growth than those who arrive in the late rounds.

    The risks are of a different kind from those of listed shares. Early-stage companies can fail outright. The ones that succeed usually need several more funding rounds, each of which dilutes existing shareholders. Your money can be tied up for years, and you have little say over what management does.

    There is also a big difference between a professionally managed venture portfolio and a single SPV that claims to offer access to a famous start-up. I would not build a strategy around individual Series A or B deals without a credible manager, real access to the underlying company, transparent fees and a clear understanding of the legal structure.

    Listed investment trusts

    For many UK investors this is the practical route. You buy shares in a listed trust whose managers have the mandate and the experience to invest in private businesses, and you leave the selection to them.

    That gets you access to private companies before they list, a spread of businesses and industries, professional selection and due diligence, and an investment you can buy and sell through an ordinary brokerage account.

    Scottish Mortgage, Schiehallion, Baillie Gifford US Growth and RIT Capital Partners are the obvious examples, along with specialist private equity trusts such as HarbourVest and Pantheon.

    Check what a trust actually holds before buying. Not every trust has much in private companies, and allocations can change a good deal over time. Find out how the private holdings are valued as well, because no exchange prices them from day to day, and remember that the trust's own shares can trade at a wide premium or discount to its assets.

    With those caveats, this is where I would start.

    Private secondary markets

    The UK now has a regulated framework for trading shares in private companies: PISCES, the Private Intermittent Securities and Capital Exchange System. The London Stock Exchange operates one venue under it, the Private Securities Market, which completed its first auction in March 2026.

    A company using PISCES stays private. It opens trading windows in which existing shareholders can sell to eligible investors. That could make ownership of private companies more transparent and offer a way in before any IPO.

    The limits are considerable. The company decides whether to take part, when trading happens and which investors can see its auctions. PISCES is a secondary market only, so companies cannot raise new money on it. And for now it is aimed at professional and institutional investors, plus individuals who meet the high-net-worth or sophisticated investor tests.

    It is not a retail market yet, but it is worth following closely.

    Industries before individual names

    The approach I find most interesting is more systematic. Start from the industries likely to produce the next set of transformational businesses, and let the individual names follow.

    AI is the obvious one. The same thinking applies to robotics and automation, biotechnology and drug discovery, defence technology, energy storage and next-generation energy infrastructure, space technology, and semiconductor manufacturing and advanced materials.

    The work is to identify promising companies in those sectors while they are still private, understand how they are financed, and establish which vehicles give legitimate access to them. That means following funding rounds, institutional investors, commercial progress, valuations and secondary sales.

    It also means telling commercial progress apart from hype. A $500 million funding round does not make a company a good investment. Entry valuation, dilution, competitive position, capital needs and the eventual exit matter as much as the technology.

    What I would take away from the Anthropic IPO

    I would not conclude that the answer is to start investing in every promising Series A, B or C. That swaps one problem for another: trying to predict which highly uncertain private companies will eventually succeed.

    Private markets deserve more attention from retail investors. The old split, private companies for institutions and public companies for individuals, now has real consequences for returns. We should understand the vehicles that give access to private businesses before they list.

    Access alone is not an investment strategy, though. Being able to buy into a private company does not make the valuation attractive, and a good company can still be a bad investment at the wrong price. The structure, fees, shareholder rights, dilution and liquidity need as much scrutiny as the business.

    And it pays to think in portfolios. Anthropic may be an extraordinary winner, but a sensible venture strategy cannot depend on finding that one company. It needs to be spread across companies, industries, investment stages and funding vintages. For some people that will mean combining ordinary public-market investments with listed investment trusts and carefully chosen private exposure. The balance depends on your circumstances, your time horizon and how much capital you can afford to lose.

    The next Anthropic

    The morning after a spectacular IPO is probably not the best time to make an investment decision. It is a good time to think about how investing is changing.

    Anthropic shows that some of the largest gains can be made long before a company reaches the public markets. It also shows that ordinary investors are not shut out of all of them, even if the early funding rounds remain largely closed.

    Five years from now we may well be asking the same question about another company. Next time I would like to have done a little more homework.

    Sources

    Disclaimer: This article is for informational and educational purposes only and does not constitute investment advice or a personal recommendation. Private-market investments carry significant risks, including the potential loss of the entire investment, dilution and limited liquidity. Investment trusts can trade at a discount or premium to their net asset value. Past performance is not a reliable indicator of future returns.

    Weekly newsletter

    The UK alternative investment market, weekly.

    New listings, FCA status changes, and market moves — every week.

    By subscribing, you agree to receive the weekly Other. newsletter. Unsubscribe anytime. See our Privacy Policy.