How to Ride the AI IPO Wave Without Paying the Day-One Price

On Monday, Anthropic filed publicly to list its shares on Nasdaq under the ticker ANT, with the offering expected before Thanksgiving.¹ The company expects the deal to match or beat the size of SpaceX’s record June listing, and it could begin formal marketing as soon as the week of November 9.² OpenAI, which filed confidentially in June, is working on a private raise of up to $30 billion at a $1.25 trillion valuation ahead of its own listing.³

Put those three names together and you have the largest wave of new stock the market has seen in years. SpaceX alone raised a record $86 billion.¹ Anthropic says it wants to top that. OpenAI is being valued at more than a trillion dollars before it has sold a single public share. For an ordinary investor, this looks like the moment the companies in the headlines finally become companies you can own.

That is true. But SpaceX has already run the experiment, in public, and the results are worth reading before the next prospectus lands. The stock closed its first day at $160.95, up 19.2% from its offering price.⁴ By July 21 it was at $123.54.⁵ The business went on to post its first profitable quarter as a public company.⁶ The first-day buyers spent the summer waiting to get back to even. This report is about the gap between those two facts, and what to do with it before Anthropic’s roadshow begins.

The Wave

The scale of what is coming is new. SpaceX priced its offering on June 11 at $135 a share, selling 555,555,555 shares, with the underwriters holding a 30-day option on 83,333,333 more.⁷ Trading began on June 12 on the Nasdaq Global Select Market, and first-day volume topped 503 million shares.⁴ With the over-allotment counted, the deal came to just over $86 billion, distributed by 23 underwriting firms.⁸

Four months later the company gave the market a reason to believe the hype had substance. In the third quarter, reported on October 20 after the close, SpaceX booked $14.6 billion of revenue, up 161% from $5.6 billion a year earlier, and $4.4 billion of net income, its first profitable quarter since going public. Adjusted EBITDA was $8.8 billion. Starlink reached 13.6 million consumer subscribers. The company gave no numeric guidance.⁶

Anthropic’s filing shows a different kind of business at a similar scale. Revenue for the first half of 2026 was about $16.3 billion, and the prospectus says revenue grew 1,368% in the quarter ended June 30 compared with a year earlier. More than half of the Fortune 500 were customers generating at least $100,000 each over the twelve months to September 30. The same filing shows a net loss of $80.7 billion, most of it from a rise in the fair value of bonds that convert into the company’s shares, and it warns that compute and infrastructure will remain its largest operating expense for the foreseeable future. It did not disclose a share count or a price range.¹

The valuation is already a moving target. Anthropic closed a $65 billion Series H round on May 28 at a $965 billion post-money valuation.⁹ By Monday, Reuters reported that some current and prospective investors were valuing it at about $1.5 trillion, down from roughly $2 trillion only a week earlier.¹ A half-trillion-dollar swing in the private market, in seven days, before a price range exists, tells you how much of this wave is being priced on sentiment.

OpenAI is the third wave behind the first two. Its last round in March valued it at $852 billion post-money; the new round under negotiation, reportedly led by Coatue and investors from the United Arab Emirates, would value it at $1.25 trillion before the new money goes in.³ The question for an individual investor is not whether to pay attention. It is how to take part without being the last money in.

What SpaceX Taught First-Day Buyers

SpaceX opened at $150.00, up 11.11% from its $135 offering price. It touched $174.99 around 1 p.m. and closed at $160.95, a gain of 19.22% on the day.⁴ Anyone who was allocated shares at $135 was up nearly a fifth by the closing bell. Anyone who bought in the open market that day paid somewhere between $150 and $175.

Then the air came out. On June 18 the stock stood at $185. On July 21 it was $123.54, a fall of 33.22% in about a month.⁵ The day after its first earnings report as a public company, the shares fell about 14% to close at $108 on August 5.¹⁰ From the June 18 price, that was a drop of more than 40%. From the first-day close, it was a loss of roughly a third.

✓ Trusted partner One analyst with decades on Wall Street says there is a simple trade, available in any regular brokerage account, that positions you for this wave without buying a single share on day one →

What happened next is the part that matters. The stock rebuilt. It ended October 12 at around $163, about 50% above the August close, and that recovery came through six separate insider share releases.¹⁰ Then came the third-quarter report, with its first profit.⁶ The business did not let first-day buyers down; the price they paid did. Measured from the offering price, the stock was up about a fifth by mid-October. Measured from the June 18 price, it was still down.

The business did not let first-day buyers down; the price they paid did.

That is the pattern the SEC has been warning individual investors about for years. Its bulletin on IPOs says plainly that “purchasing shares in the market immediately following an IPO can be risky,” that the offering price “may bear little relationship to the trading price,” and that underwriters may buy shares to support the price in the first few days, after which the price can fall well below the offering level.¹¹ SpaceX followed the script almost exactly: an opening pop, a few weeks of support, a slide, and a slow recovery that only rewarded people who paid less.

Note what the lesson is not. It is not that SpaceX was a bad company, or that the next listing will be. It is that the first day of trading is the one day when demand is at its maximum and supply is at its minimum, which is the worst possible combination for someone placing a buy order. Every other day of that stock’s life has offered a better balance.

Why Retail Rarely Gets the Offering Price

The reason individual investors so often end up buying on the open market is structural, not personal. In a conventional IPO, the underwriters decide who gets shares. The SEC’s bulletin notes that if you are a client of an underwriter involved in the deal you may be offered a chance to participate directly, and that for most people the route in is to buy the shares “when they are resold in the public market in the days following the IPO.”¹¹ In SpaceX’s case, the lead-left underwriter, Goldman Sachs, allocated the vast majority of the shares on its own authority, with the institutional portion split among the joint book-runners.⁸

SpaceX was, by the standards of the business, generous to small investors. Fortune estimated that about 30% of the shares went to the retail side, against the usual 5% or less, through platforms including Charles Schwab, E*Trade and Robinhood, with a further 5% of the shares reserved at $135 for employees, friends and family of executives, and people the company does business with.⁸ Even so, more than 503 million shares changed hands on day one, most of them at prices well above $135.⁴ An unusually large retail allocation still left most of the retail demand buying from someone else at a higher price. In a normal deal, where the retail slice is a twentieth of the book, the squeeze is far tighter.

The other half of the structure is the lock-up. Existing holders typically agree not to sell for a set period, often 180 days, so the shares trading on day one “are generally only shares that were sold in the IPO.” When lock-ups expire, the SEC notes, a large volume of shares can become available at once and push the price down.¹¹ That is the supply side of the first-day squeeze: a company with 13.6 billion shares outstanding, of which only the roughly 640 million sold in the offering, over-allotment included, could trade at the start.^10, 7

SpaceX staggered its releases rather than opening the floodgates at once. About 911 million shares were freed after the second-quarter results in August. Up to 1.3 billion shares, worth roughly $200 billion, became eligible to trade on Thursday, October 22, two days after the record third-quarter report.¹² Another 328.4 million unlocked on Saturday, October 24, with Monday, October 26 the first session they could trade. Up to 797.6 million more follow after the close on December 8. Elon Musk’s own shares are locked until June 2027. By early December, roughly 40% of the company’s shares could be free to trade, up from about 21% in mid-October.¹⁰

Read that calendar from a buyer’s point of view. The person who bought at $160 in June was competing with every other retail buyer for a thin float. The person buying in December will be choosing from a float nearly twice as large, with the company’s first two public quarters on the record. Supply is arriving, on a schedule, and it will arrive for Anthropic and OpenAI as well, on the same lock-up clock.

The Backdoor That Got Crowded

Some investors saw all of this coming and looked for a side door: publicly traded funds that already held SpaceX stock before the listing. The idea was sound. The execution got expensive. One closed-end fund that holds stakes in private technology companies, with SpaceX as its largest position at about 16.2% of the portfolio, was trading at $40.06 on the day SpaceX began trading, a premium of roughly 100% to its last reported net asset value of $19.97. Short interest in the fund was about 17%, with bears betting the premium would compress once SpaceX had a public price. As Benzinga put it, “As a closed-end fund, there is no automatic mechanism to bring that premium back in line.”¹³

The bears were right. As of Monday, the same fund closed at $28.11 against its last reported net asset value of $34.30, as of June 30, which the site shows as a discount of nearly 20%.¹⁴ Investors who bought the backdoor in June paid two dollars for every dollar of assets and now hold shares worth less than the assets behind them. The underlying holdings did fine; the fund’s own net asset value is up sharply over twelve months.¹⁴ The premium is what vanished, and the premium was the price of access.

There were cheaper side doors. One exchange-traded fund held about $292 million of SpaceX exposure through a special purpose vehicle as of May 21, charging a 0.75% expense ratio with no carried interest, and the same report noted that the closed-end fund “has historically traded at large premiums to its net asset value.”¹⁵ But the cheaper route gave only a sliver of exposure, and the richer route charged a toll that ate the gain. Either way, the pattern repeats the first-day lesson in a different costume: when the whole market knows what you want, the price of getting it rises until most of the reward is gone.

Follow the Money, Not the Ticker

Here is the better question. A listing the size of SpaceX’s or Anthropic’s does not only create one new stock. It moves money through a set of businesses that are already public, already profitable, and already available in any brokerage account. None of them require an allocation, a premium, or a first-day scramble.

✓ Trusted partner See the research behind this story →

Start with the banks. SpaceX negotiated a gross spread of 0.75% or less, tying the lowest percentage on record for a conventional IPO, and still generated an estimated $646 million in fees for its underwriters.⁸ Anthropic has named Goldman Sachs and Morgan Stanley as its lead banks, with 20 other underwriters behind them and Morgan Stanley acting as stabilisation agent.¹ The same firms sit at the top of both deals, and they are paid at pricing, not on the first day’s close.

Then the exchange. Nasdaq’s rulebook sets the entry fee for a new listing at $325,000, and the annual fee for a company with more than 150 million shares outstanding at $199,000.¹⁶ Those are small numbers. The large number is the trading that follows: 503 million SpaceX shares on the first day alone.⁴ An exchange gets paid every day a popular stock changes hands, and it is paid whether the stock goes up or down.

Then the index managers, who in this wave are among the largest buyers of all. SpaceX joined the Nasdaq-100 on July 7, less than a month after its debut, which Reuters said was among the fastest-ever additions to the index. On the same day, BlackRock launched a new iShares Nasdaq 100 ETF, IQQ, with a 0.12% fee waived to 0.10%; its existing Nasdaq-100 strategies already held more than $41 billion.¹⁷ Vanguard reported that as of June 30 its Total Stock Market Index Fund held $3.2 billion of SpaceX, 0.14% of the fund, and its Extended Market Index Fund held another $1.16 billion, while the S&P 500 kept SpaceX out under its rule requiring a profit in the most recent quarter and over the trailing four.¹⁸ SpaceX’s third quarter cleared the first part of that test.⁶ Index inclusion is forced, price-insensitive buying, and the firms that run the indexes and the funds collect fees on every dollar of it.

Then the companies on the other side of the giants’ spending. SpaceX is in talks to raise $40 billion, about $10 billion in bank loans and $30 billion in investment-grade debt, to buy Nvidia artificial-intelligence chips, with Apollo expected to lead the financing and PIMCO in discussions as well.¹⁹ Anthropic’s own filing says compute and infrastructure will remain its largest operating expense for the foreseeable future.¹ The capital raised in these listings does not sit in a vault. It is spent, quickly, on chips, data centres, power and the lenders who finance them. Those suppliers and financiers are listed companies with earnings, and a giant’s IPO is, for them, a purchase order.

The categories, then: the underwriters who price the deal, the exchange that hosts it, the index and fund managers who must buy it, and the suppliers and financiers who receive the money it raises. Each is a way to be paid by the wave without owning the wave’s most expensive hour. The work is in choosing among them, which is exactly why most investors do not bother and buy the ticker instead.

The Rate Backdrop, and the Honest Risks

All of this is happening into a tightening cycle. The Federal Reserve began a two-day meeting on Tuesday and announces its decision on Wednesday at 2 p.m. Eastern, with Chairman Kevin Warsh’s press conference at 2:30. The target range is 3.75% to 4% after a quarter-point hike in September, and CME FedWatch puts the odds of no move this week at about 90%.²⁰ A Reuters poll found 94 of 100 economists expect a hold in October, every economist polled expects a hike at the December meeting, and markets price about a 73% chance of another quarter-point increase before year-end.²¹ The next meeting, on December 8 and 9, comes with a fresh set of projections.²⁰

Rates matter more to this wave than to most. A company valued at a trillion dollars on expected future profits is, mathematically, the asset most sensitive to the rate used to discount those profits. SpaceX fell a third in the summer with policy on hold.⁵ Anthropic will be marketed and priced in the weeks before a Fed meeting the market expects to deliver a hike.^2, 21 The companies that get paid by the wave, by contrast, are being paid now, in fees and orders, not in discounted promises.

The risks run in both directions and deserve a plain statement. SpaceX’s third quarter shows a hot listing can grow into its price; a buyer who held through the summer and ignored the noise was ahead of the offering price by mid-October.¹⁰ Anthropic could price conservatively and trade up for months. The December hike could fail to arrive. The supply from lock-up expiries could be absorbed without a wobble. On the other side, Anthropic’s $80.7 billion reported loss, its reliance on compute spending, and a valuation that moved by half a trillion dollars in a week are exactly the ingredients of a volatile first month.¹ And the suppliers, banks and index managers described above have their own cycles; a stalled wave would hurt their fees and order books too. Nothing in this report removes risk. It is about choosing which risk you are paid for.

The Calendar

The next eight months are unusually well mapped. Wednesday, October 28: the Fed decision and Chairman Warsh’s press conference.²⁰ The week of November 9: Anthropic could begin formal marketing of its offering.² Late November, before Thanksgiving: the Anthropic IPO is expected to price and begin trading on Nasdaq as ANT.¹ December 8: up to 797.6 million more SpaceX shares are released after the close.¹⁰ December 8 and 9: the Fed meets again, with new projections, and a hike is the consensus call.^20, 21 June 2027: Elon Musk’s own SpaceX shares come out of lock-up.¹⁰

Each date is a moment when the crowd’s attention will spike and the price of the obvious trade will spike with it. The lesson of June is that the obvious trade, at the moment of maximum attention, is the one with the least room left in it. The hardest part of a hot IPO is not getting in. It is getting in at a price that leaves room to make money, and the first day is the day that room is smallest.

There is another way to read the same calendar: as a schedule of payments. Every date on it sends fees to an underwriter, volume to an exchange, forced buying to an index fund and orders to a supplier. Those businesses can be bought this afternoon, at a price set by their own earnings rather than by a roadshow. Finding the right one takes work, and almost nobody is doing that work, because the ticker is so much easier to type.