We're six days out from SpaceX's first-ever earnings print, and a Morgan Stanley analyst just slapped a $300 target on a stock trading around $118. Somebody's math is aspirational. Here's how we got here: SpaceX priced its IPO at $135, opened at $150, ran north of $200 — then slid back below the offer price. The next pressure point was already circled: first public earnings on Aug. 4. Two business days later, a wave of pre-IPO holders could become eligible to sell. This is AI IPO Watch. Today — one bank touching three AI deals at once, an OpenAI filing that says the quiet part out loud, and a lockup clock nobody's talking about. Let's start with the receipts. This story isn't over: SpaceX post-IPO repricing. Follow us wherever you're listening, and the next chapter comes to you. Daniel Sparks, writing in The Motley Fool:
The rocket and satellite company went public in June at $135 per share, saw the stock climb as high as $225.64, and has since watched it fall to about $118 as of this writing -- roughly 48% below the high, and under the IPO price itself. And under the company's own lockup rules, the report starts a clock: two trading days after results are released, up to 20% of eligible insider and employee shares become free to trade for the first time.
Aug. 4, after the close — SpaceX's earnings debut. But the part that matters comes two trading days later, when up to 20% of insider and employee shares unlock. So those numbers become the evidence greeting the first real wave of people allowed to sell. That's the setup on the 4th. And here's the tape they're selling into — priced at $135 in June, ran to $225, sitting around $118 now. Under the IPO price. Forty-eight percent off the high. The revenue holds up — $19.3 billion trailing, Starlink at 11.4 and 10.3 million subscribers. The business story is intact; the price just ran ahead of it. Right, and whoever on that pricing team argued against a fatter range looks awfully smart today. The pop got the headlines; the current holders are underwater. The retail buyers who took dual-class paper for the access — this is the week they find out what that access costs when the lockup opens. So you've got growth that compounds, then a supply valve opening two days later. If Starlink prints anywhere near 50% again, that's the only thing standing between the tape and the sellers. This one's from OpenAI:
We recently submitted a confidential S-1. We expect it to leak so we’re just announcing it. We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company. But it’s a complicated set of tradeoffs and this gives us the option to go public sooner if that ends up being best.
OpenAI filed a confidential draft S-1 back on June 8, and they're announcing it themselves under Rule 135 — because, in their words, they expect it to leak. Confidential only describes the stage of the process; it doesn't guarantee secrecy. The company just proved that by beating the leak to the punch. Read the timing line, though. They said it may be a while because there are things they want to do that are, quote, easier as a private company. That's a board looking at where consumer and industrial deals are pricing this year and deciding this window punishes them more than it rewards them. Which is the honest version, actually. A confidential S-1 keeps the option open without locking in a date. There's no price range, no timing — but there's now a real document at the SEC, which is more than we could say a week ago. And notice the hedge at the end — quote, sooner if that ends up being best. They've written optionality right into the press release. The filing's real, the calendar isn't, and every private investor trying to get in before the door closes just got told the door has a lock they don't control. Compare that with SpaceX, six days out from actual earnings. One company's living under the disclosure regime; the other's still window-shopping it. June 8 is the date that matters here — everything else is a decision they haven't made yet. Mariyam Shajil, writing in International Business Times Singapore:
Morgan Stanley analyst Adam Jonas said SpaceX's recent stock selloff has pushed shares toward a level that implies investors are assigning no value to its artificial intelligence business, calling the drop an attractive buying opportunity. In a note to clients Friday, Jonas set a $300 price target on SpaceX, with more than half of that valuation coming from the company's AI operations.
So Adam Jonas puts a $300 target on SpaceX, with more than half of it — north of $150 a share — pinned on the AI business. And it lands the same week a 20% lockup slug comes free, two trading days after Aug. 4. An analyst upgrade right before a supply event. Institutional buyers know exactly what that framing is for. Retail sees a $300 number and a bank logo. Here's my line every time: you won't find $300 in any filing; it comes from Adam Jonas's model. The filed number is $135, and the tape says $118. Now read the syndicate. All four IPO underwriters — Goldman, BofA, Citi, Morgan Stanley — have buy ratings. Four ratings from the syndicate aren't four independent reads. You're hearing the house talk its own book six days before earnings. This is the repricing we flagged: $135 at the offer, $225 at the peak, underwater now. The headlines went to the pop; the holders ate the haircut. Today, the pricing team that resisted a higher range looks pretty smart. And the argument goes, near $100 the market's assigning zero to AI. Or maybe the market's just repricing the mythology, and Jonas is naming a floor before the insiders can sell into it. From Financial Post:
Now that the hordes who crowded into chipmakers are heading for the exits, some investors and strategists are approaching the theme from a different angle: Focusing on shares of the firms that are helping provide all the funding for the AI buildout. For Ohsung Kwon, an equity strategist at Wells Fargo & Co., the investment cycle makes big banks an “AI-adjacent sector” and that could provide a key reason for further outperformance in their shares.
Here's the number that matters in this piece — banks just booked their highest equity-offering advisory revenue since 2021. The AI IPO cycle has moved beyond chips; for banks, it's a fee event, and the house is already getting paid. And look who's on the leaderboard. That Bloomberg lender index is up 14 percent this year — beating both the S&P and the Nasdaq 100 — and Morgan Stanley's sitting right in it. And that completes the map we've been building. One bank is touching three deals at once. Morgan Stanley put out the SpaceX analyst note, shows up in OpenAI's hiring pipeline, and collects the underwriting fees. That's a lot of relationship capital concentrated across every competing AI capital story at once. And every one of those mandates pays whether the retail buyer wins or not. SpaceX at $118, below its IPO price, six days from earnings — the current holders are underwater and the advisory fee already cleared. That asymmetry is the business model. Wells Fargo's strategist is calling big banks an “AI-adjacent sector,” Eric. Which is a polite way of saying they figured out how to get long the buildout without buying a single data center. Right — no capex or lockup, and none of the dual-class paper. Just the toll booth. When the pop happens, the bank keeps the spread; when it doesn't, they keep the fee. That's the trade nobody puts in a roadshow deck. Value Add Pulse writes:
The 2026 IPO window is genuinely open -- Nasdaq reported a record $129.3 billion raised in new listings during the first half of the year -- but nearly every consumer and industrial debut this cycle has priced well below the valuation chatter that preceded it, a pattern worth naming explicitly as this week's deals come to market.
Record window, record haircut. Nasdaq books $129.3 billion in the first half, and almost every one of these deals prices below the number floating around before the roadshow. Jersey Mike's is the cleanest tell — chatter said ten to twelve billion; the actual range came in around eight. That two-to-four-billion gap was never in any document. It was a negotiating position wearing a valuation costume. And that's exactly the data an issuer's board reads. It puts that line from OpenAI's confidential S-1 in a different light — “things that are easier as a private company.” You're hearing a board look at the discount and decide the window punishes them right now. The CXMT deal gets me the most — priced at 8.66 yuan, then popped 471 percent. Underwriters deliberately set that floor low, and the market did the repricing they wouldn't. Nobody was blindsided by demand; that money went straight to the day-one allocation. SpaceX gives you the same chart — priced at $135, ran to $225, sitting at $118 now. The pop got the headlines; the holders are underwater. The team that resisted the higher range looks smart in hindsight, and nobody's writing that piece. Have feedback, story ideas, or corrections? Email us at aiipowatch at lantern podcasts dot com. We’d love to hear from you.
What we’re watching next: SpaceX reports quarterly results after the market closes on August 4, its first as a public company. Two trading days later, up to 20 percent of eligible insider and employee shares become free to trade for the first time.
You’ll find links to every story in the show notes if you’d like to read further.
That’s AI IPO Watch for today. This is a Lantern Podcast.