← AI IPO Watch

SpaceX IPO Scrutiny Widens to Valuation and Allocations (July 02, 2026)

July 02, 2026 · 7m 47s · Listen

SpaceX wants a valuation north of Meta and Berkshire combined — on less revenue than Macy's. Let that sit for a second. If you're just joining: reports said Mirae Asset Securities walked away from the SpaceX book with zero allocation, after Korean client demand never made it into the order book. That left the brokerage explaining itself to clients, drew local regulatory attention, and turned a hot AI-adjacent listing into a test of whether global IPO distribution actually works under pressure. This is AI IPO Watch. Today, the Mirae mess goes to lawyers, an index rule quietly commits four billion dollars, and that Macy's comparison? It explains more than the S-1 does. Cassidy, start with the number. IBTimes UK writes:

SpaceX is urging New York investors to value the aerospace giant at more than $1.7 trillion, a figure eclipsing the combined market capitalisation of Meta and Berkshire Hathaway, despite currently generating less annual revenue than the US retail chain Macy's. The company's latest S-1 filing, which posits a staggering $28.5 trillion 'SpaceX economy impact' based largely on future AI dominance, has triggered a fierce debate on Wall Street.

The number's on the front page now. SpaceX is telling New York investors to value it north of 1.7 trillion — more than Meta and Berkshire combined — on revenue that comes in under Macy's. Less than Macy's. And the S-1 fills that gap with a 28.5 trillion dollar 'economy impact' number, most of it pinned to future AI dominance. Right — and only 370 billion of that 28.5 trillion is tied to actual rockets and satellites. The rest is the story they want you to buy, printed on SEC letterhead so it reads like disclosure. Here's what a sophisticated buyer does with the Macy's line, though — they strip it out. The pressure point is the Starlink margin holding this whole thing up, because part of that profit is a gift from the launch division, with a transfer price the documents never actually price. So the margin that justifies the trillion-dollar multiple leans on a zero-cost launch transfer. Value the launch fairly and Starlink's economics look very different. That page isn't in the filing. Here's Park Jung-hwan at Businesskorea:

A truth battle has flared up between Mirae Asset Securities and foreign media over the initial public offering (IPO) of SpaceX, which is evaluated as the largest ever. Bloomberg reported that Mirae Asset failed to be allocated even a single public offering share because it misunderstood the order submission procedure with the lead underwriters, but Mirae Asset completely denied this, stating, “We submitted the order normally according to the procedure guided by the lead underwriters and also received official confirmation.”

So the Mirae story went adversarial overnight. Bloomberg said they fumbled the order procedure and got zero of a $1.1 billion Korean allocation — and now Mirae's calling it a malicious false report and taking legal action. Right, so we've gone from process embarrassment to a truth battle, their phrase, in about seventy-two hours. Mirae says they submitted normally and got official confirmation. Bloomberg says multiple anonymous sources say otherwise. And here's the part that actually matters to me — a syndicate-adjacent player is now threatening to sue the financial press over its allocation. Think about the next book-building where Korean demand needs to show up. Bookrunners remember who made noise. What I can't get past is that both versions can't be sitting cleanly in the documents. Somebody's version isn't in the order book — and on a four percent float, one miscommunication is unrecoverable. There's no back-end reallocation when supply's this thin. Do Hyunjung at The Herald Business is tracking this. So here's the Herald Business version — Mirae takes a demand inquiry, treats it like an order submission, ends up with zero shares. Clean, mundane, embarrassing. Except the version we just aired from Businesskorea has Mirae calling that account a malicious false report and reaching for lawyers. So we've got two documents, one incident, and no agreement on what happened. For me, that's the tell. If a syndicate member is publicly disputing its own allocation story, and threatening the press over it, every bookrunner on the next Korean deal is watching. Making noise like this doesn't get you a bigger book next time. Right — and that's why a single miscommunication is unrecoverable here. Supply. On a float this thin, there's no clean-up trade. You miss the window, you're just out. That's what makes one email fatal. A process error that on a normal deal is a phone call and a fix. Here it's a lawsuit. This one's from MarketBeat:

SpaceX NASDAQ: SPCX will bypass traditional public market seasoning requirements to enter the Nasdaq-100 index on July 7. This regulatory shift triggers an estimated $4.3 billion in forced institutional buying just weeks after the initial public offering. Paired with a rumored terrestrial backhaul partnership that positions Starlink Mobile to immediately challenge legacy telecom providers, SpaceX commands a near-term liquidity catalyst capable of temporarily overriding structural valuation headwinds.

July 7. Fifteen trading days after the debut, SpaceX drops into the Nasdaq-100, and J.P. Morgan says that forces about $4.3 billion of passive buying into it. And QQQ doesn't read the S-1 we've been picking apart all week. It doesn't care about revenue less than Macy's. Its whole job is to match a weight. Right. The catalyst here is Nasdaq changing its own eligibility rules, so mega-cap IPOs can get in after fifteen days instead of months. The demand comes from a rule change. It doesn't come from buyers deciding the price makes sense. And it lands on a four percent float. So when SPCX prints that first real price, you're seeing forced buyers reach into an empty room before you get anything like normal price discovery. Same trick as a cornerstone book, just running backwards. Cornerstones manufacture demand before the deal; index inclusion mandates it after. Retail's going to anchor to that July 7 number like it means something. And the governance kicker — passive funds now have to hold the dual-class structure whether they like it or not. Index inclusion takes away the one vote you always had: walking away. If AI IPO Watch helps you stay ahead of the market, take a second to subscribe and leave a review wherever you're listening. It helps other people find the show, and it really supports our daily work.

What we're watching next: SpaceX's Nasdaq-100 inclusion is scheduled for July 7, when benchmarked funds would need to adjust exposure. Then its first public earnings release is slated for August 6, after which MarketBeat says about twenty percent of insider shares become eligible for sale.

You'll find links to every story we covered in the show notes, so if one caught your ear, you can dig in there.

That's AI IPO Watch for today. This is a Lantern Podcast.