SpaceX is drifting back toward one-thirty-five. SK Hynix cracked the day after its debut — turns out the supply math doesn't care how good the roadshow was. If you're just joining: SpaceX has moved from IPO spectacle into valuation digestion. Investors clearly like the brand. Now they have to decide whether the market cap, the sales multiple, the thin public float, Starlink's profit concentration, and those staggered insider lock-ups can actually hold up the price that followed the offering. This is AI IPO Watch. Today — a chip debut that plunged, a rocket that flew home, and who's left holding the bag when the glow fades. Investing.com, with Isaac Olaosegba:
However, the euphoria didn’t last. SPCX has since shed more than $800 billion in market value from that peak, sliding over multiple sessions in a row, including a 16.4% single-day drop, setting new all-time lows. As of Monday’s session, shares were trading around $139. If the stock breaks below the $135 IPO price, it would mean every investor who bought at the offering price is running at a loss.
Quick update on SpaceX's repricing — the stock is now pressing right up against its $135 offering price. It opened at $150, ran north of $200, briefly made the company worth more than Microsoft. And now here we are. Every dollar of that drop from $200 back toward $135 is stock changing hands from the offering buyers to the people who chased it above $150. Whoever bought at $161 on day one is underwater now. And notice what the article hangs the whole story on — a debt sale and lockups. The equity is round-tripping while there's a bond deal and a staggered unlock calendar still sitting out there unhit. $75 billion raised, the largest offering ever recorded, and a month later the market's asking whether $135 is a warning sign or a buying opportunity. If you're the issuer, that tells you the bankers nailed the price. The narrative always gets this backwards. Pricing at $135 and watching it climb to $200 didn't mean they left money on the table. Fast money was renting the stock, and the rent's coming due before the lockups even open. Here's what the Securities and Exchange Commission is reporting. Churchill Capital Corp XI filed an 8-K dated July 14 — a Rule 425 written communication. Which means Klein's eleventh blank-check vehicle is out here talking to somebody about a deal, quietly, on a Tuesday. The eleventh. Cayman Islands shell, 640 Fifth Avenue, and it's still hunting in 2026. That tells you the SPAC machine never actually died — it just went on mute. And for anyone new to the acronym — a SPAC is a listed shell that raises cash first, then goes shopping for a private company to merge into. Faster runway to the ticker than a full S-1. Faster, and a lot less disclosure on the way in. Here's the important part: a Rule 425 filing means they're communicating about a target. But an 8-K cover page doesn't name it or price it. The interesting number lives in whatever this merger becomes, and it isn't on this page yet. So we've got a whisper-quiet SPAC filing the same afternoon a PE shop is loudly slapping 'AI-enabled investing' on a press release. One of those is a marketing document and one is a legal one — guess which one gets the headline. From City A.M.:
Clearlake Capital Group, L.P. (“Clearlake”), a global investment firm managing integrated platforms spanning private equity, liquid and private credit, and other related strategies, today announced a partnership with Databricks, the Data and AI company, and West Monroe, a global business and technology consulting firm, to accelerate Clearlake’s portfolio companies’ adoption of data, analytics, and AI capabilities.
So Clearlake announces a partnership with Databricks and West Monroe for 'AI-enabled investing and portfolio value creation.' Read that press release closely and there's no fund, no filing, no dollar figure — it's a PE firm branding its portfolio-monitoring workflow as an AI play. And there it is, Cassidy. Clearlake's putting the AI story where the actual AI money is being made right now — inside portfolio operations, deal origination, due diligence — instead of on a Nasdaq ticker retail can touch. On the same day, Churchill Capital XI quietly files an 8-K — a blank-check shell still hunting for a target in 2026. One firm files a SPAC in the dark, another brands itself around AI with a bullhorn. Same exit pressure, different costumes. Smart money is acting like the AI monetization story is still being built behind the curtain. West Monroe connecting 'investment, operational, financial data in a secure environment' — that's a private edge, deliberately not disclosed to public markets. Mateusz Kaczmarek, over at TechStock2, has the details. Thirty-two days to refly a booster, and the stock's drifting back toward $135. The hardware turnaround is the fastest thing at SpaceX right now — faster than the profit case for anything except Starlink. And this piece keeps circling back to Starlink economics for a reason: Starlink is the segment you can actually see. Starshield's still the black box. Right, the bull and the bear were arguing over Starshield and neither could see it. Now the stock's back near IPO price, and the only numbers anyone can price are the visible ones: boosters getting reflown and dishes getting sold. Coming down toward $135 from above — that's a round trip. The people who priced this at $135 did their job. Every dollar it's given back since came out of somebody's pocket. And I can tell you whose pocket. Somebody bought at $152 reading an $800 target as a promise. That buyer's underwater now — before a single lock-up share has hit the tape. This one's from AOL:
This position helped generate intense demand when SK Hynix brought its shares to Nasdaq through a $26.5 billion American depositary receipt offering. But the stock’s first few days in the U.S. market have also shown how quickly enthusiasm around AI chips can reverse.
Priced at $149, opened at $170 — call it 14% above the range. And then it gave it all back. The seven-times oversubscription looked like conviction on paper. That was momentum capital dressed as conviction. Momentum capital doesn't hold the position through day two — it hands the position to whoever's slower. And the slower money read a $26.5 billion HBM story as a growth guarantee. But look at the same tape today — IBM down 22% on a memory capex squeeze. That's the customer side of this exact trade getting hit. SK Hynix's whole pitch was: we're the memory that feeds the AI processors. If your biggest customers are suddenly squeezing capex, that demand curve isn't as vertical as the roadshow drew it. Right — you can be the best HBM maker on earth and still have a bad week when the people buying your chips get told to slow their spend. Even a great component with real customers and factories behind it only works if the orders keep coming. IBM's number is the first crack in that assumption, showing up in real time. Got a company we should be watching, a filing we missed, or a correction to share? Send us a note anytime at aiipowatch at lantern podcasts dot com. We read every message.
You'll find links to every story we covered today in the show notes, so if something caught your ear, you can dig into the original reporting there. That's AI IPO Watch for today. This is a Lantern Podcast.