A Korean broker apologized to its own clients this week because it promised them a piece of the SpaceX IPO and delivered exactly zero shares. Today, demand meets the fine print — and someone got hurt in the margins. This is AI IPO Watch. On deck: a $2.5 billion humanoid robotics SPAC, fourteen cornerstone investors piling into one Hong Kong listing, and advisors on record calling the whole scarcity rush a trap. One tap on follow, and we'll be back in your ears before you know it. Yara over at Gasgoo has the details. Fourteen cornerstone investors locked up ahead of the Momenta book — that's a Hong Kong specialty. The whole point of cornerstones is to manufacture demand: commitments are lined up before the book even opens, so the deal looks pre-validated. Right, and Gasgoo is the one framing it as 'why are they snapping up shares.' Fourteen names committing early can be engineered into a confidence signal before the market has really spoken. And it works — retail sees fourteen institutions in the prospectus and reads it as scarcity. Those investors, though, negotiated terms and a lockup. They got in on the ground floor; they weren't chasing the deal. Which is the whole trick. Cornerstone allocation soaks up float before the offering opens, so whatever's left prices into thinner supply. That whole 'snapping up' framing becomes a pricing lever. In a Hong Kong tape where the index has been sliding, that concentration matters. The appetite is narrow, parked in a handful of new issues with cornerstones stapled on. Here's The Edge Malaysia:
The brokerage inadvertently treated an early request to indicate investor interest as the point at which it had submitted binding orders, the people said. As a result, more than US$1.1 billion (RM4.48 billion) worth of Korean demand was never entered into the IPO order book, they said.
Okay, this is the one. Mirae Asset — Korea's biggest brokerage — was one of 23 underwriters on SpaceX, and the only one that walked away with zero shares. And the reason is almost stupidly mundane. They treated an early indication of interest as if it were a binding order. Over one-point-one billion dollars of Korean demand never even made it into the book. 'Project Apex.' Somebody named it Apex and it ended with an apology letter and a regulator walking through the door. I keep coming back to this: a 4% float had to meet global demand. When supply is that thin, one miscommunication doesn't get smoothed over. There's no slack in the allocation to absorb a mistake. It just breaks. Right — the thin float didn't just create a retail narrative problem. It broke the distribution chain at the brokerage layer. The clients at Mirae weren't naive retail rubes. They were just downstream of supply that couldn't stretch. Simon, writing in TopSecretStocks:
Agility Robotics’ landmark $2.5 billion go-public merger with Churchill Capital Corp XI (ticker: $CCIX) solidifies the transition of humanoid robots from venture-backed capital into viable, publicly traded commercial assets, providing massive scale and transparency to the robotics sector.
So the first pure-play humanoid company to hit a North American exchange, and it's coming in through a blank-check vehicle. Churchill Capital Corp XI — $2.5 billion pre-money on a company still burning venture money. And note the ticker confusion right in the source — CCXI in one line, CCIX in the next. When the promoter can't keep the shell straight, that's a pretty good tell about how carefully this got assembled. Backed by Nvidia, Amazon, SoftBank, Foxconn — and Foxconn's leading the $200 million PIPE. So the strategic partner building the robots is also underwriting the float. That's a supply-chain bet wearing an investor's coat. The number that actually means something is the $300 million in multi-year committed orders. Everything else is a valuation the sponsor negotiated. Committed orders, at least, can be tested against a proxy — if they disclose the counterparties. And a SPAC skips the roadshow grind, which means less scrutiny on those orders, not more. Retail sees '$300 million committed' and '10x' in the headline and never asks who signed, or when they can walk away. The headline literally asks 'Positioned to 10x?' A pure-play humanoid stock going public through a shell, framed as a ten-bagger before the merger proxy's even effective. I'll wait for the proxy, thanks. From Memeburn:
Chinese tech companies have raised $3.1 billion from mainland listings this year to June 18, more than five times last year’s pace. - Nearly 50 companies have applied for IPOs in Shanghai and Shenzhen, with planned fundraising of at least 126.1 billion yuan, or $18.7 billion.
$3.1 billion from mainland listings by June 18 — more than five times last year's pace. Beijing is routing these chip and robotics firms toward Shanghai and Shenzhen as policy, not as a gentle nudge. And that's the distinction I care about. When the state designs the pipeline, an onshore listing serves self-reliance first and capital-raising second. Nearly 50 applicants and the queue reads like an industrial plan. That's a completely different pressure from the SpaceX dynamic we've been talking about all week. A founder here is aligned with a mandate instead of managing a board's exit anxiety. That changes the whole filing calculus. It also changes what the price signal means. If the buyer base is directed toward these names for reasons beyond returns, the pop tells you even less than it usually does. And it usually tells you nothing useful for retail. InvestmentNews, with Gregg Greenberg:
The calls are coming in fast. Clients see SpaceX on the news, or hear that Anthropic or OpenAI may be heading toward a public listing, and they want in — immediately. For financial advisors, it is becoming one of the defining client communication challenges of 2026: how do you redirect genuine enthusiasm about a real investment theme toward a disciplined strategy, without dismissing the opportunity entirely?
This one's useful: three advisors are on record, names attached — Boswell, Malone, Gottbehuet — saying the smart money was in long before the bell. That's the buy side saying the quiet part out loud. And Boswell gives you the mechanism, not just the vibe. Twenty-five years ago a company went public four years after its first round. Now it's ten years and seven rounds of funding before retail ever sees a share. Ten years and seven rounds. By IPO day, the value creation already happened — for the people who got in at round two. Retail's arriving right as the early money gets its exit. And 'scarcity' is the word selling the deal here. You're being sold rarity right as the asset becomes more widely available than at any point in the company's life. Right. The stock is available. The good entry point closed years ago. Got a tip, correction, or company we should be watching? Send it our way at aiipowatch at lantern podcasts dot com. We read your notes, and they help sharpen the coverage.
You’ll find links to every story we covered today in the show notes, so if something caught your ear, you can jump straight into the source material there. That’s AI IPO Watch for today. This is a Lantern Podcast.