Morgan Stanley initiates on SpaceX and leads with... AI compute at four bucks a watt. Funny how the shiniest number is never the one in the filing. If you're just joining, SpaceX's path to market got bigger than a simple IPO watch a while ago. Before and after the Nasdaq debut, investors chased exposure any way they could: MEXC-linked products, tokenized securities, tracker certificates, subscription campaigns, perpetual futures, and actual shares. The question through all of it: did those wrappers broaden access, or just fragment price discovery around what SpaceX is really worth? This is SpaceX IPO Watch. Today: a Morgan Stanley cost figure, a two-hundred-billion IPO wave threatening a sell-off, and the question of who's really setting SpaceX's price. Let's get into it. SpaceX public listing isn't over. Follow us wherever you're listening, and the next chapter comes to you. This one's from blockchain.news:
SpaceX achieves infrastructure capex costs of about $4 per watt through vertical integration, less than half the industry average of around $9 per watt according to Morgan Stanley's semis and internet team analysis. • Vertical integration spans liquid cooling systems, battery installations, substations, and planned expansions into turbines, solar manufacturing, networking, and compute via Terafab to eliminate AI scaling bottlenecks.
Adam Jonas just put a number on the thing I've been circling all week — $4 a watt versus a $9 industry average. That's the vertical-integration moat with a price tag on it. Or it's the sexiest possible story bolted onto a coverage initiation. Jonas leads with AI compute at $4 a watt — but is that number in the S-1, or is it Morgan Stanley's semis team modeling backward from a price target? Meanwhile, the Starlink margin question — the one that actually determines whether this business holds up — just got shoved under the shinier headline. Right on schedule. But that's the point, Cassidy — it's a second revenue leg that didn't exist in anyone's model six months ago. 10.3 million Starlink subs, and now a compute business the sell-side is finally underwriting. Sure. And the projection is $319 billion in revenue by 2030, climbing to $3.3 trillion by 2040. Eric, one of those numbers is a forecast. The other is fan fiction with a decimal point. Fine — the 2040 number is a slide. But the $4 versus $9 spread is a bottom-up cost read, not a vibe. That's the piece institutions will actually anchor to. It's also the piece that quietly reframes a government-dependent launch company as a commercial AI play. Convenient, given how much of the revenue still leans on NASA and DoD. When we hear SpaceX is 'worth' hundreds of billions — or now, apparently, close to two trillion dollars before it even goes public — who's actually setting that number? And how much should a regular person trust it as a preview of what the stock will actually be worth? Great question, because that number comes from a very different place than a real stock price. In private markets, valuations are usually set by insider tender offers and late-stage funding rounds — a small group of institutional buyers and existing shareholders agreeing on a price for a slice of the company, not millions of investors trading it out in the open every day. For SpaceX, the IPO target range reported by media outlets is roughly $1.75 to $2 trillion, per BNN Bloomberg. The catch is that SpaceX is selling only about four percent of the company in this offering, according to analysis from NC State's Poole College of Finance, so that tiny slice gets used to extrapolate a valuation for the whole enterprise. And some serious analysts think that math is stretched. Morningstar put its fair-value estimate around $780 billion — roughly 55 percent below the $1.75 trillion IPO target — and called the company 'significantly overvalued,' partly because of risks from Musk's AI venture xAI, which they described as a 'material threat of value destruction.' So yes, private-market valuations are real data points. They're also set by a thin market with limited price discovery, and professional analysts can land in a very different place. So if only four percent of the shares are actually hitting the market on day one, does that small float make the opening price even less reliable as a signal of true value? Exactly. A thin float can amplify early price swings in either direction. That's why the NC State analysis flagged SpaceX's offering structure — including a fixed IPO price and staged insider lock-ups — as an unusual setup that could reshape how investors interpret the opening trade. The people watching most closely on day one are the ones who've held for years: firms like Fidelity, Ark Invest, and Baron Capital, who accumulated stakes long before any of this was obvious, per reporting on early SpaceX backers. For everyone else, the cleaner read comes in the first weeks of trading, once more shares are unlocked and the market gets broader. That's when we'll see where the crowd actually prices this thing. AOL writes:
SpaceX raised $75 billion in its Nasdaq debut on June 12, pricing shares at $135 each and valuing the company at nearly $1.77 trillion. Total proceeds later climbed to $85.7 billion after underwriters exercised their option to buy additional shares, making it the largest offering ever recorded, CNBC reported.
So here's the number that should make everyone sober up. SpaceX pulled $85.7 billion out of the market on June 12 — Renaissance says that's roughly two-thirds of every IPO dollar raised in the U.S. this year. And now AOL's running the inelastic-markets angle: $200 billion in new supply, maybe a trillion in market cap at risk. Every dollar pulled to fund new shares can erase five in existing value. Right, but the hypothesis treats SpaceX, Anthropic, and OpenAI as interchangeable supply. That's where I get off the bus. Sovereign-critical infrastructure doesn't get absorbed the same way a software float does. You can't model a launch monopoly like it's another SaaS ticker. Fair distinction — but the mechanism doesn't care whether the buyer respects the moat. The dollars still come out of somebody's portfolio to fund $85 billion of new paper. And this is exactly why I keep coming back to a Starlink-only structure. A smaller float means less of this drain-the-pond pressure. The full-stack $1.77 trillion listing is what puts you at the front of that supply queue. If you're tracking SpaceX’s path to the public markets, you might also like The Data Center Daily, a daily briefing on AI compute, hyperscaler capex, power, chips, and energy markets. Find it wherever you listen to podcasts.
Links to every story we covered today are in the show notes, so if one of them deserves a closer read, start there. Thanks for listening. That’s SpaceX IPO Watch for today. This is a Lantern Podcast.