The three hard-tech moonshots fueling SpaceX’s unbelievable IPO

SpaceX’s reported $75 billion stock offering is said to be heavily oversubscribed, but the article argues the company’s implied valuation is far above sober third-party estimates: Morningstar values it at about $825 billion and Damodaran at $1.2 trillion versus bankers’ near-$1.8 trillion assessment. The core debate is whether SpaceX can execute on an orbital data center strategy that requires a reusable rocket, a U.S. chip foundry, and satellite production at roughly 6,666 units per year. The piece is cautionary overall, highlighting major execution risk despite strong investor enthusiasm.

Analysis

The market is effectively pricing two different businesses into one equity: a cash-generative launch/network monopoly and a venture-style AI infrastructure option. That structure tends to suppress realized value because the mature cash flows get dragged down by the execution discount on the speculative leg, while the speculative leg is also mispriced because customers may use SpaceX as a stopgap, not a strategic home. The subtle second-order issue is that any meaningful AI compute buildout competes directly for capital, management attention, and launch cadence with the core business that actually funds the story.

The biggest near-term risk is not demand; it is industrial throughput. If SpaceX cannot ramp satellite, power, and launch integration on the stated timeline, the equity will likely re-rate on a software-like multiple of the existing businesses rather than a platform multiple of the whole vision. That creates a classic path dependency: a single Starship or production slip over the next 3-6 months can compress the implied option value sharply, because the market is paying today for a 12-24 month step-function in capability that may not be physically achievable.

For competitors, this is a mixed signal. Hyperscalers and frontier labs may benefit if SpaceX’s AI ambitions absorb scarce capital but fail to scale, because customers will keep buying external compute while the market learns that space-based inference is a long-dated project. Conversely, if the thesis works even partially, the real pressure lands on the economics of terrestrial data-center expansion, not on model labs themselves; power-constrained colocation, grid-equipment, and chip supply chains could see a temporary relief rally if investors extrapolate orbital compute as a future demand sink. The contrarian miss is that the market may be underestimating how valuable the core monopoly already is, but overestimating how quickly it can be converted into an AI stack with durable margins.

Key Decisions for Investors

• Short MORN tactically on the premise that the market will continue to discount the company’s fair-value framework if the IPO trades above the sober estimates; target a 4-8 week horizon with tight risk control around any evidence of sustained aftermarket support.
• Use a pairs expression: long XLK / short small basket of AI-infrastructure enablers with exposed capex intensity if SpaceX-like orbital compute hype pulls forward sentiment too far; the thesis is that real monetization is delayed while valuation multiples get ahead of throughput.
• Buy short-dated downside protection on the new issue if options are liquid: 1-2 month puts financed by selling far-out calls to express the view that execution headlines, not demand, are the first crack point.
• If you want to own the core monopoly, prefer a barbell in the ecosystem instead of the IPO itself: long select launch/satellite supply chain beneficiaries with near-term cash flows, while keeping the high-variance AI option exposure capped.
• Set a catalyst watch on the next Starship/production update over the next 30-60 days; if cadence slips or FAA/mission timing gets pushed, fade the IPO on the grounds that the market is paying for a 12-24 month acceleration that is unlikely to survive contact with manufacturing reality.