SpaceX IPO'd at $1.77 trillion and closed its first day at $2.1 trillion, the largest public offering in history. The company lost nearly $5 billion last year, has never turned a profit on rocket launches in 24 years of operation, and is sitting on AI infrastructure it does not yet know how to monetize. The valuation is not justified by the books. It is justified by something else.
That something else is the tolerance for improbable bets. Starlink, the satellite internet business that actually generates SpaceX's real revenue, was once an implausible idea that got funded anyway. The original article makes the uncomfortable argument that financial overexuberance is not a market failure but a mechanism: it is what converts moonshots into infrastructure. The same logic applies to the current AI buildout, where frontier model price wars are already emerging, data center projects are collapsing mid-construction, and SpaceX itself is selling compute to rivals like Google rather than using it internally.
The risk is not evenly distributed. Retail investors took 20 to 25 percent of the SpaceX IPO allocation, compared to the typical 5 to 10 percent, meaning everyday buyers are absorbing a larger share of the downside if the bet fails. Read the full piece for the specific breakdown of where SpaceX's money actually comes from, and why the AI parallel holds even if you think the valuation is absurd.
[READ ORIGINAL →]