Behind the video
SpaceX At $1.83 Trillion Might Be Cheap
Overall Thesis
The host concludes that SpaceX appears underpriced, emphasizing its compute infrastructure, construction speed and ability to serve growing AI demand. The discussion presents revenue comparisons and the guest's earnings-based upside scenarios, while acknowledging market volatility, unlocks, merger uncertainty and AI bubble concerns.
Narratives
NVDANvidiaNvidia is discussed as the supplier of chips needed for SpaceX's compute expansion. The discussion concerns supply availability and customer priority rather than Nvidia's share-price outlook.
Key Arguments
- The host identifies Nvidia Rubin chips as part of the proposed infrastructure expansion.
- The guest argues that Nvidia manufacturing capacity will support deliveries.
Risks acknowledged
- Timely chip availability is raised as a potential dependency for SpaceX's expansion.
SPCXSpaceXThe host views SpaceX as apparently underpriced because its compute business and infrastructure advantages may justify more than its current valuation. He supports its competitive position and expanding AI demand, but the numerical stock-return forecasts come from the guest or an outside investor rather than an explicitly adopted host target.
Key Arguments
- The host presents an outside investor's calculation that valuing compute at Anthropic's proposed revenue multiple could account for SpaceX's entire valuation.
- The host argues that SpaceX has secured land, power and construction capabilities needed to deploy compute rapidly.
- Faster construction and capital recovery could give SpaceX an advantage over competitors facing financing and power constraints.
- The host endorses Grokbot's distribution and agent interface as potential advantages.
- Growing agent adoption and increased spending by existing users could expand demand for compute.
- The guest argues that future earnings could support substantial valuation upside, but those numerical forecasts remain attributed to the guest.
Risks acknowledged
- The stock has fallen despite the optimistic fundamental discussion.
- Unlocks, merger uncertainty and AI bubble concerns complicate valuation.
- Revenue-multiple comparisons may be weaker than an earnings-based valuation.
- The guest acknowledges uncertainty about the earnings multiple investors will assign.
- The host acknowledges that Grok's underlying model trails leading competing models.
Predictions (1)
MSFTMicrosoftThe host uses Microsoft to illustrate financing needs and potentially idle AI infrastructure awaiting power. This is an operational and competitive concern, without an explicit forecast that Microsoft shares will decline.
Key Arguments
- The host cites a report of a planned approximately $25 billion bond sale.
- He discusses the cost of chips and facilities remaining idle while awaiting power.
- He presents another commentator's theory that unused chips could be deployed through SpaceX infrastructure.
Risks acknowledged
- The chip-sharing scenario is attributed to Brian Wang rather than established as a completed arrangement.
- The guest argues that infrastructure deployment is principally a question of timing.
Hedges & Caveats
- The host explicitly introduces the opening valuation calculation as another investor's math rather than a price target.
- The host says the current environment makes the apparent undervaluation difficult to assess.
- Unlocks, merger discussions and AI bubble concerns may weigh on the stock.
- The guest's fivefold and more-than-triple upside forecasts are not explicitly adopted by the host.
- The guest acknowledges that valuation multiples can vary and that terrestrial infrastructure expansion faces constraints.