SpaceX's Thesis Sounds Far-Fetched. Is There Any Way This Company Could Ever Make Money?

The article argues SpaceX's $1.75 trillion IPO valuation is built on a flawed AI thesis, centered on moving power-intensive workloads into orbit to exploit solar energy. It highlights major execution and cost risks, including slower satellite transmission, poor maintainability, and more efficient terrestrial solar alternatives. Overall, the piece is critical of SpaceX's AI growth strategy and suggests the company is unlikely to generate shareholder value from it.

Analysis

The market is likely overestimating the monetization path here and underestimating the option value embedded in terrestrial bottlenecks. Even if orbital compute remains a long-shot, the narrative still reinforces a multi-year capex pull-forward into power, cooling, networking, and domestic data-center buildouts — the real beneficiaries are the picks-and-shovels names, not the headline AI aspirants. On that basis, the setup is more bearish for speculative “future platform” stories than for the AI supply chain itself.

For NVDA and INTC, the first-order read-through is positive only if the market continues to view compute scarcity as the limiting factor; the second-order risk is margin pressure if customers start demanding lower-cost architectures and more efficient power deployment. Space-based compute is so operationally complex that it actually widens the gap between the cost of leading-edge hardware and the cost of deploying it, which favors incumbents with the balance-sheet to absorb slower depreciation cycles and more frequent refreshes. That said, any incremental capex shifting into custom silicon, interconnect, or thermal management is still constructive for the broader semiconductor ecosystem over the next 6-18 months.

The bigger tradeable miss is the market’s tendency to conflate “technically imaginable” with “economically scalable.” That usually creates a local top in narrative-driven IPOs and adjacent comps, but it can also trigger a valuation reset across late-stage private AI infrastructure names if investors start demanding proof of unit economics. The catalyst path is fast: poor post-IPO commentary, weak disclosure on orbital economics, or a broader de-rating of moonshot capex stories could hit in days to weeks, while the supply-chain beneficiaries remain intact over months.

Key Decisions for Investors

• Short the IPO / newly listed name on first tradable liquidity if the opening valuation remains >20x forward revenue equivalent; cover on any post-deal-day selloff of 15-20% as narrative air comes out.
• Buy NVDA and INTC on 3-6 month weakness only if the market sells the article as a semiconductor-negative event; the better risk/reward is that compute scarcity remains intact and any orbital compute is too remote to matter near-term.
• Pair trade: long SMR/CEG/NEE or data-center power infrastructure proxies vs short speculative AI-in-orbit narrative names for a 3-12 month horizon; thesis is that the real capex goes into power and cooling, not launch-dependent compute.
• If options are liquid, use put spreads on the IPO name for the first 30-60 days post-listing; the setup favors implied-volatility compression once the market digests the execution gap.
• Avoid chasing any private-market AI infrastructure SPVs tied to exotic energy/space claims for now; require at least two quarters of disclosed unit economics before underwriting the valuation.