SpaceX Could Soon Be Worth $2 Trillion. This $66 Metric Is the One to Watch

SpaceX’s S-1 highlights that more than 60% of sales come from Starlink, where ARPU fell from $99 in 2023 to $66 in Q1 2026 and management expects further declines as international expansion accelerates. Offseting that, SpaceX said Connectivity operating income jumped 120% to $4.4 billion and adjusted EBITDA rose 86% to $7.1 billion in 2025, while Starlink customers surged 347% to 10.3 million. The article frames the upcoming IPO as potentially massive, but emphasizes execution risk around falling ARPU versus operational efficiencies.

Analysis

The important signal is not that SpaceX is scaling, but that it is choosing lower-ARPU geographies and still claiming margin expansion. That implies the business is increasingly a throughput machine: launch cadence, terminal manufacturing, and satellite deployment efficiency matter more than price optimization. If that math works, the real competitive moat is not broadband pricing power but the cost curve in manufacturing and launch logistics, which should pressure every LEO and satellite-adjacent competitor with weaker vertical integration.

Second-order, falling ARPU is actually a feature if customer acquisition cost and deployment cost drop faster than revenue per user. The market should be watching terminal subsidies, churn, and enterprise mix: consumer ARPU compression can be offset only if higher-value aviation, maritime, government, and backhaul contracts scale faster than residential mix deterioration. If enterprise penetration stalls, the story shifts from “growth at scale” to “share gains bought with margin dilution,” which is a much harsher underwriting framework for an IPO priced like a scarcity asset.

For public comps, the read-through is mixed: TSLA benefits only indirectly through Musk halo and launch-cost innovation, while NVDA and INTC are more relevant as potential beneficiaries of satellite/network compute demand over years rather than quarters. The contrarian miss is that the market may be over-anchoring on ARPU decline as a negative when the more material variable is unit economics per launched satellite and per terminal sold. If SpaceX’s cost-down claims are real, the IPO could still work even with falling ARPU; if they are not, the first place it shows up is slower launch-backed constellation expansion, not headline revenue.

Key Decisions for Investors

• Do not chase the IPO at any price: require a post-listing lockup-window reset or a 20%+ discount to modeled implied valuation before initiating exposure; at $1.75T-$2T, the margin of safety is thin if ARPU keeps slipping.
• For optionality, consider a long-dated call structure on TSLA only if the SpaceX listing prices exceptionally well; the asymmetric benefit is narrative/sentiment, but the trade should be sized as a momentum kicker, not a fundamentals bet.
• Pair idea: long NVDA / short INTC on the premise that satellite-network expansion and edge connectivity increase high-end networking/compute demand faster than legacy CPU share recovery; use a 6-12 month horizon and cap risk if launch cadence disappoints.
• Avoid shorting the IPO on ARPU alone; wait for evidence that customer growth or enterprise mix is rolling over. The better short trigger is a 2-quarter deceleration in subscriber adds or a failure of operating leverage to hold while ARPU falls.
• Set a watchpoint on SpaceX-adjacent public comps after pricing: if the deal prints rich and the aftermarket is strong, expect private-market multiples for frontier infrastructure names to re-rate upward for 1-2 months; if it fails, expect a quick compression in all mega-cap private tech marks.