This Semiconductor Stock Could Be a Better Way to Invest in AI and Satellites Than SpaceX

STMicroelectronics is positioned to benefit from SpaceX-driven AI and satellite expansion, with management expecting data-center revenue to reach $1B in 2026 and more than double in 2027. The company also forecasts $3B of SpaceX-related satellite revenue during 2026-2028, versus total revenue of $11.8B last year. STMicro guided for $18B in revenue by 2028, while EPS is expected to nearly triple from this year through 2028; the shares trade at roughly 37x forward earnings, though semiconductor cyclicality remains a material risk.

Analysis

The STM thesis hinges less on headline AI exposure than on whether optical interconnect and power-management content can scale before its legacy industrial/automotive recovery fully materializes. If management converts its stated data-center trajectory, that business becomes large enough to alter the market’s view of STM from a cyclical auto/industrial semiconductor supplier to a mixed-cycle connectivity and power franchise. The valuation re-rating opportunity is therefore real, but only if incremental revenue carries materially higher gross margin than the company’s current mix.

The key non-obvious risk is customer concentration disguised as end-market diversification. A large satellite-related revenue target tied to one launch/customer ecosystem creates procurement, qualification, launch cadence, and pricing-power risk; it should not receive the same multiple as broadly distributed cloud capex. Competitive leakage is also plausible: Coherent (COHR), Lumentum (LITE), Marvell (MRVL), and Broadcom (AVGO) capture more direct optical-connectivity upside, while Infineon (IFX) and onsemi (ON) compete for power content. STM needs design-win evidence and backlog conversion, not aspirational end-market TAM, to justify a sustained premium.

Over the next 1-3 months, the stock will likely trade primarily on utilization, auto/industrial inventory normalization, and gross-margin guidance rather than space/AI narratives. Over 6-18 months, disclosed data-center revenue, satellite program order visibility, and factory-loading recovery could drive earnings-estimate revisions. Contrarian view: the market may be underpricing the earnings torque from utilization recovery, but is likely overpricing any revenue figure that remains dependent on a single private customer without contractual volume, pricing, and margin disclosure.

Key Decisions for Investors

• Watch-list STM for a long entry only after the next earnings release confirms sequential gross-margin expansion and data-center revenue/order growth; target a 6-12 month position sized for a cyclical recovery, with thesis invalidated by a further cut to 2027 revenue or utilization outlook.
• Prefer a pair trade long STM / short IFX for 6-12 months if STM demonstrates data-center revenue conversion: STM has more potential mix-shift upside, while IFX remains more exposed to a conventional automotive/industrial normalization. Exit if STM’s data-center backlog fails to expand for two consecutive quarters.
• Do not underwrite satellite-related revenue at full value until STM discloses customer concentration, committed purchase volumes, and segment margin. Treat any announcement of a multi-year supply agreement or capacity reservation as the near-term catalyst for upgrading the position.
• For direct AI-networking exposure, retain NVDA/AVGO/MRVL as cleaner beneficiaries; STM is a higher-beta satellite-plus-power adjunct rather than a substitute. Avoid chasing STM solely on SpaceX-linked commentary absent independently verifiable order data.