Elon Musk's SpaceX applies to launch 1m satellites into orbit

SpaceX has filed with the FCC to deploy up to one million solar-powered low-Earth-orbit satellites (operating roughly 500–2,000 km altitude) as “orbital data centres” to provide AI compute capacity, significantly expanding beyond its existing ~10,000-strong Starlink constellation. The company argues the network would be a greener, scalable solution to meet accelerating AI processing demand and serve billions globally, but offers no timeline and faces technical, regulatory and operational risks including high launch costs, cooling/power infrastructure, space-debris and astronomical interference — factors that make the plan transformative if realized but highly uncertain for investors.

Analysis

SpaceX's million-satellite proposal shifts potential winners toward launch and small‑sat supply chains, radiation‑hardened components and AI‑accelerator makers (beneficiaries: RKLB, LHX, RTX, NVDA) while posing long‑term competitive pressure on GEO sat operators and parts of the data‑center REIT complex (EQIX, DLR). If credible, the plan implies a multi‑year increase in launch demand (order‑of‑magnitude uplift: 5x–10x launches over a decade in a base case) that will compress per‑unit launch economics and force consolidation among small‑sat OEMs and insurers.

Tail risks are regulatory blocks (FCC/ITU/European restrictions), a Kessler‑type debris cascade or major collision that could create >$5–$20bn in industry losses; these are low‑probability but systemically disruptive events. Immediate market impact will be muted (days–weeks) while technical and permitting work continues; the material re‑rating of suppliers versus incumbents is a 2–7 year story driven by capex, insurance capacity and spectrum allocation.

Practically, this favors exposure to launch integrators, rad‑hard component suppliers and select AI hardware vendors with 6–24 month trade horizons while cautioning against pure GEO operators and marginal data‑centre landlords until regulatory clarity emerges. Options and structured call spreads provide asymmetry to capture upside from contract/manifest wins without full equity exposure, and credit spreads on high‑capex aerospace issuers may widen near news flow.

Consensus underestimates the non‑linear costs of operations in LEO (power, cooling, thermal cycling, collision avoidance): even halving launch cost per kg may leave a million‑sat fleet marginal economically, so upside may be over‑priced. Historical parallels (satcom hype cycles) show infrastructure suppliers—not incumbent operators—capture most long‑term value. Unintended consequences include astronomer/regulatory backlash, stricter spectrum/visibility rules, and insurance capacity shocks that could strand inventory and reprice the supply chain.

Key Decisions for Investors

• Establish a 2–3% portfolio long in Rocket Lab (RKLB) via a 9–12 month call spread (buy near‑ATM calls, sell ~30% OTM) within 3 months to capture increased launch cadence; target 40–80% upside, cut the trade if RKLB misses two consecutive manifest targets or option premium falls >50%.
• Add a 1–2% position in NVIDIA (NVDA) via 6–12 month call spreads (buy ATM, sell 25–35% OTM) to play sustained AI compute demand tied to orbital + terrestrial growth; take profits at +30–50% or trim if implied volatility drops >30% within 3 months.
• Initiate a dollar‑neutral pair: long L3Harris (LHX) 1.5% weight, short Viasat (VSAT) 0.75% weight (6–24 month horizon) to express rad‑hard/component upside vs GEO incumbent erosion; stop‑loss: LHX down 15% or VSAT up 20%.
• Reduce combined exposure to data‑centre REITs Equinix (EQIX) and Digital Realty (DLR) by 1–3% over the next 3 months and rotate into aerospace/defense suppliers; explicitly monitor the FCC docket and any ESA/NOAA petitions—if the FCC signals authorization >100k satellites within 90 days, increase launch/supplier allocation by an additional 1–2%.