Morgan Stanley initiates coverage on SpaceX: opening long-term space through AI infrastructure and low-orbit networks

Overview

This is Morgan Stanley’s first coverage of SpaceX. The report argues that SpaceX is combining near-monopoly launch capability, the world’s largest low-Earth-orbit satellite network, and a rapidly scaling AI infrastructure stack into a single platform, creating an opportunity to integrate orbital assets, global connectivity, and compute capacity into a next-generation AI infrastructure stack. In the base case, the report forecasts company revenue to rise from $44.9bn in 2026 to $318.6bn in 2030, reaching $3.342tn by 2040.

Core views

The core views are organized around four investment debates: first, whether Starship can significantly lower $/kg through rapid reusability and thereby make Starlink, mobile satellite, orbital compute, and space infrastructure markets more attractive; second, whether Starlink can achieve broad TAM penetration using V3 broadband satellites and Mobile Gen 2 satellites; third, whether SpaceXAI can leverage vertical integration, ground compute buildout speed, and eventual orbital compute to deliver industry-leading $/W and time-to-power; fourth, whether enterprise AI can transition from neocloud compute leasing to managed infrastructure, tools, and end-to-end enterprise applications. The overall long-term tone is constructive, but the report acknowledges near-term engineering, regulatory, financing, and execution risks.

Analysis framework

The report uses segment-level SOTP 15-year DCF as its primary valuation framework, applying an 11.1% WACC and a 50% discount to AI business value, with cross-checking against valuation multiples from high-growth large-cap technology peers. Forecasting is structured around four KPIs: revenue per watt, cost per watt, launch cost per kg to orbit, and the number of Starlink users or connectivity nodes. The model covers number of launches, share of internal missions, Starlink capacity, connectivity user penetration, mobile direct connectivity, ground and orbital compute deployment, enterprise AI contract pricing, and long-term margins.

Methodology notes

• Valuation methodsSOTP DCF

  15-year DCF by segment

  The report uses segment-level 15-year SOTP DCF as the valuation base for the target price, assumes 11.1% WACC, and applies a 50% discount to AI business value, then validates with multiples from high-growth large-cap technology peers.

• operating_modelKey Debate KPI Framework

  Four key KPIs mapped to four investment debates

  The report maps revenue per watt, cost per watt, launch cost per kg, and Starlink connectivity nodes to enterprise AI commercialization, compute cost and time-to-power, Starship economics, and broad Starlink TAM adoption.

• market_sizingTAM and Penetration Model

  Connectivity, mobile direct-connect, and robotics device penetration model

  The report estimates Starlink’s long-term connectivity revenue across home broadband, government and enterprise, mobile direct-connect, robotics, and embodied AI device scenarios, assuming capacity expansion drives down prices and raises penetration.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Space Exploration Technologies Corp. (SPCX.US)

  Core coverage target

  Strengths

  Has reusable launch capability, low-Earth-orbit satellite network scale, vertical integration strength, and potential to scale AI compute infrastructure.

  Weaknesses

  Very large capital expenditure, complex commercialization path, and multiple critical technologies not yet fully proven at commercial scale.

  Comparison

  The report cross-validates against valuation multiples of high-growth large-cap technology peers and notes that 2028 EV/EBIT/Growth is 0.41x, below the 25th percentile of comparable AI-enabled companies.

  Risks

  Starship reusability, orbital compute, direct-to-device connectivity, financing capacity, regulation, geopolitics, and key-person concentration risk could all affect valuation.

• Starlink

  Core of SpaceX connectivity business and long-term TAM expansion

  Strengths

  V3 broadband and Mobile Gen 2 satellites are expected to expand capacity, improve speed and latency, and support consumer, enterprise, government, and mobile segments.

  Weaknesses

  Capacity expansion must translate into actual penetration gains, and ARPU is expected to decline as the share of international markets rises.

  Comparison

  Compared with terrestrial networks, Starlink’s opportunity lies in serving devices and regions difficult for ground infrastructure to reach.

  Risks

  Spectrum, regulation, competition, user adoption, capacity utilization, and pricing pressure.

• SpaceXAI / xAI / Grok Enterprise

  AI compute and enterprise AI commercialization vehicle

  Strengths

  Fast ground compute buildout, deep vertical integration, and potentially lower long-term costs as orbital compute ramps.

  Weaknesses

  Near-term revenue is weighted toward neocloud leasing, while higher-value software and end-to-end enterprise services remain to be validated over time.

  Comparison

  The report views its ground infrastructure as about half the industry average cost per watt, with deployment speed about 6–8 times the industry.

  Risks

  Shifts in compute supply and demand, GPU and power supply constraints, customer concentration, AI model competition, orbital compute feasibility, and AI regulation.

Key data

• Rating and target priceOverweight; $300.00 target; $160.42 currentFirst-time coverage; implied upside around 87%.
• Revenue forecast2026e $44.9bn; 2030e $318.6bn; 2040e $3.342tnBase-case revenue model, with long-term growth primarily from connectivity and AI infrastructure.
• Free cash flow pressureFCF not expected to turn positive before 2035The report projects capital expenditure needs could reach about $300bn annually by 2031, with average annual external financing needs of about $84bn from 2027 to 2034.
• Starship launch costAbout $500/kg in 2030; below $200/kg in 2035; below $150/kg in 2040Cost declines are treated as a core assumption for expansion in Space, Connectivity, and AI.
• Number of Starship launches46 in 2027; 375 in 2030; 6,019 in 2040Between 2027 and 2040, over 75% to 90% of Starship launches are expected to be for internal missions.
• Connectivity revenue2025 $11.4bn; 2030 $120.6bn; 2040 $687.7bnGrowth is expected from Starlink broadband, enterprise, government, mobile direct connect, and robot connectivity.
• Starlink capacity expansionAbout 2.5x in 2027; about 19x in 2030; about 600x in 2040Based on increased saleable capacity through Starship and V3 satellites.
• AI compute deployment160 MW in orbital compute by 2028; 2.7 GW in 2030; 111 GW in 2035; 364 GW in 2040The report expects orbital compute to become the majority of total compute capacity in 2032.
• Ground AI infrastructure cost$4-$5/W, excluding chipsBelow industry average of about $9/W; deployment speed described as about 6–8x the industry.
• Scenario valuationBear case $75; base case $300; bull case $600The bull case gives greater value to orbital AI upside; the bear case reflects material failures in space programs.

Impact & implications

If the report’s assumptions play out, SpaceX’s investment narrative would evolve from a pure space or satellite-internet company into a combined infrastructure platform spanning launches, connectivity, compute, and enterprise AI applications. Starship cost reductions could amplify the economics of Starlink capacity, mobile direct-connect, and orbital compute; Starlink could become a data transport layer for areas beyond terrestrial network coverage; and the AI segment could benefit short-term from compute supply-demand tightness, while long-term value depends on converting low-cost compute into high-value enterprise services. For investors, the key is not linear growth in any single business, but whether these businesses form a mutually reinforcing infrastructure flywheel.

Risks

• Full Starship reusability, recovery, refurbishment costs, and high-cadence launch capability have not yet been validated at commercial scale.
• Orbital AI compute, direct-to-device connectivity, and large-scale AI infrastructure all carry significant execution and technology risk.
• High capital spending means FCF is not expected to turn positive before 2035; average annual external financing needs of about $84bn from 2027 to 2034 are a major risk.
• Regulatory and geopolitical factors such as launch rules, spectrum, orbital debris, export controls, AI regulation, cybersecurity, and anti-space threats may constrain business development.
• The company has key-person dependence on Elon Musk and may be affected by Tesla-related conflicts of interest or resource allocation.
• Space is hard; major Starship or on-orbit anomalies should be expected over time, and a major failure could trigger the bearish case.

What to watch

• SpaceX Starship Flight 13 in July 2026, Flight 14 by the end of Q3 2026, and the first commercial Starship payload in Q4 2026.
• Reliability of Super Heavy booster and ship recovery, reusability, refurbishment costs, turnaround time, and launch cadence.
• Starlink broadband and DTC user updates, capacity expansion, ARPU trends, and international market penetration.
• New neocloud deals, contract pricing from customers such as Anthropic and Google, and the degree of compute supply-demand tightness.
• Government contracts and connectivity agreements across aerospace, maritime, enterprise, mobile, and public-sector applications.
• Buildout speed and $/W cost of Terafab, Solarfab, COLOSSUS, and COLOSSUS II ground AI infrastructure.
• Cost curve, reliability, and commercialization demand for orbital compute after deployment begins in 2028.