Bernstein reiterates Outperform on SpaceX, focusing on Starship reusability and the orbital data center pathway ahead of its first earnings report

Overview

This report reviews the major investor debates surrounding SpaceX's fundamentals following its IPO and presents Bernstein's view ahead of the first earnings release. It emphasizes that short-term quarterly results should not be the investment focus; what truly matters is whether the company can convince the market that its orbital data center and Starlink connectivity businesses have long-term expansion pathways.

Core views

Bernstein's core view is that SpaceX's multitrillion-dollar valuation opportunity is primarily a question of whether it can be achieved, rather than when it will be realized in the short term. The most critical variable is whether Starship can achieve rapid, full-stage reusability, because the model requires approximately 3,600 Starship launches in 2031, implying an almost daily launch cadence. Semiconductor supply, FAA/FCC and global regulation, and the sustainability of AI computing demand are the other three key issues determining whether the growth model can work. Chinese competition, Starlink broadband penetration, and the direct-to-device business are legitimate areas of debate, but the report considers their impact on the overall value proposition relatively secondary.

Analysis framework

Based on frequently raised questions in post-IPO investor communications, the report divides the issues into core questions that determine whether the growth model succeeds or fails, reasonable debates that do not change the value proposition, and issues that the authors believe are technically solvable. The valuation uses a sum-of-the-parts approach based on segment EBITDA in 2031, applying late-stage venture-capital discount rates of 25% and 35% to Connectivity and AI, respectively, while assuming a one-year delay in orbital data center launches.

Methodology notes

• Valuation methodsSum-of-the-parts valuation

  SOTP valuation based on EBITDA for each business in 2031

  Bernstein values SpaceX's Space, Connectivity, and AI businesses separately and applies late-stage venture-capital-style discount rates to different businesses to reflect high growth and high execution risk.

• Technical feasibilityHeat-transfer physics analysis

  The Stefan-Boltzmann law and orbital thermal balance

  The report evaluates the cooling capacity of orbital AI data center satellites using radiative cooling, sun-synchronous dawn-dusk orbits, and dedicated radiator area, concluding that a computing load exceeding 100 kW is physically feasible.

• Scenario analysisInvestor debate stratification

  Classifying debates into core success-or-failure factors, secondary reasonable debates, and technically understood issues

  This framework distinguishes variables that genuinely affect valuation from short-term market noise, helping investors focus on Starship reusability, capacity, regulation, and computing demand.

• Supply-chain analysisSemiconductor capacity estimation

  Inferring wafer capacity and fab requirements from orbital data center capacity

  Assuming approximately 20 GW of new orbital computing capacity in 2031 and a computing load of approximately 120 kW per satellite, the report estimates that SpaceX may require annual wafer capacity of approximately 2.8 million wafers and around five dedicated fabs.

Asset mapping & comparison

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

• SPCX.US

  Core covered security

  Strengths

  Starship, Starlink, orbital data centers, and a government/enterprise customer base form a long-term growth narrative; the target price indicates significant upside potential.

  Weaknesses

  The valuation depends heavily on rapid Starship reusability and orbital data center scaling that have not yet been fully validated.

  Comparison

  Compared with Chinese aerospace competitors, SpaceX remains clearly ahead in reusable rockets and launch cadence; compared with terrestrial data centers, orbital data centers may have lower potential costs but greater execution complexity.

  Risks

  Regulation, semiconductor supply, the AI computing cycle, space debris, capital expenditures, and share-price volatility.

• Starlink

  Core asset of SpaceX's Connectivity business

  Strengths

  Global network, low marginal expansion costs, high customer satisfaction, enterprise and government customer coverage, and a constellation that is currently significantly larger than those of competitors.

  Weaknesses

  International expansion may lead to lower ARPU, while penetration in cities and areas covered by fixed-line networks remains debated.

  Comparison

  It has advantages over terrestrial broadband in rural, maritime, aviation, emergency, and government applications, while creating competitive pressure on U.S. telecom operators.

  Risks

  Telecom regulation in various countries, spectrum approvals, pricing pressure, and terrestrial broadband competition.

• Orbital data centers/AI business

  Core source of long-term valuation upside

  Strengths

  If AI inference demand continues to grow, orbital data centers may provide low-cost computing capacity with high solar utilization.

  Weaknesses

  The business depends on large-scale launches, satellite manufacturing, semiconductor supply, and in-orbit operations and remains at an early stage commercially.

  Comparison

  The report believes training is better suited to large terrestrial data centers, while inference is better suited to orbital data centers; low-latency co-location needs such as high-frequency trading still favor terrestrial facilities.

  Risks

  An AI computing bubble, insufficient capacity, technical and engineering delays, and regulatory uncertainty.

Key data

• Rating and target priceOutperform; 239.00 USDCorresponding to the July 30, 2026 closing price of 112.20 USD and implying 113% upside.
• Market capitalization and enterprise valueMarket capitalization of 1,476.46 billion USD; EV of 1,490.10 billion USDBased on the figures disclosed in the table, dated July 30, 2026.
• First earnings dateAfter market close on August 4, 2026The report considers the quarterly results themselves unimportant; management's confidence in the growth path is more important.
• 2031 Starship launch assumptionApproximately 3,600 launchesAchieving this target requires full-stage rapid Starship reusability and an almost daily launch cadence.
• 2031 orbital computing capacity assumptionApproximately 20 GW of new capacityBased on a computing load of approximately 120 kW per AI1-specification satellite, requiring the launch of approximately 170,000 satellites.
• Semiconductor capacity requirementApproximately 2.8 million wafers of annual wafer capacity; approximately five dedicated fabs; costs potentially exceeding USD 160 billionThe report considers the scale very large but not impossible; the timeline and supply-chain execution are key risks.
• Starlink satellite networkApproximately 9,600 satellites currently, with planned capacity increasing by more than 10 timesThe report believes its global network enables low-marginal-cost expansion and pricing flexibility.
• Starlink broadband latencyApproximately 23 ms round tripThe report believes this is not a major issue for most gaming and consumer applications.
• Orbital data center latencyApproximately 2 ms at an altitude of 600-800 kmThe report believes this is not an issue for most inference applications, while training remains better suited to terrestrial data centers.
• Chinese competitionChina has submitted plans to the ITU for more than 200,000 low-Earth-orbit satellitesThe report considers China the most serious long-term competitor, but estimates it is approximately ten years behind SpaceX in reusable rockets and launch cadence.
• 2031E total revenue forecastUSD 554,251 millionSourced from the financial forecast table in the appendix; the table extraction contains some noise, but the direction indicates substantial revenue expansion.

Impact & implications

For SPCX, the report supports a long-term bullish view but emphasizes extremely high execution risk. If the Starship heat shield, recovery and reusability, and launch-site expansion make substantial progress, the market may increase its confidence in the orbital data center pathway. If semiconductor capacity, regulatory approvals, or AI computing demand slow, the valuation foundation would come under pressure. Starlink's consumer, enterprise, and government businesses are viewed as high-growth and already profitable sources of support, while the direct-to-device business is considered unclear and not a core investment theme. The report also suggests that Starlink poses competitive pressure on U.S. telecom operators.

Risks

• Rapid reusability of the Starship V3 booster and upper-stage Starship has not yet been fully demonstrated, and heat-shield damage assessment remains critical.
• The model's approximately 3,600 launches in 2031 require a substantial simultaneous increase in the number of launch sites, regulatory approvals, and operating cadence.
• Orbital data centers may require enormous semiconductor capacity and capacity investment exceeding USD 160 billion, creating significant supply-chain execution challenges.
• FAA, FCC, national telecom regulations, and data-sovereignty rules may constrain large-scale launches, satellite communications, and orbital data center deployment.
• If AI computing demand experiences cyclical oversupply or a bubble burst, orbital data center revenue assumptions would be affected.
• China is a serious long-term competitor and may narrow its gap with SpaceX with state support.
• The path for Starlink's direct-to-device mobile business is unclear, with uncertainty surrounding both MNO partnerships and the construction of its own terrestrial network.
• Space debris and micrometeorite impacts are real risks, particularly for large solar panels and radiators.
• High capital expenditures and cash-flow pressure may create financing risks.
• The share price has been highly volatile following the IPO, and short-term market sentiment may diverge from long-term fundamentals.

What to watch

• Whether management demonstrates high confidence in the growth path following the first earnings release on August 4, 2026.
• Whether Launch 14 attempts and successfully completes second-stage Starship "chopsticks" recovery.
• The results of the Starship heat-shield damage assessment after Launch 13.
• Progress on the selection of additional launch sites and launch approvals, particularly FAA environmental reviews and flight-path approvals.
• FCC approval progress for the large Starlink V3 constellation and AI data center constellation.
• Whether SpaceX discloses semiconductor capacity, fabs, or supply arrangements with existing foundry or memory manufacturers.
• Starlink's global subscriber growth, ARPU changes, and enterprise/government customer expansion.
• Whether AI inference demand continues to support large-scale, low-cost computing deployments.
• Whether direct-to-device secures U.S. or overseas MNO partners.
• Progress in the reusability of China's Long March 10B and the deployment pace of its low-Earth-orbit constellation.