Bernstein initiates coverage of Space Exploration Technologies Corporation (SPCX.US), assigns Outperform, and sets a target price of 239 USD

Overview

This report is Bernstein's initiation coverage on Space Exploration Technologies Corporation (SPCX.US). Bernstein segments SpaceX into Space, Connectivity, and AI: Space is core capability and strategic moat, Connectivity is led by Starlink with high growth, and AI is positioned as the key long-term valuation driver via orbital data centers. The analysts believe that the key to a multi-trillion-dollar valuation is not traditional launch revenue, but whether SpaceX can deploy AI computing to orbit with Starship's large-scale low-cost launches and achieve a relative cost advantage versus ground data centers.

Core views

Core views include: first, SpaceX has established leading global orbital launch market status through Falcon 9 and reusable technology; if Starship achieves full two-stage reuse, it will further expand the lead. Second, orbital data centers are expected to save power, cooling, land, and maintenance costs, and with scaled deployment they could create several-hundred-billion-dollar EBITDA potential. Third, Starlink's consumer broadband, enterprise/government connectivity, and mobile-direct use cases have global expansion potential, but their scale alone is insufficient to independently support a multi-trillion-dollar valuation. Fourth, the investment conclusion is highly sensitive to Starship launch cadence, semiconductor supply, long-run AI compute demand, and regulatory conditions.

Analysis framework

The report uses a sum-of-the-parts valuation framework, with 2031 EBITDA as the core valuation input, and applies different EV/EBITDA multiples and discount rates to AI, Connectivity, and Space. The AI segment uses AI peer multiples, Connectivity uses forward-looking EV/EBITDA multiples above S&P, and Space uses aerospace and defense multiples. Because orbital data centers become the main profit source in later years, the report argues that valuing with earlier years would understate the eventual growth driver.

Methodology notes

• Valuation methodssum-of-the-parts

  Sum-of-the-parts valuation

  The report splits SpaceX into AI, Connectivity, and Space and values each segment using 2031 EBITDA with segment-specific multiples and discount rates, then aggregates them to an implied base valuation of about 3.1 trillion USD.

• Valuation methodsforward EV/EBITDA multiple

  Forward EV/EBITDA multiple

  The AI segment references NVDA, AVGO, ASML, CRWV and peers using 23x; Connectivity uses 15x; Space uses 13x.

• risk_adjustmentlate-stage venture discount rate

  Late-stage venture discount rate

  The report applies higher discount rates to segments still in high-growth and high-uncertainty phases: 35% for AI, 25% for Connectivity, and 10% for Space.

Asset mapping & comparison

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

• SPCX.US

  Core covered stock

  Strengths

  It has world-class reusable launch capability, a global Starlink connectivity network, and a potential pathway into AI compute infrastructure through orbital data centers.

  Weaknesses

  The business model is highly dependent on Starship launch cadence, full-reuse execution proof, and semiconductor supply, and may remain free-cash-flow-negative through 2031.

  Comparison

  Compared with traditional aerospace peers, SpaceX's Space segment resembles platform infrastructure; compared with ground AI data centers, the potential advantage of orbital data centers comes from potential savings in power, cooling, land, and maintenance costs.

  Risks

  Starship launch incidents, reuse failure, semiconductor bottlenecks, AI compute demand below expectations, regulatory constraints, and extreme solar activity.

• Starlink

  Core asset in the Connectivity segment

  Strengths

  It has global coverage, and consumer broadband as well as enterprise and government connectivity are expected to grow quickly, with relatively high customer satisfaction.

  Weaknesses

  It is still relatively more expensive and slower than wired broadband, and direct-to-phone offerings require ground network components, increasing commercialization uncertainty.

  Comparison

  The report argues that it should not be compared simply to traditional satellite operators or network carriers such as Verizon and T-Mobile.

  Risks

  Regulatory issues, spectrum, competition, ground-network partnerships, and uncertain unit economics.

• Starship

  Critical infrastructure for scaling AI and connectivity

  Strengths

  If full two-stage reuse is achieved, it will significantly reduce cost per kilogram to orbit and raise payload capacity to roughly four times that of Falcon 9.

  Weaknesses

  As of the report date, full reuse has not yet been fully validated, V3 launch history is short, and it requires sustaining an extremely high cadence of several thousand launches per year.

  Comparison

  The report sees peers including New Glenn, SLS, and China's Long March 10, but notes that none are close to SpaceX in reusable commercialization and launch scale.

  Risks

  Launch accidents, first-stage or upper-stage reuse failure, regulatory approval delays, and insufficient capacity expansion.

Key data

• RatingOutperformBernstein initiation coverage rating.
• Target Price239.00 USD12-month target price based on the base-case SOTP valuation.
• Base-case Valuation3.117 trillion USDExhibit 107 shows segment discounted valuation totaling about 3.117 trillion USD.
• 2031E Total Revenue554.3 billion USDThe report expects revenue to grow roughly 30 times versus 2025, with AI contributing about 60%.
• 2031E Adjusted EBITDA403.5 billion USDFinancial tables show 2031E adjusted EBITDA of about 403.5 billion USD.
• 2031E AI EBITDA284.2 billion USDThe SOTP table shows AI 2031 EBITDA of 284.224 billion USD; the text also cites AI EBITDA of about 305.0 billion USD.
• 2031E Starship Launches3,543The report expects about 3,543 Starship launches in 2031, with nearly 90% dedicated to orbital data centers.
• 2031E Compute Capacity37 GWThe report expects 37GW of total compute capacity in 2031, with 27GW being orbital compute.

Impact & implications

If full Starship reuse and orbital data centers are implemented as assumed in the report, SpaceX's valuation engine would shift from launch services and satellite communications to AI infrastructure, potentially reshaping the competitive landscape for AI compute, satellite communications, and commercial space. However, this thesis is highly execution-sensitive; in a three-year delay bear case, the target price is 160 USD, and in a severe case where orbital data centers fail and AI remains ground-only, the target price is 119 USD.

Risks

• Starship may fail to achieve rapid reuse or high-frequency launches, delaying orbital data center deployment.
• Starship launch accidents could affect regulatory approvals, capital market confidence, and commercialization progress.
• Semiconductor or other critical material shortages, especially where chips required for orbital data centers may require support from multiple new fabs.
• Long-term AI compute demand may underperform expectations, weakening orbital data center revenue and profit assumptions.
• Launch frequency, satellite deployment, spectrum, or antitrust regulatory constraints may raise expansion complexity.
• Extreme solar events could significantly disrupt satellites, orbital data centers, and communication infrastructure.
• High capital expenditure before 2031 may keep free cash flow negative.

What to watch

• Subsequent Starship launches, especially the 13th, 14th, and 15th flights referenced in the report, and full-reuse validation.
• Whether Starship V3 launch cadence approaches the annual multi-thousand-launch level required for orbital data center deployment.
• Whether Terafab and third-party semiconductor supply plans can support AI compute scaling.
• Whether orbital data center cost, cooling, radiation shielding, laser communications, and latency performance match report assumptions.
• User growth and margins in Starlink consumer broadband, enterprise, and government connectivity.
• Whether direct-to-mobile can be commercialized through MVNO or required ground network components.
• How regulators' stance shifts on launch frequency, satellite deployment, and market dominance.