SpaceX Is 'Cheap and Getting Cheaper' Morgan Stanley Says

Morgan Stanley reiterated an Overweight rating and $300 price target on SpaceX at about $159, arguing its 2028 growth-adjusted EV/EBIT multiple of 0.3x is roughly 40% below the 0.5x median for mega-cap AI peers. The analyst sees AI valued at only $32 per share by the market versus $165 in the firm’s sum-of-the-parts model, while compute pricing could materially alter revenue: each additional $10 per watt would add more than $40 billion to consensus revenue. Shares rose 7.35% Friday to $158.96 and were up 3.6% to $164.63 Monday morning; upcoming catalysts include Starship Flight 15 and third-quarter results. Morgan Stanley disclosed it co-managed a SpaceX offering in the past year and received investment banking fees.

Analysis

The bullish case rests less on a conventional launch-business re-rating than on an unproven conversion of compute capacity into high-priced, utilized, cash-generative AI infrastructure. The key diligence gap is not just contracted $/watt: verify delivered capacity, utilization, contract duration, customer concentration, power and networking constraints, and the capex/depreciation required to earn those revenues. A headline revenue sensitivity is not an earnings sensitivity. If pricing falls as capacity scales or deployment slips, the growth-adjusted multiple can deteriorate quickly; comparisons with mature mega-cap platforms also depend heavily on distant estimates and are not directly comparable.

Near term, a successful Starship milestone could improve sentiment and the perceived cost curve, but a ship catch is not itself proof of commercial economics. Earnings provide a more useful test if they disclose AI revenue quality and spending. Over 6–18 months, power availability, compute utilization, and repeat customer demand matter more than model-release cadence. NVIDIA could benefit from incremental accelerator demand, while existing neocloud providers face potential pricing pressure if SpaceX brings substantial supply online. There is no clear basis here for a directional AMZN, GOOG, or META trade.

Positioning evidence from one analyst’s 40-client poll is weak, not a reliable measure of broad ownership or short interest. Morgan Stanley’s prior banking relationship warrants discounting its valuation conclusion. Overall, the setup is positively skewed only if compute economics validate; avoid treating the analyst’s sum-of-parts value as independently established.

Key Decisions for Investors

• SPCX: Favor a small, staged position over chasing the sharp reaction; add only on confirmation of deployed capacity, realized pricing, utilization, and manageable investment needs. Treat Starship Flight 15 as a sentiment catalyst, not the fundamental thesis.
• For event exposure, consider a defined-risk call spread rather than outright near-dated calls; size for a failed test or delayed earnings disclosure. Avoid specifying strikes until liquidity and implied volatility are checked.
• Monitor the next earnings disclosure for AI revenue and contract economics, plus capex, power constraints, customer concentration, and guidance. A revenue beat without evidence of utilization or cash conversion would not validate the valuation argument.
• Falsification: reduce or exit the thesis if deployment or customer uptake stalls, realized compute pricing materially undercuts the cited contract range, investment requirements rise without revenue conversion, or a severe Starship setback undermines the expected cost curve.