SpaceX's First 100 Days Are in the Books. Here's the Report Card.

SpaceX closed at $154.72 on Sept. 22, roughly flat versus its $150 June 12 IPO opening price, after falling more than 50% peak-to-trough in July and subsequently recovering. Q2 revenue surged 92% to $7.8B and adjusted EBITDA nearly tripled to $3.5B following AI-compute rental deals with Anthropic and Google, while Starlink connectivity revenue rose 66% as subscribers doubled. However, SpaceX posted a $541M Q2 GAAP loss and remains valued at roughly $2T, leaving the author cautious despite orderly lockup expirations and potential Nasdaq-100 ETF buying support.

Analysis

The key underwriting issue is revenue quality, not the reported growth rate. Monetizing idle AI capacity can lift utilization and near-term EBITDA, but it should be valued closer to CRWV/NBIS infrastructure revenue than to recurring connectivity or launch economics: customer concentration, rapid GPU depreciation, power/network costs, and hyperscaler bargaining power all limit durability. If AI rentals become material enough to support the equity narrative, SPCX risks a multiple reset toward neocloud peers; if they remain immaterial, the apparent growth acceleration fades from comparisons within 1-3 quarters.

The more durable earnings lever is connectivity scale, where incremental subscriber additions should carry materially better contribution margins once satellite and ground-network fixed costs are absorbed. The market is likely underpricing the tension between this long-duration optionality and the capital intensity required to maintain constellation capacity, spectrum rights, launches, and terminal subsidies. Watch whether operating cash flow improves alongside EBITDA: persistent GAAP losses or rising capex per incremental subscriber would signal that reported profitability is being purchased through infrastructure spend rather than converted into equity value.

Technically, completed early unlocks reduce one overhang but do not establish durable institutional demand. The asserted QQQ-flow catalyst should be treated skeptically until Nasdaq-100 eligibility, index weight, and rebalance mechanics are independently confirmed; lockup expiry alone does not mechanically require passive buying. Near term, SPCX is vulnerable to any deceleration in AI-rental bookings or customer concentration disclosure, while a 6-18 month upside case requires connectivity ARPU, churn, and capex efficiency to validate a higher-quality recurring revenue mix.

Key Decisions for Investors

• Maintain SPCX as a watch, not a core long, until the next earnings release discloses AI-rental revenue concentration, contract duration, and capex required to serve it. A long becomes actionable only if connectivity growth translates into improving operating cash flow rather than EBITDA alone.
• Express relative valuation through long META / short SPCX over the next 1-3 months, sized beta-neutral. META offers AI monetization funded by a mature cash-generative advertising base, whereas SPCX retains execution and capital-intensity risk; cover the short if SPCX demonstrates sustained positive free cash flow or raises connectivity margin guidance.
• Avoid using anticipated QQQ flows as a standalone SPCX catalyst. Set an event-driven alert for confirmed Nasdaq-100 inclusion or a disclosed index-weight change; absent confirmation, treat any pre-rebalance rally as potential short-entry liquidity rather than fundamental demand.
• For AI-infrastructure exposure, prefer a selective CRWV/NBIS basket only after verifying contracted backlog, power availability, and financing terms; SPCX's capacity-rental activity may validate demand but also increases competitive supply. Reduce exposure if GPU lease pricing weakens or hyperscaler customers shift workloads in-house.