SpaceX (SPCX) Q3 2026 Earnings Call Transcript

SpaceX reported Q3 2026 revenue of $14.6B, up 161% year over year, and net income of $4.4B versus a $1.1B loss a year earlier; adjusted EBITDA rose 306% to $8.8B. Management guided to Q4 capex of $35B–$40B and Q1 2027 revenue of $33B–$37B, while targeting 2.5–3 gigawatts of compute capacity by year-end and projecting $500B–$1T in ARR by end-2027. The company cited $63B in backlog, with $42B expected to be recognized within 12 months, but highlighted compute infrastructure constraints and uncertainty around Flight 15 and a possible SpaceX–Tesla transaction.

Analysis

The market is likely to price this as a compute-scarcity winner, but the important distinction is between installed nameplate capacity and monetizable, powered capacity. If customers are on short-duration, spot-linked contracts, SpaceX retains upside when compute prices rise—but also carries utilization and repricing risk if the market loosens. The capex ramp makes cash conversion, customer concentration, and funding terms more important than adjusted EBITDA; verify operating cash flow, debt issuance, and committed-versus-optional capex before assigning value to management’s return claims.

There are material source-quality flags: the supplied takeaways and transcript give sharply different GPU deployment rates, while the ARR trajectory is exceptionally aggressive relative to the near-term revenue outlook. Treat both as unverified until reconciled with filings and customer/contract disclosures. A miss in powered capacity, utilization, or gross margin would expose the equity to multiple compression even if demand remains strong.

Second-order, near term: accelerated infrastructure orders could support ASML, TSM, and LRCX, but no Terafab equipment orders are confirmed; do not capitalize a 2028 fab ambition into current earnings. Longer term, successful in-house chip production could reduce external sourcing and complicate the supplier upside. Starlink’s airline proposition may aid differentiation for Copa Holdings and Lufthansa, but installation economics and incremental passenger revenue are undisclosed. The possible Tesla combination is not a catalyst until committees agree terms and shareholder approval is secured.

Contrarian view: the bullish case may be underweight execution and capital intensity, while overvaluing headline EBITDA as distributable cash. The immediate upside can persist on momentum, but the 1–3 month test is capex-to-cash conversion; the 6–18 month test is whether powered capacity, utilization, and realized pricing scale together.

Key Decisions for Investors

• Do not chase SPCX solely on the reported growth or compute targets. First reconcile the GPU deployment discrepancy against the earnings release/SEC filing and confirm cash flow, net debt, customer concentration, and contract duration.
• Set a conditional SPCX entry watch: reconsider a long only after verified reporting shows new compute capacity is energized and generating revenue without a material deterioration in gross margins or balance-sheet metrics. Falsify the thesis on a guidance cut, persistent power delays, or capex rising faster than operating cash generation.
• Keep ASML, TSM, and LRCX on a supplier watchlist, not as an earnings trade: seek confirmed Terafab tool orders or disclosed procurement commitments before attributing meaningful incremental demand; monitor whether eventual insourcing threatens external chip demand.
• Treat the TSLA combination as event risk, not base-case value. Reassess only on agreed transaction terms and a credible path to the required shareholder vote; absent that, avoid merger-arbitrage positioning.