Goldman Sachs Just Made a Bold Statement About SpaceX. Here's Why I'm Not Buying It

Goldman Sachs projects SpaceX's AI division could generate $322 billion in revenue by 2030, up from $3.2 billion in 2025, implying 151% average annual growth. The bank also estimates total SpaceX revenue could reach $474 billion by 2030 versus $18.7 billion last year, but the article argues these forecasts may be used to support a lofty IPO valuation. SpaceX plans to price shares at $135, raise $75 billion, and value the company at about $1.77 trillion.

Analysis

The market is being asked to underwrite two narratives at once: a frontier industrial platform and an AI software compounder, both priced as if execution risk is near zero. The most important second-order effect is not whether the operating business can grow fast, but whether the capital structure and index float dynamics create self-reinforcing demand after pricing; that can matter more for the first 30-90 trading days than any fundamental forecast. Goldman’s involvement is a credibility amplifier, but also a reputational overhang if the post-IPO multiple compresses quickly.

The AI revenue assumptions imply an extremely steep ramp that would require not just customer adoption, but sustained access to frontier compute, model training economics, and monetization at enterprise scale. That creates indirect winners elsewhere in the AI stack: any company selling picks-and-shovels to compute, networking, memory, power, and cooling can benefit if investors rotate from "one winner" narratives into broader infrastructure exposure. By contrast, if the market starts to doubt the pace of AI monetization, the first read-through is likely to be multiple compression in high-beta AI names rather than any immediate hit to underlying operating results.

For GS, the immediate trade is less about exposure to SpaceX fundamentals and more about underwriting economics and reputation. A successful IPO could modestly support fee narratives and IB franchise optics, but if the deal is perceived as promotional, it risks feeding skepticism around future large-cap tech offerings and could weigh on sentiment toward lead managers in follow-on deals. The catalyst window is days to weeks around pricing and the first lockup/secondary headlines; the real fundamental test is 12-24 months out when actual revenue realization must converge with the launch story.

Key Decisions for Investors

• Avoid chasing GS into the IPO date; use any strength in the next 1-2 weeks to fade the underwriting halo via a tactical short or underweight, with a 3-5% downside target if the deal prices aggressively and then trades down on valuation skepticism.
• Express a relative-value long in the broader AI infrastructure basket versus speculative AI platform stories: long NVDA and/or MSFT on 3-12 month horizon, funded by trimming exposure to names whose valuation depends on long-dated monetization assumptions; risk/reward favors established cash conversion over narrative optionality.
• If there is a tradable IPO pop, consider a short-term put spread on the new listing after the first few sessions, targeting 1-3 months post-deal when the market typically re-prices exuberant growth assumptions; structure for limited premium outlay and defined risk.
• For GS holders, use call overwriting into the event if positioning is already long-only: upside to underwriting fee headlines is likely smaller than the reputational risk if the market questions the valuation, making capped upside more attractive than open-ended exposure.
• Watch for a rotation into semiconductor and AI infrastructure suppliers on any disappointment in the IPO narrative; if the market starts distinguishing "AI monetization" from "AI picks-and-shovels," add NVDA on weakness rather than the speculative issuer.