Goldman Sachs Just Predicted SpaceX's AI Revenue Will Be This Number By 2030

SpaceX has begun marketing a planned IPO targeting at least $75 billion in proceeds at a $1.77 trillion valuation, potentially making it the largest IPO ever. Goldman Sachs is reportedly pitching aggressive 2030 revenue projections, including total revenue of $474 billion and AI-unit revenue of $322 billion, but the article questions their realism given the unit’s $6.4 billion loss in 2025 and more than $20 billion of capex. The piece is primarily about valuation and investor expectations rather than confirmed operating results.

Analysis

The setup is less about whether SpaceX can clear a headline valuation and more about how much incremental capital the IPO will siphon into adjacent winners and “picks-and-shovels” names. If the road show succeeds, the immediate beneficiaries are the transaction-adjacent balance-sheet providers and infrastructure suppliers, while the real loser is any public-market AI peer whose multiple is being benchmarked against an implausibly high long-dated revenue curve. In practice, that often compresses the private-to-public valuation gap for every AI-adjacent asset, but only at the price of making execution risk more visible over the next 6-18 months.

The biggest second-order effect is capital intensity. A narrative built on sovereign-AI vertical integration implies persistent burn before any operating leverage arrives, which means the market is effectively underwriting continuous external financing for several years. That dynamic can pressure return assumptions across the broader AI supply chain: compute, power, networking, and advanced packaging names may see more demand in the near term, but the longer the market leans into this model, the more it will scrutinize capex efficiency and cash conversion elsewhere.

For GS, the near-term catalyst is less trading revenue and more reputation risk: if investor pushback is strong, it will read as a vote against the most aggressive growth assumptions in the market, not just this deal. For TSLA and INTC, the linkage is more subtle: any real semiconductor-manufacturing partnership would be a multi-quarter positive for semiconductor tool/material ecosystems, but the market will likely discount it until orders become visible. The contrarian miss is that the market may be over-focusing on the unicorn math and underpricing how much of the story depends on access to power, wafers, and data-center capacity rather than app-level adoption.

Key Decisions for Investors

• Avoid chasing GS into the IPO headline; instead wait for indications of investor demand. If book quality is weak, fade GS on a 1-3 week horizon as the market reassesses underwriting credibility rather than fundamentals.
• Long a basket of AI infrastructure enablers on weakness, not the speculative platform story: prefer MSFT over any single-model narrative and use a 3-6 month horizon. Risk/reward is better because enterprise distribution and capex discipline matter more if AI monetization proves uneven.
• Pair trade: long semiconductor/industrial capex beneficiaries, short the most expensive AI narrative names. The setup works if the market starts valuing real revenue visibility over TAM-driven projections; target 2-4 months with asymmetric downside in the short leg.
• If TSLA is pressured by the 'sovereign AI' capital sink narrative, use any rally to buy downside protection rather than outright short. The risk is that Musk-driven hype can keep the stock elevated for weeks even if the fundamental case weakens.
• Watch INTC as a higher-beta beneficiary of any actual fabrication partnership, but only on confirmation of orders. Entering pre-confirmation is a low-conviction trade; the better expression is calls or call spreads over 6-12 months if the terafab narrative gains credibility.