News /

2026-06-07

The S&P 500 index committee has declined to waive profitability rules for SpaceX, a decision that also forecloses near-term entry for major AI firms.

S&P 500 Rejects SpaceX, Closing a Path for OpenAI and Anthropic

The S&P 500 index committee has declined to grant SpaceX an exception to its profitability requirements, a decision that carries direct consequences for the AI sector. OpenAI and Anthropic — both of which have been pursuing or considering public market entry — now face the same structural barrier, as neither company currently meets the sustained profitability thresholds required for S&P 500 inclusion.

The ruling ends a period of speculation about whether index managers might accommodate the new class of high-valuation, cash-intensive technology companies that dominate private markets but operate at persistent losses. The answer, at least for now, is that they will not.

The S&P 500's eligibility criteria require a company to report positive as-reported earnings in its most recent quarter and positive cumulative earnings over the four most recent quarters. SpaceX, despite its scale and strategic significance, does not meet that standard. The committee's refusal to waive the rule signals that market capitalization and strategic importance are not sufficient substitutes for financial performance under the index's current framework.

For AI companies, this matters beyond symbolism. S&P 500 inclusion triggers mandatory purchases by index funds and ETFs that track the benchmark, creating a substantial and predictable demand event for a company's stock. It also functions as a legitimacy signal for institutional allocators operating under mandates that reference the index. Being excluded from that mechanism means foregoing one of the most reliable post-IPO capital inflows available in public markets.

OpenAI's anticipated transition to a for-profit structure and any eventual IPO had been viewed by some analysts as a potential fast-track to index inclusion given the company's valuation, which has exceeded $300 billion in recent private funding rounds. Anthropic, similarly, has raised capital at valuations that would place it among the larger constituents of the index if it were public and eligible. Neither company is profitable at an operating level in any conventional sense — both are running substantial compute and staffing costs against revenue that, while growing, has not converted to sustained net income.

The broader implication is that the AI sector's most prominent private companies exist in a valuation category that the public market's most important benchmark was not designed to absorb. The S&P 500's profitability rule was constructed to filter out speculative listings, but it now also filters out companies that are deliberately deferring profitability to accelerate infrastructure buildout and market capture — a strategy that is rational at scale but structurally incompatible with index eligibility.

This creates a two-tier dynamic in institutional capital. Investors who want exposure to frontier AI development at the infrastructure layer — the companies actually training and deploying the most capable models — cannot access that exposure through the dominant passive investment vehicle. They are instead limited to adjacent plays: cloud providers, chip manufacturers, and enterprise software companies that benefit from AI adoption but are not themselves the primary operators of it.

The S&P 500 committee could revisit its criteria. Index methodology is not static, and there is historical precedent for rules being adjusted to reflect the composition of the modern economy. But the committee's explicit refusal to grant an exception, rather than simply declining to act, suggests the current framework will hold for the near term.

For AI companies evaluating IPO timelines, the calculus now includes not just when public markets will value them fairly, but how long it will take to achieve the earnings profile that makes index inclusion possible. That timeline, for companies still scaling compute infrastructure at the rate of OpenAI and Anthropic, is measured in years, not quarters.

Sources: — Ars Technica (https://arstechnica.com/tech-policy/2026/06/sp-500-blocks-fast-spacex-entry-wont-waive-rule-for-unprofitable-ai-firms/)