4 Lessons for Future OpenAI and Anthropic Investors, Following SpaceX's Historic IPO

The article uses SpaceX’s June 12 IPO as a cautionary template for the upcoming OpenAI and Anthropic IPOs, highlighting frothy valuation risk: SpaceX priced at $135 and peaked at $225.64 before trading around $160, with valuation at ~95x 2025 revenue ($18.7B) and ~112x last year’s sales. It argues “profits matter” because Anthropic’s operating losses are narrowing and it expects its first adjusted operating profit this year, while OpenAI is described as posting a $38.5B net loss on $13.1B revenue in calendar 2025. Finally, it notes both OpenAI and Anthropic plan to float only ~5%-10% of shares versus SpaceX’s ~4%, implying potentially volatile debuts dominated by retail/increased hype rather than institutional steering.

Analysis

The immediate market effect is mostly about supply, not fundamentals. When a deal floats only a sliver of equity, the first-week tape is driven by scarcity and retail momentum, which can create a violent mark-up that later reverses when secondary supply, insider selling, and SBC dilution become visible. That makes the real risk-reward unattractive for late buyers even if the underlying business quality is strong.

The bigger dispersion is between profitable and cash-burning AI franchises. A company with a credible path to operating leverage deserves a structural premium, but that premium can vanish quickly if growth decelerates or if management uses the public currency to fund acquisitions and compensation. For public-market AI infrastructure names like NVDA, the spillover is more positive than negative: hot IPOs can reinforce the “AI spend is still accelerating” narrative, but only for as long as capex budgets stay intact.

Contrarian view: the consensus is likely underestimating how fast the story shifts from narrative scarcity to dilution math. The first 1-2 earnings prints after listing, plus any post-IPO insider selling or expanded float, will matter far more than launch-day valuation debates. NDAQ may get a modest volume/issuance tailwind, but this is not a durable earnings step-up; the more durable winners are the picks-and-shovels providers, while the most vulnerable assets are the richly priced IPOs themselves.

Key Decisions for Investors

• Stand aside on SPCX and any OpenAI/Anthropic-like IPO proxy for the first 2-4 weeks post-listing; fade strength only after the market has to reprice dilution, with a hard stop above the debut-day high.
• Relative-value long NVDA vs short late-stage AI narrative exposure (OTAI proxy) over 3-6 months if IPO enthusiasm peaks; thesis is that infrastructure monetizes AI spend with less balance-sheet risk and less float overhang.
• Treat NDAQ as a low-conviction, tactical beneficiary only if the IPO calendar broadens beyond a one-off headline; otherwise, the fee/volume boost is too small to matter.
• Set an alert for any IPO filing that shows a float above 10% or retail allocation above 20%; that is usually a signal of distribution pressure rather than pure institutional demand.