Prediction: This Upcoming IPO Will Be Even Bigger Than SpaceX

Anthropic, the maker of Claude AI, is reportedly targeting a late-2026 IPO valuation above $2.0T (with some forecasts of $2.2T+) and a timing around October. Revenue growth is cited as accelerating from a $47B annualized run rate in mid-May to $65B by end-July, with insiders expecting $100B–$120B by year-end, alongside a reported $65B May funding round at a $965B valuation (up from a $380B valuation in February). The article suggests the IPO could raise at least $150B, potentially breaking SpaceX’s IPO valuation and capital-raise records, with the main uncertainty being profitability and audited financial details.

Analysis

This is less a one-name IPO story than a signal that frontier-model economics are still being financed at extreme duration. If a private AI lab can command that kind of valuation while still leaning on huge external compute spend, the market is effectively saying the winners are the infrastructure toll collectors, not necessarily the model layer itself. That is constructive for NVDA and, second-order, for the networking/optics stack (AVGO, ANET), but it is not automatically bullish for every AI software name — capital tends to concentrate at the scarce layer, and smaller application vendors can get multiple compression if they cannot prove moat or pricing power.

Near term, the trade is mostly sentiment-driven until audited filings disclose revenue quality, customer concentration, and how much of the run-rate is usage versus contractual. If the filing confirms sustained hypergrowth and a credible margin path, AI infrastructure can re-rate for 1-3 months; if not, the headline valuation becomes a liability because the market will infer that growth is being bought with ever-higher capex and dilution. The key falsifier is any sign the revenue run-rate stalls below the implied trajectory or gross margin deteriorates as compute intensity rises.

Contrarian view: the consensus is fixated on the valuation record, but the more important variable is financing need. A very large IPO raise can be read as strength, yet it also means the company still needs external capital to fund compute, which is structurally negative for long-duration equity returns even when headline growth looks spectacular. Over 6-18 months, the more durable implication is a stronger AI oligopoly: hyperscalers and GPU vendors gain bargaining power, while less differentiated AI software and private-market comps face tougher fund-raising and higher multiple scrutiny.

Key Decisions for Investors

• Do not preemptively chase the IPO story; wait for the S-1. Act only if audited disclosures confirm revenue run-rate above $80B and no gross-margin deterioration — otherwise treat the move as sentiment, not fundamentals.
• Long NVDA / SMH on S-1 filing strength, with a 1-3 month horizon. Best case is a clean read-through to continued GPU demand; downside is limited if you size for a 5-8% pullback on any filing disappointment.
• Pair trade: long NVDA or AVGO vs short IGV for 1-3 months if the filing confirms compute-heavy growth. Thesis is capital concentration toward infrastructure and away from application-layer monetization; exit if software guidance inflects higher.
• Add ANET on any post-filing dip. If Anthropic’s spend trajectory is real, backend networking should remain one of the highest-quality second-order beneficiaries; use a 5-10% stop and look for 2:1 upside if AI capex stays elevated.
• Avoid using the 'record IPO' narrative to buy illiquid private-market proxies unless there is a public, investable vehicle with clear NAV support. In this setup, liquidity and mark-up risk matter more than the headline valuation.