Anthropic is targeting an IPO the size of the largest ever

Anthropic says it could file for a public IPO as soon as the end of this month and expects a valuation/raise that matches or exceeds SpaceX’s record $75B fundraising. The company reported a net loss of almost $42B in 2025—about 5x the prior year—implying significant ongoing cash burn. While it targets a debut that could match or beat the largest IPO ever, the scale of losses keeps the near-term outlook cautious.

Analysis

The most immediate read is not “AI is winning,” but that late-stage private capital is still willing to pay up for scarcity, which should keep the upper end of the AI funding market open for another 1-3 months. That is supportive for the adjacent supply chain: chip vendors, HBM memory, advanced packaging, and hyperscaler capex narratives (NVDA, AMD, AVGO, ASML, MSFT, AMZN, GOOGL). The second-order risk is that an eye-popping IPO target becomes a valuation anchor for every private AI startup, increasing the chance of a sentiment reset if the filing shows burn outrunning monetization.

The bigger issue is balance-sheet physics. If the business is truly compounding losses at that pace, the market will eventually demand either much faster revenue conversion, pricing power, or a financing structure that pushes dilution and governance risk onto public holders. That makes the likely catalyst path asymmetric: the IPO window itself can lift AI beta in the next few weeks, but the S-1/roadshow is where scrutiny on margin structure, customer concentration, and compute commitments can reverse the trade quickly.

Contrarian take: consensus may be underestimating how much this kind of mega-valuation crowds out other late-stage software names and overestimating how easily “AI story” translates into public-market support. If the offer is priced to a record multiple, the post-listing air pocket could be severe unless there is unusually strong order-book depth. For public comps, this is more useful as a sentiment barometer than a fundamental signal; the tradable angle is likely in the suppliers of AI infrastructure, not in chasing the IPO itself.

Key Decisions for Investors

• Stay long the AI infrastructure basket into the filing window: NVDA / AMD / AVGO / MSFT on a 1-3 month horizon, because a marquee IPO can reinforce capex expectations and support multiple expansion. Use a tight trailing stop if AI hardware lead indicators flatten or if hyperscaler capex guides down.
• Do not chase the IPO at the open unless the S-1 shows clear unit economics; if public, wait for first-day stabilization and only consider participation after the gross margin and customer retention data are disclosed. The key falsifier is evidence that losses are still scaling faster than revenue for the next 2 quarters.
• Pair trade idea: long NVDA or SMH vs short a basket of high-multiple unprofitable software/AI-adjacent names; the market usually rewards picks-and-shovels first while punishing vague AI monetization claims if risk appetite tightens.
• Watch for a sector rotation tell: if 30-day implied vol and secondary activity spike across late-stage AI names, reduce exposure to the weakest private-to-public comparables. That would indicate the IPO is becoming a liquidity event for insiders rather than a durable valuation reset.
• If the filing lands before month-end, treat it as a catalyst alert rather than a conviction trade; the first real entry point is after the book shows whether demand is broad-based or concentrated in crossover funds.