The AI trade isn't dying — it's just changing addresses from chipmaker to model builder, and investors are about to find out if the rent is worth it.
The Summary
- Anthropic is prepping a blockbuster IPO with demand pushing pricing to the top of the range or higher, even as the company burns through $2.7 billion annually
- Nvidia shares dropped 17% amid CEO succession concerns after Jensen Huang's departure, with analysts questioning if the board moved too fast
- Edward Yardeni calls the AI trade "alive and well" but "in transition" as market focus shifts from infrastructure providers to model builders
- Blackstone's Viral Patel says the AI investment boom is only accelerating, signaling institutional confidence despite the rotation in leadership
The Signal
Wall Street is running a live experiment on what matters more: the picks or the shovels. Nvidia is down 17% while navigating life without Jensen Huang at the helm, and investors are suddenly asking whether the chip giant can maintain its stranglehold on AI infrastructure without its visionary CEO. Meanwhile, Anthropic is commanding blockbuster IPO pricing despite burning $2.7 billion a year. The message: the market believes the value is migrating upstream to the companies building the intelligence, not just the hardware running it.
This isn't a repudiation of the AI thesis. Edward Yardeni frames it as a transition, not a collapse. The capital isn't leaving AI, it's just shopping for different exposure. Nvidia's stumble tells you two things. First, even the most dominant infrastructure plays are vulnerable to succession risk when their entire market position is tied to one person's technical vision and customer relationships. Second, the market is maturing past the "build the rails" phase and into the "who runs the fastest trains" question.
"The AI trade is alive and well, just moving from Nvidia's earnings and CEO search to the Anthropic IPO."
Anthropic's IPO demand is so strong it's pushing pricing to the top of the range or higher, which is remarkable given the fundamentals. Burning $2.7 billion annually isn't a bug in the pitch deck, it's the entire story. Here's what investors are buying:
- The belief that frontier model development is a winner-take-most market
- Early positioning in the race to build agents that can actually do work, not just answer questions
- Access to a company that could become the AWS of intelligence infrastructure
Bank of America's Brian Moynihan noted the mega IPO is reshaping Wall Street's deal pipeline. Translation: if Anthropic can command this valuation while hemorrhaging billions, every other AI lab with a credible model and enterprise contracts is updating their S-1 drafts. The market is telling founders that progress toward AGI is worth more than profitability, at least for now.
But there's a tension here that nobody's really addressing. Anthropic founding investor Anjney Midha says the US is falling behind on AI buildout. If that's true, and if Anthropic is the crown jewel of American AI, then this IPO isn't just a liquidity event. It's a referendum on whether capital markets can fund the compute, talent, and infrastructure needed to stay ahead of international competition. The $2.7 billion annual burn isn't waste, it's the table stakes for playing in the AGI game.
The Implication
Watch what happens to Anthropic's stock in the first six months. If it holds or climbs, expect a wave of AI lab IPOs in 2027 as OpenAI, xAI, and others decide they'd rather tap public markets than keep raising private rounds at escalating valuations. If it craters, the entire "build now, profit later" thesis for frontier models gets tested, and capital flows back toward profitable infrastructure plays or companies with actual revenue models.
For anyone building in this space, the lesson is clear: the market will pay for speed and capability, even without profitability, but only if you're in the top tier. Second-place models won't get Anthropic pricing. They'll get acqui-hired or shut down.