Peter Schiff Called 2008. Now He’s Calling the AI Trade a Bubble.

Peter Schiff’s read on where 2026 markets currently stand, described at length in a recent interview: AI-related stocks alone have added roughly $6 trillion in market capitalization this year, while every other sector in the S&P 500 combined has actually lost value over the same period. The index headline is positive. The market underneath it, in Schiff’s framing, is not — a small handful of companies are carrying an otherwise flat-to-negative index on their own.

He points to two well-known valuation measures as evidence the concentration has gone too far: the Shiller CAPE ratio sits above 40, comparable to levels last seen at the peak of the dot-com bubble, and the Buffett Indicator — total market capitalization relative to GDP — sits well above the level that marked the 2000 peak. Neither measure is a precise timing tool, but both have historically been reliable warnings that valuations have drifted far from anything supportable by underlying earnings growth.

Importantly, Schiff’s argument isn’t that AI itself is fake or that the technology lacks real value. His distinction is between the companies currently providing AI infrastructure and the (often different) businesses that will eventually capture the economic benefit of using AI more cheaply and efficiently. Historically, in comparable technology build-outs — the railroads of the 1880s are his go-to example — the infrastructure builders themselves often weren’t the biggest long-term winners; the businesses that used the new infrastructure to reach new markets were.

It’s worth noting Schiff has a well-documented track record of being early rather than perfectly timed — he called the 2008 housing collapse well ahead of the crash, but has also flagged bubbles that took years longer to unwind than his initial calls suggested. That track record is itself a useful caution: even a well-reasoned, well-supported thesis from a credible analyst is not a substitute for staying diversified rather than making a single large bet in either direction.