AI stock valuations echo 1999 as strategists flag bubble risks

Wall Street strategists are drawing fresh comparisons between today's AI stock rally and the dot-com bubble, warning that valuations in semiconductor and hardware names have surpassed levels seen at the peak of the 1999 mania.

Applied Materials Inc. this week crossed the price-to-sales valuation it held at the peak of the dot-com bubble in April 2000, as strategists warned that AI stock valuations now rival — and in some cases exceed — the excesses of the 1999 mania.

"The parallels to 1999 are becoming harder to ignore," said Victor Dergunov, an investing group leader at Seeking Alpha. "The current AI-driven rally is the most significant bubble in market history."

The Shiller cyclically adjusted price-to-earnings ratio for the S&P 500 stands near 42, approaching levels last seen during the dot-com era. SanDisk Corp. has surged 4,800% over the past 12 months, while Micron Technology Inc. gained 875% and Seagate Technology Holdings PLC rose 705%, according to market data. The S&P 500's concentration in a handful of AI-linked megacaps has reached levels that exceed the tech-heavy index composition of early 2000.

The comparison matters because the dot-com bust erased $5 trillion in market value from the Nasdaq between March 2000 and October 2002. With hyperscaler capital expenditure commitments exceeding $700 billion in 2026, any slowdown in AI infrastructure spending could trigger sharp corrections in memory, hardware, and cloud stocks that have priced in years of uninterrupted growth.

Valuations Stretch Beyond Dot-Com Peaks

The valuation debate centers on whether AI represents a genuine productivity revolution or a speculative mania detached from earnings reality. Applied Materials, the semiconductor equipment maker, now trades at a price-to-sales multiple that exceeds its dot-com peak, according to Investor's Business Daily. The company's market capitalization has swelled past $400 billion as AI chipmakers race to secure manufacturing capacity.

Micron's gross margin expanded to 74.4% in its fiscal second quarter from 36.8% a year earlier, with revenue surging 196% to $23.86 billion. Chief Executive Officer Sanjay Mehrotra called memory "a strategic asset" in the AI era — unusual language for a commodity business that has historically been defined by boom-bust cycles. The company guided the current quarter to $33.5 billion in revenue at an 81% gross margin.

The Picks-and-Shovels Rotation

Billionaire investor Stanley Druckenmiller sold his entire Alphabet Inc. stake in the first quarter of 2026 and rotated the proceeds into SanDisk, Seagate, Micron, Broadcom Inc., and Arm Holdings PLC, according to a 13F filing. The trade reflects a thesis that whichever foundation model wins the AI war, the hardware layer gets paid regardless.

Broadcom, the laggard of the group with a 58% 12-month gain, still pays a dividend and offers the cleanest retail entry point among the five, analysts said. But buying SanDisk after a 4,800% surge is a different trade than the one Druckenmiller made earlier this year, and the entry price now does the work the thesis used to do.

What Happens When CapEx Peaks

The risk, strategists argue, is that AI infrastructure spending is inherently cyclical. Hyperscaler capital expenditure commitments are unprecedented, but eventual slowdowns in data center buildouts could leave memory and hardware suppliers with overcapacity. The Roundhill Memory ETF fell 15% in a single session earlier this month after a guidance miss, illustrating how quickly sentiment can shift in the memory trade.

Seagate's Mozaic HAMR drives are qualified at five of the world's largest cloud customers, with nearline production largely spoken for through mid-2026. But once those contracts are fulfilled, the question becomes whether demand can sustain the current production run rates. "The bull case is already priced in," said Kenio Fontes, an analyst at Seeking Alpha.

For investors, the bubble comparison carries real portfolio implications. A rotation out of high-growth AI names into value or defensive sectors could accelerate if the narrative shifts from "AI revolution" to "AI overvaluation." The S&P 500's year-end target of 8,000, maintained by some strategists, assumes the rally continues — but the margin for error is narrowing with each new all-time high.

This article is for informational purposes only and does not constitute investment advice.