Tech stocks hit lowest valuations since OpenAI launched ChatGPT

The AI trade that defined markets for nearly four years has fully unwound, at least on paper. The S&P Global tech sector now trades at 21x forward earnings, a level last seen before ChatGPT debuted on November 30, 2022, according to Truist analyst Sam Grelck.

That’s a steep fall from 32x forward P/E in October 2025, representing a roughly 34% compression in how much investors are willing to pay per dollar of expected earnings.

The paradox: earnings are actually rising

What makes this selloff unusual is that it’s happening against a backdrop of improving fundamentals. Forward earnings estimates for the tech sector have climbed 19% over the past three months alone. That growth rate nearly doubles the next strongest sector.

Semiconductors tell a similar story. The PHLX Semiconductor Index, the benchmark for chipmakers, sits at roughly 20x forward P/E. That mirrors where it traded in January 2023, well before the AI infrastructure buildout kicked into high gear. For context, the SOX peaked at 30x forward earnings in mid-2024, meaning chip stocks have shed a third of their premium valuation.

Capital expenditures among hyperscalers and semiconductor manufacturers remain robust. Revenue growth continues to accelerate across the AI supply chain.

Why the mood shifted

The compression isn’t random. A convergence of public statements from the people building AI has spooked investors. Dario Amodei, CEO of Anthropic, Sam Altman of OpenAI, and Elon Musk have each called for greater caution in advancing AI technology.

Overreaction or appropriate caution

Truist’s Grelck and other analysts have flagged the possibility that markets have overcorrected. The logic is straightforward: if earnings estimates are rising by 19% while valuations compress by 34%, the gap between price and underlying business performance is widening fast.

Much of the uncertainty around AI development timelines may already be baked into current prices. What hasn’t been absorbed is the continued growth in actual demand for AI infrastructure, the servers, chips, and cloud capacity that generate real revenue regardless of whether artificial general intelligence arrives in 2028 or 2035.

For crypto markets, the tech valuation reset carries indirect but meaningful implications. Bitcoin and major digital assets have historically shown strong correlation with tech stock sentiment during macro-driven selloffs.

The semiconductor angle matters here too. Companies manufacturing AI chips are the same ones whose products power crypto mining operations and blockchain infrastructure.