Is there an AI bubble? No, but some tech companies show signs

• Alphabet currently shows strong overvaluation evidence with stock prices up more than 70% in the past year, outpacing the Nasdaq Composite's 20% growth
• Researchers developed a new SV-ADF statistical framework that can identify bubble dynamics at individual stock level rather than mislabeling entire sectors
• Nearly all semiconductor companies were in bubbles after ChatGPT's November 2022 release, and Tesla showed overvaluation signs in 2020, but these bubbles have since collapsed

Market Summary: AI Bubble Analysis

Key Findings

Cornell researchers have determined that while the AI sector overall is not experiencing a bubble, certain individual tech companies exhibit speculative bubble characteristics. The study, published in July 2026 in *Frontiers in Mathematical Finance*, introduces a new statistical method (SV-ADF framework) to identify genuine bubble dynamics at the company level.

Companies Showing Bubble Signs

Alphabet (Google) stands out as currently showing strong overvaluation evidence, with stock prices surging over 70% in the past year—significantly outpacing the Nasdaq Composite's 20% growth.

Historical bubble episodes identified include:

• Semiconductor companies: Nearly all showed bubble characteristics following ChatGPT's November 2022 release
• Tesla: Exhibited overvaluation signs in 2020
• Cryptocurrency: Bitcoin and Ethereum displayed exuberance from December 2020 onward

All these historical bubbles have since collapsed, though some cases show renewed exuberance.

Methodology

The research analyzed AI-exposed stocks from 2020 through April 2026, including the "Magnificent Seven" (Apple, Amazon, Nvidia, Meta), semiconductor firms (TSMC, Broadcom), and cryptocurrencies. The new statistical approach distinguishes between normal volatility and genuine bubble dynamics, addressing limitations of existing models that often mislabel entire sectors.

Market Implications

The research emphasizes heterogeneity across AI companies, warning against treating AI as a monolithic investment category. Lead researcher Robert Jarrow notes that identifying bubble stocks enables better investment decisions, as bubbles inevitably burst, causing dramatic price declines. This selective analysis provides investors with more nuanced risk assessment tools.