Insider Brief
- A recent academic study finds signs of speculative, bubble-like behavior in the Nasdaq and several leading artificial intelligence-linked stocks during the current AI-driven market surge.
- The analysis shows that speculative dynamics are concentrated in a subset of firms — particularly Nvidia, Microsoft, and Tesla — whose stock prices have diverged from both the broader market and fundamentals.
- While the research does not conclude that all of AI is in a bubble, it offers several signals to look for as investors and market participants navigate the current AI market.
Concerns about an AI bubble have grown louder as valuations climb, chip shortages deepen and a small cluster of companies pull further ahead of the market. Investors worry that AI’s promise — real as it is — may be inflating expectations faster than revenues, echoing other periods when transformative technology triggered speculative behavior.
Business leaders, for their part, must decide how aggressively to deploy capital in AI infrastructure, while workers inside the sector are feeling the volatility in hiring, compensation, and project priorities.
A study examining the past decade of stock market data suggests that today’s AI market shares traits with earlier technology booms that ended abruptly. Researchers from Yale University and Macquarie University applied formal bubble-detection tools to the Nasdaq and the so-called Magnificent Seven stocks to test whether explosive price dynamics have taken hold.
Their findings may offer a research-backed framework to help firms and individuals understand whether today’s AI-driven surge is behaving like a sustainable growth story or something more volatile.
Explosive Growth Patterns
Markets can rise for many reasons, but bubbles follow a distinctive statistical pattern. In bubbles, prices do not simply increase, they accelerate in a nonlinear, self-reinforcing way. Using the PSY and GSADF tests, the study finds that the Nasdaq entered such explosive regimes repeatedly between 2017 and 2021. After a brief correction in 2021–2022, speculative behavior returned in December 2022 and persisted through January 2025.
For investors and operators inside AI companies, this suggests enthusiasm is no longer tracking fundamentals alone.
Once markets begin behaving in mathematically explosive ways, history shows the odds of overshoot rise sharply.
Stocks Breaking Away From the Market
A second signal emerges when individual companies begin rising faster than the sector that contains them.
That may sound confusing. Here’s an example: If the Nasdaq rises 15% over a year but a single AI stock rises 80% over the same period without a comparable jump in revenue or profits, that divergence signals the company is being priced on enthusiasm rather than on sector fundamentals.
By comparing each stock to the Nasdaq, the researchers show that AI winners are not moving in lockstep with the broader market. In face, some are running far ahead of it. NVIDIA and Microsoft diverge most dramatically, posting long stretches where their growth outpaced the index by wide margins. Amazon, Apple and Tesla experience shorter episodes of similar behavior, while Alphabet and Meta show only brief deviations.
This matters because when a stock no longer behaves like part of the market, but instead becomes its own momentum engine, investors are no longer pricing the company — they are pricing the narrative.
Prices Detaching From Fundamentals
To determine whether a company’s rise is driven by performance or exuberance, the paper decomposes returns using the Fama–French three-factor model. This isolates the “idiosyncratic” component — the part not explained by market, size, or value forces — and tests whether that residual shows explosive behavior. The results again point to NVIDIA and Microsoft as the most speculative, with Tesla also displaying repeated explosive episodes.
A few implications stand out:
- Market-adjusted speculation appears concentrated among companies supplying AI compute and infrastructure.
- Software and platform firms show more mixed or episodic signs of exuberance.
- Apple shows no statistically significant bubble behavior in the most recent window.
When a stock continues flashing speculative signals even after accounting for fundamentals, the remaining force is typically sentiment.
Mildly Explosive Growth, Not Just High Momentum
Not all bubbles form through dramatic spikes, the study suggests.
Some develop as prices drift into a zone economists call “mildly explosive,” a transitional pattern between normal persistence and outright mania. Using near-unit-root inference, the study finds clear signs that NVIDIA and Tesla entered this territory between 2017 and 2021. In the 2022–2025 period, all seven companies exhibit behavior consistent with explosive or near-explosive dynamics, though with different intensities.
For decision-makers inside AI firms, this suggests that price-driven volatility — and the hiring, spending, and supply-chain pressures tied to it — may prove more structural than cyclical.
Echoes of Past Technology Booms
The researchers also compare today’s AI surge to earlier eras of technological enthusiasm, including RCA in the 1920s and Cisco during the dot-com boom. NVIDIA’s rise — more than 570% between late 2022 and mid-2024 — fits the pattern set by those high-flying investments. Those prior episodes followed a familiar trajectory where technology breakthroughs justified early optimism, valuations then climb even faster than adoption or revenue, and eventual corrections hit once expectations race ahead of reality.
The AI boom shows similar features. Compute supply has become the defining chokepoint, early corporate interest has evolved into capital-intensive commitments, and the biggest players are widening their lead over the rest of the industry.
What This Means for the AI Economy
The study does not declare the entire AI sector a bubble.
Instead, it shows that speculative dynamics appear unevenly across the industry, concentrated especially in firms that supply the infrastructure enabling rapid AI growth. The Nasdaq exhibits repeated signs of exuberance, suggesting AI enthusiasm is influencing broader market sentiment as well.
For investors, managers and workers, the takeaway is not to brace for collapse but to recognize that AI’s financial cycle may be entering a more volatile phase. The signals emerging now resemble those that preceded past technology corrections. Understanding them may help decision-makers separate durable opportunity from momentum-driven noise– a distinction that matters even more in a market defined by speed.
Important to note that this isn’t meant to be investment advice, but to serve as a way to increase awareness of factors that might affect the AI industry and its participants.
