An artificial intelligence bubble may already be forming in the US stock market, but a major correction does not appear imminent, according to new research from Yardeni Research.
In a report published Wednesday, the investment research firm introduced an AI Bubble Meter designed to assess speculative market conditions and identify potential triggers for a stock market crash.
Although the indicator suggests that AI-related valuations have reached elevated levels, historical comparisons indicate that the conditions typically associated with a major market downturn have not fully developed.
Yardeni Research Introduces AI Bubble Meter to Track Market Risks
Yardeni Research developed its new AI Bubble Meter using Anthropic’s Claude Fable 5.1 artificial intelligence model.
The analytical tool features two separate indicators that evaluate different aspects of financial market risk.
The first, known as the Conditions dial, measures whether the stock market displays characteristics commonly associated with speculative bubbles.
To make this assessment, the model examines conditions resembling those preceding the dot-com bubble collapse of 1999–2000 and the US housing market crisis of 2006–2007.
Meanwhile, the second indicator, called the Triggers dial, evaluates whether the economic and financial pressures that historically caused speculative bubbles to collapse are emerging.
These potential triggers include excessive borrowing, tightening credit conditions, and sustained increases in financing costs.
According to Yardeni Research, the latest readings suggest that an AI bubble exists, but the immediate warning signs of a major correction remain relatively limited.
AI Bubble Conditions Reach 83 Out of 100
The firm’s latest AI Bubble Meter readings reveal a significant difference between speculative market conditions and potential crash triggers.
The Conditions indicator currently stands at 83 out of 100, placing it firmly in the high-risk category for bubble formation.
However, the Triggers indicator registers just 34 out of 100, suggesting that the forces capable of ending the current market expansion remain comparatively weak.
Yardeni Research believes that speculative conditions have persisted for some time, particularly as investors assign elevated valuations to companies benefiting from artificial intelligence development.
Nevertheless, the broader financial environment has not yet experienced the same combination of credit excesses and rising borrowing costs that contributed to previous market collapses.
One warning signal has already emerged.
The yield on the two-year US Treasury note has increased by approximately 115 basis points from its December low, indicating a meaningful rise in short-term financing costs.
Despite this development, Yardeni noted that credit conditions outside AI-related industries have not become excessively stretched.
Consequently, the research firm believes that the ingredients for a speculative bubble are present, while the mechanisms that could trigger its collapse remain only partially developed.
Historical Data Suggests Limited Near-Term Stock Market Crash Risk
Yardeni Research also examined historical stock market performance to determine how similar conditions affected the S&P 500 in previous decades.
Its analysis covered market environments dating back to 1965, focusing on periods when speculative conditions were elevated but potential crash triggers remained relatively low.
According to the firm’s findings, these circumstances were followed by an S&P 500 bear market within the next two years in only 9% of historical observations.
Furthermore, the benchmark index recorded a median inflation-adjusted gain of 18.3% over the subsequent two-year period.
The historical picture changed considerably when both indicators reached elevated levels.
In those circumstances, the proportion of observations followed by an S&P 500 bear market within two years increased to 42%.
The median inflation-adjusted market performance also deteriorated to a decline of 12.5%.
These findings suggest that elevated stock valuations alone have not consistently resulted in major market downturns.
Instead, the combination of speculative valuations and deteriorating financial conditions has historically represented a more serious warning signal.
However, historical relationships do not guarantee future market performance, and the AI Bubble Meter should not be interpreted as a precise prediction of an upcoming crash.
Rising Treasury Yields Could Become a Major Warning Signal
Although the firm’s current assessment does not point toward an imminent AI bubble collapse, Yardeni Research emphasized the importance of monitoring changes in the Triggers indicator.
In particular, the research team is watching the direction of short-term US Treasury yields.
The two-year Treasury yield has already risen significantly from its December low, but the increase remains below the levels observed before certain previous financial market collapses.
Yardeni Research identified increases of approximately 200 to 350 basis points as historical reference points associated with the periods preceding the 2000 technology bubble collapse and the 2006 housing market downturn.
A further sustained increase in borrowing costs could therefore create additional pressure on highly valued technology companies and businesses financing major AI infrastructure investments.
Higher interest rates can make debt-funded expansion more expensive while also reducing the present value investors assign to anticipated future earnings.
As a result, companies whose valuations depend heavily on long-term growth expectations could become particularly sensitive to tighter financial conditions.
AI Infrastructure Spending and Hyperscaler Earnings Under Scrutiny
Beyond Treasury yields, Yardeni Research plans to examine third-quarter earnings reports from major cloud computing companies and AI infrastructure providers.
Particular attention will focus on whether hyperscalers are beginning to report unused computing capacity.
These companies have invested heavily in data centers, advanced processors, and other infrastructure required to support artificial intelligence applications.
However, concerns remain over whether demand for AI services will expand quickly enough to justify the scale of these investments.
Evidence of excess capacity could suggest that infrastructure development is progressing faster than commercial demand.
Such a development could weaken expectations for future revenue growth and increase pressure on companies whose valuations depend on continued AI investment.
Conversely, strong utilization rates and sustained demand could help support the current expansion.
For this reason, upcoming corporate earnings announcements may provide important evidence about the financial sustainability of the AI investment cycle.
Is the AI Stock Market Bubble About to Burst?
Yardeni Research’s latest analysis presents a distinction between the existence of a speculative bubble and the likelihood of an imminent market collapse.
With the Conditions indicator at 83 and the Triggers indicator at 34, the firm’s model suggests that AI-related market valuations remain elevated while the traditional catalysts for a major downturn are less pronounced.
Historical observations also indicate that periods of high speculative activity have sometimes been followed by further stock market gains rather than immediate corrections.
Nevertheless, rising Treasury yields, tighter credit conditions, and potential overinvestment in AI infrastructure remain important risks.
Looking ahead, the direction of borrowing costs and the financial performance of major AI companies could help determine whether today’s elevated valuations remain sustainable.
For now, Yardeni Research’s assessment suggests that the AI bubble may continue expanding, even as investors face growing uncertainty over its eventual outcome.
