Federal Reserve Vice Chairman: Four Dimensions for Assessing the AI ​​Bubble

Author: Zhang Feng

Artificial intelligence (AI) is reshaping the global economic and financial landscape at an unprecedented pace. As capital markets continue to show great enthusiasm for AI-related companies, an unavoidable question arises: Are we witnessing a speculative frenzy similar to the dot-com bubble of the late 1990s?

In 2025, at the Cleveland Federal Reserve Bank Financial Stability Conference, Vice Chairman Philip N. Jefferson systematically elaborated on his comparative analysis of the current AI boom and the dot-com bubble era, and proposed four key indicators for judging whether an AI bubble exists.This speech not only reflects the cautious observation of the world's most important central bank on emerging technologies, but also provides market participants with a clear framework for rationally assessing the AI ​​boom.This speech not only reflects the cautious observation of the world's most important central bank on emerging technologies, but also provides market participants with a clear framework for rationally assessing the AI ​​boom.

I. The Federal Reserve's Observation Basis: Dual Mandate and Financial Stability

All of the Federal Reserve's policies and observations revolve around its statutory "dual mandate"—maximizing employment and price stability. Jefferson clearly pointed out that assessing the impact of artificial intelligence must start from this fundamental mission. This means that the Federal Reserve's focus on AI is not only on its technological breakthroughs or market performance, but also on how it affects overall employment levels, labor productivity, economic growth potential, and inflation trends. From an employment perspective, AI exhibits a dual effect. On the one hand, it promotes employment by improving work efficiency and creating new jobs (such as AI research, deployment, and maintenance); on the other hand, its automation substitution effect may lead to the shrinkage of some occupations, especially impacting younger, less experienced workers. Jefferson pointed out that if AI only replaces existing labor without simultaneously creating new jobs, it may trigger a short-term economic slowdown. This dynamic balance of "substitution and supplementation" is the core of judging the structural impact of AI on the labor market. From a price stability perspective, AI's increased productivity helps reduce production costs, putting downward pressure on prices. Applications such as efficient resource allocation, supply chain optimization, and decision support can all help curb inflation. However, at the same time, the construction of AI infrastructure (such as data centers) drives up the prices of inputs like land and energy, and rising salaries for AI talent may also lead to cost-push inflation. This dual impact makes the net effect of AI on inflation highly uncertain and requires continuous monitoring. To achieve this dual mission, a robust and resilient financial system is crucial. The Federal Reserve continuously monitors systemic risks through its semi-annual Financial Stability Report (FSR). A recent survey shows that 30% of market contacts consider a "shift in attitudes toward AI" a significant risk to the financial system, a sharp increase from 9% in the spring. This seems to warn that a sudden reversal in market optimism regarding AI could trigger tighter financial conditions and an economic downturn. Therefore, the Fed's inclusion of AI in its financial stability monitoring framework is precisely to prevent asset bubbles and financial vulnerabilities that may arise from the technology boom. II. Monitoring Framework: FSR and Market Sentiment Tracking The Fed's monitoring of AI is not isolated but embedded within its overall financial stability assessment system. The FSR not only focuses on traditional risks such as leverage ratios, asset valuations, and financing risks but also incorporates the structural changes brought about by emerging technologies. Jefferson emphasized that policymakers must distinguish between "cyclical fluctuations" and "structural changes," and AI likely falls into the latter category. This means that productivity gains brought about by AI could alter the relationship between employment and inflation, thereby affecting the transmission mechanism of monetary policy. Market sentiment is one of the key focuses of the FSR. Surveys show that nearly one-third of market participants are aware of the potential risks of a reversal in AI sentiment. This consensus itself could become a "self-fulfilling prophecy"—once the optimistic narrative shifts, a rapid capital exodus could lead to a sharp adjustment in asset prices. Compared to the dot-com bubble era, today's speed of information dissemination and the widespread adoption of algorithmic trading may amplify market volatility. Therefore, the Fed's tracking of sentiment indicators is essentially an early warning of potential systemic risks. Furthermore, the application of AI within the financial industry itself also brings new monitoring challenges. While AI tools such as high-frequency trading, robo-advisors, and risk models improve efficiency, they may also trigger new risks of homogenization and procyclicality. The Federal Reserve is strengthening the identification and assessment of these emerging risks by expanding its analytical toolkit (including utilizing AI technology itself).

III. Four Core Indicators: A Touchstone for Judging the AI ​​Bubble

Third, four key differences were identified by Jefferson by comparing the current AI boom with the dot-com bubble of the late 1990s. These differences can serve as core indicators for judging whether there is a serious bubble in the current AI field.

(I) Profitability Basis: From "Story-Driven" to "Profitability-Supported"

During the dot-com bubble, many companies went public based solely on the ".com" concept, lacking a sustainable profit model, with meager or even zero revenue, relying on external financing and market frenzy to maintain operations.

In contrast, leading companies in the current AI field (such as some tech giants) generally possess solid and diversified profit channels. They not only generate revenue directly through AI services but also deeply embed AI into their existing product systems, enhancing their core business competitiveness. This "profit-driven" development model makes AI investment more fundamentally grounded, reducing the space for pure speculation. However, Jefferson also points out that the activity in the private equity market may partially mask the profitability difficulties of early-stage AI companies. A large influx of venture capital into AI startups, despite their unlisted status, has resulted in high valuations, and if they fail to achieve profitability in the future, they could still become a source of risk. Therefore, the observation of profitability indicators needs to consider both the public and private markets. (II) Valuation Level: Relatively Restrained Price-to-Earnings Ratios During the peak of the dot-com bubble, the price-to-earnings ratios of internet companies often reached hundreds or even thousands of times, reflecting the market's irrational optimism about long-term growth. Currently, although the stock prices of AI concept companies have risen sharply, their price-to-earnings ratios are still far below historical peaks. This indicates to some extent that while investors are chasing AI, they are still anchoring their valuations to the actual profitability and cash flow of companies. Of course, the rationality of valuation needs to be judged comprehensively in conjunction with industry characteristics and growth stage. As a general-purpose technology, AI has huge long-term value creation potential, and a moderate premium is reasonable. However, if the valuation rises too quickly and deviates from the fundamentals, it may still breed a bubble. The Federal Reserve focuses on valuation metrics precisely to distinguish between rational elements and overheating signals within market enthusiasm. (III) Number of Listed Companies: Limited Speculation In 1999-2000, over 1,000 internet companies went public, creating a speculative frenzy where even changing names to include ".com" could inflate stock prices. Currently, there are approximately 50 listed companies explicitly categorized as "AI core enterprises" (based on specific indicators), a number far fewer than during the internet bubble. This indicates that market speculation is relatively concentrated and has not yet spread to the entire market. However, Jefferson also cautioned that the private equity market may harbor a large number of AI startups that, while not publicly traded, are actively raising funds. If these companies were to go public en masse or the financing environment were to change drastically in the future, they could become new destabilizing factors. Therefore, the "number of companies" indicator needs to be dynamically monitored, covering both public and private equity sectors. (IV) Financial Leverage: Low Debt Dependence During the dot-com bubble, many companies relied on equity financing, resulting in limited debt leverage, which to some extent reduced the direct impact of the bubble's burst on the financial system. Currently, AI companies also rely less on debt financing, which helps limit risk transmission. However, recent trends show that to support the massive investments in AI infrastructure (such as data centers and computing clusters), some companies are beginning to increase bond issuance and credit financing. Jefferson specifically points out that as AI expands from software to hardware infrastructure, the demand for capital investment is rising sharply, which may lead to a gradual increase in leverage ratios. If the AI ​​sentiment reverses, highly leveraged companies will face greater debt repayment pressure, and thus spread the risk to a wider economic sector through credit channels. Therefore, the evolution of leverage indicators needs to be closely monitored. IV. Implications for Market Practitioners Jefferson's arguments not only provide an analytical framework for policymakers, but also offer important insights for investors, businesses, and researchers: First, observing problems must start from the observer's fundamental task. Investors should move beyond short-term market sentiment and delve into the substantive impact of AI technology on a company's fundamentals (profitability, cost structure, competitive barriers). Companies, in turn, need to focus on how AI can improve their productivity and long-term competitiveness, rather than blindly chasing concepts. Secondly, distinguish between cyclical fluctuations and structural changes. AI represents a technological revolution that may last for decades, and its impact is structural. In market fluctuations, it's crucial to differentiate between long-term trends and short-term noise, avoiding misjudging structural opportunities as cyclical bubbles, or vice versa. Thirdly, pay attention to overall market reactions and systemic risks. The rise of a single company or sector does not necessarily constitute a bubble; it is necessary to assess the overall market valuation level, capital concentration, leverage, and consistency of sentiment. Particular attention should be paid to signs that the AI ​​narrative is shifting from "profit-driven" to "story-driven." Fourth, make good use of analytical tools, including AI itself. AI technology can be used to more accurately assess market risks, corporate value, and economic impact. Practitioners should actively utilize data analysis, machine learning, and other tools to improve decision-making quality, while being wary of new risks that may arise from model homogenization. V. Continued, Multi-dimensional, and Dynamic Participation with Rationality and Enthusiasm Jefferson's final conclusion was relatively cautiously optimistic: Based on a comparison of four dimensions—profitability, valuation levels, number of companies, and financial leverage—the current AI boom differs significantly from the dot-com bubble, and the possibility of a repeat of the dramatic collapse of the late 1990s is low. AI development is rooted in a group of mature companies with stable profits, and the overall financial system is relatively resilient. However, uncertainties remain. The long-term impact of AI on employment, inflation, and productivity still needs time to be verified; market sentiment may reverse; activity in the private equity market may mask risks; and the possibility of infrastructure investment driving up leverage warrants attention. Therefore, the Federal Reserve will continue to monitor AI development to ensure it unfolds in a stable and resilient financial environment, ultimately serving the fundamental goals of maximizing employment and price stability. For the market, Jefferson's analysis provides a toolbox for rationally assessing AI investment. Amidst the wave of technological revolution and capital enthusiasm, remaining clear-headed, distinguishing between essence and appearance, and focusing on long-term value may be the best approach to avoid bubbles and embrace change. Is AI a bubble? The answer lies not in a simple yes or no, but in continuous, multi-dimensional, and dynamic observation and judgment.