Federal Reserve Vice Chairman: The AI craze is not "Internet Bubble 2.0," but we need to be wary of debt risks

Federal Reserve Vice Chairman Philip Jefferson stated that the current surge in artificial intelligence-related stocks is unlikely to repeat the scenario of the late 1990s internet bubble burst, mainly because AI companies have established actual profitability and the financial system remains robust. On Friday, Jefferson spoke at a Federal Reserve conference in Cleveland, stating that unlike the speculative internet companies of that time, current AI-related enterprises are mature companies with relatively mature and growing sources of profit. He pointed out: These companies have not yet heavily relied on debt financing, which may reduce the extent of broader risks transmitted to the economy through credit markets when AI market sentiment shifts. However, he also warned that the leverage in the AI industry may rise, and losses could increase once AI sentiment changes. He will closely monitor this trend. AI Companies Are Profitable, Valuations Are Lower Jefferson elaborated on three key differences between the current AI boom and the internet bubble era. First and foremost, the core difference lies in the fundamentals of the companies. He noted that, unlike many companies in the late 1990s that went public easily based on concepts and speculative prospects with almost no profits, today’s companies closely related to AI technology “generally have mature and continuously growing revenue streams.” Secondly, due to a solid profit foundation, the price-to-earnings ratios of AI-related companies are currently far below the peak levels during the internet bubble. Jefferson stated that during the last tech boom, investors were clearly more willing to bet on the “promise” of the internet, leading to stock prices at that time containing a higher premium than today’s AI stocks. Finally, from a market breadth perspective, the speculative enthusiasm of that time was also more widespread. Data shows that at the peak of the late 1990s, over 1,000 companies were classified as “internet companies,” with many boosting their stock prices simply by adding “.com” to their names. In contrast, the current market boom is more concentrated, with statistics indicating that only about 50 publicly listed companies are considered core AI enterprises. However, Jefferson also acknowledged that the rapid development of the private capital market in recent years may have masked some of the investment enthusiasm, as many unlisted companies also position themselves as AI companies. Debt Leverage Is Key Jefferson emphasized that the AI industry has so far had limited reliance on debt financing. He believes that this limited use of leverage “may reduce the extent of risk transmitted to the broader economy through credit markets when AI market sentiment shifts.” However, future trends are worth monitoring. Jefferson cited some analysts' predictions that future investments in AI infrastructure will require more debt financing. He mentioned that just a few weeks ago, major players in the industry had issued a large amount of corporate bonds to support AI infrastructure development. Jefferson warned: If this proves to be the case, the leverage in the AI industry may rise—if market sentiment towards AI changes, losses may also increase. I will closely monitor this trend In the latest financial stability monitoring work, the Federal Reserve conducted a routine survey of market participants. The latest survey found that 30% of respondents cited the mainstream sentiment shift towards AI as a significant risk to the U.S. financial system and the global economy, a substantial increase from 9% in the spring survey. Jefferson stated that the financial system remains robust and resilient, with the high-risk appetite and leverage levels of hedge funds balanced by strong household and corporate balance sheets and high capital levels in the banking system. The impact of AI on the economy and monetary policy remains unclear Regarding the profound impact of artificial intelligence on the overall economy, Jefferson maintained a cautious stance. He acknowledged that AI technology could dramatically and "bumpily" change the world, but it is "too early" to determine its exact consequences on the labor market, inflation, and monetary policy. Jefferson analyzed that AI could suppress inflationary pressures by enhancing productivity on one hand, while on the other hand, competition for AI talent, data centers, energy, and land could also drive up costs in specific areas. He believes: The productivity gains brought by AI may affect the relationship between employment and inflation, thereby influencing the implementation of monetary policy. Jefferson concluded that policymakers need to distinguish between the structural changes that AI may bring and cyclical factors. Given that there are still too many unknowns, it is premature to assess the impact of AI on monetary policy. He emphasized that it is crucial to remain humble when evaluating this rapidly evolving technology