The Bank of England warns: AI valuations are "overvalued," and data center "bottlenecks" may trigger a market correction

The artificial intelligence boom is pushing technology stock valuations to historic highs, but the Bank of England has issued one of its clearest warnings to date: valuations of AI-related stocks are already "overvalued," and the key infrastructure supporting their development (such as data centers) is facing "bottleneck" risks, which could trigger a "significant market correction."

The Bank of England pointed out after its latest Monetary Policy Committee meeting that market optimism about AI has pushed stock valuations to "tense" levels. The central bank warned that the increasing market concentration in the AI sector makes the stock market particularly vulnerable once the optimistic sentiment surrounding AI diminishes.

The central bank further emphasized that the "substantial bottlenecks" facing AI progress, such as limitations in electricity, data, or commodity supply chains, could also impair corporate valuations, especially for companies whose revenue growth relies on high investment expectations in AI infrastructure. "The risk of a significant correction in asset prices remains very high."

This official warning injects new urgency into the already heated market debate. So far this year, the Nasdaq 100 index has risen by 19%, with its forward price-to-earnings ratio reaching 28 times, significantly higher than the ten-year average of 23 times.

Valuations at "tense" levels, infrastructure bottlenecks as key risks

The Bank of England believes that from multiple indicators, stock market valuations appear quite tense, especially for technology companies focused on artificial intelligence. The high valuations combined with the increasing concentration within market indices amplify the market's fragility. If investors' expectations about AI potential become less optimistic, the downside risks for the market will be significant. Currently, the AI boom has become the single largest driver pushing stock markets to their highest valuations since the dot-com bubble.

Market optimism is largely built on expectations of continued large-scale investments in AI infrastructure. The central bank warned that the real-world resources that AI development relies on—electricity, data, cooling systems, and related commodities—may not keep pace with the explosive growth in demand.

This concern has been corroborated by investment banks. Goldman Sachs analysts recently viewed the power grid as a "vulnerable link" and warned it could become the next "bottleneck." The issue is not only the lack of idle capacity in the power grid or the backlog of approvals for data centers connecting to local grids; Morgan Stanley analysts also pointed out another emerging problem—the supply of water resources used for cooling data centers.

Market giants debate: Industrial revolution or value destruction?

There are significant differences in opinion among market giants regarding whether AI constitutes a bubble. Amazon founder and executive chairman Jeff Bezos believes that this could be an "industrial bubble," different from a pure "financial bubble," as it can drive social progress rather than lead to financial destruction. He stated:

"Investors are now willing to give a six-person team with no product billions of dollars, and that is happening today." However, other seasoned investors hold a more cautious view. Ray Dalio, founder of Bridgewater Associates, stated earlier this year that the current market has a "bubble" similar to that of 1998 or 1999. David Einhorn, founder of Greenlight Capital, also mentioned that spending on AI infrastructure is "so extreme" that "a reasonable possibility is that this cycle will bring about massive value destruction."

Despite the warnings, as the economist Keynes famously said, "The market can remain irrational longer than you can remain solvent." The complex financing mechanisms and strong government support may allow this wave of AI enthusiasm to continue to expand in the short term