Bitget App
Trade smarter
Buy cryptoMarketsTradeFuturesEarnAISquareMore
ECB Vice President sounds the "bubble alarm": AI asset valuations are "very high," stock markets are prone to corrections

ECB Vice President sounds the "bubble alarm": AI asset valuations are "very high," stock markets are prone to corrections

智通财经智通财经2026/09/15 06:56
Show original
By:智通财经

European Central Bank Vice President Vujičić warns: Asset rallies driven by artificial intelligence are prone to correction risks.

According to Zhitong Finance APP, European Central Bank Vice President Boris Vujcic has warned that a surge in market valuations, primarily driven by the artificial intelligence boom, is making the stock market more vulnerable to correction risks.

"Valuations like these—P/E ratios, forward P/E ratios—if not unprecedented, haven't been seen for a long time," the Croatian official said on the ECB's podcast "Euro Matters," released Tuesday. "These valuations may eventually prove to be reasonable, but they might not."

He warned that "the exposure is substantial and still growing," adding: "We must be extremely cautious and closely monitor this situation, because with such a large exposure and so much investment pouring in, it will inevitably create risks for stock market repricing."

Central Bank Governors in Unison

Vujcic's remarks further reinforce the growing consensus among central bank officials, regulators, and investors: the valuations of major AI companies may be excessive, and a sudden drop could trigger a broader global market correction. Just one day earlier, ECB President Christine Lagarde also candidly stated that asset valuations in the AI sector were "very high" and that a correction was "entirely possible"—although the timing is impossible to predict.

This is not an isolated stance within the ECB. According to reports, five ECB economists (Malin Andersson, Johannes Breckenfelder, Stefano Corradin, Kalin Nikolov, and Maria Antonietta Viola) concluded in an analysis published on August 17, "A correction of current stock market valuations is possible," and for the Eurozone, "This would be an issue of financial stability, not just a private concern"—in other words, the shock would not only be borne by technology stock investors alone.

The Timing of These Warnings Is Noteworthy

These comments are emerging just as AI developers like Anthropic PBC and OpenAI are clashing with the Donald Trump administration over their calls for an industry-wide slowdown in AI development. This stance has already triggered a sell-off in tech stocks and is raising questions about the sustainability of the global data center construction boom.

In fact, last night’s US market action offered a vivid illustration. At Monday’s close, all three major indices fell. The hardest hit sector was semiconductors—the Philadelphia Semiconductor Index plunged 5.86% in one day, its biggest drop since July 1.

On the same night, the yield on the US 10-year Treasury passed 5% during intraday trading for the first time since October 2023. What’s more, after the brief tariff panic of 2025, margin debt in the US stock market has soared 77% in 14 months, surpassing $1.5 trillion, and now faces the triple shock of "historic valuations, rising risk-free rates, and an AI slowdown."

Just How Extreme Are Current Valuations?

Vujcic’s "rare in years" comment is underpinned by a set of historic valuation data. According to data compiled by industry information platform NextFin, the S&P 500’s Shiller CAPE (Cyclically Adjusted Price-to-Earnings Ratio) has remained above 40 since May 2026—the only other time it held at this level for several consecutive months was just before the dot-com bubble peaked in March 2000. The current reading is just over 40, approaching the record 44.2, and is about 2.3 times the long-term average of 17.

Overheating is not limited to earnings multiples. The Buffett Indicator—US stock market cap to GDP—has climbed above 237%, far exceeding Buffett’s own 200% "playing with fire" warning threshold. Since this bull market started in October 2022, the S&P 500 has risen 127%, the Dow 95%, and the Nasdaq 161%, with gains largely driven by AI infrastructure spending.

ECB Vice President sounds the

The rally is also highly concentrated: the "Magnificent Seven" (Apple, Microsoft, Nvidia, Google, Amazon, Meta, Tesla) make up over 35% of the S&P 500’s market cap; the top 10 companies account for about 38% of the index’s value, but only 31% of its earnings (according to NextFin, as of early 2026). This concentration means a stumble in a few stocks becomes a market-wide event, not just manageable sector rotation.

For Europe, the exposure is very real. According to the above-mentioned ECB economist team, Eurozone households hold around 440 billion euros worth of the top seven US stocks, with pension funds and insurers having similar exposure—much of it the result of "involuntary" concentrated positions via passive index tools.

This Time Is Different: The Cushion Is Thinner

Compared to the bursting of the dot-com bubble in 2000, the critical difference may not lie in the bubble itself, but in the available safety net. In 2000, the Federal Reserve had ample space to cut rates, and the government could provide support. But analysis suggests the current starting point leaves much less policy room—interest rates are already relatively low and public debt is high. Should a correction and broader market instability occur together, policymakers would have a hard time restoring calm—this is precisely why the ECB now considers it a "financial stability issue."

On the podcast, Vujcic also listed geopolitical risks and fiscal policy as major threats to the financial system. The former "can rapidly change prices and market attitudes," while on the latter, some countries have "fiscally unsustainable conditions in the long run."

"One thing we've learned from past experience is that unsustainable things are just that—unsustainable," he said, adding that it’s better to address these problems sooner rather than later. "This is also something we must monitor very closely, because these markets could also be repriced—and potentially quite rapidly."

Geopolitical risk is not imaginary: Brent crude was already above $107 a barrel on Monday, and the conflict in the Middle East is adding a second pressure on both inflation and the bond market, further weighing on valuations. On the interest rate side, CME’s FedWatch tool shows the market sees a nearly 90% chance of a 25 basis point Fed hike in September; Macro Risk Advisors warn that rate hikes possibly starting this week could trigger a roughly 10% correction in the S&P 500.

Central Bank Restraint: Warnings Are Not an Outright Bearish Call

It’s worth noting that the ECB remains quite measured even as it raises these warnings. Reports say their analysis clearly states it is not predicting a crash, and that the timing of corrections "cannot be known in advance"—only in retrospect; moreover—and this is important for investors rushing to exit—"this does not mean current prices are necessarily the ceiling." If AI ultimately proves transformative, even after a reset, future valuations could be "much higher."

In other words, the ECB is trying to untangle the mistaken linkage between "AI’s success" and "the safety of today’s prices" created by the current narrow but relentless rally. The quality of AI technology is one thing; the price investors pay for it is another—which is what Vujcic and his colleagues are truly concerned about.

0
0

Disclaimer: The content of this article solely reflects the author's opinion and does not represent the platform in any capacity. This article is not intended to serve as a reference for making investment decisions.

Understand the market, then trade.
Bitget offers one-stop trading for cryptocurrencies, stocks, and gold.
Trade now!

You may also like

Meta (META.US) strengthens self-developed AI chips! Plans to deploy data centers in the first half of next year, aiming to reduce inference costs and energy consumption

Meta Platforms plans to deploy its next-generation self-developed artificial intelligence chips to data centers in the first half of 2027, aiming to reduce the energy consumption and costs required to run AI models through customized chips.

智通财经2026/09/15 16:16
Meta (META.US) strengthens self-developed AI chips! Plans to deploy data centers in the first half of next year, aiming to reduce inference costs and energy consumption