Set Trending Topics as a preferred source on Google.
A team of economists at the European Central Bank recently published a warning on the ECB Blog about the consequences of a possible correction in AI-driven stock markets. The core message from Malin Andersson, Johannes Breckenfelder, Stefano Corradin, Kalin Nikolov, and Maria Antonietta Viola is that a setback is likely—and the euro area would be affected through households’ fund holdings far more strongly than many investors may realise.
Valuations at dot-com-era levels
The analysis starts with valuation metrics in the US market. According to the authors, the cyclically adjusted price-to-earnings ratio (CAPE ratio) of the S&P 500 is close to its historical high—at a level last seen during the dot-com bubble. Valuations have also risen in the euro area, though to a lesser extent.
The ECB experts outline two research-based explanations for why technological shifts so often end in a boom-bust cycle. The first is a rational interpretation: with a new technology, future productivity is highly uncertain, while potential gains are almost unlimited in the best case. This option value pushes up the valuations of early adopters. The example in the article is Nvidia, whose share price has increased twentyfold since 2022.
Even if the technology delivers on its promise, this logic suggests that prices can fall. The reason lies in the nature of the risk: as long as AI is an experiment undertaken by individual companies, the risk can be diversified within a portfolio. Once it becomes widespread across the economy, it turns into an economy-wide risk that cannot be diversified away. Investors then demand a higher risk premium, which historically has generally had a stronger effect than the positive impact of rising cash flows. The second, behavioural-economics interpretation assumes that overly optimistic market participants drive prices beyond fundamentals. When sentiment shifts, the correction is even sharper.
Both explanations lead to the same conclusion: at some point, a setback will follow. According to the ECB Blog, it is impossible to predict when in advance, since boom-bust patterns only become clear in hindsight. The authors also stress that today’s prices need not be an upper limit: if AI proves transformative enough, valuations could still be higher after a correction than they are today.
€440 billion through funds and ETFs
The most relevant part of the article for European savers concerns who in the euro area is actually exposed to the Magnificent Seven. The ECB’s answer: mostly indirectly, through investment funds and ETFs, rather than through directly held shares.
Using look-through data on fund investors (SHS look-through, market values for the third quarter of 2025), the ECB estimates euro-area households’ exposure to US technology stocks at around €440 billion. The blog post explicitly notes that those affected may not necessarily be aware of this concentration risk, as inflows mainly occur through low-cost, broadly diversified index products. Insurers and pension funds also hold substantial positions, according to the analysis.
From the ECB’s perspective, this fund structure is itself a transmission channel. In a sharp correction, funds must sell assets to meet redemptions—first liquid securities and, under sustained pressure, harder-to-sell positions as well. This pushes valuations down further and triggers additional outflows. What begins as a problem for individual portfolios thus becomes a question of financial stability.
The analysis considers a scenario in which a stock-market correction coincides with broader market instability particularly sensitive. Unlike during the dot-com era, the scope to counteract such a shock with interest-rate cuts or fiscal measures is more limited today.
Europe’s market is more sober—but tightly linked
The experts largely reject the idea that the euro area faces a domestically generated correction of its own. Valuation metrics are clearly below US levels; productivity and margins are rising in Europe’s ICT sector; business sentiment in digital services appears steady; and AI adoption among European companies has increased noticeably only a few years after the launch of ChatGPT. Digital investment in the euro area has risen by more than three times cumulative GDP growth over the past decade.
However, this is immediately qualified: European stock exchanges are dominated by old-economy companies, which limits the potential for a domestic downturn. At the same time, stock markets on both sides of the Atlantic have historically moved very closely together, meaning that a US correction would reach the euro area. According to the blog post, the effects could extend beyond financial markets and affect sentiment, financing conditions, and hiring behaviour in the euro area. The authors’ conclusion: an AI-related setback in the United States would not remain a purely American problem.
Aus Datenschutz-Gründen ist dieser Inhalt ausgeblendet. Die Einbettung von externen Inhalten kann in den Datenschutz-Einstellungen aktiviert werden: