Frankfurt, 21 August 2026 — EBM Newsdesk Analysis —Katie Winearls
Economists at the European Central Bank have warned that soaring stock market valuations driven by enthusiasm for artificial intelligence are likely to suffer a correction — even if investors are ultimately right about AI’s ability to transform the global economy.
In an ECB blog published this week, five economists argued that history suggests technology-driven investment booms frequently end in sharp market reversals, regardless of whether the underlying innovation eventually delivers substantial productivity gains.
Join The European Business Briefing
New subscribers this quarter are entered into a draw to win a Rolex Submariner. Join 40,000+ founders, investors and executives who read EBM every day.
SubscribeUS equity valuations are now close to historical peaks when measured by the cyclically adjusted price-to-earnings, or CAPE, ratio. European valuations have also increased, although less dramatically. Much of the surge reflects expectations that artificial intelligence will deliver higher corporate profits and reshape large parts of the economy.
AI can succeed while stocks fall
The ECB economists’ central argument challenges the idea that a market correction would necessarily mean investors were wrong about artificial intelligence.
They point to previous technological revolutions including railways, electricity, radio and the internet. Each generated genuine economic transformation while also producing periods in which shares of companies associated with the new technology rose rapidly before falling heavily.
One explanation is what economists describe as the “option value” surrounding emerging technologies.
When the ultimate impact of a technology is extremely uncertain, investors may rationally assign very high valuations to companies capable of becoming dominant winners. Nvidia’s extraordinary rise since 2022 is cited by the ECB economists as an example of investors pricing in potentially enormous future gains.
But as a technology spreads through the wider economy, the nature of the risk changes.
Instead of being concentrated in individual companies, uncertainty becomes economy-wide. Investors can no longer diversify away the risk as easily and may demand a higher risk premium. That can push valuations lower even while profits continue growing.
The alternative explanation is less benign: investors become excessively optimistic and push prices beyond levels justified by fundamentals. When confidence fades, the resulting decline can be even more severe.
Europe has €440bn exposure
The warning has significant implications for European investors.
Eurozone households have around €440 billion invested in the so-called Magnificent Seven US technology companies — Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla. European pension funds and insurers have roughly the same level of exposure, according to the ECB analysis.
That leaves Europe vulnerable to any major Wall Street technology sell-off.
European stocks are not as highly valued as their US counterparts, but the two markets remain closely correlated, meaning a sharp decline in America would probably spread across the Atlantic.
The risks could also be harder for policymakers to manage than during previous market crashes.
The ECB economists warned that governments and central banks have less room to respond through interest-rate cuts or fiscal stimulus than they did during the aftermath of the dot-com collapse. A correction combined with wider financial instability could therefore have significantly more serious consequences.
Timing remains impossible to predict
None of this means an AI crash is imminent.
The economists stress that the timing of any correction cannot be predicted reliably, while stronger-than-expected profit growth could continue supporting valuations.
But their broader message is uncomfortable for investors: artificial intelligence does not have to disappoint for technology shares to fall.
AI could transform productivity, create enormous corporate profits and become one of the defining technologies of the century — while investors who bought at the highest valuations still suffer substantial losses.
That distinction may become increasingly important as markets attempt to decide how much of AI’s promised future has already been priced in.
