The ECB puts EUR 440bn on Europe's AI trade, then drops the bubble question

Five economists at the European Central Bank published an analysis on Monday arguing that stock valuations are likely to correct, and that the conclusion holds whether current prices are fundamentally justified or driven by investor exuberance ^[1]. The number attached to that argument is the part operators should file: euro-area households hold roughly EUR 440 billion of exposure to US technology equities, much of it through funds and ETFs that can amplify losses when redemptions come in size ^[5].

The post is by Malin Andersson, Johannes Breckenfelder, Stefano Corradin, Kalin Nikolov and Maria Antonietta Viola, and the views are the authors' own rather than the ECB's ^[2][3]. It places US valuations close to their historical peak on the CAPE ratio, which compares prices with inflation-adjusted earnings averaged over the preceding decade, and notes euro-area valuations have also risen, by less ^[19][28]. "Economic research on past technological revolutions points to a worrisome conclusion: a correction of current stock market valuations is likely," the economists wrote, citing the 19th century railway boom, the spread of electricity and radio in the 1920s, and the 1990s internet ^[4][27].

The mechanism is worth reading carefully because it does not require anyone to be wrong now. On the rational account, extreme uncertainty about a new technology's payoff creates an option value that lifts price-to-earnings ratios for early adopters, with Nvidia offered as the case where investors are paying for the chance of the next Google, building on work by Pastor and Veronesi in 2009 ^[20][21]. While the technology sits in a few firms, failure is diversifiable; as adoption spreads, the uncertainty becomes economy-wide and investors demand a higher risk premium, which historically prevails over cash-flow gains ^[7]. Prices can fall while profit growth stays robust ^[8]. The behavioural account gets to the same place by a shorter route: overconfident investors bid past fundamentals, and the market falls when that confidence fades ^[9]. Both paths imply a boom then a pullback, possibly followed by recovery, and the authors are explicit that the timing is unknowable in advance and identifiable only with hindsight ^[11][10].

The transmission channel is the fund wrapper itself. Most euro-area holdings of the Magnificent Seven sit in mutual funds and ETFs rather than direct shareholdings ^[12], and a correction can force those funds to sell to meet redemptions, starting with the most liquid positions and ending with the impaired ones, pushing prices lower and triggering further redemptions ^[13]. That is why the authors treat a Mag7 correction as a euro-area financial stability question rather than a portfolio question ^[14]. Insurers and pension funds also hold significant positions, so total euro-area exposure runs above the EUR 440 billion household figure, which is measured as of the third quarter of 2025 ^[15][22][6]. Retail investors are exposed through the prevalence of those names in global index funds and pension funds, potentially without knowing it ^[18].

Europe's own market looks less stretched: price-to-earnings ratios are considerably below US levels, ICT productivity and markups are rising, and the digital services business climate does not look exuberant ^[23]. Regional digital investment over the past decade grew more than three times cumulative GDP growth, with firm-level AI adoption picking up notably a few years after ChatGPT's 2022 launch ^[24][25]. The exposure, in other words, is imported.

What to watch is the policy backstop, which the authors say is thinner than in 2000: "Unlike in the dot-com episode, today's starting point leaves markedly less room to cut interest rates or use fiscal policy to cushion the fallout" ^[17]. Their own framing is that the real risk is not the correction but a correction into constrained monetary and fiscal space ^[16], and the blog tells investors to prepare for it ^[26]. The practical task is look-through concentration in default pension and index allocations, and gating or liquidity terms in the cross-border equity funds holding the position.