As a growing part of the financial community questions whether artificial intelligence has created a bubble, Fabiana Fedeli takes a different view. The CIO for Equities, Multi-Asset and Sustainability at M&G Investments argues that AI is not the source of a fundamental market imbalance, but rather acts as an accelerator of price adjustments in an environment that has become extremely fast-moving.
According to a Reuters survey conducted in November 2025, more than half of asset managers believe investment in artificial intelligence constitutes a bubble, while nearly 45 percent cite AI as the main tail risk for global markets in 2026. Earlier this year, Jeremy Grantham, co-founder of GMO, also warned about excessive market concentration around a small number of large technology stocks.
Less volatility, more market speed
Fedeli did not deny high valuations or the potential for excesses in certain segments. She argued, however, that the diagnosis is often misplaced. In a conversation with Investment Officer, she presented data pointing to a clear acceleration in market dynamics rather than a structural increase in volatility. In other words, markets have not become fundamentally more unstable, but they do react more quickly.
One figure illustrated this shift. In April 2025, the market for S&P500 index futures experienced intraday swings so large that the total value of those price movements amounted to nearly 2.5 trillion dollar within a single trading day. This occurred without any change in economic prospects or earnings expectations. Such episodes show that markets are now capable of generating very large moves over extremely short time horizons. “Moves that would once have been interpreted as a strong signal may today be nothing more than noise”, Fedeli noted. The speed of these adjustments, she argued, makes short-term fundamental analysis increasingly difficult.
This dynamic is not confined to US markets. Since 1975, the Japanese market has experienced only five episodes of extremely high short-term volatility. Two of these occurred between the summer of 2024 and the spring of 2025, at levels comparable to those seen during the 1987 stock market crash or the global financial crisis. The clustering of such extreme episodes in such a short period points to a regime shift in market behaviour: markets have become faster and more sensitive to dominant narratives.
Widespread AI, highly divergent outcomes
It is within this framework, Fedeli argued, that the role of artificial intelligence should be understood. AI has become deeply embedded in the real economy. For investors, the key question is therefore no longer whether they have exposure to technology—which has become almost universal—but which companies are able to translate AI into tangible economic results. This distinction is clearly reflected in equity performance. In 2025, the median return stood at around 13 percent, while Nvidia rose by 38 percent and Google performed even better.
Even among the largest US technology companies, markets are now making much sharper distinctions between firms, depending on the credibility of their earnings trajectories. This dispersion in performance directly accelerates price adjustments. Earnings releases, or even limited signals about the use of AI, can quickly trigger arbitrage activity, sometimes out of proportion to what has actually changed at a fundamental level.
The winners are not always obvious
The beneficiaries of a technological breakthrough are not necessarily those that dominate the narrative at the moment it emerges, Fedeli stressed. During the dotcom boom, Walmart successfully used the internet to transform its logistics, inventory management and operational efficiency. Yet at the time, it was not regarded as a defining technology company.
Artificial intelligence could produce similar outcomes. Companies that currently sit outside the prevailing technology narrative may still achieve significant productivity gains from AI—gains that are not yet fully reflected in their valuations.
In markets capable of such rapid movements, Fedeli argued for strict investment discipline. During very short-lived corrections, such as those seen on “Liberation Day”, the issue is not whether to reassess an investment thesis, but whether anything fundamentally has actually changed.