Global equity markets are showing growing signs of bubble-like behaviour, but the core artificial intelligence trade still appears to have further upside, according to Bank of America Global Research.
In its latest Global Equity Volatility Insights report, the bank argues that while pockets of the market are already displaying instability consistent with late-cycle excess, the main AI-linked segments of U.S. equities remain well short of conditions typically associated with an imminent bubble peak. Bank of America’s Bubble Risk Indicator (BRI) suggests that speculative pressure has intensified in select areas, including nuclear- and quantum-themed stocks and some Asian equity markets, notably South Korea’s Kospi.
By contrast, the central AI trade — spanning the S&P 500, Nasdaq Composite and the so-called Magnificent Seven — continues to show relatively subdued bubble signals. That divergence underpins the bank’s view that AI-related stocks may still have room to extend gains into 2026, even as broader market risks rise.
At the same time, Bank of America cautions that the overall trajectory of U.S. equities is increasingly reminiscent of past technology-led boom cycles. Analysts draw a parallel between the Nasdaq’s rally following the launch of ChatGPT in late 2022 and its climb after the release of Netscape in the mid-1990s — a period that ultimately culminated in the dot-com bubble.
The bank argues that dismissing bubble risks entirely would be complacent. While AI stocks have not yet reached extremes, the broader market is steadily moving toward a more fragile state as valuations stretch and volatility dynamics shift. In that context, AI is seen not as an exception to bubble dynamics, but as a potential catalyst for a larger, more prolonged asset-price boom.
In Bank of America’s assessment, the relative restraint in AI-related volatility measures does not signal safety, but rather suggests the trade may still be in an earlier phase. As enthusiasm spreads more widely across markets, the risk of excess is likely to rise — even if the AI core continues to lag the froth seen elsewhere
