- A record 38% of investors believe companies are overinvesting, and 52% expect an AI hyperscaler to cut capex within six months, while 40% do not anticipate such a reduction.
- AI bubble risks now rank as investors' biggest market concern, with hyperscaler credit events a close second.
- Net technology overweight plunged to 12% from 37%, the largest monthly reduction since June 2006.
Investors cut their exposure to technology stocks at the fastest monthly pace since June 2006, according to Bank of America (NYSE:BAC)'s November fund manager survey, as growing concerns over artificial intelligence spending prompted a sharp reassessment of positioning. The survey, which covered 169 investors managing $461 billion, showed net tech overweight plunged to 12% from 37%—a 25-percentage-point decline that still leaves investors net overweight the sector, but far less enthusiastically than a month earlier.
AI capex cut expectations jump
A sharp increase in expectations for an AI spending slowdown has emerged, with 52% of investors now expecting a major AI hyperscaler to announce a capital expenditure cut within six months, up from 30% in October. The remaining 40% do not anticipate such a reduction. This shift highlights growing investor concerns about the sustainability of AI infrastructure spending and potential pressure on technology stocks.
The retreat comes as a record net 38% of respondents say companies are overinvesting, while 52% expect an AI hyperscaler to announce a capital-expenditure cut within six months, up from 30% in October. Only 16% want companies to increase capital spending, the lowest share since May 2020. AI bubble risks now rank as investors' biggest market concern, selected by 34% as the biggest extreme market risk, followed by 31% citing a hyperscaler credit event.
AI bubble tops global market risks
The survey shows investors increasingly concerned about risks tied to the AI investment boom. Some 34% identify an AI bubble as the biggest tail risk to markets, while 31% cite an AI-related credit crisis as the second-biggest threat. The findings highlight growing concerns over AI valuations, heavy capital spending and debt exposure. Investors increasingly see AI-related shocks as the most significant potential threats to global financial markets.
A reassessment, not an abandonment
The survey points to a recalibration of expectations rather than a wholesale exit from technology. Semiconductor stocks remain the most crowded trade, though selected by 39% of respondents, down from 82% in July. That combination—less enthusiasm but still-heavy positioning—helps explain continuing concern about further selling. Average cash holdings actually declined to 3.6% from 3.9%, suggesting investors are not battening down the hatches entirely.
"Investors want a pause in AI capital expenditure, not another acceleration," said Michael Hartnett's team at BofA, which remains "tactically risk-off" and argues the survey still does not show enough investor fear to justify buying.
The results underscore a broader debate about the financial returns from the AI boom. Goldman Sachs (NYSE:GS) has estimated the largest technology companies were on course for roughly $800 billion in capital expenditure in 2026, with analysts expecting $1.2 trillion in 2027. Those figures represent estimates, not completed spending, but they highlight the scale of investment that investors are now questioning.
BofA economists have warned that a sustained AI-market correction could weaken the economy both directly, through reduced investment, and indirectly, through a reversal of stock-market wealth gains that support spending. The IMF raised a similar concern in October, noting that disappointing AI investment or profitability could trigger an asset-price correction, with leverage and financial interconnections transmitting losses beyond the sector.
Credit quality in focus
The concern is not limited to equity valuations. Some 38% of respondents identify hyperscaler bonds as the likeliest source of a credit event, ahead of private credit at 30%. That shift in focus comes as the IMF's review of 14 AI-related transactions since 2023 identified about $300 billion in equity investments and $650 billion in debt linked to $1.4 trillion in future revenue commitments. On Monday, IMF Managing Director Kristalina Georgieva warned that the global economy was relying heavily on the "AI leg," with circular AI financing—investment that returns to the investor through purchase commitments—exceeding $1 trillion, roughly two-thirds debt-funded.
The survey also showed investors reduced their bond underweight by 13 percentage points, the largest monthly shift toward bonds since April 2025, although they remained net underweight. Half of respondents still expect a soft landing, compared with just 3% expecting a hard landing, suggesting the evidence points more clearly to a reassessment of AI spending and market leadership than to wholesale withdrawal from risk assets.
A separate development on Monday added a layer of regulatory risk to the AI infrastructure theme, as the Trump administration announced restrictions intended to prevent Chinese companies from bypassing chip controls by renting computing capacity in overseas data centers. The measures, which affect advanced-chip shipments to several Southeast Asian countries and Brazil, do not explain the survey responses themselves, as the survey closed on November 12, before the announcement.
Bank of America did not respond to a request for comment on the survey findings. The report is based on responses collected November 6–12.