Fund Managers Cut Technology Exposure to 12%, Biggest Reduction Since 2006
Key Facts
Fund managers cut their net overweight in technology stocks to 12% in November from 37% in October, the biggest monthly reduction since June 2006, according to the Bank of America survey cited by Reuters. That means investors still have a net positive exposure to the sector, but have clearly pared back the size of their bet. The survey covered 169 investors managing $461 billion in assets. The shift came as investors focused on whether heavy spending on artificial intelligence would continue to support earnings growth or face a slowdown. The survey results showed that concern extends beyond stock valuations to the pace of capital spending and the returns expected from new infrastructure.
Reuters said concerns about a slowdown in data-center spending after the U.S. midterm elections weighed on technology stocks. The Nasdaq lost more than 5% in October and was down a further 3% year to date through November 17. In the November 16 session, the S&P 500 fell 1.9%, marking its third consecutive loss, while the Nasdaq Composite declined 2.5% and the Dow Jones Industrial Average fell 1.4%, according to the Associated Press. The agency attributed the drop to technology stock selling led by Nvidia, while Treasury yields retreated after rising the previous week on inflation concerns. The moves reflect the timing of the position cuts alongside an actual wave of pressure on the indexes, even as net exposure to technology remained positive in the November survey.
The survey recorded a record net 38% of investors saying companies were overinvesting. In addition, 52% of respondents expected one of the AI giants to announce a cut in capital spending during the six months following the survey. The AI bubble was the biggest tail risk respondents feared, followed by a credit event at a major company in the sector, Reuters reported. Still, responses did not point to broad expectations of an imminent recession: only 3% of respondents expected one. Their cash holdings were at a three-month low, suggesting that reduced technology exposure came alongside continued investment in other assets.
Goldman Sachs Research's earnings estimate puts this shift in the context of the index's reliance on AI investment. The firm said nearly half of S&P 500 earnings-per-share growth in 2026 would come from that investment. The index's earnings per share grew 51% year over year in the second quarter of 2026 and 26% over the previous four quarters, compared with average growth of about 7% over the past 30 years. Goldman expected the largest U.S. cloud computing giants to spend $800 billion on capital expenditures in 2026, up 94% from 2025. That level of spending makes company results and forecasts for data centers an important factor in investors' assessment of the market's earnings trajectory.
Financing risks stand alongside concerns about returns. The Federal Reserve estimated that the largest U.S. banks have nearly $1 trillion in financial exposure to technology companies, data-center developers and other participants in the AI infrastructure boom, according to Axios, citing the November 2026 Financial Stability Report. The report said companies borrowed more than $680 billion to fund data centers and advanced computing infrastructure during the first ten months of 2026, about five times the amount raised throughout 2023. The share of respondents in the Federal Reserve's financial risk survey who cited AI as a threat to financial stability also rose to 87% in fall 2026 from 50% six months earlier. Federal Reserve Chair Kevin Warsh said disappointment in AI could lower expected returns and valuations, and perhaps lead to an economic slowdown and a more pronounced rise in unemployment, but described that as "an illustrative scenario, not a forecast."
The results of this spending are expected to show up in the earnings trajectory over the coming years. Goldman expects S&P 500 earnings per share to grow about 11% in each of 2027 and 2028, to $415 and $460, respectively, while estimating that cloud computing giants will spend about $1.2 trillion in 2027 and $1.4 trillion in 2028. But it expects AI investment's contribution to earnings to slow and become a marginal drag in 2028 as the impact of spending fades and productivity gains become more important. It estimated that a $250 billion upside or downside surprise in capital spending over the following year could move S&P 500 earnings growth by about 6 percentage points in the same direction. Nvidia's earnings report was the anticipated event after the market close on November 17, in interim CEO Colette Kress's first appearance before Wall Street since taking the role, according to the Associated Press.