Fund Manager Survey: Investors Cut Tech by the Most Since 2006
Professional investors pulled money out of tech at the fastest pace in 20 years, and a majority now expect at least one AI spender to slow down. The survey lands the same day Nvidia reports.
·4 min read
Bank of America's November fund manager survey, published on Tuesday, November 17, shows professional investors cut their technology holdings by the most since 2006. The net share of managers overweight tech fell 25 percentage points in a month, CNBC reported, and 52% of respondents now expect at least one AI hyperscaler to announce a cut to capital spending within six months, according to Fortune.
The shift matters because these are the investors who carried the AI trade, and they are now asking companies to spend less, not more. The survey arrived after the Nasdaq Composite fell 2.49% on Monday and hours before Nvidia's quarterly report.
- -25 pts
- Monthly drop in net tech overweight, the biggest since 2006
- 52%
- Expect a hyperscaler capex cut within 6 months
- 16%
- Want companies to raise capex, lowest since May 2020
- 3.6%
- Average cash level, still very low
The biggest tech exit since 2006
The November cut was the largest one-month reduction in tech exposure since June 2006, according to both CNBC and Fortune. It does not mean managers have turned bearish on the sector. Startup Fortune, which also covered the survey, put the remaining net overweight at 12%, down from 37% in October. Fortune noted that tech is still an overweight position even after the cut.
In other words, investors are trimming a large bet rather than reversing it. That leaves room for more selling if the news on AI spending gets worse, and it is why strategists watch the next few surveys as closely as this one.
Fund managers now want less AI spending
The clearest change is in what investors want companies to do with their cash. Only 16% of managers said they want companies to raise capital spending, the lowest share since May 2020, Fortune and Startup Fortune reported. Fortune described the message from the survey, as framed by BofA strategists led by Michael Hartnett, as a call to pause AI capex.
A majority, 52%, expect at least one of the big AI spenders to announce a capex cut within the next six months. That is a direct challenge to the plans the hyperscalers laid out in October, which we broke down in our analysis of AI capex and cash flow. Fortune also cited Goldman Sachs work showing hyperscaler capex expected to rise 58% next year to about $1.4 trillion, based on FactSet estimates, so the gap between what investors want and what companies have planned is wide.
Hyperscaler bonds named the top credit worry
Asked where a systemic credit event is most likely to start, 38% of managers pointed to AI hyperscaler bonds, ahead of private credit at 30%, according to Fortune. Startup Fortune reported the same 38% figure. A year ago that answer would have been hard to imagine: the largest tech companies were seen as the safest borrowers in the market.
The answer reflects how much debt now funds the build-out. We tracked that shift in the AI debt bubble and in CoreWeave's bond yields above 16%.
Still bullish on the economy, wary of AI
The survey is not a recession call. Only 3% of managers predict a recession, both Fortune and Startup Fortune reported, and the average cash level was 3.6%, a level CNBC and Fortune both described as uber-low. Fortune reported that survey respondents expect the S&P 500 to end 2027 at 7,926, compared with Monday's close of 7,416.59.
“We stay tactically risk-off.”
Startup Fortune quoted Hartnett as summing up the mood as euphoria on the boom and AI being over. The combination of low cash and a fast exit from tech is what makes the reading unusual: investors are still invested, but they are moving money away from the part of the market that led the last three years.
What the AI Bubble Index says
The AI Bubble Index read 61.4 after Monday's close, in its Normal band, down from 63 a week earlier. Sentiment and attention remains the hottest of the five families at 99.7, near the top of its range, which fits a week when AI bubble talk is everywhere. Momentum has cooled to 47, and Fragility, at 68.7, stays above Euphoria at 61.5. The systemic risk family reads 69.1. You can follow each family on the dashboard.
Surveys like this one are a positioning signal, one of the classic bubble indicators covered in our guide on how to detect a market bubble. A crowded trade that is being unwound can fall faster than fundamentals alone would suggest, and it can also stabilize once positioning is clean.
What to watch next
Nvidia's results and outlook after Tuesday's close, Anthropic's IPO pricing and its first trading day, expected Thursday, November 19, any hyperscaler that trims its 2027 spending plan, and whether the December survey shows tech moving to an underweight.Questions people ask
- What did the November 2026 BofA fund manager survey show?
- Managers cut tech exposure by the most since 2006, with the net overweight down 25 points in a month. 52% expect a hyperscaler to cut capex within six months, and only 16% want companies to spend more.
- Are fund managers underweight tech stocks now?
- No. Tech is still a net overweight after the cut. Startup Fortune put it at a net 12%, down from 37% in October.
- What is the biggest credit risk in the fund manager survey?
- 38% of managers named AI hyperscaler bonds as the most likely source of a systemic credit event, ahead of private credit at 30%, according to Fortune.
- Do fund managers expect a recession?
- Very few do. Only 3% predict a recession, and average cash holdings were a low 3.6%.
Sources
- Is the bubble popping? Fund managers just cut tech by the most since 2006 as Goldman sees AI having pumped profits 25% above trend · Fortune · November 17, 2026
- Investors lower tech stock exposure in biggest exit since 2006 · CNBC · November 17, 2026
- Fund managers cut tech the most since 2006, but they still own it · Startup Fortune · November 17, 2026
- Nasdaq falls more than 2%, extending slide to third day as Nvidia drags down AI trade · CNBC · November 16, 2026