Key takeaways
- Wells Fargo upgrades AI infrastructure outlook to Neutral
- Semiconductor stocks have fallen 38% from June highs
- AI spending could stop boosting earnings in 2028
Wells Fargo has upgraded its outlook on US artificial intelligence infrastructure stocks from Negative to Neutral after a brutal semiconductor selloff erased much of the sector's previous valuation premium.
In a November 9 research note, analysts led by Ohsung Kwon estimated that semiconductor stocks had declined approximately 38% from their June highs, leaving valuations more attractive despite concerns about slowing AI investment.
The change follows a broader semiconductor valuation collapse, as investors increasingly question whether technology companies can generate sufficient returns on their enormous computing investments.
Wells Fargo's latest stance is less pessimistic, but it does not amount to a broadly bullish recommendation.
Wells Fargo Sees Improved AI Stock Valuations After 38% Crash
The bank argues that semiconductor stock prices already reflect a substantial slowdown in anticipated earnings growth.
According to its analysis, current valuations imply approximately 8% year-over-year earnings growth, considerably below the expectations that previously supported the AI investment boom.
The reassessment follows an earlier Nasdaq 100 correction, during which concerns about AI capital expenditures and financing costs triggered widespread selling.
Separate valuation data show how dramatically market expectations have changed.
The broader technology sector recently traded at 17.1 times forward earnings, compared with 17.9 times for the S&P 500, marking its first valuation discount since 2017.
That does not necessarily mean every AI stock is undervalued. However, investors are now paying substantially less for expected future earnings than they did earlier in the year.
AI Spending Added Eight Percentage Points to S&P 500 Earnings Growth
Wells Fargo also highlighted how heavily corporate earnings have benefited from the AI infrastructure boom.
The bank estimates that AI-related capital spending contributed approximately eight percentage points to S&P 500 earnings growth during the past year.
That support is expected to continue through 2027, but analysts warned that its contribution could largely disappear in 2028 as investment growth slows and depreciation expenses increase.
The concern is particularly relevant as companies commit billions to servers, electricity infrastructure and data centers.
Those investments have supported semiconductor suppliers such as Nvidia and AMD, but they also create significant future expenses.
The risk is already visible among AI infrastructure companies financing expansion through borrowing.
For example, CoreWeave's elevated default protection costs illustrate how investors are assessing financial risks alongside demand for computing capacity.
