Chip Stocks Slide as Election Results Cloud Data-Center Outlook

Marvell shares fell more than 5% in Wednesday morning trading as election results triggered a broad selloff in chip stocks tied to the AI data-center buildout. Intel and Sandisk each dropped more than 2%, Micron Technology slipped about 2%, and Nvidia traded down fractionally along with the PHLX Semiconductor Index.

The trigger was political, not corporate. A blue wave swept both chambers of Congress, according to the source report, and Democrats won key governors' races after campaigning on tougher rules for data-center construction. Wall Street's forecasts for semiconductor demand rest heavily on new data-center capacity coming online, so investors are now pricing what the new political backdrop means for chips and infrastructure names.

What does the election actually change?

Cody Acree of StoneX told MarketWatch he doubts the election outcomes will hinder data-center buildouts in the near term. His reasoning is concrete: the money is already committed.

"While the election raises the risk of tougher permitting, power and water restrictions, and less favorable tax treatment, it does not unwind projects already funded, contracted or under construction," Acree said.

Texas sits at the center of investor attention. Acree calls the state "the clear epicenter of the U.S. AI data-center buildout, with the state likely accounting for roughly 30% of national capacity additions" through 2028. That's nearly a third of the country's planned capacity concentrated in one state.

Why Texas matters so much

Democrat Gina Hinojosa won the Texas governor's race after a late surge. During her campaign, she promised a moratorium on data-center construction until state legislators set new rules for the industry. A governor pledging to pause the buildout in the state carrying the largest share of national capacity is a genuine risk factor for chip demand.

Acree still sees the pipeline holding. "The Texas AI buildout remains intact and exceptionally large, even if the political headwinds are shifting toward more uncertainty regarding planned project approval," he said.

He characterized Wednesday's market action as a "reset" of valuations within the tech sector, "not evidence that 2027 data-center demand or build rates have suddenly collapsed."

Where the long-term risk sits

The bigger issue, Acree said, could emerge over time if investors start believing that policy uncertainty around costs or approvals will constrain the AI buildout farther out. The stakes for chip stocks are direct.

"That matters because most chip stocks are valued based on the assumption of several years of nearly uninterrupted growth," he said.

D.A. Davidson analyst Gil Luria told MarketWatch the election outcomes mean data centers will become more expensive to build. In some states, he warned, construction may become "prohibitively expensive."

Luria pointed to how these decisions actually get made. Data-center approvals happen at the local level, and each municipality negotiates its own terms with companies pursuing projects. That fragmentation means regulatory friction can surface state by state, city by city, rather than through one federal lever.

"Overall, the additional hurdles for data centers may act to slow down the buildout in the US, which could limit the deployment of the chips that go into those data centers," Luria said.

The bottom line for investors

Wednesday's declines hit the semiconductors most exposed to AI infrastructure hardest. Marvell's drop of more than 5% outpaced Intel, Sandisk and Micron, while Nvidia and the PHLX Semiconductor Index moved only fractionally — suggesting the market is discriminating between names levered to future data-center builds versus those with broader revenue bases.

The near-term picture, per both analysts, is intact: funded and contracted projects proceed regardless of election results. The open question is whether permitting fights, utility restrictions and tax changes in states like Texas raise costs enough to slow projects planned but not yet financed — a risk that matters because chip valuations already assume years of uninterrupted AI-driven growth.