Bernstein has recommended buying shares in companies exposed to artificial intelligence infrastructure, arguing that concerns about political opposition to data centre construction in the United States may have created an attractive opportunity ahead of Tuesday’s midterm elections.
The brokerage identified Quanta Services, Eaton, Vertiv and Equinix among its preferred stocks. It warned that shares could initially come under pressure if candidates supporting tighter restrictions on data centre development perform strongly, but said the market may have overstated the risk of projects being halted.
Bernstein estimates that 244 gigawatts, or about 56% of the 433 GW US data centre project pipeline, is exposed to political risks based on election probabilities and candidates’ policy positions. However, it expects political debate to focus primarily on requiring developers to cover a larger share of electricity and grid infrastructure costs rather than blocking construction altogether.
The brokerage maintained Outperform ratings on Quanta Services, with a price target of $804, Eaton at $534 and Vertiv at $376. It also favoured Equinix, with a $1,270 target, and Digital Realty, with a $226 target, citing their exposure to data centre infrastructure demand.
Other preferred US stocks included nVent, with a $233 price target; Johnson Controls, at $194; Trane Technologies, at $555; Forgent, at $48; and Legence, at $103. Bernstein also maintained an Outperform rating on GE Vernova, with a $1,298 target, highlighting opportunities across the electrical equipment and power infrastructure supply chain.
In Europe, Bernstein favoured Siemens, with a €330 price target, Siemens Energy, at €210, and Schneider Electric, at €340.
The recommendations come as public resistance to data centre expansion grows, with electricity costs and environmental concerns becoming significant political issues. Bernstein said data centre construction represents a major economic contribution, accounting for around 8% of state-level GDP, while craft labour wages account for approximately 3%.
The brokerage expects developers to face greater financial obligations for power infrastructure and grid upgrades. Nevertheless, it believes those additional costs would be relatively immaterial and unlikely to derail the wider investment cycle in AI infrastructure.
Sources: Investing.com News.