2 Hidden AI Stocks to Buy in the S&P 500 Sell-Off

In This Article:

It's no secret that artificial intelligence (AI)-related stocks have sold off in the recent market correction. It's partly due to investors locking in gains, competitive concerns relating to Chinese start-up DeepSeek, and, most notably, valuation concerns. Consequently, focusing on companies with relatively secure market positions that trade on tempting valuations makes sense if you want to buy on the dip.

Retail electricity and power-generation company, Vistra (NYSE: VST), and data center equipment maker, Vertiv (NYSE: VRT), fit the bill. Here's why.

Vistra, data centers, and AI

Vistra was one of the best performing stocks on the S&P 500 (SNPINDEX: ^GSPC) last year, so it's no surprise that investors took some profit on it this year. Still, that doesn't mean the power company isn't an alluring stock to buy right now.

The excitement around the company is represented in the table below, which demonstrates how Vistra increased its generating capacity from nuclear and renewable energy last year. As a result of the acquisition of Energy Harbor and the acquisition of its outstanding 15% interest in Vistra Vision (the subsidiary that houses its nuclear business, and renewables and storage projects), Vistra significantly increased its nuclear powered generating capacity and also added to its renewable generating capacity.

Fuel Source

Type

Net Capacity 2023 (MW)

Net Capacity 2024 (MW)

Growth

Natural Gas

Combustion, combined cycle, steam turbine

24,313 MW

24,120 MW

(0.8)%

Coal

Steam turbine

8,428 MW

8,428 MW

0%

Uranium

Nuclear

2,400 MW

6,448 MW

169%

Renewable

Solar/Battery

1,358 MW

1,474 MW

8.5%

Fuel Oil

Combustion turbine

203 MW

187 MW

-(8)%

Total

N/A

36,702 MW

40,657

10.8%

Data source: Vistra SEC filings. MW=megawatts.

It's good news because the market, and more importantly, the hyperscaler data centers (large data centers often optimized to support data-intensive AI-application growth), have warmed to the idea that nuclear-powered electricity is the solution to their long-term power needs to support AI growth.

All three of the largest cloud service providers -- Microsoft's Azure, Amazon.com's Amazon Web Services, and Alphabet's Google Cloud -- inked deals last year to procure power for their data centers from nuclear-powered plants. Nuclear power is carbon-free, reliable, available 24/7, and doesn't have the downside issues of intermittency that characterize renewable energy sources.

It's a good solution for hyperscalers that want to solve their power needs and meet their emissions goals simultaneously. In addition, investors shouldn't overlook the ongoing electrification-of-everything megatrend (electric vehicles [EVs], web-connected devices, smart buildings/infrastructure, data centers, heat pumps, EV charging networks), which continues to drive electricity-demand growth before and after AI became a common talking point.