Inside S&P 500 AI boom, industrials are getting as rich as tech stocks

As the AI infrastructure buildout and global geopolitical developments spur massive spending, an old-guard sector of the economy is getting a level of attention from investors that rivals the action in tech. The industrials sector of the S&P 500 is trading at a price-to-earnings ratio above 30, a level investors more often associate with high fliers, and a P/E ratio that is well above industrials’ long-term average, which is closer to 20.

“If you look at [Industrial Select Sector SPDR] XLI from State Street, its valuations are really high relative to the S&P 500,” Cinthia Murphy, director of research at VettaFi, said on the most recent “ETF Edge.”

“It’s as high as tech, so it really is a sector that has really had its moment in the sun and picked up a lot of attention,” Murphy said.

The race to construct AI data centers in hopes of gaining a durable business advantage in a rapidly-shifting digital infrastructure has gripped the corporate world and the tech sector. Alphabet, as part of its earnings report on Wednesday, forecast capex for this year of $195 billion to $205 billion. Prior guidance was for spending of $180 billion to $190 billion.

McKinsey & Company estimates suggest that this investment won’t be stopping anytime soon – with reports showing that global spending on data centers could reach nearly $8 trillion by 2030. The vast majority of this spending will be dedicated to data center infrastructure and IT equipment. 

Alphabet warned those spending numbers could go even higher in 2027.

“AI is a tech play, but nothing happens without the build out of the infrastructure,” Murphy said. “There’s a whole backbone infrastructure that needs to be built, and that has really pushed up industrials.”

“We have only just begun this buildout. We are a few hundred billion dollars into it. Trillions of dollars of infrastructure still need to be built,” Jensen Huang, Nvidia CEO, said in a March blog post. “This is becoming the largest infrastructure buildout in human history.” 

The vast majority of projects in the next stages of a nationwide data center expansion are expected to take place in rural areas. In order to accommodate the staggering power demand associated with data centers, expanding rural power grids may prove essential. Peak production capacity for rural power grids in the U.S. remains comparatively limited, on average. Now, there exists a push to develop facilities that regularly consume up to twenty times their existing power capacity, intensifying corporate demand and the need for industrial support. This demand comes despite rising public animosity towards new data center construction, driven by local concerns surrounding strained power grids and rising electrical costs.

Constructing this AI backbone across the U.S. requires new electrical substations, strengthening high-speed fiber internet capabilities, and developing…