It’s Time to Shift Your AI Bets

Utility stocks are the next great beneficiaries of the artificial intelligence boom.

Where do you go in a stock from a one-year gain of 240 percent? At least some investors are betting on a repeat performance for leading 3D graphics processor Nvidia Corp (NSDQ: NVDA), which is already up another 34 percent year-to-date.

Unfortunately, such high expectations are all too easy to disappoint. And now trading at a price of 85 times trailing earnings—almost three times the Nasdaq 100 and four times the S&P 500 multiple—Nvidia is priced for perfection.

That’s not to say the company won’t grow its sales and earnings quite a bit the next few years. Artificial intelligence is top priority for technology giants like Amazon.com (NSDQ: AMZN), Meta Platforms (NSDQ: META) and others. And Nvidia has 90 percent share of the AI training market, as well as 50 percent of the AI inference market.

Based on that dominance, a major investment bank this week raised the stock’s upside target to $800. Even that, however, would only be an additional gain of about 20 percent for Nvidia. And with 58 investment banks already rating the stock buy, it’s fair to ask who the new investors will be to push it higher. Company directors and executives, for example, are heavy sellers so far in 2024.

The artificial intelligence boom is still in its very early stages. But the higher Nvidia has climbed, the more difficult future gains will be. The best AI bets are no longer components makers. They’re energy utilities, without which deploying this technology is simply not possible.

Utility stocks were big time underperformers last year and have lagged so far in 2024 as well. The reason is the Federal Reserve policy of the past two years. By relentlessly pushing up interest rates, they’ve increased the attraction of money market funds and convinced some that utility investment-led growth plans are doomed.

We’re still early days for utilities’ Q4 earnings reports and guidance updates. But from what we’ve seen so far, management is actually ramping up spending on power grids as well as generation sources. And fueling AI’s need is a major driver.

Put simply, AI is a power hog. Demand for electricity is soaring from the giant data centers needed to enable widespread applications. And utilities are meeting it.

Industrial research firm McKinsey forecasts electricity demand from US data centers will more than double by 2030. Rival BCG predicts demand will reach 45 gigawatts by then, or 7.5 percent of America’s total power consumption. And facilities are literally sprouting up in every region of the country, meaning opportunity for companies across the board.

Much of the infrastructure to serve data centers will be in utilities’ rate base earning a guaranteed return. But much is also individual PPAs—power purchase agreements—between producers and individual corporations. And the result is a parallel burst in renewable energy demand, since these buyers tend to prefer zero CO2 sources that also zero out potentially volatile future fuel costs.

Companies like AES Corp (NYSE: AES) and NextEra Energy (NYSE: NEE) are at the center of the PPA boom. And judging from NextEra’s earnings guidance in late January, demand for projects is accelerating into 2024. In fact, management called its ongoing AI related projects just the “tip of the iceberg.”

Rising electricity demand from data centers means higher revenue and earnings for utility companies. But the bar for a major leg up in sector stocks this year is considerably lower: That’s simply affirming previous growth projections many on Wall Street have come to doubt.

AI demand is the basis for another leg up entirely. And that’s not even taking into consideration the potential efficiency benefits from AI applications.

Power grids were extremely complex animals even before residential solar adoption began creating massive volatility in 24-hour demand patterns. And even as the public’s tolerance for outages has shrunk to practically zero, external challenges are arguably greater than ever, from cyber security to increasingly intense storms, temperature extremes and wildfires.

AI adoption will enable a quantum leap in utilities’ ability to collect and process data. And that adds up to greatly enhanced preparation, prevention and response when grid-level events do occur.

AI could also eventually enable utilities to control intermittency of solar and wind energy, by vastly improved integration of storage and adapting to demand patterns to get energy where it’s needed. That means potentially dramatic reductions in operating/maintenance costs and fuel costs, which are by far the sector’s largest line item expenses.

There’s quite literally no other industry with more to gain and less to lose from AI’s massive, ongoing adoption. And unlike the Nvidia’s of the world, utilities face absolutely no burden of high expectations at this point: AES, for example, sells for less than 8 times its last 12 months earnings.

For more on AI and utilities, check out the upcoming issue of www.conradsutilityinvestor.com, which will post on February 12.