Why the data center slowdown concern is overblown. Two stocks to buy amid the fear - CNBC

The demise of data centers has been greatly exaggerated. So says every utility CEO and energy-focused investor we spoke with at the Edison Electric Institute's annual conference in Arizona. Here's why...

There is a lot of misleading, misdirected, or straight-up incorrect information floating around about data centers and the AI buildout. I'm not going to dive into the old water-use story — you can do your own research on that — but rather focus this week on what the insiders know: Some newly elected politicians either don't have the power to stop the buildouts or, if they do, the buildout will likely just continue somewhere else.

Ohio and Georgia are important states to watch with their new incoming political class, but for the sheer importance in size and scope, I'll focus on the energy and data center capital: Texas.

Incoming Governor Gina Hinojosa and incoming Senator James Talarico both campaigned on promises to go after what they believe to be affordability issues around data centers. I say "what they believe to be" because it's unclear how much data centers may raise your electricity bill, if at all. 

First and foremost, data centers are largely a state issue, meaning officials elected to federal office have little control over outcomes, at least for now. In Texas, any cancellations or moratoriums on new builds will come from the statehouse in Austin. That's where current Governor Greg Abbott has already, and rather quietly, put his own "pause" on data center permitting. That delay has been underway for a few weeks. Also, in Texas, the Governor has less direct power than in some other states. The power the governor does have largely derives from the authority to appoint those in charge of the all-important permitting process.

In the Lone Star State, that power lies with two agencies: the Public Utility Commission of Texas (PUCT) and the even lesser-known Texas Commission on Environmental Quality (TCEQ). The PUCT regulates Texas' electricity, telecom, and water industries. It helps oversee growth of anything electricity-grid-related. So if a data center wants to connect to the public utility grid, it needs the commission's approval. Its chair — a pro-growth chap named Thomas Gleeson —is also on the board of Texas' electric grid powerhouse ERCOT and will serve his term with the utility commission through September 2029. Yes, three years from now.

That agency gets involved when a data center or power provider wants to connect to the electricity grid. But what about facilities that are off the grid and have their own dedicated power sources? Texas has less control over what happens then. That said, it does have to authorize certain environmental permits, and that responsibility largely lies with the TCEQ. The TCEQ was "BBI" (my own little term for "boring, but important") until data centers came along. It's not exactly the Alamo, but this obscure group — run by a regulator named Brooke Paup — is largely the "last stand" for permitting on those new builds. Paup was also appointed by Gov. Abbott and serves through next year. In other words, unless she and Gleeson step down or there is some way to oust them, the regulatory environment in Texas may not change much.

The bottom line in this brief foray into the Texas regulatory swamp is to highlight that there are things for investors to consider around the data center/energy world that may not be at the top of the headlines.

Important Note → Be mindful that I am an East Coast-based financial broadcaster and journalist, not an expert on Texas regulation or how the state operates. There may very well be things I don't know, things my sources aren't aware of, or other ways in which the new Governor and state House can assert power. For one, she recently said she may try to freeze electricity rates. Hint: that would be a terrible idea unless you want to damage the growth of badly needed new power sources like wind and solar.

Disclaimer is done. Let's assume I am wrong on Texas and there is a larger push to slow or stop data center growth in the state. Conversations with some pretty smart people lead me to believe that if those facilities aren't built in that state, they will be built in another: Louisiana, Oklahoma, or another state. Sources tell me that Louisiana Governor Jeff Landry and his team are already working the phones, telling companies that his state would love to welcome new business. Same for Oklahoma. Both are gas-rich, have plenty of power, and are only too happy to grab any data center dollars.

Let's acknowledge that not all data centers are going to be built… and never were. There are way too many plans to make them all happen. Some DCs were shelved before the election. Some others may soon be canceled, either for reasons of public pushback or otherwise. It's also not clear if we even need all the planned compute. There will be a time when companies realize they have enough — or at least enough for now — and turn off the spending taps.

"The midterm election results pose only modest risks to AI investment, for now. We expect newly elected Democratic governors in states that account for over a third of planned US data center capacity to impose moratoria on new data centers. However, these will likely only impact projects not yet permitted, leaving the vast majority of planned 2027 construction unaffected. We therefore expect data center construction spending to remain strong next year, though moratoria that are stricter or longer-lasting than we expect pose downside risks."

The firm adds that it still sees spending on AI-related hardware coming in around $1.4 trillion by the end of next year. In other words, the spending boom is going to keep going, at least for now.

Another key point for all you energy- or utility-focused investors out there: While data centers are one huge driver of electricity demand, they aren't the only one. Electric cars, reshoring manufacturing, and the "electrification of everything" will also grow the need for more gigawatts. This is why we asked every utility CEO we spoke with at the EEI conference if they saw any slowdown in their internal electricity growth estimates. All gave the same answer: no. Not one hinted at a downturn. Not Exelon's Calvin Butler. Not Entergy's Drew Marsh. And not Drew Maloney, head of EEI. Of course, power demand is about more than just data centers, but it's still important for investors to hear it from the top.

Side note: Many of you think I'm wrong about my "hot take" on data centers and the election. If you missed last week's Power Insider, this is how you responded to my X poll on the election and energy.

So… why am I writing this week about AI and data centers? Because, as I've said before and will keep saying, AI and data centers are just energy stories in disguise. If they grow, energy demand grows. Demand for solar, wind, and natural gas grows. The economy grows. Because, after all, energy is the economy.

Oh, and in the past week, I've flown from Newark to Rome to Chicago to Phoenix and back to Newark. My brain may be a little mushy, so bear with me.

So now that my big take on data centers and regulation is over, let's get granular. What energy stocks do you buy right now?

Let's get the "1-2-3" on six stocks: one in solar, two in power, and three in oil and gas.

One: William Blair is upgrading First Solar (FSLR) to Outperform. The election result is part of the story. Analyst Jed Dorsheimer and his team believe the new Congress and certain state governorships will benefit local support for utility-scale solar, and that First Solar is best-positioned, though he does caution investors that it may take "some time to show up in the bookings." He adds that Texas is the largest builder of solar, so the shift in state-level leadership is a big deal. Fair warning, however: upgrades do not mean the stock will go higher. Deutsche Bank upgraded FSLR back in early July — an upgrade I wrote about — and the stock has fallen 22% since.

Two: Staying with the macro power theme, Barclays' Nicholas Campanella is particularly optimistic on two utility names. He tells Power Insider that he "really likes NiSource (NI) and Xcel Energy (XEL) right now." It has to do in part with what I just wrote about above. Campanella says the political outcomes in states such as Texas and Georgia may drive investors into utilities in other areas, particularly those with "more stable and established data center policies in place." That would fit with Indiana-based NiSource and Minnesota-based Xcel. While Minnesota may have a Democratic tilt, the company operates in eight different states. It is also working with Alphabet (GOOGL) on a massive new data center, and CEO Bob Frenzel told us that he is confident they can successfully navigate it all.

Three: Now to the "3"… three picks on what powers some of the power: oil and natural gas.

In an interview on CNBC, Bernstein analyst Bob Brackett says he likes ExxonMobil (XOM), calling it a great "barbell" play between natural gas and refined products. Brackett — a former Exxon employee — also believes it's a relatively "safe" stock to own right now in the volatile and headline-driven world of energy. Brackett has an Outperform rating and $186 target on the stock, giving it about another 10% upside from here.

Brackett also likes the lesser-known Expand Energy (EXE). If you're not familiar with Expand Energy, don't worry; it's still relatively new. The company is the combination of Chesapeake Energy and Southwestern Energy. It's a natural gas behemoth, and Brackett says the gas story is a "great long-term story" that's trading below a "midcycle price," and Expand fits right into it. Brackett also notes that Expand has a new CEO. He has a $139 target on the stock, making him one of the more bullish analysts on the Street. Expand has a median price target of $123.54.

Meanwhile, the chart looks tasty on another energy name: Cheniere Energy (LNG). Piper Sandler technical guru Craig Johnson likes the stock, and while he isn't a fundamental energy analyst, he was one of the few to upgrade the entire sector back in February when the Iran war began. Cheniere — along with Venture Global (VG) — is a pure play on why I was just in Europe: the worsening European energy crisis. Cheniere exports massive quantities of U.S. natural gas to customers in Europe, Asia, and other parts of the world. Johnson likes Cheniere in part because it recently broke out of a "very big base," and he wants to buy stocks that are on the upswing.

Cheniere is a well-run, rather quiet company with a fantastic CEO in longtime energy executive Jack Fusco. Jack has forgotten more about running an energy business than most will know, and the company continues to expand production. As I've pointed out for over four years, without American natural gas, Europe would be in dire straits and cutting demand even more than it already is. It's not hyperbole to say that American natural gas is — literally — keeping the lights and the heat on across Germany, Italy, and other countries.

Remember this incredible stat: just one tanker of LNG can produce a year's worth of power for over 100,000 average-sized European homes. Next year, Europe is still planning to stop buying LNG from Russia. That means demand for American natural gas is only going one way: up.

This week's Inside Line is with Amrita Sen, founder and director of market intelligence at Energy Aspects. Amrita is one of the smartest and most well-connected energy experts in the world, and she took a few minutes from her constant travel to answer a couple of questions.

This week's Take A Look is a little outside the norm, but an important interview nonetheless. It's with the U.S. Ambassador to Italy, business magnate, and former star of CNBC's Billion Dollar Buyer, Tilman Fertitta. In a wide-ranging interview from the ambassador's residence in Rome, we talked about that day's stock market crash, business, the consumer, and, of course, energy. Full disclosure: Tilman and I have been friends for over 15 years.

This Washington Post article was published a few days after the election. The underlying research from Harvard University and the University of Pennsylvania that it was based on is here and is worth a read. The study came out on October 9 (and yes, I missed it as well).

Not blaming data centers for higher energy costs is definitely random, but interesting!

Catch up with more on energy including interviews and video content from CNBC and Power Insider.   

Read the last issue of Power Insider here.