Hidden gen. AI picks that should be on investors' radars

AI has been quite the buzz on Wall Street recently as top tech companies with heavy investments in AI have led the recent rally seen in the market. With more and more companies getting involved in AI, it can be hard for any investor to wade through the names and find a great entry point for their portfolio.

Morgan Stanley Equity Analyst Stephen Byrd joins Yahoo Finance to break down some of the best AI picks that have gone under the radar for many on Wall Street and what investors need to keep in mind with the AI sector.

Byrd points out one of the biggest drives behind his picks and why they have much room to grow: "We see about a 50% drop in... essentially the cost of compute, cost of computation, in these data centers. That's really all Nvidia (NVDA). It's these new chips that are just much more efficient. And that's flowing through the entire ecosystem. As those costs drop, we think demand for a variety of gen. AI applications will rise, and that's essentially the thesis of our view of gen. AI adoption across many industries."

For more expert insight and the latest market action, click here to watch this full episode.

This post was written by Nicholas Jacobino

Video Transcript

SEANA SMITH: All right, well, stocks off to a strong start of the year. Overall, the excitement really fueling that momentum to the upside. It's pushed the major averages to those all-time highs. Now, for investors looking to buy into this craze, there's more ways to play AI than maybe those obvious names when you're thinking of NVIDIA and Microsoft.

Morgan Stanley out with a recent note looking at companies that are actually powering AI, calling out a few of those hidden AI gems. Joining us now we want to bring in Stephen Byrd, Morgan Stanley equity analyst to discuss. Stephen, it's great to have you here on Yahoo Finance.

So talk to us just about your thesis here because I think a lot of investors when they're thinking about joining in on AI, they want to get a piece of this massive rally, they're thinking of those obvious names. You're, though, looking at some of those companies that's powering AI. What's the real opportunity here for investors and why?

STEPHEN BYRD: Yeah, thank you for having me. The opportunity really here is around the infrastructure. It's very understandable the first port of call is Nvidia. We understand that and support that. But the infrastructure required, everything from power generation, power electronics, data centers themselves, that growth rate is set to accelerate. Many of those stocks have not participated in the rally that the first derivative AI play is really have, and that's essentially the call.

MADISON MILLS: So talk to us then about your data center economics model. In particular you talk about how the cost of compute is falling. What is driving that, and what does that indicate to you about certain trends that investors should be keeping up to date with?

STEPHEN BYRD: Yeah, I have to say I was very surprised at the results of our assessment of data center economics. We see about a 50% drop in essentially the cost of computation in these data centers. That's really all NVIDIA. It's these new chips that are just much more efficient. And that's flowing through the entire ecosystem.

As those costs drop, we think demand for a variety of Gen AI applications will rise, and that's essentially the thesis of our view of Gen AI adoption across many industries. That's just going to spur demand higher, and that's going to create even more demand for the infrastructure for the nuts and bolts, all the power that's going to be required for all these data centers.

SEANA SMITH: Steven, how underappreciated do you think this opportunity is at this point by investors?

STEPHEN BYRD: It's interesting. There are a few obvious plays that are part of the infrastructure dynamic. But there are many, many companies that are key enablers that are not well understood. So, for example, Bloom Energy, a fuel cell company, the stock has not done well. And yet that's the type of technology that can allow data center companies to very quickly get new data centers operational.

Whereas Constellation Energy, it's a stock we like very much. Ticker is CEG. Stock has done very well. We support that. We really walk through in detail the economics. For a big nuclear player like Constellation, it's very likely that we're going to see extremely large data centers really essentially supercomputers sited at existing nuclear plants. So that idea has captured a lot of investors' attention, whereas other names like Bloom really has not gotten any attention at all.

MADISON MILLS: And when you look at Bloom, Constellation Energy, and you mentioned that they're not getting a ton of attention, which is great for our audience to hear. Maybe it's a good time to get in. But what was the biggest thing you looked at in terms of the fundamentals when sussing out those picks so that we can try and kind of mirror your understanding?

STEPHEN BYRD: Absolutely. So one thing that also surprised me about our analysis is time. What I mean by that is for data center developers, for hyperscalers, time to getting powered up is absolutely critical. It's incredibly valuable economically. So we looked at a variety of companies that could essentially speed up the time to power.

Constellation can do that with their existing nukes. You can fairly quickly cite data centers out of nuke. Bloom Energy's fuel cells can be deployed in 50 days. And so that's a very fast solution that I think the market is missing. What we're used to seeing with data centers, just connect the data center to the grid, then get some renewables, you're good to go.

That's running into problems now because the grid is so congested. So Bloom provides a solution at a low cost. It's just been missed because the conventional thinking hasn't really fully changed into this concept that, look, the grids are constrained, and time is absolutely critical. So that's really the lens we've been looking at.

SEANA SMITH: And Stephen, put this opportunity even in more perspective for our viewers. How early are we in this adoption cycle when it comes to AI? And then what does that tell us just about some of the future gains, the future growth that we could see from some of these companies like Bloom Energy that we have up on the screen?

STEPHEN BYRD: That's such an important question. We think we're very early days. And I say that because we look at the pace of NVIDIA chip evolution, evolution of chips from others as well. We look at the future compute that's going to be needed for Gen AI more broadly. We see very rapid growth. Gen AI growth in terms of electricity growth is going to be about 100% a year for many years to come.

And what's interesting to me is as the cost of compute drops, new use cases show up. So the demand for compute is only going to go up and up. We're extremely early days in this dynamic. And now investors are just really starting to look across what I think of as the value chain to providing the power to these data centers, so we're very early days in that regard.