Technology shares took a dive last week, as a broad market sell-off pummeled high-flying AI stocks led by Nvidia, Microsoft and Google.
The slump is bound to turn the spotlight back on a growing worry: that the frenzy over AI, which sent Nvidia stock soaring, is creating a bubble.
But veteran technology analyst Stacy Rasgon, who has covered Nvidia and the chip industry for more than 15 years, isn't worried. Rasgon, a managing director at Bernstein Research, thinks concerns about a bubble — or an "air pocket" as he calls them — are overblown.
That's because the ongoing AI buildout is real and "massive" and will likely go on for a few years, he told Investor's Business Daily.
"The chance of an air pocket at some point is probably 100%," he said. "But I don't think it's now."
Rasgon spoke to IBD on Friday when technology stocks took their heaviest beating in months.
This interview has been edited for clarity and brevity.
AI Stocks: High Expectations
IBD: How does the market slump change your view of where tech is headed, specifically AI?
Rasgon: Semis (semiconductor stocks) until recently had been massively outperforming most mostly on the back of AI. If you look across the industry outside of AI, it's not super. PCs and smartphones, they should be better than last year, but they're not like ripping anytime soon. I think data center, outside of AI, is pretty weak. Server CPU has been pretty weak. Networking looks pretty weak.
Outside of AI, the sector dynamics out there are not fantastic. Yet the sector itself — it's reasonably strong. Valuations are high.
Maybe things are taking a little bit of a breather. I kind of get it. I kind of understand what's going on. Does it change the long-term view of anything? Probably not. Expectations in the near term are maybe coming down a little bit and stocks are coming down commensurately.
AI Stocks: Is There A Bubble?
IBD: What do you think of the view that there is an emerging AI bubble?
Rasgon: I'll pick on Nvidia because they're seeing most of the benefits. The numbers have gotten so big so quickly, people just worry about sustainability. I get it.
At the same time, it seems like the demand is still good. What people worry about is order cancellations. It doesn't seem like we've had any order cancellations. They've got new products that are ramping. I get the air pocket worries.
I've covered Nvidia for a long time. We've seen air pockets. The chance of an air pocket at some point is probably 100%. But I don't think it's now.
Massive AI Infrastructure Growth
IBD: We're in the infrastructure building phase with AI. What's your projection on how long that's going to take and what it means for the broader tech sector?
Rasgon: I think over the long term, it's massive. I'm still convinced that in five years or 10 years, we will be talking about numbers that are materially higher than what we're talking about today.
We've done a lot of work on the compute requirements for generative AI. And it's massive. What will it look like in two years? I don't know. It's probably not a straight line. I don't know. I don't think anybody knows at this point.
How AI Compares To Past Tech Trends
IBD: How does this compare to past trends in tech, like the cloud?
Look at 2000-2001. Look at the fiber build outs. We had over builds I'm sure.
We did a whole note (at Bernstein Research) "AI infrastructure: Bonanza or Bubble." A positive compare might be, just talking about the business, when Amazon built AWS. There was a big investment phase. But the business was there and massively profitable.
The less charitable compare would be the fiber buildups for 2000. We've got dark fiber sitting in the ground.
IBD: This was leading up to the Y2K fears (when concerns of major systems disruptions with the start of the new millennium led to huge investments to beef up computer networks).
Rasgon: Yeah, the tech bubble. At least, I think, with the (AI) infrastructure that's getting built, it's not like they're putting stuff into warehouses. It's getting built in data centers. It's getting deployed.
Eventually, you've got to be able to build business models on it, where you can get a return, right? If it turns out that there's no return, that's a problem. Everything comes crumbling down.
AI Stocks: Biggest Worries
IBD: And that's the fear now, right? That companies, including startups, are building these without any clear business model.
Rasgon: Well, I don't know if that's true. Especially on the enterprise side, we're in the experimentation phase. I still think it's all good.
I'm not really worried about air pockets this year. I'm not really worried about air pockets next year. 2026? I don't know. We'll see.
IBD: What is your biggest worry?
Rasgon: The air pocket worries are probably the biggest tactical risks for Nvidia and a lot of others. I understand that numbers have gotten really big. I understand that worry.
You can follow Benjamin Pimentel on LinkedIn and Twitter/X @benpimentel
