Is the AI Trade Actually Over? Here's What the Selloff Really Means

Good morning. This is Dylan Jovine with Behind the Markets. Happy Friday. Today is Friday, July 31st, TGIF, and the last day of the month. Anyway, today I'd like to talk about basically, is the AI trade in aggregate over? Is the AI trade over? And the answer for me is clearly not. But as we talked about yesterday, we don't have a master narrative driving or carrying the torch, a story. Wall Street hasn't settled on a story. And I'll tell you, I have some notes here. Look, no matter what, despite concerns, I mean, you've seen in the past month that AI stocks have gotten hammered, specifically the infrastructure plays that were receiving all this capital spending. They had gone bonkers, and they have gotten hammered in the past thirty days. I am very grateful that before I went on my annual vacation, we closed a lot of those positions, but we still have a couple of positions there, right? Like Amkor, I think, or Arm, we still have, we sold half of them for profits. And, you know, we still recommend in the closed position portfolio, we still have those stocks. And, you know, what do I think? Everyone's saying, "Is the AI trade over? Do you still like this stock? Do you still like that stock?" Look, in my view, I view this recent decline as a matter of positioning really, really short-term investor positioning rather than deteriorating fundamentals on the whole thing. I also expect AI compute basically to significantly outpace supply for years to come. There is no  doubt about that in my view. The big picture, the big build-out narrative is still firmly in place. Just look at the Mag Seven earnings. I mean, this is happening. And I expect AI capabilities will develop in a nonlinear way because remember, you have the robots building themselves, which is a whole, you know, terminator-style nightmare. But I'm just talking about economics on this channel and finance and investing. I'm not gonna talk about the end of the world. I'll leave that to much smarter people than me. But look, what we're building out here is this kind of intelligent superhighway. And, you know, a lot of people that follow this have, one in particular who's pretty smart guy, friend of mine said to me, "Look, companies are stopping this token maxing." And token maxing is when you allow your employees at your company to use as much Anthropic or OpenAI or whatever it is that they use, or Gemini, don't worry about the cost, just figure this out, use this. A lot of companies are putting a clamp on that because token maxing, which is tokens are the price that the AI models charge to use their service, they're so expensive. A lot of people are worried about that. And there's a good point. That's a good point. But when you look at the percentage of companies that actually are using, the penetration still is under 10%, so this has so much room to grow. But again, if you look at a chart over time, this is what the chart will look like. But if you look closely, there'll be a lot of little dips along that chart. This is the nature of it. This is what it looks like when you build out infrastructure at this scale and pace. Look, I am fundamentally bullish on the improvement of AI capabilities. I am fundamentally bullish on the adoption of AI and associated capital spending. And, you know, given recent pullbacks in AI infrastructure stocks, I think Amkor looks great. I think a few really look very good right now, and I'm very, very excited. But, you know, when you actually look at AI capital spending over the next few years, 2026 we're at $800 billion from the  hyperscalers, you know, five of them. Five largest US hyperscalers. 2027, CapEx is expected to go to $1.2 trillion, and 2028 is expected to go to $1.4 trillion. So look, where are areas right now in this AI pullback that I think look really, really good? One two three four five. I'm gonna give you five areas that I think look really, really good as we've seen such a great pullback in the market. Here are the five areas. Number one, AI infrastructure bottlenecks for sure. You know, these are companies that help address labor shortages, shorten time to power, expand electricity. These stand to benefit, including fuel 📍 cell providers, Bitcoin operators that used to mine Bitcoin that are now transitioning into data in-infrastructure providers, turbine manufacturers, energy storage companies, power developers, data center REITs. These companies look good. That's number one. Number two, compute manufacturing ecosystem. Businesses that are positioned to benefit basically from growing demand of AI compute. But remember something, like semiconductor producers and the packaging companies like we talk about Amkor, as the value of intelligence rises, but supply remains constrained. It's a big, big deal. Also, you know, three leading Chinese AI solution providers. I mean, we don't know. I shouldn't even have these officially on my list, but I see that they're mispriced. So we'll see what happens there because the Trump administration might decide to block them from even entering this market. We shall see. Another area that I think is gonna do very, very well, energy security assets. Businesses that support reliable energy supply and storage are critical to obviously AI infrastructure development and of course hyperscalers themselves. Large tech companies have the scale to generate attractive returns on AI CapEx, and accelerate this adoption, and you could see that with Microsoft's earnings. It's actually a tale of two cities. Companies that are actually able to see their cloud business grow look great, but every time Meta says, "Well, we may or may not use it for cloud," you can see that Mark Zuckerberg still has glory on his mind with how he's gonna use that compute. But we shall see how that all works out. As far as I'm concerned, those things just look cheap. They sold off even before this latest AI infrastructure sell-off. So those are the five areas that I think do very, very well, and some of them we sold off for nice gains. We recommended selling for nice gains in the last four to six weeks. And now that they've really gotten crushed and way below the selling price, we might recommend getting back into some of them. Anyway, that's all I have for you today. Have a wonderful weekend. God bless you. I will see you on Monday.