5 High-Torque AI Infrastructure Stock Plays To Buy Now

Hello and welcome to being exponential. We are doing our macro episode today. So we want to start this episode by talking about the AI sell-off. Uh Luke, uh in the last episode you mentioned we should be waiting around 3 weeks or so to see what the AI hyperscalers are doing. Um so can you give us some insight on to what stocks we should be looking at or not looking at between now and then? >> Yeah, so I think that I mean long story short, to cut to the chase, this is a great buying opportunity in AI infrastructure stocks. That's my big viewpoint right now. But we're going to have to wait, like you said, to the end of July, I think, until this trade really starts to reawaken because well, I what I discovered this week is that the market doesn't give a crap about what's going on right now in the AI industry. They give a crap about, in fact, they give all the craps about where the AI industry is going over the next 6 to 12 months. Because we all know that the state of the AI industry today is sensational, right? Samsung, who is arguably one of the three to four most important players in the AI supply chain, infrastructure supply chain, reported preliminary Q2 results this week and they were genuinely excellent by any reasonable standard. Operating income was just shy of 90 trillion one. The expectation was around 84 trillion Juan. That's about $60 billion, just under $60 billion. Uh that was up 19 times yearover-year. Revenue more than doubled. Memory demand was described as being white hot. Pricing power remains extreme consistent with everything that we have heard from uh from Micron, from SK, from pretty much everybody else in the memory supply chain. So that report confirmed I mean without a doubt unequivocally that AI infrastructure demand is running hot in the present tense but the market did not care right I mean on that day I think the socks ETF dropped about 6% the SMH ETF dropped about four to 5% the whole NASDAQ complex was down more than 1% so we got these fantastic numbers telling us that today things are amazing and we still dropped you five to six% in a lot of AI infrastructure names. So what the heck? Well, that tells you the market doesn't care about what's going on right now. It tells you their focus is elsewhere. Where else could their focus be? Well, the focus is clearly on what is the market going to look like in six to 12 months because there is now finally a bunch of new supply coming online. What happens in the semiconductor world is demand goes vertical and then you got to wait around a few years for more supply to come online and then the more supply comes online and often times it comes online right as that vertical demand curve starts to flatten and then the supply exponential supply growth meets flattening demand and then all the things that drove these stocks higher reverse course and drag them lower. And so that's the fear because now we have the exponential demand and we have the supply coming online and we need confirmation that demand is not going to flatten. The fear that demand is going to flatten as supply comes online is what is driving these AI stocks lower right now. Is why Micron and Sandis dropped 10% the day after Samsung reports. Fantastic numbers. So it all comes down to the demand curve. Well, who is the demand curve? The demand curve is the hyperscalers, right? the core bull thesis on the AI trade ever since the the dawn of the AI trade, ever since the launch of chat GPT in 2020 in late 2022 and increasingly so in 2025 and 2026 has been that the biggest most powerful deepest pocketed companies in the world are emptying their entire coffers all of their cash even going into the debt markets to invest in one corner of the global economy which is the AI infrastructure supply chain. Um, if that continues, then the trade continues. Demand continues to go parabolic. If that does not continue, then the fears look legitimate and this trade is due for some bigger wobbles, if you will. But I think that when these companies report in three weeks, and we're talking Amazon, we're talking Meta, we're talking Alphabet, we're talking Oracle, we're talking Microsoft, we're talking those big dogs. When they report in 3 weeks, I think there's no doubt, at least I have no doubt in my mind that they are at the very least going to reaffirm their 2026 capex plans. If not, I think it's actually more likely that they hike their 2026 capex plans and then provide super super super bullish commentary about 2027 and 2028 capex plans. They're not going to give exact explicit numbers for that, but they will give some directional comments on that. I think those directional comments will be directionally positive. So, I think that again I think that those updates are going to be really, really, really positive. Now, why do I think that? Well, on the same day that Samsung reported these fantastic numbers, Amazon news broke that Amazon is actually reportedly raising another $25 billion in bonds to fund AI infrastructure. Now, that raise pushes global AI related debt issuance to about $335 billion this year, which is more than double the level in 2025. And the companies tapping that are not companies that need debt. Amazon's tapping it, Alphabet's tapping it, Meta is tapping it, Oracle is tapping it, Nvidia is tapping it, SpaceX is tapping it. These are cash flow positive companies and for most of them outside of SpaceX, super cash flow and Oracle, super cash flow positive companies with massive balance sheets loaded with cash, but they have already emptied all of their free cash flow. And as opposed to pumping the brakes, they're going to the debt markets to look for more funding. That does not sound like a bunch of companies that are going to say in 3 weeks, you know what, we've reached our limit. We're going to actually reduce our capex plans for the year. It sounds like a bunch of companies that are in an existential race to spend as much as they can to win the AI arms race. And throw on top of that, Open AAI just raised $122 billion. remember that Anthropic just raised I think about $60 billion. SpaceX just raised what was it? $85 billion and SpaceX is now becoming a big compute player. We've talked about that on this podcast. You put all that together and I just don't there's no fundamental ingredients here for these companies to come out the Reese and say, "Yeah, you know what? We're we're not going to spend as much as we thought." We have all the ingredients that come out and say, "We're going to spend more than we thought." And once they do, that'll confirm that the demand curve is going to continue to go vertical, even as the supply curve starts to go vertical, too. And that's going to reinvigorate bullish sentiment and reinvigorate the AI trade. And these stocks that have dropped 20 to 30% are going to roar back higher 50 to 60% in a matter of weeks. And so, I think this is kind of a good place to be accumulating these AI infrastructure stocks. I remain very bullish on the trade. I think it will be the dominating theme for the second half of the year. Roger that. Yeah. So, you mentioned the $25 billion uh tap from Amazon. Um and then even with Google, I believe last month in June, they they asked for $80 billion and 10 billion of that came from Birkshire Hathway, which is, you know, notoriously anti-tech or historically pretty anti-tech. So, is it almost safe to say that a lot of the the the sell off the the downward trend in the AI um market has just been sentiment? >> Oh, it's entirely sentiment. Well, first off, you had you had quarter end and halfend fund repositioning uh that was going on in late June and and we did some of that too, like with our model portfolio. Okay, you know what? These stocks have been on such a great run. Why would we not just take a little bit of profits and rebalance into some other stocks? Of course, you're going to do that. That's natural. That's not saying anything about a lack of confidence in the AI boom, which is saying we're up 150% on stock XYZ. Let's go, right? like let's take some let's take some of that to the bank and you know let's rebalance it maybe throw a little bit into Nike even though it's not going to do anything right like that's that's just sort of what you do as a disciplined fund manager at this stage in the game so that was going on at the end of June and then I think in the beginning of July you just have people questioning well can this continue forever right like you know it just psychologically people are like at the end of a year or the end of a quarter or the end of a half of a year they're like okay wait can that continue right now this is the time to do some house cleaning now's the time to rethink I think the the core ideologies are powering the portfolios are powering the markets. So that's what's going on right now. We're kind of in a news vacuum, right? We're in this quiet period and now people are kind of reesting can this continue, but all that all those questions will be resolved at the end of the month when the big tech companies come out and and report earnings. So, I think what you have right now is this kind of this buy zone is what I'm calling it because if you kind of look at at the technicals on where we are, um the SMH ETF, which is that semiconductor ETF we left to track, that has that's down about 14%. It's about 14% off its highs. Historically speaking, that is kind of the bottoming zone for AI semiconductor stocks, right? Every garden variety pullback in AI semi stock since the AI boom began in late 2022 has bottomed in that 10 to 15% pullback range except for there was one in late summer 2024 uh where we dropped 24% then we had liberation day where we dropped 35%. So there there were two that were different, but outside of those two, there had been, I think, almost about a dozen 10 to 15% pullbacks since late 2022, and they all bottomed in that 10 to 15% pullback range. And so we are right there right now. We're at 14%. So we're in that range where we should bottom. Now, it's also important to remember that um again, volatility is a feature of tech bull markets, not a bug of tech bull markets. This AI boom has been characterized by sharp rallies, then sharp pullbacks, then sharp rallies, then sharp pullbacks, then sharp rallies, then sharp pullbacks. Lather, rinse, repeat. We talked about those 10 to 15% pullbacks. That correction, we've had 13, this is the 13th 10% plus correction in semiconductor stocks and essentially a infrastructure stocks since the boom started in late 22. the 13th. That's pretty crazy because none of those 13 and none of the prior 12 I should say um derailed the boom, right? Semiconductor stocks are up I believe it's let me look here. We're up 565%. 2023. So, we've had we went through 12 different 10 to 15% corrections that felt like the end of the world at the time. And through it all, they weren't. and semi stocks roared 565% higher. So that tells you the pattern is buy when those semi stocks are down 10 to 15%. This is also not unusual just to get in all tech bull markets. Do boom same thing. We had several 10% corrections during that time in semi stocks. We actually had two bare market draw downs nearly 40% one in 95 96 one in 9798. And yet through it all none of them killed the semi trade. From the start of 95 to the peak in March 2000, semi stocks sore more than 1100% even considering two near 40% bare markets in that time. So volatility is a feature of hyperbull markets in tech, not a bug. You don't get the 100% plus rallies in less than 12 months without the 10% plus pullbacks every few months and the occasional 20% plus bare market every few years. You got to take the good with the bad. And I think right now we have a zone where it feels like the end of the world, but it's actually a buying zone and we're going to reawaken and these stocks are going to burst higher by the end of July with the confirmation of a big tech in those earnings reports. So that's where I'm viewing the market right now. And that's why I think everything that I'm seeing is saying don't run, don't hide, be patient, but this is this is accumulation time. Roger that. So I I gota ask my favorite question. So as of this week and last week, we've talked about some fang stocks, some pretty large cap stocks that you you know, we have pretty uh good conviction on in terms of buying the dips. Uh we talked about Amazon, we talked about Meta. So do you have any suggestions on smaller cap stocks or smaller stocks that would generate more torque that we could get into now and in in this particular phase of the AI selloff? Well, it's all those semiconductor stocks, man. Okay, so let's You want some names? The people love names. Let's Let's >> People love names. >> People love names. So, you want to know what I'm doing? I'm I'm running a I'm running a screener. I'm looking for stocks that are up a bunch this year. Stocks that are up a bunch over the past year. So, they got a lot of momentum. Uh they're more than 10% below their 52- week highs. So high momentum stocks in a correction but still above their 200 day moving averages. So maintain their uptrend. That's the world I want to live in right now because I don't think again the underlying fundamentals of the market have changed. So assuming they have not changed, I want to find those high momentum stocks that have had a lot of momentum and have been doing really, really well, which are now in the midst of very sizable pullbacks, but amidst those pullbacks are still maintaining their long-term uptrend, their long-term positive price trend. That's the world I want to operate in right now. And if you want if you want like a smaller high torque some high torque names that are in that world I think air test systems AEHR is a pretty good pick there. They're in that semiconductor equipment and materials world. They're totally part of the AF supply chain. They are they meet all that criteria. Up a bunch in the last 6 months 12 months down a bunch recently still maintaining that uptrend. I think that's an interesting one to look at here. I think uh we talked about that recently OS that's a name that has been on a massive tear as the kind of eyes of the physical AI economy but now they're in the midst of a big pullback but they're still maintaining their big uptrend. That's a name that that I really like here. Um Ultra Clean UCT is a name that I really like there. They they do a lot of the kind of like uh cleaning and test inspection equipment that's very mission critical to the afrastructure supply chain. That's a name. The chart looks really good there. I think that's one that can go a lot higher. Looks like a good buy zone for that. Excelsius Technologies, ACLS, I think that's also another one that can really get going here. ACM Research, ACMR, I also like that one a lot. So again, I'm operating in this world of up a bunch in the last 12 months, up a bunch in the last six months, down a bunch over the last few weeks from uh record highs, but still maintaining positive price uptrends. looking at the fundamentals of those companies, seeing which ones I like the best. Those are a couple of the smaller names that we're looking at right now. Again, I think a lot of them, you don't want to catch falling knives, and a lot of them aren't falling eyes right now, but as soon as they start to show good technical support, those are names I think I would be interested in. Absolutely. >> Awesome. Appreciate that, Luke. Uh so, you know, we mentioned in 3 weeks or so, end of the month, we we'll be looking at those hyperscaler earnings. Is there anything else we should be looking back at around that time in uh in terms of earnings? No, because nothing else freaking matters, >> right? [laughter] >> I mean, we can talk about Borch and we can talk about the Fed and we can talk about oil and we can talk about the war in Iran and we can talk about this and we can talk about that till we're blew in the face, but the only thing that matters are the inputs. All of which that those are inputs, but the inputs to the decisions of the big tech companies because they are spending $800 billion this year on AI infrastructure. They're going to spend over a trillion dollars next year probably on AI infrastructure and they'll probably continue to spend a trill over a trillion dollars per year 2028, 2029, 2030 on AI infrastructure. What matters in the face of a trillion dollar per year capex super cycle? Like nothing matters except for the inputs to that super cycle. So yes, if oil were to spike back to 100 and stay above 100, get to 120, 140, 150, that's an input which would re which would cause those companies to rethink how they spend all that AA infrastructure money. um if you know Borsch goes crazy and hikes rates and the 10-year Treasury E gets to six to 7% these companies aren't going to tap the debt markets at those rates and so they are going to be constrained by their cash flows which they've already hit the ceiling on. So that you have to look at everything through the lens of how it impacts the big tech spending decisions and right now I'm not seeing risk factors there. Oil is down at 70. Not a risk factor to big tech spending decisions. 10 years, four and a half, not an impact to uh big tech spending decisions. Wars has kind of said we're just going to chill until inflation gets down to 2% and inflation is heading lower because oil is down to 70. So, um he's probably going to cut by the end of the year. Not a big impact on big tech spending decisions. The consumer is still spending. Consumer confidence is kind of rebounding a little bit. the labor market's fine. So like there's really nothing out there where there's like something that's a big risk factor that's going to massively impact these big tech spending decisions. And in the absence of that, I think the big tech spending decisions are going to remain what they have been for the last several quarters and even years, which is put it on the AI infrastructure and AI compute and build build baby build. So that's the only thing I'm watching. I think that's the most important thing by a wide margin. What are these companies going to say and do with respect to their capex plans at the end of the month? >> Got it. Yeah, this has been an interesting year for this market. I want to say it's almost impervious. We had one of the worst oil crisis at the top of the year and then we get nothing really fundamentally changing with the market. Then we have amazing earnings from these, you know, large companies and again it it's not pushing the needle too much. Um, but >> you only had amazing earnings because of the big tech spending decisions. >> Right. >> Right. Like that. Again, it all comes back to the big tech spending decisions. Why did we have 30% EPS growth in the first quarter? Because semiconductors had like 80% EPS growth and everybody else was flat, right? Like it wasn't like Nike was like everybody's buying our shoes or Limon's like everybody's buying our clothes or Coca-Cola's like everyone's buying soft drinks. say no, it was Micron saying everyone's buying memory, right? Like that's what drove the record earnings and that's what's driving this market um higher is the the decision or are the decisions of the big tech companies to keep spending everything that they have uh on AI infrastructure. So yeah, we went through this oil crisis and yeah, we talked about the Fed maybe hiking rates and yeah, we saw inflation spike up, but were any of those things significant or severe or durable enough to change what are long-term capital allocation decisions by the biggest companies in the world? The answer is no. If you're going to change a these companies believe that what [clears throat] they are building today with AI compute will be a valuable monetizable asset not to mention the world's largest ever concocted business mode competitive business mode for the next hundred years. They are literally making I would say decadel long multi-deade long decisions. They're not going to change multi-deade long decisions because oil got to 100 for a month. They'll change it if oil gets to 150 and stays at 150 for six months or 12 months because then there's a real flow through there and they're going to be forced by the market to change. But that didn't happen. The rate hike stuff kind of a scare for a few weeks didn't happen. Inflation above 4% got there for a month maybe two. We'll see how the next print is but won't stay there for that long. So, none of these changes are significant enough, severe enough, or durable enough to impact big tech long-term capital allocation decisions. And that's why dips are buying opportunities in my opinion. >> Love it. So, Taylor's oldest time, follow the money. Uh, let's uh close out this episode. Luke, anything else you wanted to mention uh outside of that? I mean, I just I think it's a real big thing to remember that volatility, and I said it on this podcast on the one before, I'll keep saying it. Volatility is a feature of bull markets, not a bug. You do not get We had stocks in our model portfolio that went up more than 100%. I mean, several of them that went up more than 100% in the span of four to five months. That does not happen without 10, 20, 30% pullbacks as well. So now we're getting that it's a buying opportunity. The fundamentals haven't changed. But don't be scared of that. We've had 13 10% plus pullbacks in semiconductor stocks since the launch of CHATBT. 13. That's through 23, 24, 25, and 26. So 13 over three and a half years. These just happened. They happened and none of them changed the course, the ultimate course, the ultimate trajectory of these stocks. Why is this time different? The one thing that could change it is if the demand curve flattens, which is why late July matters. That the supply curve is now going exponential too. If supply curve is exponential and demand curve flattens, that's the end. But I don't see that happening. And so if that doesn't happen and demand curve stays exponential alongside an exponential supply curve, then the stocks rebound and you get this this trade is going to live on and going to last throughout the second half of the year. I think it's exactly what's going to happen. >> All right, and then with that, we'll close out the episode. Uh we hope you enjoyed. You definitely enjoyed recording. Make sure to like, comment, subscribe. We are always looking at your questions and looking to answer them if we can. That's it for being exponential this week. Take care. [music] [music]