Missed the Memory Boom? This is Next.

40% of data center energy consumption is for cooling. That's right. All those data centers that Microsoft and Google and Amazon are building, they need to be cooled down. And today, we're going to look at some hot stocks that are benefiting from that AI money train. Choo choo. But you need to be very careful. From sneaker makers to cannabis growers, everybody out there is trying to attach themselves to the AI trade. Now's the time to be skeptical, not climb on board every growth story out there in finat land. Here are three important things to look for when distinguishing genuine opportunities from teams with dreams. One, massive growth today, not the promise of growth tomorrow, growth today. Two, pure play exposure or a clear trajectory towards pure play exposure. We call these skate to where the puck will be opportunities. You can think of when Nvidia first started out with data center revenues. How those went from perhaps 15% of total revenues to what upwards of 90% where they sit today. You need to see that trajectory. And number three, management's stated intentions of increasing exposure usually through divevestaturures or acquisitions. In other words, they're trying to streamline their business to focus more on where the growth is coming from. That implies, of course, that they already have some of that growth in the form of strong revenues for a particular segment. And number four, this is a bonus, recurring revenues for when the data center construction mania plateaus. And for those of you that think the data center expansion is going to increase until the end of time, you can just skip this one. So, let me give you an example of exactly what I'm talking about here. Check out this first slide in the latest earnings deck from a company called Modí. Fourth consecutive record year of revenues, they say. And when we look at their last year of revenue growth, it's 23%. So fairly strong, especially when they're building that off of $2.5 billion base. So whenever you see strong revenue growth numbers, you want to see what sort of base those are coming off of. The larger you get, the more tough it is to demonstrate really strong growth numbers. That's why what Palanteer is doing is so impressive. Then you can see here they talk about completing three strategic acquisitions. And further down, most important here, they talk about the announced plan to spin off performance technologies. That's actually a huge chunk of their existing business right now. Now, before I get into that, be very careful when you're just looking at revenues and perhaps not paying attention to what margins are doing. Here you can see, well, they're actually seeing their profitability decline a bit over time. This is contrary to what you usually see with data center stocks that have a lot of pricing power and see their gross margins increase rather than decrease over time. So when they talk about getting rid of this performance technologies segment of their business, you can see here that constituted 44% of total revenues in 2025 and they're actually expecting flat to 5% growth in 2027. So that's actually dragging down this other component of their business, climate solutions. You can see a key segment in there titled data centers. So let's drill into climate solutions. Here you can clearly see that data center subsegment driving growth. 158% quarterover growth. Now you need to be careful with these quarterover numbers because sometimes there's some cyclicality embedded in them. But data center revenues in this particular quarter exceeded $400 million or around 66% of this entire segment. Here you can see where their outlook has the expectation of 60 to 80% growth for data centers. That's pretty good. Now, notice here that the guidance that they're giving actually includes that performance technologies component because they haven't spun it out yet and they're forecasting revenue growth of 20 to 35%. So, it's very strong, right? 23% last year, 20 to 35% this year. But once they spin out that performance technologies segment, the revenue growth is actually going to be a lot stronger than that. So, they're going to reissue guidance. So this is really just a matter of waiting until the dust settles to see what this business looks like when they transform themselves. So we look at valuing modin. We can use two methods here. So the first is what we call a simple valuation ratio. This divides market cap by annualized revenues. It's easy. Anybody can calculate it. It's quite responsive to the growth of a company and it's a great way to value companies that aren't profitable yet. Now, Modí is profitable and all the other companies we're going to talk about today are profitable as well. So, you could make a strong argument to say, well, we should just move right into looking at forward price to earnings ratios. So, the idea here is that you're looking into the expectations for next year. Since things are growing so fast, we want to take into account some of that future growth. That's called a forward price to earnings ratio. We use the average of what analysts are expecting for next year. So if we calculate both these numbers from modin we get an SVR of three which is what like almost a third of the average SVR is sitting right around eight I think in our catalog about eight in the NASDAQ as well. So from that perspective it's undervalued but then when we look at price to earnings as I said this is more of a mature company so we can use that metric we get about 30. So depends on where you look, right? But I think forward P typically would be in the range of 25 to 30 when looking at benchmarks. And interestingly enough, most the companies that we're looking at today sit within that range. Now, how does Modin compare to other leaders in the cooling space? Well, this is a great chart that we used from Spear Invest to produce a piece we did several years ago looking at all the stocks that would benefit from the growth of data centers. And Spear did a great job here of pointing to HBM as perhaps being the highest growth data center segment inside the rack. And they were absolutely correct. That's what you see today with SKH Highix, company we recently covered, Micron and of course Samsung, the three leaders in HBM have seen tremendous share price appreciation. And when we moved to outside the rack, the highest growth component there was expected to be thermal management or cooling, heat dissipation. And back in October of 2024, Spear had a look at which companies were leading in that space. Here they've noted the top three players controlling 45% of that market were Vertive, Schneider and Eaton. And then of course when you look at these names you want to consider how much pure play exposure you're getting. That's very important right so for Vertive what 80% or more of the revenues coming from data centers moving down the list and Vent a lot less. Now this is from 2024 but it's probably directionally accurate. And when we look at two of the names on this list, so Vertive and Invent, we can see that indeed in 2025, both of these companies saw dramatic revenue growth. So 28% or more. And when we look at valuations, of course, here I've calculated SVR for both. But I think again looking at forward PE is probably more appropriate here. And we see them vertive at 35 and vent at 28. Now when comparing these two companies, Vertive is more focused on comprehensive data center infrastructure, so power and cooling, whilst Invent has meaningful but not dominant data center exposure. Going back to that chart from Spear Invest, it's always a double-edged sword when you're too overexposed to a particular theme. Now, we like that exposure. That's what we're trying to get. but also when there's a significant downturn for whatever theme that is, then a company like Invent with broader end markets that's going to provide them some buffer against a data center capex cycle, right? Should any of the hyperscalers decide to pull back on all that money they're throwing at AI, Marie Rate is going to get out of hand, right? On the downside, I like this piece from the Mly Fool, a firm that's gone dramatically downhill over the past few years chasing clicks. They talk about the ultimate AI bet. Why Eaton and Invent are top stocks to buy in 2026. Everybody wants to tell you what the top stocks to buy. Are not going to be there to tell you when to sell them, just the top stocks to buy, right? And there's not a lot of room for nuance here because these are two very distinctly different ways to get exposure. In regards to Eaton, they appear to be showing the most consistent growth among the lot here. They have 17 years of increasing their dividend in a row. We call these two birds with one stone play. So you can invest in a value stock whilst getting some upside exposure to a growth theme. Here in this slide you can see this distinctly spelled out by eaten. Look at that. Second largest segment is data centers and distributed IT. Strong double-digit growth expected going forward. And they state that's from 2024 to 2030. So very soon that should become their largest segment. And also what's worth noting here is that they're talking about how they're going to use AI internally to boost that bottom line. This is very important because everybody's focused right now in the AI trade in terms of growth stocks. But think about value stocks and firms that may not have a lot of topline growth ahead of them, but the efficiencies that they can create with AI are incredible. So you can expect lots of bottom line improvements over time which translate into greater earnings which translate into bigger dividends, stronger dividend growth. Here you can see Eaton talking about their robust partnerships with Microsoft Palunteer noted here and of course they talk about service now and their agentic AIs. It's very interesting. This firm has a forward PE of about 30. So roughly in line with the other names that we've talked about today. Now, if you're only selling to data center customers, what happens when those buildouts start to taper off? Well, people will say things like, "Well, you got replacement parts, right? Servers, chips, networking equipment, HVAC." Well, these all have various lifespans, right? How often does the HVAC unit of your office building get replaced? Not very often, right? So, if you're primarily involved in new building construction, you may be seeing a nice boost right now. But where are the recurring revenues going to come from? One such name that's been brought up is Comfort Systems. You can see here this piece by Zach. This company was actually raised by one of the more prolific gentlemen on our Discord server, Russ, a paying subscriber who often brings interesting names to our attention, and he raised Comfort System. So, if you're involved in construction, then you're going to be benefiting from all the building that's happening. That's what we see here with Comfort Systems. This is a firm that has more than 50 operating companies under their umbrella in 184 locations across the US. They describe themselves as a leading building and service provider for mechanical electrical modular and plumbing building systems. The word modular being key there. Tell you why in a second. But data center and technology infrastructure work has exploded to represent roughly half of all their total revenue. It's made their backlog go absolutely through the roof. Here you can see their revenue breakdown and of course as I said technology what representing over half of their revenues year to date in 2026. But look here on the right it talks about new construction representing over half. And then we look above there and you have well 11% of revenue is coming from service projects, service and maintenance. That's not a whole lot off of their base. If all these data center buildings are built and that's the this massive expansion that you're seeing now, unprecedented trillions of dollars being pumped into building data centers that's not going to last forever. What happens to a company like this? Well, they need to fall back on services revenue. So, in the case of Comfort Systems, one of the reasons they're seeing a big boom is that their off-site modular manufacturing is exactly how hyperscalers are trying to shorten build time. So, you don't have to go put everything together on site. They're doing it off-site. Speeds up building time. And their backlog now is reached somewhere around 12.5 billion driven almost entirely, they say, by technology and modular data center wins. Now, it's important to note they're primarily billing for high margin engineering labor, their modular assembly, and complex installation rather than selling some sort of a proprietary cooling product line that might give them more opportunity for services and recurring revenues down the road. Now, when we look at their SVR 5.5, okay, below the average and a Ford P, interestingly enough, right in that range, right, 25 to 30 range. Now, one name that is usually brought up in the context of cooling I wanted to touch on as an example of something we wouldn't find compelling would be Munter's group. And if you're going to be claiming that you're benefiting from the AI boom, we need to see the revenue growth now, not tomorrow, today. Here you can see the quarterly revenues for Muners Group on a decline. The last piece that we did looking at data center cooling pointed to the fact that the other segments for this business are actually bringing them down. I recall food tech being one of those and at that time it was actually growing and I saw in their recent press release, well they're going to spin off food tech to sort of focus on their growth segment. That's great, but food tech was growing as well was the other segment they had that wasn't. So this isn't appealing. We don't want to invest in a company that's planning to see growth in the future when there are all kinds of alternatives here that are actually seeing growth today. So just in summary, revenue growth is what shows us that you're participating in this data center investment boom. Cooling is one component of these buildouts, a critically important component given how much energy it consumes and one that's expected to show outsized growth relative to other themes within the broader data center buildout thesis. Now, many companies are going to be benefiting from the initial buildout boom, but we'd lean more towards companies that have diluted offerings as opposed to those moving more towards concentration through M&A events for the purpose of being able to show some resilience if there's any sort of pullback because if or when the AI bubble bursts, the rerate is going to be spectacular for some of these names. Not all. That's a good segue into this piece that we recently did on what happens if the AI bubble bursts, how that might impact your portfolio. Give that a watch next. Thanks so much for taking the time to watch this today.