Comfort Systems is a D-tier serial acquirer due to excessive exposure to AI capex cyclicality, which creates risk of deceleration when spending falls off.

All right. Well, that brings us to our last serial acquirer of the day, Comfort Systems. Uh I can kick us off a little bit here because I actually originally pitched Comfort Systems on the podcast to our colleague uh to our colleague Daniel. Um really interesting company.

Uh, I was not expecting it to Has it been a 5x, Kyle? >> Yes. >> Yeah. I was not expecting it to um 5x since I called it overvalued uh in the spring of last year. So, uh you know, take what I say with uh with a a grain of salt.

Um we've got uh uh one user lay max saying that uh comfort system should be S tier. Um thank you me for having discovered it. Yeah, I'm I'm uh very glad that you discovered it because I I I discovered it and then didn't invest in it and missed out on the five baggers.

So, um I'm glad glad somebody made some money on on Comfort Systems.

Um for for anyone not familiar, um Comfort Systems is uh really in the um contracting services business and specifically like HVAC. Um and so you know it was as boring of a business truly as you could probably think of.

Uh for a long long time they would you know go around um and make these acquisitions of of small uh you know HVAC uh electrical engineering players um and uh you know roll them up uh and then you know just continue to run the same playbook that we've seen a lot of these companies run.

Um the thing with comfort systems is that you had this very boring compounder uh in the HVAC space get totally hijacked by um the AI trade um for for better or worse.

And I mean for for folks like Laymax who uh who who who who have capitalized on the the runup in the stock uh you know that's that's a great thing. Um from my perspective you know I had already thought again you know we we hadn't even seen some of the the massive spending deals that the hyperscalers were uh were going to announce later uh in in 2025 and into 2026.

Uh so you know but I I I thought we were already getting to a point of of a lot of frothiness um in the AI trade. Kind of hesitant to call it a bubble. Um but you know definitely a point of frothiness uh where there was a lot of capital being allocated and a lot of that was being extrapolated uh that these trends would continue going forward and we just didn't know whether they would be.

So that's why I thought you know at a fifth of the price I thought comfort systems was overvalued.

And I guess I should clarify the reason they got swept up in the AI trade um is because uh they're sort of uniquely able to uh cater to data centers. Um so that's become a rapidly growing uh part of the business.

Uh and so you know as you can imagine data centers uh are these massive massive operations. Um you're you know running superpowered computers uh that need a lot of cooling. um and and it's very very valuable resources.

So you uh you can't really afford for them to get overheated and destroyed. Uh so you know having um sort of a specialized uh you know HVAC teams that can come in and and support uh your data center operations is incredibly important.

And so like I said across the country um comfort systems uh you know has acquired a number of businesses that specifically cater to servicing data centers. So um as this data center boom has has taken off they've found that their services are um you know incredibly highly in demand.

Uh so you know on the the one hand you have this cyclical boom um supporting them. uh I I wasn't sure how sustainable this would be long term and I wouldn't say that I I still am uh very confident um and you know historically you see a lot of economic cycles especially around revolution revolutionary technologies uh they get caught up in these um boom and bust cycles and we're very much in the boom um considerably more boom than when I thought things were already uh topping out last year uh so you know it could be another year, two years, three years, I don't know, you know, maybe there won't be maybe this cycle will be different.

Maybe there won't be a bust. Uh but if if history is any guide, uh you know, markets tend to to overshoot in one direction and then overcompensate in the other direction. Um and I suspect that uh yeah the types of um multiundred billion dollar commitments we're seeing from hyperscalers uh at some point uh they will have uh sufficiently balanced supply and demand and won't need to keep uh building out massive data centers.

Um and then that just means incrementally uh a lot less new business for for comfort systems.

Um so I I'm mixed on comfort systems. I think it is um I think it is a great um management team. Um I think uh you know they've they've run the business very well and very disciplined for for a long long time.

Um but then at the same time um particularly the stock but also to some extent the business model um has become increasingly uh sort of a leveraged bet on um AI capex. Uh and that just makes me that makes me really uneasy.

Uh, and so I think this is going to rub people the wrong way because I I probably am gonna I would propose uh maybe making comfort systems our lowest rank here. Um, one because we need one can't everything can't be CT or above but also also two like I said um you know at the beginning this is all comparative right?

So you know if we put comfort systems as a D it doesn't mean that it's actually a Dquality business. It's a Dquality compared to Constellation Software and Birkshshire Hathway.

Um and uh yeah, the you know it's it's highest margin, you know, highest growth part of the business. um I would say is you know with constellation software there's uh some debate for whether AI is a tailwind how much of a disruption it'll be uh we pretty much know for a fact um that uh once AI capex falls off um comfort systems business is is going to to fall off uh or at least growth is is going to dramatically decelerate uh and so you kind of are looking ahead and you know that that um bump in the road is is coming.

Um, so anyways, I would probably put it I'd probably put it at a D, but how do you think about it, Kyle? >> Yeah, I mean, this is an interesting one because we we haven't really gone over cyclicality too much in in any of the uh companies that we've we we've looked at today, but here you can see that obviously Comfort Systems backlog has been incredible, especially the past two years.

Um, and it could be incredible for who knows how many more years into the future. But the fact I think that Sean was trying to make and that I I agree with is that we don't know when um when this when this investment cycle into the AI data centers is is going to stop because at some point it will who know I don't know when but it will um and but you know the point also being comfort systems is still a good company.

I mean look look at the backlog even before 2025. It's very very good growth.

But I think kind of getting to Sean's point about comparing I mean the insider ownership on this business is is is quite low um compared to I think every single business that we've gone into today.

Um their incentive program is really really good. I'll give them that. Really really really good.

Um but then you know I think if you if you're talking about a business that's exposed to cyclicality out of all these businesses this one's number one. Um, you know, you could make the argument, well, Bergkshire maybe is up there too, but the thing with Bergkshire and the thing with all these other businesses, well, not all of them, but Berkshire and the Constellation and everything, they're they're super diversified.

So, even if one part of the business is exposed to some degree to cyclicality, it has other areas that can make up for it.

Whereas Comfort Systems, I mean, if you just look at their organic growth, it it kills basically all the companies that we've talked about today just because they have this tailwind that none of these other companies really have right now.

whereas I feel with comfort systems comparatively speaking it just has way too much um exposure to cyclicality and then kind of what to Sean was Sean's point being there like in my opinion yes I think it probably belongs below everything else but if if we come back in 5 years and um the the AI the cycle is still going uh comfort systems might blow all these out of the water in terms of returns I I could I could very well see that but um yeah I think you had it at D.

I think I think that's I I I think I'm I'm that's kind of where I would put it as well.