Comfort Systems is a high-quality business with strong growth, but buying at the current elevated valuation is too risky due to cyclical AI exposure and lack of margin of safety; the speaker remains on the sidelines.

What happens when you pass on a stock for being too expensive and it goes on to become one of the best performers that we've ever covered? That's basically our story with Comfort Systems.

We had a fair value target in mind and the stock blew right through it and we sat on our hands and watched. Today, we're going to find out what exactly we missed and whether expensive was ever really the trap that we thought it was.

So today's episode we're going to cover a business that Sean covered a while back, Comfort Systems USA. For anyone who didn't listen to the original pitch, it's basically a mechanical, electrical, and plumbing installation and service provider based out of Houston.

While management doesn't like the label serial choir, I think it's probably the best label that we can use to help understand exactly what they do.

And and part of the reason I like the business was that it does have this wellestablished playbook with over $10 billion in revenue and and over 23,000 employees. So, in terms of size, it is a pretty big business.

But the real story has been their ability to continue compounding even at a large cap size. And so for a business to have a market cap of over $60 billion while compounding revenue at 13% a year and earnings per share at 22% a year for the last two decades, that is really something special.

But what I found even more incredible is that most of the time when you look at the growth rates of a business with maybe a similar market cap to the size of Comfort, growth tends to be plateaued or is even declining.

When I look at the past 10 years of comfort systems, revenue growth has actually increased to 21% while earnings per share has gone up to 37%.

So, it's actually really really clear, at least from what the numbers are telling us, that the business is just getting better. Just in the last year since you pitched this, Sean, revenue has gone up by 33% while EPS has gone up by over 50%.

So, you know, it's just at a very astounding growth story.

Now, I think part of it is definitely timing. I think when you pitched it, it was kind of right before it started to really, really heavily inflect. So you mentioned in your episode on some of our winners that you didn't actually have any regrets in missing this one because you believe that the current tailwind is going through the result of kind of these cyclical forces that were and probably still are very very unpredictable.

And before we get into the cyclical nature of comfort systems business and what they do, I do just want to give listeners a little bit more context on the original decision to pass on investing in comfort systems.

And so when I first analyzed the business, I did estimate a fair value of about $320 per share, which feels a little funny in hindsight because, you know, I was quibbling over what a good entry point would be in the stock.

And I talked about maybe wanting to buy some shares if the stock fell below $300 per share. And the stock has absolutely just rocketed from there and traded at a huge premium to my valuation.

And so I already had low conviction in what the business would be worth because its earnings are cyclical. So I I put it in the too hard pile, but I did end up missing out on a five bagger.

And so at the time of the recording, the share price is almost $1,800, which uh yeah is painful to see that this opportunity was so clearly staring me in the face and we didn't capitalize on it.

And so, you know, a huge opportunity in some sense was missed.

And again, we can talk about whether this was a mistake or simply just the occasional reality that if you're going to stick to a disciplined fundamentalbased investment approach, every now and then you're going to get these things where the casino falls in love with a stock and it flies to the moon, but it's not necessarily indicative of the types of fundamentals that some of our influences and Graham and and Munger and Buffett would find attractive.

Yeah, I think with some businesses it's just crazy to think about how much value they can create even in a pretty short time. So, I remember listening to your episode and not to insult you or anything because I never bought the business either, but you were discussing whether to buy it at 290 or 320 and with hindsight, you know, haggling over those small price differences just wouldn't have made any difference.

Instead of maybe a four to five beggar, you would have been looking at a five to six beggar. And I don't think anybody is greedy enough to really pass on either of those two options.

Now, when you you say it like that, it does feel almost embarrassing to have it be that big of a payout in such a short time period, basically from immediately after I looked at the company.

I think there's a major difference though in making an analytical mistake versus making what is maybe a correct and pragmatic call as an investor by honoring your own margin of safety and then having that just so happen to get massively overrun by an unknowable swing to the upside in demand where you made a calculated decision given the information you had on the table.

Maybe you made the best decision available. It's just that in the random distribution of everything that can happen, there's going to be these small percentage outcomes where you end up being hugely wrong.

Sure. Yeah. So the original thesis was based on comfort systems growing through a few different levers. And so the first was organic growth. And if we assumed they never made another acquisition again, then I thought maybe the business might grow somewhere around 2 to 4% annually, which would basically mean around the same level of GDP growth because demand for mechanical and electrical engineering services should be pretty steady when you really zoom out.

And if anything, they might outperform GDP a little bit.

And so, especially as you have things like data centers becoming larger and larger parts of the economy, which we'll get to, that's a big part of the Comfort Systems story today.

Comfort Systems is specialized in doing maintenance work on data centers and in HVAC work as well.

So, I think the real beauty of this business is in the fact that they're a serial acquirer, meaning they've demonstrated the ability to continue to allocate capital back into the business by snapping up regional contractors all across the country and then doing so with very high rates of returns on those investments.

So, just to give you an idea, since 2015, they've had a return on incremental invested capital well north of 20%. And so, what that means is the more money that they can reinvest, the better.

And what that means is that if they were able to reinvest 100% of earnings at these 20% rates of return and when you stack organic growth on top of that, you should expect earnings to compound at well more than 20% a year.

And the thing is, of course, it's not that simple though because often times you can't expect companies to continually find enough investment opportunities to put all their earnings into that generates 20% plus return.

So maybe you can find some incrementally very attractive investments, but you're not going to be able to reinvest 100% of your earnings into those opportunities. And typically what you more often see is diminishing returns on investments after a certain point.

So only so much capital can be invested each year at a satisfactory rate of return.

And then your expected returns as a shareholder, especially in a serial acquirer business become the byproduct of the organic growth and the underlying companies growth from whatever percentage of earnings can be invested into acquisitions times the rate of return on those investments.

And then any capital returns to shareholders that come via either buybacks or dividends. And so with Comfort Systems, they do actually have multiple different areas that they can reinvest capital into in making acquisitions with.

And so that is both in the field of electrical contracting and also mechanical contracting with mechanical contracting comprising over 70% of revenue. And so really the big question for me was what do the opportunities before this company look like going forward?

Are they running out of attractive businesses to acquire? And then also correspondingly is basically the law of large numbers kicking in where they have to find bigger and bigger deals over time or significantly more deals to be able to move the needle.

And so that is where I ended up being quite wrong.

I I thought that the business was much more mature than at least it's proven to be in the last year. Yeah. So as a major fan of serial choirs, I can actually attest to this exact business model.

I think when it's done correctly, the business essentially just turns into a compounding machine. And when you have one of those, you just hold it. And good things tend to happen if the business is very, very wellrun and if you're just willing to be patient.

And I think Comfort Systems is one of those really kind of boring but highquality businesses that investors love to talk about, but unfortunately tend to rarely own. And one reason that many investors probably don't own it is simply because they're just too boring.

But even though Comfort Systems was boring and probably still is boring, it definitely had a very interesting angle to it that you kind of already alluded to here, which was the company's connection to the AI craze by servicing these data centers.

That turned an incredibly mundane business just two or three years ago into one of the hottest names in the entire market. And correspondingly, Comfort Systems backlog absolutely ballooned in 2025 and is still carrying some momentum into 2026, albeit at a lower growth rate.

So just so you have some idea of how the backlog has actually grown. It has grown by over eight times since 2020. Just in an insane growth figure and it actually more than doubled between 2024 and 2025.

So looking at it in 2025 like you did. I think you might have already thought that you were probably a little bit too late. What might have been an attractive boring business to own at a fair price had already received this massive tailwind from the AI spending.

And so, you know, there was a risk that had already been swept up in the hype and become massively overvalued based on just these temporary boosts to earnings. Is that kind of fair to say that?

Yeah, I would say that's exactly what I thought.

And I did see that backlog double over the course of just a year for what had otherwise been a pretty slow growing and plain business, at least in more recent years up until that point.

And so I assumed that comfort systems would still be the beneficiary of incremental spending on data centers. But I definitely underestimated just how much further the AI capex cycle could go and is still going.

And I very much recognize though that AI would be a net positive for the business and could boost their earnings power for years because once these data centers are built, you need routine HVAC maintenance to keep the buildings functioning.

So new buildings can become customers for many years and revenues from HVAC services on those centers tends to be much higher margin than the other type of contracting work that Comfort Systems does because it's very specialized.

you're talking about cooling very very expensive in advanced computer chips and so you can see how the rise of AI was becoming a boon for comfort systems business in the physical world but I think we've all been shocked by just how much money the big tech names have been spending on data center capex I mean really it's it's unprecedented historically asset like businesses are breaking records for the amount of money invested in a single year

And at the time I looked at comfort systems, it wasn't that long after we had this so-called deepseek moment where it seemed like LLMs could be trained at a fraction of the cost.

But spending has just kept growing nevertheless. And so even when you identify a lasting positive tailwind, as I think I did, I've very much been humbled at how difficult it is to account for just how asymmetric those impacts can be on a given industry.

I mean, data center spending has grown exponentially. And exponential growth is a really hard thing to confidently model because if you're even slightly off, the margin of error in your intrinsic value can just become massive versus underwriting growth in a more stable industry that's more linear in nature.

So I was pretty much wrong by an order of magnitude about just how much comfort systems could benefit from AI capex. I would say with it is still an open question of what percentage of these inflated earnings are actually going to be sustainable for them and maybe it is a new paradigm shift and these earnings will be more sticky but this is really true for any cyclical business.

You need to have an opinion on where you are in the industry's business cycle and then be aware of what multiple you're paying for that business relative to where it's at in the cycle.

So when things are going abnormally well, earnings will rise explosively and then the PE ratio can actually look cheap at a time when the stock is actually more expensive.

And so if you buy a cyclical business at the top of the cycle, the multiple is going to look cheap on paper because Wall Street is not giving the company full credit for what it recognizes as a temporary earnings boost.

And then once the cycle turns down and pessimism around the stock bottoms out, then the stock can actually offer the best value despite the PE ratio on paper looking much much more expensive because the denominator and earnings are really depressed.

And so point being, the rule of thumb that you'll hear people say when it comes to investing in cyclicals is that when cyclicals are cheap, they're actually expensive. And when they're expensive, they're probably actually cheap. That's right.

And the thing that I found interesting about Comfort Systems, at least your original thesis, was kind of the shift just actually in terms of the operating margin. So in the original thesis, the model projected that operating margins would steadily decline over 5 years to around 8 12% from around 10% range when you first looked at it.

But today, operating margins have expanded all the way up to 16 a.5%.

And this is kind of a case when you run into some really interesting problems that you kind of just discussed with cyclicality. You know, look at comfort systems operating margins from 2005 until 2022 and they were a pretty nice range that was quite tight kind of in like around 5%.

But since 2022, operating margins have skyrocketed to nearly 17% today.

So it kind of begs the question, you know, what are the normalized operating margins going to be in the future? Yeah. No, I mean the margin expansion has been really incredible for a business where it's not like software as a service or manufacturing where you can think of how economies of scale directly translate into more operating leverage.

If anything, that's one of the criticisms of this serial acquirer model with regional contractors because you can really only grow by acquiring other contracting businesses and they're very labor intensive and so it's very onetoone of you know if you want another dollar of earnings you need to spend a dollar to acquire that and you're not necessarily getting you know massive operating leverage because you don't have

So that's what really has made this margin expansion that Comfort Systems has pulled off all the more incredible but also all the more hard to believe at least from my perspective that it's going to be lasting.

for the business to really increase by about five times in price. Margins aren't the only reason for that. So, let's go quarter by quarter here and walk through how things have played out because it's not often you find a business that can 5x in a single year.

Well, the PE was somewhere in the mid-30s when the episode came out and today it's sitting at about 43 times. Now, that's obviously a really nice multiple expansion, but I don't think that's nearly enough to explain a 5x in share price without a massive earnings growth, which to be fair, there's been a fairly decent amount of.

The first is in revenue growth. So, in 2024, the business did 7 billion in revenue. In 2025, they did 9.1 billion in revenue. And in the last 12 months, they've done 11.2 billion in sales.

Since fiscal year 2024, revenue has compounded at 34%, which is no small feat for a business of this size.

Now, when we add the margin expansion and revenue growth together, we get Comfort Systems net income, which is nearly tripled in the last 12 months versus their 2024 earnings and is up by almost 10 times from their 2020 earnings number.

And so I think the crazy part with this growth in some sense too is that it was fully telegraphed at least for part of their revenues. And what that means is to say you could have seen a lot of this growth coming by just simply keeping tabs on the company's reported customer backlog that we mentioned earlier.

But then at the same time, if the AI bubble, if you want to call it that, had popped, then that whole backlog could have evaporated overnight. So when I think about modeling comfort systems, I clearly underestimated not only how much the backlog would grow, which is definitely hard to predict, but again, what percentage of that backlog would convert to real sales dollars.

The mechanical segment, which has historically been the larger, more traditional HVAC segment, has accelerated its growth to about 40% over the last year, which is really incredible.

But the real story has been the electrical engineering services segment, which has grown by 81% in the past year.

I went back and looked at Comfort Systems percentage of revenues from the electrical segment, and it's moved drastically over the year. So in 2017 it accounted for just4% of revenue and since then it has risen very very sharply to become nearly 29% of revenue.

And in Comfort's case they have steadily made more and more acquisitions since 2022. Nothing here seems super super splashy to me as they've deployed about $319 million over the last year over a handful of kind of these smaller acquisitions.

Now Sean, you mentioned in your episodes that they made an average of about one and a half to two deals per year when you first pitched the business, but in 2025 they really really stepped up their acquisition pace to around four and two of them were made after your pitch in FY and Ziffra and Meisner Electronic.

Yeah, I think the great thing about Comfort Systems is that they make these acquisitions from a position of strength. They're very disciplined on price. And in this case, like most of their deals, these acquisitions were funded almost entirely from cash with little need for debt.

And so, I think both of them look like pretty good deals, but the Fine and Zilstra deal in particular stands out. And so, the purchase price is about 10 times operating profits or 0.7 times revenue. So, that is a pretty reasonable price.

And I'd say FZ is also strategic for Comfort Systems in the sense that it adds to Comfort's portfolio of electrical contractors, which we've mentioned as being this major growth area for the company.

So if anything, I think it signals just how strong demand has continued to be in this area where FC brings more contractors supporting data center buildouts and also more exposure to the healthcare industry as well, which I think is generally a good area for comfort to try and diversify into.

Right? Hospitals have huge HVAC needs and they don't really go out of business or have downturns because people are always sick. So they make for great customers from Comfort Systems perspective as an HVAC contractor and serer.

So yeah, I like it mostly because comfort is really sticking to what they do best and what's been working and also they're not venturing well beyond their circle of competency or making massive unprecedented deals.

One thing I found interesting about Comfort Systems was the increase in same store sales growth. So, when you look at most serial acquirers, they tend to have organic growth rates kind of in the low single digits, but looking at the numbers for fiscal year 2025, they had about 3.4% 4% in contributions from new acquisitions and an outstanding 26% growth in same store sales driven by demand in both technology and data center sectors.

But this is a double-edged sword with historical same store sales growth being in kind of that 3 to 4% range. I would say that current growth rates seem very highly probable to regress to their mean.

And again, depending on whether it becomes clear that we've been in an AI spending bubble, you could see more than just a reversion. You might see a revenue decline for several years into the future.

But I will say, you know, that doesn't look imminent.

If we look at the mix of acquired versus same store growth on the backlog this quarter it actually increased by $1.6 billion and 1.4 billion of that was in same store growth. And so when we think about whether all this AI capex spending comes to fruition and if that is true then comfort systems remains uniquely positioned to capitalize on that spending.

And we can speak more about that. But that would effectively mean that the stocks run is not entirely unjustified.

And if AI is really everything that technologists are promising and more, then comfort systems could very well be in the midst of a new growth paradigm for their business that really nobody saw coming at the scale 5 years ago.

And then on the other hand though, if we're to have a stock market crash, a bunch of AI startups go under, maybe lenders bail on financing data center buildout, maybe Anthropics IPO doesn't go very well this fall.

I think that's when it's planned for. Well, then much of Comfort Systems backlog would evaporate, and we mentioned that earlier. And then the jump in revenue and earnings that they've seen would look like a flash in the pan.

Well, let's talk a little bit more here about how AI spending has really translated into this just massive boon for Comfort's business.

So if you look over the last year at all of Comfort Systems customer categories, there's one that really stands out very obviously over the other and that is the industrial segment.

So this segment has historically been the largest. But the growth over the last year has been near parabolic at 74%. And this segment alone accounts for nearly 3/4 of comfort systems revenue in the first half of 2026.

Now, the reason that I highlighted the industrial segment is that technology is included in this segment. And since data centers and chip manufacturing is a big reason that Comfort Systems has had this meteoric rise upwards, I definitely think it's very important to highlight it here.

Just to give you an idea of how large this part of the business is, 56% of Comfort Systems year-to-date revenue was just from technology, up from just 30% in 2024.

And they also just added another electrical business, Hunt Electric, which could add a decent amount of industrial and technology based business as well.

in per capita energy consumption is finally starting to grow again thanks to AI and data centers and so comfort systems is more of a second order thinking bet on this and that's an idea that comes from Howard Marx who talks a lot about that in investing and so thinking about the second derivative effects essentially and so increased electricity demand leads to more copper demand but less obviously it also leads to greater demand for HVAC technicians

I knew that data centers and EVs were dramatically increasing demand for electricity and and therefore electrical engineering services as well. And that's because my wife works in renewable energy.

So I saw through her work how energy demand had just been exploding. And yet, I didn't connect the dots enough to not underestimate how much this segment could continue to grow and benefit Comfort

Not to say that comfort systems management is untrustworthy by any means, but just a general view of how I tend to look at management.

I would say Comfort Systems really does have a highass management team. I mean, the fact that they were so reserved about the opportunities ahead of them probably limited my own optimism and businesses that tend to conservatively underestimate their prospects are actually probably in a stronger position than the companies that have to sell investors on just how great their business model is.

It's one of those things where if you know you're good or in a good position, you know, the numbers will eventually show that. So, you're happy to let the numbers do the talking for you.

And that's definitely more of comfort systems style.

And so, I think in my initial thesis, I probably should have spent more time on the data center and chip manufacturing part of the [clears throat] business and thinking about just how impactful that could be.

But now that we're revisiting it, we do have some more time to provide a fresh perspective on the topic. And so if we look at the company's latest presentation, management has discussed large project bookings in this segment in states like Texas, North Carolina, and Indiana.

So it's not a surprise that Comfort Systems is therefore focusing on Texas in particular as a growth avenue in the future. And so in other words, I think Texas is very important to the future of comfort systems business.

And that means that any investors looking at comfort systems today, I think you would want to pay close attention to the local regulatory environment there and the sentiment around data centers, not just in the US, but in Texas in particular because we've seen in many places data center construction and in the impact on the energy grid is very much being protested and I suspect it will become an increasingly controversial issue in the state of Texas which could ripple across to affect comfort systems. businesses.

Yeah, these are all great points and I think it really shows us that there are some real legs behind the tailwinds that Comfort System has, but it might be a few years before things begin to level off.

What we can tell from Comfort Systems disclosures is that they're willing to look out about a year or so. So, their backlog disclosure represents revenue that's expected over the next 6 to 12 months, but it probably actually underestimates the real number because they also have some shorter term projects inside of the next, say, year that are not included in the backlog.

And since the backlog is continuing to swell, obviously that's a good signal that things are getting better, right? And as we said, management is not crazy aggressive about the guidance that it tends to provide.

But they did recently note that things are likely to continue going well for the remainder of 2026.

Now we really have no way of knowing whether Alphabet is even a customer of comfort systems but if we use Alphabet and the hyperscalers as kind of proxies for data centers and chip spending I think it's really obvious that spending is likely to continue to increase over the next few years.

So I would say the chances are pretty good that comfort systems will continue to benefit from these trends.

But again, as we keep coming back, it's just hard to know how long these trends are going to last.

Uh how about we shift here though and discuss whether comfort has any competitive advantages and clearly comfort systems is riding a big tailwind that is lifting entire industries that that the company is exposed to.

And so in theory that should attract more competition and I do think that it inevitably will. But in the meantime, we've already discussed that the company's profit margins are very much trending up and to the right.

And that implies that demand is growing fast enough to more than offset any new competitive pressures.

And because profit margins have been expanding, we know that this is not a case where Comfort Systems is out there just trying to bid on any project they can find, even if it comes at a loss, just to pump up their revenue figures in the short term.

And to make things even better, the electrical contracting business is very much tied to having more longerterm contracts than some other forms of contracting services. So basically, every incremental dollar they're earning in this electrical segment is arguably worth more than some of their other business segments because those cash flows are on paper at least contractually more likely to continue for longer. They're more locked in.

this is a business that's I would say not even close to the same amount of quality as Comfort Systems, but it's an interesting opportunity for me.

But when demand is high like it is right now and obviously like Comfort Systems is exposed to, it's a really really good thing which helps boost margins.

So you brought up in your initial pitch that you thought comfort systems wasn't necessarily the strongest business in terms of having a moat and I agree with you then actually and I still do but the one competitive advantage that you highlighted was that they are highly specialized and they have this kind of geographic advantage that helps them stay sticky with their current customer base.

Now one additional advantage that I think comfort systems has is developing the ability to work on just larger larger projects. So at the end of 2025, they had about $24 billion in aggregate contract value spread across nearly 8,500 projects and one project was reportedly worth about $470 million. So just massive in scale.

Now if you're looking at that from the point of view of let's say a local or even a regional competitor, if they don't have a balance sheet that Comfort Systems has, they're just not going to be able to bid on a project anywhere close to that size.

So if we use Hamilton Helmer's framework, I would consider this to be kind of a corner resource. Since competitors don't have access to the same balance sheet, they simply cannot participate in those same contracts that comfort systems can easily pursue.

This is for sure becoming a much more powerful advantage for them as they begin working on larger and larger projects and as the financial scale of those projects grows.

So I'd like to just briefly cover how this advantage really works because when I first read it, it was kind of confusing. So they have something called shity bonding capacity which allows them to secure these larger projects.

When a potential customer of comfort systems wants a project to be completed, they need to know that the project will be guaranteed to be completed. If the company, let's say, messes up, let's say they walk off the job, or if the company, worst case scenario, goes bankrupt during the building process, well, then the project is going to be heavily delayed, or it might even be canceled completely, and no customer is obviously going to want that.

So, you can think of shity bonding as a type of insurance policy against that risk.

A company like Comfort Systems makes a promise to their customer that they're going to finish the job. And if they don't, well, then they have to pay for the job to be completed.

And to date, Comfort Systems doesn't have any losses, nor are they expecting any in the future.

So again, Comfort Systems strong balance sheet and track record very much give them a step up against the competition in securing shity bonds at a reasonable price. But since Comfort Systems obviously has the capacity to see these projects through, unlike really most of its peers, they can get bonded for increasingly large jobs pretty easily.

That allows them to, as we've seen, pretty uniquely be able to scale their margins as the business has grown to an extent that goes farther than probably what you would have expected a regular contracting business to be able to accomplish.

I think when we're looking at this kind of quarter resource advantage, it's probably widening. But, you know, if I'm looking at other competitive advantages, I'm not really sure there's one that really jumps out to me.

And I think that comfort systems model has been compounding over time with them being publicly now for over 30 years. And the business was founded all the way back in 1917.

one of the more compelling aspects of the company structure, and I know Sean, you originally commented on this, was like how it's kind of like Bergkshire in the fact that they're both decentralized business models.

Another part of the business's process that I think has been improving is their move towards pre-fabrication and modular construction. This allows them to build parts of the project off-site, which helps improves their ability to finish projects on increasingly more and more complex projects.

And specifically, I looked at returns on incrementally invested capital between 2015 and 2014. And the numbers were really good, over 25% returns. To put that simply, what that means is that the business was doing an exceptional job at allocating capital.

And given how well the business has done recently, I'm going to guess that their incremental returns are even higher and will continue to be so for some time. I wouldn't put Comfort Systems in the same vein as those two companies, but they certainly aren't doing so bad either.

And with that I got a number that was very impressive because our incremental ROIC has more than doubled to nearly 60%. And due to this increase the compounding rate of the business has also increased up to about 39%.

This is something I think is a great exercise because it shows that they're actually allocating capital even better and the market is currently rewarding shareholders because of it.

Yeah, it's clear that they've been doing an exceptional job on the capital allocation front with the usual caveat that I think we've discussed at length today being that much of the organic growth appears to be part of a tailwind that I don't think will last forever.

And so when I first pitched Comfort Systems, they were averaging somewhere around one and a half to two deals per year. And since the start of 2025, they've completed six deals across their mechanical and electrical segments.

And so those six deals had a combined purchase price of over $540 million. So they are deploying an unusual amount of capital, I would say, compared to what they've done in years past.

And so Comfort Systems is not going to be doing any deals that generate those kind of headlines. But when it came to Comfort Systems, I felt like I really had no idea what they were looking for. or at least in terms of their own disclosures.

Now, this doesn't necessarily mean the company is putting up a giant red flag. It just means that perhaps they don't want to attract any potential copycats and give them the blueprint to success, which would end up harming Comfort Systems over the long term.

One thing I do appreciate before I attempt to look at the multiples they pay, is that they are using notes to sellers as part of their acquisition payment. Now, I like this because it means that Comfort Systems doesn't actually have to front the entire purchase price itself.

We know from Comfort Systems disclosure that they expect to recognize about 75% of the backlog as revenue over the next 12 months. Then I applied Comfort Systems own margins to its mechanical and electrical segments to estimate future operating earnings.

And this gives us a directionally correct view of what they're paying for in these businesses.

You know, the average multiple was around nine times operating earnings. I found this interesting because usually when I look at serial choirs, this multiple seems a little bit high to me. not really high, but maybe a little higher than what I usually see.

But you also just have to consider how good Comfort Systems organic growth has been. So, in that case, I'm okay with them paying a little bit higher if they think that they're going to continue to get some organic growth.

But I will say to be honest, I'm kind of torn because over the long term, what kind of organic growth can you expect? You know, 26% organic growth from an explosion in data center spending is completely unsustainable over the long term.

Now, one final area that I wanted to highlight is a small hiccup they had in their latest acquisition, which was called Summit. So, if we look at the contingent earnout obligations for fiscal year 2025, they actually decreased by 62% caused by a lower earnout expense for Summit.

They also wrote this decrease was primarily caused by lower earnout expenses for Summit driven by larger changes in their forecast results in the prior year and as a result of them reaching their maximum cumulative earnout target.

So perhaps Summit when purchased was maybe a little too bullish on next year's numbers, but generally you don't like seeing this given how well everything has gone for the business.

And so if these decrease, it seems like that might be a good thing because costs are declining for Comfort Systems.

It's actually more of a bad thing because it means the new business they acquired is underperforming expectations. Another area that you pointed out that was a major strength for Comfort Systems in your original analysis was their debt profile.

So around the time that you were researching them, they had about $131 million in debt, but also had $860 million in cash on the balance sheet.

On top of that, the business had just generated over $700 million in operating cash flow over the last 9 months. So, you know, it's pretty obvious the business was incredibly safe in terms of its debt profile, but given the fact that they've spent nearly $540 million on acquisitions over the last year and a half, I thought it would be a good idea to maybe revisit their debt and see how things have played out since then.

To the fact that this business has negative net debt was a real positive for me. And with comfort systems, you know, they have a lot of financial leeway. There is a margin of safety built into the business there.

And I would say that bankers are probably falling over themselves to lend to a company like Comfort Systems with such a high quality balance sheet because you know that there's such a good credit.

So, you'll be happy to know that Comfort Systems has actually decreased its long-term debt since you looked at them. So, as of Q2 of 2026, they currently have 1.8 billion in cash and only $53 million in long-term debt.

Last summer, they amended their credit facility to increase borrowing capacity up to about $1.1 billion. But from what I can tell, this is completely undrawn.

So all the debt that they're currently carrying is actually from the notes to former owners as part of their acquisition strategy. And these notes carry really really good interest rates of just about four to 5 1.5%.

(59:00) And these notes carry really really good interest rates of just about four to 5 1.5%. Doesn't really surprise me. I mean given how much money this business has made and how its cash position has swelled, it does seem obvious that they wouldn't really need liquidity.

(59:23) But the only real reason I can see that they would need to reach into that bank debt is if they found a monster acquisition target with a multi-billion dollar price tag, but that doesn't really fit into their wheelhouse of making these small regional acquisitions of contracting businesses.

So, I wouldn't expect it, but you can't completely rule it out.

Now, there really isn't too much more to say when it comes to debt simply because the business just doesn't have that much. (59:47) The business simply has a beautiful balance sheet as it's well below its debt covenants and is very very well protected by both equity and cash that it has on its balance sheet.

And when I first looked at Comfort Systems, Brian Lane was their CEO and president and he had clearly done a very good job of setting the business up for success.

(1:00:07) So, interestingly, Lane gave up the title of president in December 2025 to Trent McKenna, who previously had served as the business's chief operating officer, as well as serving some other roles since he joined the company all the way back in 2004.

Yeah. So, this seems like Brian Lane has perhaps found his successor. (1:00:30) You know, he's now about 69 years old, and I don't really blame him for wanting to maybe give up some responsibilities to a possible replacement for him in the future.

So the title of president carries a lot of responsibility and this is probably a good test for Trent McKenna to see if he's the right guy to continue leading this business into the future.

(1:00:46) Since I think that culture is just an heavily heavily underrated part of a company's competitive advantage, I'm always happy to see businesses that internally promote.

So I actually really like this assignment for Trent.

It's hard to really give him an honest review yet because he's been in the position for such a relatively short period of time. So all we can really do is look at some of the operating metrics since he's been there.

(1:01:02) And I would say so far they're very good.

But to be fair, I'd probably attribute more of that to Brian Lane, their current CEO. It feels like a move in the right direction.

And given how long Brian Lane has surrounded himself with this current version of the management team, I I can only guess they've also made a large contribution as to why this business has had so much success.

(1:01:22) And I think another part of the business I highly respect has been just how stable the leadership team has been. Right. The stability in leadership is always kind of a positive because I think it denotes that a business does have this kind of good culture.

But unfortunately, I think the market realizes that too, which is why I think stable management of a good company tends to carry a premium price tag on the business.

(1:01:44) So, you probably won't be surprised that this business currently trades on about 31 times operating earnings, which is very high. But, you know, given the recent performance, it's probably not all that surprising.

Another area worth exploring in regards to management is in insider ownership. So on this end, the numbers don't exactly scream alignment, but I think they're adequate enough.

(1:02:05) So since 2024, insider ownership has been on a slow decline, going from about 1.8% in 2024 down to about 1.2% currently, considering one of the business's founders, William George, who is the CFO, is still on the team.

I would have liked to see maybe a little higher insider ownership, but it's definitely not a game-breer for me. (1:02:27) And and how about we discuss maybe the incentive plan in more detail then?

because in my initial analysis, I outlined how management compensation at all levels is tied to earnings per share growth and free cash flow.

And so that's actually one of the better incentive plans we've probably ever covered on this show where there's just a real degree of alignment that helps shareholders feel like they will be enriched alongside management.

Yeah, I absolutely love this incentive plan. (1:02:46) It's probably one of the better ones that I've ever had the privilege of analyzing. Another area I like about this business is that the base salaries all seem to be very decent for a company with a market cap of $60 billion.

So all executives are making between $300,000 and a million.

(1:03:02) Now, kind of getting back to your point there on the incentives, they're basically just sticking with the exact same thing, which I think is brilliant. Um they're sticking with earnings per share as well as this kind of free cash flow incentive.

So the other thing that I really really respect for them is that they give targets for earnings per share as well as for free cash flow. (1:03:18) Most businesses that I've come across don't actually tell you what those targets are publicly.

I assume they just keep them internally so management knows what they need to hit, but for whatever reason, they don't tend to tell shareholders. But in Comfort Systems case, they actually do disclose these numbers.

So listen to this. EPS target in 2023 was about $5.50 and in 2025, it's up to $15.75. (1:03:38) The free cash flow target over the same period went from about $24 million to $32 million.

Now to be fair, the actual numbers are quite a bit higher. So I think it's a target that they have very high conviction in that management can hit. Nonetheless, I feel that the simple act of hitting these targets will create a lot of shareholder value.

(1:03:54) And obviously, if they exceed them, which they have by a country mile so far, then shareholders are obviously going to be even happier. Now, just to give you an idea of how well management has done, they've received the full 200% of their target, which is the maximum they're allowed to for each of the past 3 years.

(1:04:08) It's pretty safe to say management is top-notch here and the proof is right there in the numbers. Executives have earned very nice performance bonuses, but they've been welld deserved.

They've also earned shareholders a pretty nice rate of return, right, with shares compounding at over 100% annually since 2023.

(1:04:26) I completely agree. So, the only kind of yellow flag I found here is in the declining ownership rates. You know, I think this is a signal that management is maybe converting their options into cash upon vesting rather than holding them.

But, you know, I guess I'm I'm okay with this for the most part. (1:04:42) I'd obviously prefer that they keep insider ownership above 1%. So, there is a risk it drops below that over the next few years if insiders keep selling.

But, you know, if you look at this from a strategic standpoint, management might just think that the timing right now is probably when they will get most of the cash from their vested options.

(1:04:58) Perhaps maybe when this data center chip manufacturing cycle ends and shares become more reasonably priced, they'll be more likely to hold on to them, but we obviously won't know until that actually happens.

So, when we first looked at the company, our co-host Daniel asked some really great questions about the risk inherent to this business. (1:05:14) And so I think we looked at the risk in the business cycle simply because you know this was to some extent a company that was so closely correlated with the construction cycle more broadly across the US and that was and has been a pretty volatile industry.

But I made the point that Comfort Systems is not only reliant on new builds. (1:05:38) So even if new builds are in a rut, about 40% of its business comes from existing buildings, things like renovations, expansions, maintenance, and repairs.

And in an economic slowdown, if there's a cash crunch, some of that stuff can be deferred, but it can only be deferred for so long. And so that sort of serves as a natural hedge for the business, right?

And you can actually argue that the company has further diversified, which I think is another edge for that business. So, you know, now that they have the multiple segments and are working more on technology, which obviously has its own cycles, they're somewhat more protected.

(1:06:11) But again, once that cycle turns, I can't see how this business maintains its current momentum. And you can probably argue that they're trading more traditional construction related cyclicality for more exposure to just data center or chip manufacturing cyclicality.

(1:06:27) So the other risk that I'd point out here is more of an inherent risk in this business that I personally lost money on. And that's just looking at a business that publishes their backlogs than trying to extrapolate future numbers from them.

So I've had two very small businesses that had backlogs that I felt were very large for their market cap size. And the fascinating part was that I ended up actually doing well on those bets for a very short period of time.

(1:06:49) But then the longer I owned the business, I started running into some issues.

And you know, some of those issues might be that the projects take longer than you think. So the revenue that's recognized originally in a year, maybe that takes 18 months, or maybe that takes two years.

(1:07:04) Or you run into the problem where they're working on just a bunch of different projects and they have deadlines that come out immediately after they record their earnings and therefore they don't get that revenue until the following quarter.

So my experience with businesses with growing backlogs, to be honest, hasn't been great. (1:07:20) But I will say, you know, looking at this business's revenue, earnings per share, and free cash flow, it's absolutely outstanding how steady their growth has been over the last few years.

The only KPI that seems to be a little bit more volatile from quarter to quarter is in their free cash flow number. Well, it is after all a business that constructs things and so therefore, you're going to have some fluctuations quarterly in working capital needs.

And yeah, that that shows up in free cash flows, but it doesn't bother me hugely.

(1:07:38) I think I'm okay with the lumpy cash flows in the short term because over the long term it does seem to balance out. And since free cash flow has compounded at 22% a year for the last two decades, I have a lot of faith that they'll manage these working capital needs pretty well.

Well, I think it's the time of the episode where we look at the intrinsic value of comfort systems. (1:07:59) So, since we're revisiting the business, usually we wouldn't see much of a shift in our model, but I actually had to make some pretty big adjustments to the model given the five times in share price appreciation that we've seen. so far.

Yeah. Yeah. (1:08:16) I mean, the original model was built using 2024 numbers and well, a lot has changed in that time in the world of AI. So, it's definitely due for an update.

(1:08:32) So, I went through and rebuilt the model using the trailing 12 months numbers from the first half of 2026. Now outside of the revenue base number changing substantially the other part of the model that needed adjustment was obviously on the margins.

So the original model had operating margins gently declining from about 10 1/2%. (1:08:48) Now instead of the margins declining the business has actually scaled dramatically and operating margins are now running at about 16 1.5%.

So we have to account for that in the model as well.

But if I'm being honest here, modeling this business, it's pretty tough. (1:09:03) And I think it might actually be even tougher now than when you first modeled it, Sean, simply because I feel the business is now running somewhere around the top of the cycle.

And to be quite honest, I'd be lying if I said I had any insights on to how much longer this AI cycle is going to last. Well, I'm right there with you on that. I'd prove to be pretty premature in saying that, though.

(1:09:22) But how did you uh think about the base case when you tried to model out comfort systems fair value? So, I assume that revenue continues to compound at about 13% per year on average, reflecting some slowdown, but also a new reality where Comfort Systems has more opportunity for growth thanks to AI than it did over the last few years.

(1:09:38) Now, with the assumptions that the company's valuation multiple comes down once it's clear we've reached the top of the AI spending cycle, I also try to approximately account for this by using an exit multiple that is about 2/3 of the company's current valuation.

Evidently, these are all very subjective inputs where you're trying to be directly correct, even if you're inevitably going to be precisely wrong. (1:09:53) So, you know, with all these assumptions, I get a fair value of about $1,600 for Comfort Systems, which is a small discount to the current stock price.

And this is before applying any margin of safety to determine what might actually be an attractive entry price.

The takeaway to me is that Comfort Systems has legitimately seen its intrinsic value increase dramatically thanks to some external factors and their savviness in capitalizing on those opportunities.

And so it really is a great business set after this episode. I (1:10:22) I would be even more excited to own than when we first looked at it, but the range of outcomes here are really really wide.

And so nobody knows where this AI super cycle will lead us. And yet it's become so critical to the intrinsic value of this business, especially at this valuation.

(1:10:45) And I just couldn't get comfortable owning it at a price that didn't more significantly hedge some of the downside risks, which I don't think are being priced in fully today.

I don't think hundreds of billions of dollars are going to be spent indefinitely on AI capex and infrastructure. And so I'm sort of in the same place as last time where I'm thinking the business is really good and has continued to exceed my expectations, but buying it at these prices feels like really a momentum bet on AI spending being able to continue as it has.

And fortunately for Comfort Systems in the last year, it has very (1:11:16) much accelerated, but now we're another year into the cycle and we're at a five times higher stock price.

And it would just feel really speculative to buy in at these prices. it would feel like chasing after an opportunity that has already gone by.

And so, honestly, just recognizing my own biases, too. I have a lot of FOMO from this one. And I think that clouds my thinking. (1:11:39) And so rather than potentially doubling down and making a catastrophic mistake by chasing after comfort systems at five times the price from when I first looked at it, I feel like it is probably better to just remain on the sidelines and take a more cautious approach and see where things balance out in the coming years and and maybe look for a moment where the sentiment sharply turns against the business and the market is discounting perhaps the quality of the new data center revenue streams that it will continue to get as those facilities need to be maintained and that might create a (1:12:11) more attractive buying opportunity.

Then on the flip side where sentiment is sort of at an extreme in the other direction but who knows if or when that will happen. (1:12:28) So as painful as it is, I do think the right decision for us is to continue to stay on the sidelines here and recognize that it's okay that this one got away from us.

We don't have to chase it and there are always going to be more opportunities.

But to be honest, when I looked at it, the more and more I understood the business, the more I actually think he made completely the right decision and not buying it because simply put, there's just so many (1:12:50) unknowns in this business.

In which case, to be honest, comfort systems probably would be really, really cheap right now. But again, I have zero conviction and I think Sean also has zero conviction in making that call.

And therefore, I think the right call also is to just skip the business.

And I think when it comes to Comfort Systems, it's very obvious that Brian Lane has excelled at both, which is why it's been such a big winner.

Company CEO Brian Lane has said there's a long runway to continue making attractive acquisitions, one that extends well beyond his lifetime. In his word, many of the businesses they acquire are run by founders who have maybe recently turned 60 and are thinking about retirement.

In the old days, their kids might have taken things over, but that's increasingly less common and has created an opportunity for companies like Comfort Systems to come in and enable people to sell the business they've been building over a lifetime to a trusted partner with a good reputation who they know won't run in their life's work into the ground.