EMCOR Group, Inc. ($EME)

Good afternoon, everyone. We're just going to get started. So a quick introduction on my part. I'm Manish Somaiya. I cover Power, Tech, and Infrastructure for Cantor. Super excited to have Tony Guzzi and Jason Nalbandian, CFO. Tony, of course, as you probably know, is the CEO of EMCOR. We initiated on EMCOR yesterday, a price target of $848. So if you haven't had a chance, check out the piece. But of course, super excited to have the conversation with the team this afternoon. So maybe I'll kick it off with you, Tony. Obviously, you had a great 2025.

What is one aspect of your business that investors still underestimate in terms of the potential and in terms of market standing?

Yes. I think we've been at this a long time with a long record of success. And I think -- and look, we don't take that for granted either, right? In the kind of business we're in, you keep your humility over that period of time. But we've been, I think, a very good compounder, right? We -- I guess if you were comparing us, we don't chase fads. We execute well. We -- you can argue we're a very broadly diversified portfolio of projects and service base. With the data -- like everybody wants to talk about with the data center option today. We -- it's not like we turn away data center work to do other work. I think part of it is if they actually looked at the numbers and realize our penetration in the data center space, they realize they're getting great exposure to data centers. But because of our size and breadth, you don't necessarily just see data centers, even though the business is up 100% in the mechanical and 70%, 80% electrical year-over-year and the RPOs have climbed. So I think that's one aspect of it. I think the other aspect of it is we are very disciplined capital allocators. I've never been the kind of person that says, well, my multiple is here. And so anything that comes up to that multiple, I can just go buy. That's never been our schtick. We've been very disciplined about where we buy, which is mechanical and electrical construction for the most part, mechanical services, which is really HVAC services, building controls, and fire life safety within mechanical. And you have to be patient. And the other thing -- just because a market is hot or you see concentration -- the old rules of investing still stand, right? Watch for concentration, watch for customer concentration on an acquisition target, watch for sector concentration, watch for geographic concentration, a little less relevant in our business because almost everything we buy is geographically concentrated because that's one of the reasons we're buying is to get into a market like we did Batchelor & Kimball years ago in 2019 and like we did with Miller this past year, a great franchise across the Southeast and that helps us fill out our portfolio. And then we can learn from each other and build a business broader than that. And I think if you look at us over a long time -- and you looked at the last investor deck, I would say, go to Page 7. Our RPOs are about $13.2 billion at the end of the year. If you look at the end of '19, they were about $4.1 billion. Today, our RPOs are $4.4 billion in data centers today. So -- and underlying that's been a lot of growth in institutional, industrial, hard industrial, manufacturing, institutional. And these are all compound annual growth, water and wastewater, north of 10% or 15%, some as high as 25% or 30% markets. In any other world, you'd say, you've done a heck of a job growing out those markets and we're going to continue doing all that. So the diversification, data center option, disciplined operators, pretty darn good capital allocators.

Maybe I'll go to Jason. You gave your margin guidance for '26. I'm sure everybody wants to know what kind of gets you to the upper end versus the lower end. Maybe if you can just talk about that.

Yes. So to me, a lot of it comes down to mix and execution. And so if you look at where we are or where we were rather for 2025, we finished the year with a record margin for EMCOR at 9.35%. Our guidance for 2026 is 9% to 9.4% on the operating margin side. And essentially, what we're saying there is if you look at that high end of that guidance, that 9.4%, we believe that we can continue to execute at the levels we're executing at today. Our mix should remain strong, and our market position and the mix we have in our RPOs should also remain strong. And so when you consider all those factors, what we believe is we have the potential to execute just as well as we did in 2025 when we move into 2026. The lower end of that range, the 9% to 9.3%, let's say, really just reflects what we think could happen if there's a shift in mix. Not to say that any of the underlying factors or fundamentals of our business are changing. But if we have more GMP work or more T&M work or more work in some of the other sectors where the margin profiles are just inherently different because some of the jobs are -- carry a different risk profile. That just shows what margins could do if mix shifts a little bit. So for us, it really comes down to mix, timing, and execution.

Yes. And the way we think about guidance, I think if you listen to our call, we think of the low end of our range, which is where the 9% sort of encapsulates. We're pretty confident about that. And I think that's how you should give guidance. And the midpoint, fairly confident. And as you get to the higher end of the range, some mix needs to work our way. We got to execute well. You're always going to have some jobs that turn out less than you expected in contracting. So keep those to a minimum and maybe we'll get there to the top end.

So maybe just related to that, what do you think the sell-side investors get wrong about how your business works and how margins work? Maybe if you can just talk about it.

I'll do it sort of like at an operating level. I think just go macro level. This is not a manufacturing company. So I've been a manufacturer, right? You lock in standards at the beginning of the year, you know your market position. You have a pretty good idea of project pricing unless there's a dislocation, and you execute in the market you're in for that year. That's -- and if you have a new product, maybe get a little bit of organic from there, maybe get a little more price in markets because your commodity costs are up. In our business, we very rarely get to repeat something exactly. And even if we're doing the same kind of facility or building like we're building the next building on a data center campus, well, what might have changed? The project team might have changed on their side, probably not on our side, but on their side. The design might have been tweaked just a little bit. The mechanical contractor on the job, if it's not us, maybe not as good, and therefore, we get held up electrically. There might be a stacking of trades. These are lots of people with lots of variables, right? And so you're running a portfolio approach to that. So we've always thought of our margins in bands quarter-to-quarter. And Jason will go through that in a minute. But the point is this is not a quarter-to-quarter business. And if you look backwards over 12 to 18 months, it gives you a pretty good idea on average how the business is operating. And unless there's a big mix shift or you get some big volume upside that you didn't expect, you sort of know within relative -- and anybody thinks we're that accurate, like one investor asked me, we thought it would be 10 or 20 bps higher on the top end. I'm looking at him like, wow, if you've got that level of precision, then you should come and join us, and I'll go do what you're doing, because we absolutely don't have that level of precision.

Yes. I think Tony's point -- that we're not a quarter-to-quarter business -- is key. We always say take a look at a rolling 12- to 24-month average. And I'll just give an example. Our Electrical segment -- for them, that average would be 12% to 12.6%. But in those 8 quarters, margins have been as low as 11.1%, 11.3% and as high as 15.8%. So they'll vary quarter-to-quarter just based on mix, timing and execution, but you really have to look over a period of time. And I think that's the thing that maybe gets lost from time to time.

Or the same thing can happen with bookings, RPOs, book-to-bill ratio. Literally, a week could change a lot, book-to-bill. What I do know, if you think about today, whether it be margins, bookings or anything like that, that's quarter-to-quarter, the fundamentals of our business haven't changed over the last couple of years. In fact, they've gotten more positive. And as we gave our guidance for 2026, we didn't anticipate a big change in fundamentals. But with our revenue guidance that we have out there, we obviously believe we're still in a growth market in the most important sectors that we play in.

Interesting. And I just want to kind of stick to the topic of margins because obviously, it is prominent, and I want you to have the opportunity to explain yourselves -- the other question I have gotten is -- which is I think a fair one. Tony, as you are looking at all the different projects that are going on, what are some of the things that you look at to make sure that a program is on track versus, a, something is flaring up and it needs to be addressed.

Yes. So it starts back here when we actually decide that's something we really want to do. And these are things like our guys are running the sort of less than $5 million business every day. And unless there's a term or condition that's crazy in there. And even they will raise their hand and say, look, we don't like the terms here, and our lawyers will come in and try to negotiate that. But if you think macro level, right, how do you manage risk in a business like this? And then how does that manifest itself in the things we watch both numerically and on the soft side. There's -- whether it's a data center project, a semiconductor project, a health care project or we're building a school, right? And on the aftermarket side, it's all about access and customer relationship and everything else because there we're just trying to stay out of the way and do a great job and get them the savings or get the facility back up and running. Let's focus on the new construction side for a minute. If you're thinking about the business, and this is sort of EMCOR Operating Discipline 101. The first part of it is -- what do we know about the owner? What do we know about the general contractor, CM we're going to be dealing with. Ultimately, in most cases on the construction side, even if the owner is directing it, we're working for a general contractor or construction manager and in our case, to a lesser extent, an EPC. Do we know who's going to be on their team? Do we know the owner? Have we had good success with that combination? And if it's new, has someone else that does benefit EMCOR has, has someone else at EMCOR had success. A great example of that is it's happening every day in the data center world. But when Wynn was bringing their casino to Boston -- we had built quite a bit for Wynn in Las Vegas. So actually, our Las Vegas team, we hadn't built a casino in a while, but we still had the people. We actually flew to Boston and sat down, went through the numbers, brought the Boston team out to Las Vegas. So let's think about how we did some of these things that are peculiar to Wynn, how they build, and how the owner behaves. And then we actually coached up the general contractor because, quite frankly, the Boston-based guys -- those East Coast guys weren't Vegas contractors. So okay, that's one thing. Next thing we look at the engineer and the architect. Have we worked with them before? In most cases, we've worked with everybody in both those cases -- GC/CMs, engineer. What are their designs like? Are they good designs? Are they designs where we're going to have to do a lot of design-assist to get it the constructability. Are they responsive to request for information as we validate the design? Are they responsive to value engineering that we may recommend as time goes on? If there's a change in the job or they click on the turn with the drawings back to us, so we can get back to work and not have a lot of dead time. Then we get to the simple question: Have we built one before? Take those first 2, try to get as much intelligence as we can. These are bigger jobs like $10 million plus. Have we built one before? And if we haven't built one in that geography or with that subsidiary, have we built one somewhere in EMCOR? And can we take that learning and bring it to bear? Or have we built something similar? So, is there similar lessons we can apply. A lot of the learning that comes today in the data center world came from our manufacturing jobs. And in -- from our data center, we did, but that was only 2 or 3 subsidiaries versus 17, with manufacturing jobs or health care, hospitals. What are the economics of this project? If it's a big enough project, it almost looks like an acquisition to us. What's the cash flow going to look like on it? What are the terms and conditions? What are we actually signing up for on liquidated damages or consequential damages? Are we going to be able to cap them? So how do we frame that? Then it gets to the final stage of, okay, once we build this, if you're a construction guy, the first thing you want to do is get off the job, right? I mean you want to be gone. And if we have an aftermarket team, they'll come in -- and they're very different people. How do you commission this building? How hard is it going to be commissioning? Did we -- if we're on the mechanical side, did we do the building controls as a part of our contract? Or did we hire the sub there? And then how we -- because these are the kinds of things. If you're doing a data center, what's the flushing look like? Are we going to be able to get the system flushed, especially on the AI data centers and get it up and running. So we take a full-scale view of it, make sure we protect ourselves as best we can in the contract. And then we go into planning if we get the job. And if we're going to do all that work on a large job, we have an idea that we have a pretty good chance of winning the job. Nothing is 100% but we're not wasting our time on something significant to ask all those questions. And then Jason has some very hard metrics that tells them once this gets going, whether the job is going well versus the WIP and whether it's not.

Yes. And there's a number of ways you can do that. But I think the most simplistic is looking at the cash position of the projects, right? Jobs that are overbilled tend to be performing well. The jobs that are underbilled, we tend to say have risk. And sometimes the risk could just be on timing. Other times, it ends up being potential overruns in the estimates. But if you look at EMCOR overall, we are in a significant net overbilled position, and I think that is a testament to our execution.

So over the next year then, where do you see the greatest margin benefit potential? And where do you see margin dilution...

Yes. I think that's a hard question to answer -- everybody gets excited and says, data center work is driving the margin. I'd say that's fair because it's a big part of what people are doing. But it's only driving the margins in some ways because you're executing well and they're really demanding jobs. And you really have to know what you're doing to sign up and do it, and you really have to be able to scale up and get there. And you better have a great prefab. So, maybe there's margin opportunity there. But we do well in other sectors, right? We do well in basically replacing the chiller room, design-build. We do very well there. So it really does go back to answering those first 6 questions well, 7 questions well. And then once you answer those questions -- well, the one I forgot that is: Who are we going to put on the job? And do we have the resources to be successful from a leadership and supervision standpoint. We start there. That's why I probably haven't mentioned it. That's a given. But if we get those right, it almost doesn't matter where we're working. We have a good chance. I always tell people, one of the ways you do well in our kind of business is the absence of badness. So it means we pegged the margin right with the right contingency, didn't have a lot of write-downs and didn't have a lost job that was significant. And Jason's point about the cash is important, right? Sometimes there's an image in people when we are net overbilled like that, a, we're collecting all the cash that could, a portion of it, we are. But also, you just don't send someone a bill in our business. You've negotiated that you're going to send them the bill, and here's what the bill is going to say. And you've hit pre-arranged milestones. So, you don't get net overbilled unless you're performing well on the job in most cases. And that's why it's such an important one. I think the other thing they watch very carefully, like these guys have great analytics that they've developed and this is probably one of the areas that AI will be. I think we're doing a fair amount of quasi-AI now on our WIP is the one thing we manage and control is labor. So, we spend a lot of time focusing on what the manpower loading looked like on a job versus what we expected, how do the hours track versus where we thought it would be, how we're doing against the labor contingency we built into the job. We don't miss materials very often, right, unless there's a big change in commodity prices. Even then we don't miss it very often.

So you talked about -- maybe just shifting to backlog. You talked about record backlog, RPOs which, of course, is very supportive for visibility. But then how do you assess quality of the backlog?

Yes. I think some of it goes to what Tony was talking about when we're assessing projects, right? It's asking ourselves those same questions about the nature of the job, our ability to execute on that job, and that will give us a sense of where we think those projects can perform. Then I think some of it is just looking at the margin and the backlog at bid, the amount of contingency in projects when we're beginning projects. And all of those things paint a picture for us about the potential of those jobs and how they compare to similar jobs that we started a year ago or 2 years ago. So it's very much a benchmarking exercise when we're looking at our backlog to see how does this stack up against historical periods. And I think what we believe right now is our backlog at the end of 2025 going into 2026, is just as good as our backlog was a year ago. So we still think we have some really good quality projects ahead of us, and I think it gives us an opportunity.

And it supports our guidance range with as much visibility as we've had in a long time. And we had pretty good visibility over the last 2 or 3 years. What I think people don't appreciate what's happening is how actively we're managing the mix, maybe less so on the upside, because go back to the way we think about qualifying things. That's agnostic, no matter what sector it is in. But then we, as a management team down through the segment level, we make sector calls sometimes and geography calls. Now, you don't see it, but there's -- we've done a fair amount of restructuring in our portfolio. Where you do see it -- if you look at Page 7 is we've taken transportation work down quite significantly. Now some of that is because the work can be lumpy, but most of it is because we basically have got out of -- we have one more year to do in one place. We basically have exited the road transportation lighting market. And we were pretty good at it at one time. Dynamics change in local markets, contracts change, and we -- the risk change, it's not worth it anymore. And so we've moved away from that. And these are from the guys who did the Tappan Zee Bridge electrically and did quite well. The dynamics here, especially in New York and some other big cities, just not worth it anymore on some of that work. And so we've decided to de-emphasize that part of our business. It's not that we weren't executing well is the -- it's that when you do an analysis of the mission and what you're going to do in the terrain and everything you're dealing with, you sit there and say, not where we should be putting our capital right now. So if you think about that, we probably over a couple of years, growing like we are organically, if you were to take snapshot, think of '21, and we've got out of some commercial markets in some of a place like San Francisco, we're no longer there in a significant way mechanically. And if you snap that line and look at it today, we probably take $300 million, $400 million out of this business on a run rate that wasn't very profitable. And that's the other way you help build up margins. But also, it's not like we made a short-term decision when we did that. We took a 5-, 7-year look at the market and said, fundamentals aren't going to change here. The risk is -- the risk reward ratio for that line of work puts us in the wrong box, much like we look at acquisitions, right? We're very disciplined acquirers, and that hasn't changed because there's a couple of hot markets right now.

And I guess maybe just a last one on backlog. How should we think about shadow backlog that you might have? And what is one sector that you're most bullish about over the next 12, 18 months?

That's pretty straightforward. That's like making a 2-inch putt, right? Obviously, we're all excited about what's happening in the data center market and some of the re-shoring opportunities we see. And really, semiconductors is both a growth story and a re-shoring story. So -- and it's beyond semiconductor, anything high-tech manufacturing and data centers, I think you'd have to be excited about if you're a highly skilled specialty contractor right now.

Right. So we keep hearing about the power availability, which is a big gating factor for everybody. How do you sort of adjust the sequencing and timing in terms of the workflow because of that challenge? And geographically, how do you position yourselves?

That's very rarely a challenge for us because of where we are in the food chain. Most times, before we come on a job or they let us loose with a contract, we've -- they've already done all that. And it might be a little slower getting the substation hooked up, but they know they're going to happen within a very short period of time. We're not -- now, we may know they want to build in a year or 2 years, and they haven't figured all that out yet. But to be honest, we haven't committed any resources other than some pricing upfront to just get them in the ballpark. Fair statement...

I think that's fair. And the other thing for us, too, is by the time we're -- as Tony said, by the time we have a contract in hand, we know that contract is going to go. And for us, we're usually on-site within 1 to 3 months, and that work begins and the majority of our work is done in 12 months.

So go back to one of the things you asked earlier, I think what you just said that's important for people to understand. Our RPOs are -- the accounting definition of RPOs, Remaining Performance Obligations. Other people report backlog, which then piles a bunch of other stuff on top of that. So we may know that we're going to do other work on a site or we may know that it's going to happen. They haven't committed to us yet formally. You could look at our RPOs as there's a formal commitment, a contract that we're going to do the work. And the only portion that's in that RPO is, if it's a service agreement, even if it's a 5-year service agreement is the non-cancelable portion, which could be anywhere from 90 to 120 days. And usually, on the service agreement, those go both ways. We can cancel them too. So I think that's a really distinct difference with us. And I think the reason I'm building off of that is that's the reason our work usually doesn't get canceled.

Yes. Our RPOs are firm and cancellations have historically not had an impact on us. The exact reason.

I can count on less than 2 hands over 15 years any project of size that's got canceled. And you could guess when they got canceled, 2008, '09, and 2020.

So just going back, maybe turning to capital allocation because we have about 5 minutes. You've said that you've been fairly balanced. What is your framework for allocating the next dollar between organic, inorganic or returning it to shareholders? How do you guys think about that?

So we have a dividend, right? $0.40 a quarter.

$0.60.

$0.60 a quarter, now screw that up $0.40.

It's $0.40, Tony...

So $0.40 a quarter, that's a given, right, put that aside. And then clearly, our priority is organic growth. And with the amount of cash we generate, we can't invest enough in organic growth because we're a capital-light business. Even though we doubled and Jason will go through these in a little bit, the amount of capital expenditure we have, mainly because we've increased our fabrication space and we've increased our software spend. We still can't spend enough, right? And we try to keep our cost structure as variable as we can. So we tend to lease buildings even when we build a fab shop. And outfitting them and we're pretty good buyers of equipment and products. So then you get to the next one, which given our druthers, right now, we would love to be able to repeat 2025 and 2026. Are we going to be able to do that? Deals happen when they happen. We talk to people over long periods of time because the people that are selling us their business typically, even like last year, we did a bunch of small ones, Jason will go through that. The bigger ones, they're selling us their family's life work and their reputation. So these aren't simple decisions. These aren't people that are looking to multiple up with the private equity firm. Typically, if PEs involved. And then we go to share buyback. I don't think we're trying to be hedge fund managers, typically, although we've been pretty good buyers of our stock. But that is there where we end up with excess cash. And we don't try to buy when the thing is rapidly accelerating. The flip side is after that, it pretty much looks like dollar cost averaging, I think, unless there's a big dislocation like there was last year after DeepSeek panic.

We tend to do share repurchases, both programmatically and then opportunistically, depending on the market. But yes, when you look at -- to Tony's point, when you look at the organic growth, it's how do we make our existing operating companies as productive and efficient as possible. So for us in recent years, it's been looking at ways to expand our prefab capabilities or use of VDC and BIM. And I think Tony referenced some numbers before. And really, what we've seen is still a capital-light business for us. Our CapEx is like 0.6%, 0.7% of revenues. But we've expanded our CapEx in recent years to the point that if revenue is growing, let's say, over a 3-year period at a 15% CAGR, CapEx is growing at a 30% CAGR. And that's those investments. And then you look to the acquisitions, and there's a number of different ways we look to acquire. Some of it is looking to buy a platform like we did with Miller in 2025. Some of it is looking to fill in a geography like we did with Danforth in 2025. And then we did 8 other acquisitions, which were all really bolt-ons or tuck-ins to existing operating companies to enhance their capabilities and further their position.

Maybe over to you, Tony, as to maybe recap everything that we've talked about. If you have to sum up EMCOR's strategy for the next 2, 3 years, how would you do that?

We're going to continue to grind it out, take it up, take advantage of great growth opportunities. Play the best team that I think is in the industry in the field because that is our core product, right? Our core product is the best field leadership in the industry. We just happen to be in this industry. We have the best field leaders. We spent a lot of time training. We spend a lot of time learning from each other, and we're committed to the things we talked about, disciplined capital allocation, Mission First, People Always. With our people, transparency, a lot of respect for each other. And that's been a long-term recipe for success. And we're not going to alter out of that one bit. We're not going to be someone we're not. We're not going to chase fads. Data centers are not a fad to us. It's a great sector to operate and build projects and do great things for our customers. So that's what it is for us.

Well, thank you so much. Just in time. Thank you, Tony. Thank you, Jason. We appreciate your participation and enjoy the conference. Thank you.