AAON, Inc. (AAON) Earnings Call Transcript & Summary

Okay. Good morning, everybody, and thank you for joining the Sidoti & Company March 2025 Small Cap Conference. My name is Julio Romero, and I cover the building products, industrials, engineering and construction sectors at Sidoti & Company. We're really pleased to be able to host AAON Inc. The ticker is AAON. With us today is Joe Mondillo, Director of Investor Relations; and Matt Tobolski, COO and President and incoming CEO effective May of this year. If you have any questions for the company -- this will be a fireside, but if you have questions, feel free to type them into the Q&A section at the bottom of your screen, and I'll ask on your behalf if time permits.

Matt, Joe, really appreciate you guys being here. Maybe to start, if you could give a brief overview of AAON and what the company does for folks who are a little less familiar with the story.

Fantastic. And good morning, Julio. Thank you very much for hosting us. So when we look at AAON as an organization, looking back to the original formation of the company back in '88, the whole premise of why AAON was launched in the first place was really to bridge the gap at the time of the industry where there were really only 2 ways to buy kind of a commercial rooftop product, and that was a catalog solution with little configurability and little option set to it or the other extreme, a full custom product. And the goal of AAON was to really bridge that gap with a software-driven, automation-driven semi-custom configurable product where you get a lot of the value that you would historically get out of a custom solution. But due to the automation and due to the software enablement tools that are in place, do it at a much more cost-effective price point. And that really was the premise of AAON at its initial launch. That's really been one of the key differentiators for AAON in the marketplace is that semi-custom configurable kind of product portfolio. The company has always been driven by innovation. And so a lot of innovation is really around providing best-in-class product performance, providing unique operating strategies that can provide tremendous value to the marketplace in application. And all of that is kind of what's led AAON to be a really premium brand in the marketplace within the commercial nonres market and the rooftop market. With that, obviously, we always came at a price premium. And so historically, the price was in that mid-teens kind of perspective. But as time has evolved, regulatory requirements have really stepped up industry expectations around energy efficiency, which has required a lot of the industry to really kind of make their way towards AAON from an overall efficiency standard perspective, and that's helped kind of close some of that gap. So where AAON used to be in the mid-teens price premium, we're now in a high single-digit price premium that's allowed us to continue becoming more exposed to more of the marketplace and really acquire more market share in the space, growing from 5% market share to 11% market share where we stand today in the commercial rooftop segment. On top of the AAON brand, through an acquisition a little over 3 years ago, the BasX branded product kind of came into the AAON portfolio. And the BasX branded product is really a data center-centric as well as semiconductor cleanroom-centric product portfolio. And unlike AAON, which operated in that sort of semi-custom environment, the basic branded product really is focused and has focused on more of a true custom engineered product, but manufactured at scale. And so when you think about data center opportunities, the quantity of units that we deploy in a given data center are substantially larger than what you would see in a kind of commercial nonres type market. And so because of that, we're able to create custom solutions that really provide a tremendous amount of value for the application in terms of fine-tuning performance, the durability of the equipment, the maintainability of the equipment. So we're able to really provide a tremendous value to customers without having the same historic price premium you would see on a custom piece of equipment because we're manufacturing at such large scale for these applications. And so you kind of get that benefit of customization with the cost effectiveness and manufacturing scale that's really allowed our value proposition of the data center market to really expand both in the cloud compute data center side with airside products as well as in the liquid cool AI space. But going forward, it's kind of the way we're operating this organization is really around those 2 primary brands with the AAON brand being that kind of commercial nonres, rooftop and indoor air handling systems and the BasX branded product segment being more on that data center and cleanroom applications.

Excellent. Great rundown, Matt. Much appreciated. So you spoke to the differences between yourself and some of your peers, semi-custom solution being one major difference. The go-to-market is another kind of key difference. Can you maybe go into that a little bit? Any other kind of major differences to yourselves versus the peers?

Yes. Definitely, the big differentiation with AAON compared to a lot of our peers is our go-to-market strategy. And especially in the rooftop segment, we sell exclusively through an independent sales channel. And if you look at a lot of our competition and peers, companies such as Trane historically has basically converted from an independent sales channel back in the '70s and has really been converting that to a factory office kind of model. You've seen that also accompanies like Daikin and with JCI, who JCI has kind of bounced back and forth a little bit between independent and factory stores. But the sort of differentiation there is, number one, independent sales channels -- independent sales reps maintain a portfolio of products. And so why that's critically important is when a sales individual goes into a customer conversation, there's not just rooftop units sitting outside of a project. There's exhaust fans, there's louvers, there's dampers, there's chillers. There's a variety of products that are going to go into an overall building application and the AAON rooftop product is just one piece of that puzzle. And so having independent sales channels that have a comprehensive line card provides opportunities to engage with customers from a variety of angles and get exposure to a lot more projects and be able to provide a higher level of value in the sell-through of a comprehensive solution versus just a bunch of one-off products kind of inside there. And so the independent sales channel has a huge amount of horsepower and being able to engage with customers and opportunities to sell solutions, comprehensive solutions that really helps AAON be a part of that large scope and really kind of specialize where our product best fits. When you think about why some of our competition has gone down the route of factory offices, a lot of it comes down to the product portfolio that they offer. So using a company that makes a centrifugal chiller as an example, AAON doesn't make centrifugal chillers. But if you're a company that does, a huge portion of the overall kind of revenue model in a centrifugal chiller is the aftermarket service aspect of that chiller. Most manufacturers of centrifugals, there is a requirement for that maintenance and service of that to be done through a factory certified offering. And typically, that is done through the factory itself. And so in a lot of opportunities, you'll see companies that manufacture chillers wanting to minimize the price point of the initial sale of that chiller to be able to guarantee that long-term annuity in the back end on the service contract. The same doesn't hold true for rooftop units. There's no long-term service contract required. There's service and maintenance required, but not to the same level as a chiller. And so that model that may work for some companies like that doesn't necessarily work for the overall kind of rooftop market that we do. So for our goal, we don't -- we're able to allow our sales channel to effectively price and position our products to motivate those sales channel members to sell the product for the initial value of that product versus companies that want to reduce the initial sales price, therefore, reduce commissions for the sake of getting a long-term annuity off that contract. So we're a much more attractive product to be able to sell for an independent salesperson given the opportunity from a profitability standpoint.

Makes sense, and I appreciate you kind of touching on all aspects of that, and the holistic answer definitely helps and ties into the refrigeration stuff, which we'll talk in a second. But maybe just starting with data centers, just wanted to see if you could speak to what you're seeing on that front. On the last call, you were pretty adamant about not really seeing a slowdown in data center related to construction demand. But just help us just understand kind of the disconnect in terms of the broader DeepSeek related headlines that we all see and concerns related to CapEx slowing for data centers versus what AAON is seeing on the ground?

I mean certainly, DeepSeek has created a tremendous amount of noise in the industry and really around just trying to understand the CapEx cycle. And really the DeepSeek case, I guess, to start with, really, a lot of the noise it would create would be around how much it costs to truly make an AI large language model. So what is the development cost? What is the machine learning investment required to kind of help generate these comprehensive AI models. And in DeepSeek, certainly, the initial news on DeepSeek showcased what -- the initial news was an extremely inexpensive kind of methodology, which, again, it's since been sort of not debunked, I'd say, but it's certainly -- we say an OpenAI model costs this much. The initial news on DeepSeek was way over here. The reality is that somewhere in between. It is more cost effective to develop than some of the other AI models. And it's created a lot of question marks around, oh, is the CapEx plans completely overshot and then we're going to have this massive clip of an investment. And we certainly don't see that. I mean if we look at the engagement we're having with our customers in terms of pipeline visibility around the machine learning CapEx investment, we continue to see strength. We continue to see that strengthening, not weakening on the heels of that DeepSeek conversation. So we haven't seen a shift in conversation and pipeline visibility and order cadence around DeepSeek. But what I'd also kind of take is while it creates noise on what it costs to develop, let's just take a hypothetical and let's say it is extremely cost-effective model, and we can reduce the AI/ML aspect on the CapEx cycle, that would create a lower price point of entry to create AI models. And it would then create more cost-effective AI models for someone to then leverage for their own business. And so a benefit of more cost-effective models would be easier price point, easier justification to begin capitalizing or utilizing AI models. And what that would do then is drive CapEx. It would drive CapEx because utilization of AI models needs inference compute data centers to be able to actually use the models. You've got the machine learning side, which is developing the models, the inference is really the application utilization. And so if you did all of a sudden have this substantially cheaper model, and we then have more demand for AI, there's been more demand to build inference compute data centers therefore, strengthening the overall AI investment cycle. So there's -- kind of the way we look at it is a lot of this DeepSeek conversation is really near-term noise around how much money we're going to throw at developing new models. But the long-term kind of investment thesis on the AI world is around this tail that's going to be on the inference side, which is still a very good growth driver inside the AI CapEx cycle and it's also a longer tail because the more we utilize AI, the more we're going to recognize the value of AI and drive more utilization. So we see it really as near-term noise. But we look at the fundamentals, number one, of being very strong on the machine learning side, but also being even stronger on the overall inference cycle as we begin to really leverage the power of AI going forward.

Super helpful. And is that point you made about a more cost-effective model making the monetization conversation easier, driving greater inference usage. Is that playing out from what you're hearing from customers and just folks across the value chain?

And we're really early in the cycle. I mean the kind of conversation I've said is we talked about it -- someone asked me a couple of days ago, are we in the first inning of the AI cycle? And I said we're in the batting cages warming up. I mean it's -- we're really -- this is a very early kind of portion of the AI sort of growth driver that we're seeing. And so I mean, really like starting to see the inference compute investment, I mean, we're early in that as well, kind of on how it starts to really monetize and kind of come into fruition.

Super helpful and a good frame of reference there. Turning to like the fourth quarter results, which were mixed. You had the, I guess, what we call the Oklahoma segment kind of dragging down the overall results, and there was a little bit of noise with regard to segmentation. Maybe first, touching on the refrigerant transition, the Oklahoma piece and kind of that part of it, if you could?

Yes. I mean we messaged 18 months ago that '24 was going to be messy, with impacts from refrigerant transition, but also interest rate environments, a soft macro with ABI, leading indicators showing softness plus an election season. So we had a lot of things we said going into '24 that was going to cause disruption. And as we entered '24, we were very vocal that we were the leader in the marketplace with having commercially available 454B equipment. We led the way in the product development cycle. We were sitting here very proud to have available products they want. The challenge is the 454B transition is the first in the history of the HVAC industry that had an EPA-driven environmental driver for why we're changing refrigerants, but also an implication to the building code. And so the fact that this refrigerant is an A2L, a mildly favorable refrigerant, we had to modify the building code to allow the use of this refrigerant. And so while we entered 2024, very proud to have commercially available product, we have sort of mapped out this mindset that you're going to see this gradual adoption of 454B equipment throughout the year. And the premise of that was AAON's core customer is very much a total cost of ownership conscious customer. And so if you're a customer that looks at what's the total investment cost of my equipment, and you say, I can buy this new equipment, which is the new refrigerant versus the old refrigerant equipment that's being phased out, when I think about the maintainability of the equipment and the overall cost to deliver that equipment, we're sitting here saying our core AAON customers are going to look at that and say, of course, I'm going to buy 454B equipment because for 410 equipment over time is going to get more expensive to maintain with component availability, refrigerant cost and a variety of drivers as time goes on. The challenge to that was the building codes were much slower to be adopted than we anticipated. And so while we anticipated this nice gradual transition of refrigerant adoption, we sat there in July assuming we're going to have 25% of our orders being 454B and we were low single digits. And that was just because at the time, only 18 building codes were actually -- only 18 jurisdictions had adopted the new building code. And so it really slowed the adoption of that new refrigerant throughout the calendar year. And as we got to our September date where we had to basically stop accepting orders for 410 equipment to allow us to manufacture that through the end of the year, we knew there'd be some noise. And so we closed off the 410 products, we go into October, and we saw a soft book -- order book down about 10% year-over-year in October. And we said, okay, that makes sense, like it's just a little bit of a lull around this transition. Well, what caught us off guard was November book gains were down 30% year-over-year. And as that materialized, we come to find out a lot of that because some disagreements between the building code and the ASHRAE design standards, some general adoption uncertainty, plus that was around when the election happened. So there was a lot of just general pause that kind of happened in that November time frame. But what that cost for us is we had to slow down in production because as a manufacturer to order product, there's a certain cycle in our kind of order to production that requires a bill of materials kind of clarification for that exact configuration that has a set lead time to it. And so we can't just eat down backlog and have our lead time equivalent to basically get really small. And as we saw that order book slow, we had to slow production to balance that lead time backlog dynamic. And that net result was volumes were down in Q4. And with that volume being down, the big driver on margin was just absorption of fixed costs. That margin is just the math equation at that point on a lot of that drop that we saw in Q4 off of that volume. We kind of guided that drove -- the high level, that's what drove the Q4 slowdown, but it's also what drove our Q1 guidance, which is a soft Q1 because that October, November, early December order cadence, that's really what we're building in Q1. And so as we saw that softness in kind of the order book, that's also what guides that softness in Q1. But as we look past that, and we look at the order cadence we saw at the end of December, plus what we talked about in our call, which has been day-by-day strengthening the order book in Q1, that's what kind of sets up that confidence that Q2 was getting back to that normalized behavior once we get through this kind of short-term disruption around that refrigerant transition.

Got it. Very helpful there. And then maybe if you could talk about the outlook going forward on the kind of rooftop portion of the business, touch on price and the price premium and also national accounts and heat pumps.

Yes. When we look at the kind of the driver and we look at the outlook, certainly, if you step back and just go to the 30,000-foot view and read the headlines and look at the leading indicators, it would signal a soft non-res market. I mean we've got ABI continuously down. We've got a lot of disruption in the marketplace just with some sort of macroeconomic turmoil, a lot of which is politically or policy driven right now. And so you've got a lot of this noise. But what would indicate there's going to be a down market or at least a soft market, we've got a few things that really are strong tailwinds for us to kind of offset any of that disruption. And then the 2 are price and product differentiation. And on the price front, there's 2 unique things that are happening around price right now. Those 2 things are cost competitiveness with the new refrigerants and cost competitiveness in a tariff-driven environment. And so on the new refrigerant side of things, we voiced consistently that based on the input cost justification, AAON does not have a price difference between a 410 product and 454B product. I mean, minus a little bit of noise in there, I mean it's more or less the same input cost perspective. Whereas the market has signaled anywhere from an 8% to a 15% cost increase with this new refrigerant. Now we haven't really gotten to a point that we have true empirical data to say what the market is really putting in for price. And so what we're hearing right now is anecdotally, AAON is the most competitive it's ever been. But we don't have hard data to say what does that look like. And so we're just now entering the traditional K-12 booking season. And the benefit of K-12 bookings is it's primarily public funded, which means there's a lot more transparency in the bid process, which tends to be where we really get a lot more clarity around what price is doing. And so as we enter this, we're keeping a keen eye on what our price point looks like and what that price premium looks like relative to competition to determine what we might have to do on a price. If you factor on top of that tariffs, AAON is benefiting from the fact that we are a vertically integrated U.S. manufacturer. So what that means is our exposure to tariffs is less than a lot of our competition because a lot of this vertical integration exists in a lot of our supply chain is in the U.S. Now there's still cost drivers. I mean it's not like it's a neutral environment. But from our perspective, we've put into the marketplace a 6% surcharge on equipment. And that's to offset input cost dynamics around tariffs that are driving costs as well as some general macro things such as steel price increases around tariffs. And so that 6% increase compared to some of our competition, that's more in the 8% to 10% range. So when you factor in the refrigerant price conversation plus tariffs, all of that would indicate that AAON's price position would be in the best place it's ever been in the market. We don't want to get overly optimistic because we want to really see how this materializes. A lot of this is sort of, I'll say, early rhetoric, and it's not really materialized into true what's getting put to the street. But every bit that, that price premium that AAON offers shrinks, the ability to sell the value that AAON offers becomes that much easier. And so it puts us into an opportunity where we can balance taking some market share, which taking market share means increasing volumes, which increased absorption and naturally increases margin just on a volume play. So we have that potential to grow volumes, top line plus margin, but we also -- if we start seeing too much come in from an order cadence perspective, such that our backlog grows and our lead time grows, we would use price because our goal is to maintain our lead time and therefore, our backlog at a 12- to 16-week kind of time frame for a commercial rooftop product. And that's really around buying habits in the commercial space. We don't want to have this uptick in orders driven by a price dynamic that is quickly offset by the fact that our lead times go out, and we no longer are competitive on lead times. And so price would be a lever we're pulling to basically balance that kind of price premium, lead time and overall kind of dynamic. And so there's definitely upside that exists around price. We're just really waiting to see how that materializes. But that aside, price is certainly an exciting piece. But the bigger piece that's exciting is really more, I'll say, the fundamentals of why AAON is a great company, and that is the innovation side of the business. That's the product differentiation that we offer. And so price certainly is a unique thing driven by a lot of more macro factors and geopolitical factors. But the bigger opportunity for us is really around a continued focus on decarbonization, on heat pump innovation that AAON offers. If you look at calendar year '24, our heat pump sales were about $100 million of our overall revenue, and that grew 40% year-over-year. So there's definitely secular trends that are driving heat pump adoption and really a focus on decarbonization, and AAON has always led the way in innovation on rooftop products. And we continue to do that with heat pumps. And so innovation we have around low ambient omni-climate heat pumps is a huge differentiator. And as we think about strategically going forward, it's an opportunity with national accounts for AAON to really get exposed to a larger base of customers than we historically have because as these large national accounts look to decarbonize or at least reduce carbon emissions, our heat pumps that we offer are the best-in-class at being able to manage that desire to decarbonize and do it in a very cost-effective manner, in a very logical manner in terms of some of the operating strategies that we offer in that system.

Extremely helpful. And just a quick refresher on the time frame of that K-12 selling season and when you would expect to have enough of that kind of K-12 public data to be able to have more clarity on industry pricing?

I mean really by the next time we release earnings, we'll have a good amount of information kind of true empirical evidence of where we were at competitively in the marketplace.

Very helpful. You touched on tariffs. I want to ask about any other kind of nontariff impacts on the new administration as it relates to AAON?

Yes, I mean there's certainly plenty of noise on policies that kind of come and go. Realistically, there's some noise right now about some of that EPA deregulation last week that would actually undo the regulation that drove us into a refrigerant change. But the reality is, even if they try on doing that, the whole industry has changed the manufacturing process around new refrigerant. So even if that regulation changes, the concept that the industry is going to all of a sudden bear the cost to switch back in the whole supply chain that is building compressors for new refrigerants. I mean there's a lot of inertia of that while there's certainly some policies and things that are creating noise, some of these, I think, ultimately will settle out and not greatly impact kind of the overall buying habits and really the regulation habits around our equipment.

For sure. We have a couple of minutes left. I want to talk about the capacity additions you've been doing. Are you -- Longview was just completed. You have the facility in Memphis that's expanding. Just talk about the capacity additions you're bringing on and where overall data center equipment production capacity should be once everything is kind of fully operational?

Yes. I mean if you think about a year ago, 12 months ago, I mean, primarily, the Redmond, Oregon facility was data center manufacturing. That was the primary area. We had a little bit of manufacturing being done in Longview, under the Coil Products segment. But realistically, it was Redmond. Well, Redmond, obviously, if you look at the backlog growth, a lot of that has been driven by data centers. And so that facility was oversubscribed and as we looked at bringing on capacity, the first step in that equation was to leverage the land that we had and the building flexibility that we had in Longview, to complete that expansion on sort of Phase 2 of a building we had built a few years back. So that was kind of the first step in our plan that we put in motion a couple of years ago. We turned that building over in basically the end of December of last year. So we're just now starting to leverage that from a manufacturing standpoint, but we are doing it very aggressively. So of that, we've already got 3 production lines up and running, building liquid cool product for AI data centers, and we have an additional 2 production lines coming online in the next few months. So we'll have 5 production lines at the facility building AI-specific products, and we'll have an additional production line building traditional cloud compute data center products at that facility. When we look at that kind of from a revenue kind of perspective, we've kind of said that the Four Walls in Oregon, so the actual original BasX building, those Four Walls, you can really get yourself to about $250 million of capacity. That's kind of just the fundamental constraint of that space. Where Longview, you can probably get to more of a $500 million capacity range. So in those 2, the manufacturing capacity kind of around data centers is really in that $750 million range and then you bring on Memphis. And if you fully dedicated Memphis and fully built up Memphis to be a data center product facility, that would add roughly another $750 million. So all in, when those are up and running and fully built out, it's not a flip of a switch, but when they were fully built out, that would represent about $1.5 billion in manufacturing capacity.

Excellent. So we're right about at the end of our time. There's some topics. We didn't get to touch on, capital allocation and resegmentation and a couple of other items. But just before we wrap up, just what aspects of the AAON story are kind of overlooked and any other key messages you'd like folks to take away from today?

Yes. I think the biggest thing is when we look at getting some of the AI noise aside for a second, when we look at the fundamentals of both segments of this business, we've got AAON core rooftop legacy business, that is at the best positioning it's ever been in the history of the company in terms of the price dynamics, but also product innovation and market share potential with heat pump innovation. So we've got a company that in the last cycle, from 2012 to 2019, grew with a 7% CAGR ahead of the industry with little price. So really just volume gain, volume share acquisition. And there's nothing to say that can't continue to happen with all of these dynamics we have to make that very competitive. And you layer on top of that, the data center side, which even if we get rid of some of the short-term noise, that segment is going to be growing in the teens at least from a CAGR perspective going forward. And so when you look at this and get rid of some of that short-term noise, if you look at this business on a 5-year window, the fundamentals on both segments really support that good growth that we've talked about in that kind of mid-teens perspective on a consolidated basis going forward for that next 5-year cycle, and doing so with improving margins and really kind of on a continued execution improvement as we go forward. So the fundamentals have never been stronger and very exciting opportunity with just obviously some short-term noise that we had in Q4 and Q1 around some of this transition implications.

Excellent. Well, Joe and Matt, thanks so much for taking the time.

Thank you all.

Thanks, Julio. I just want to say one last thing. I noticed that there's a handful of questions that maybe were not answered. So please find my e-mail on the Investor Relations site of our website, and I'll be happy to answer those.

Great. Thanks so much, guys.