Good morning. My name is Will Shelmire , and I am an account manager here at Three Part Advisors. I want to thank you all for joining us for our 17th Annual Midwest IDEAS Conference. To start things off today, we have Powell Industries, which trades on Nasdaq under the ticker symbol POWL. Representing the company today is their Executive Vice President and CFO, Michael Metcalf.
Thank you, William, and it is good to be back in Chicago again. We have not been doing this for 17 years, but I think this is our third or fourth year consecutively at the conference, and we really like the conference, so it is good to be back. Again, my name is Mike Metcalf, CFO at Powell Industries, and I will take you through where we came, where we grew up, where we come from, what we sell, and a little bit of the financial perspective. We are 80 years old, nearly 80 years old, 79 years old to be exact.
We make low and medium voltage switchgear, predominantly based in the U.S. I will get into that in a few slides. Last year, we generated a little over $1 billion of revenue. As we sit here today, we just reported our third quarter results, and we are at about $860 million, so we are well on our way to eclipse what we did in fiscal 2025. We have a global footprint. Most of our manufacturing facilities are in the United States, five in the U.S. We have a large Canadian operation and an operation in the U.K. as well.
As I will go through in a little while, our balance sheet is pristine. No debt, great liquidity position. So what is it that we make and what is it that we do? Our intellectual property is that product you see on the left side. That is a circuit breaker. So in your house, your apartment, wherever your place of residence, you have a panel full of circuit breakers. That is a 500 lbs version of what is in your house controlling your circuits. That circuit breaker goes into what is called switchgear, and that is the metal enclosure.
That is the middle picture. So we make the metal enclosure. We have the intellectual property, the circuit breaker. We actually have two medium voltage circuit breakers. We bought GE's circuit breaker in 2006, and Powell has its own medium voltage circuit breaker. So our wheelhouse is really in medium voltage, 1,000 volts up to 38,000 volts. In a fully integrated solution, we will put our switchgear and our breaker in what is called a power control room or a module. What you see on the right-hand side is this happens to be an offshore module, big, heavy, two-story blast-resistant module.
This one looks probably like it is going to an offshore rig to distribute the power on a rig. So kind of the specifications on where we play in the electrical spectrum. I just mentioned voltages. We have low voltage switchgear, so we make the cabinets. We do not have a low voltage breaker. We will typically buy our breaker from an Eaton, a Siemens. We do a lot of medium voltage where we build the switchgear, and we have two medium voltage breakers. So we have a full medium voltage line, really do not get into high voltage at all.
Our sweet spot is really between 480 volts and 38,000 volts. From a product complexity standpoint, when I was here last year, I would tell you that we are almost entirely engineer-to-order, highly configured electrical applications. As you can see, with the onset of the data center volume, this is what we call configure-to-order. So it is highly engineered. You engineer the first configuration, but you just build a lot of it thereafter. There is not an engineering thread throughout the entire project like there is on an LNG project or a utility project.
So we are doing a lot of what is called configure-to-order, and that is a lot of what is in our backlog today, which I will go through. We have our traditional engineer-to-order. Customer relationships historically have really what has gotten us to the table and won the jobs. We are very well respected in the core industrial markets. So just when you hear the grid resiliency, what is the grid and where does Powell play? It is important to understand that Powell is really agnostic to how the power rather is generated and where the power is going.
So whether it is coming off a solar field, whether it is coming off a gas plant, it does not matter. Once it goes from high voltage to medium voltage, we distribute the power. From there, it goes to either a residential neighborhood, an industrial plant, a subway, CTA, we have a lot of gear on the CTA here. So really it is important to think about Powell as just really important in the supply chain of distributing the power irrespective of where the power is coming from or where it is going.
The other important thing to understand in our business is the fact that there are two electrical standards around the globe. There is what is called ANSI, and that is where we grew up. Powell grew up in the Gulf Coast of the U.S., and we are very strong in ANSI and the ANSI market. So ANSI is electrical standards for all of North America, some parts of Latin America, some parts of the Middle East, like Saudi Arabia, is a lot of ANSI gear, and some parts of Southeast Asia. Everywhere else, it is what is called IEC, so it is the international standard.
We have one plant in the U.K. that is 100% IEC, but when you take a view of the entire TAM of IEC, very small. You can see that play out in where our geographic revenues lay out. I mentioned earlier, we have modules on the right-hand side, which I talked about, but more commonly is what we call a power control room. Some people call them e-houses, substations. These things can get up to 10,000 sq ft inside. So they can be very large, or they can be 500 sq ft. They can be small. They are engineered to the application that they are going to be suited for.
This one happens to be on the left-hand side, looks like it is in a refinery or a petrochem facility. If you can envision two dozen of these things sprinkled across the site of an LNG facility distributing power across that facility, that is kind of the solution that Powell brings to these large industrial applications. We are starting to see a lot of these in the data center world as well. We are today on the outside of the data center, so we do not do anything on the inside, but a lot of these data centers are requiring the higher-end voltages, so the 38 kV with PCRs to distribute the power into the data center.
This is just another visual of the entire solution that we provide. We have the switchgear inside the PCR, and the PCR distributing power where it needs to distribute the power. Automation is a key component for us. It is a high-margin business. We are growing this arm of our business pretty rapidly. Over the last five years it has grown quite rapidly. It is still not a reportable segment yet. We did do an acquisition, which I will talk about in a minute, in this space. When you think about automation, we think about it as predictive analytics, preventative maintenance.
When you think about one of those PCRs with lineups of switchgear in it, and there are two dozen scattered across a plant, it is much easier to monitor your equipment with automation. If there is an abnormality, it is running hot, it is dirty, whatever the case may be, our sensors and our automation can pick that up as opposed to having operations teams manually check this gear once a week or once a month, whenever it needs to happen. This is a fast-growing space across our entire industry. Schneider is doubling down on it, Siemens as well.
This is an acquisition that we closed last year, and it is a controller. It was bespoke in the U.K. for utilities, and this was a small little company, didn't want to sell to PE. The principals had built it over 40 years and wanted it to go to a good home with a culture similar to theirs, and they agreed to sell it to us. It is small but mighty. We closed it in August of 2025, and in the first month that we had the business, we were able to apply it to one of these large data center orders that we won in the U.S. This acquisition has been a great success.
Very well integrated and we are looking forward to more of these down the road. The end markets that we play in predominantly, we grew up in the core industrial markets, so petrochemical, oil and gas. I would define oil and gas refineries, pipelines, heavy industrial applications. Over the last probably five to seven years, LNG export terminals has really taken off as well, and it is still very active. Most recently, over the last couple of years, clearly the data center volume has really taken an exponential step up in our business.
It is about a third of our backlog today as we sit here. Really the same product, it is not any different, it is just, as I said earlier, you design it once, but there is just a lot of it. They just need a lot of the gear. The utility space is also. There is clearly a second derivative in utilities with the data centers. How much that is, I couldn't tell you, because when we sell a utility job, we don't ask them where the power's going, whether it's going to a residential application, infrastructure upgrade in that certain region, or a data center.
Our utility commercial activity is very robust, continues to be very robust as well. We get a lot of questions, especially when we talk to some of the European investors, how do you compete with ABB, and Siemens, and Schneider, and Eaton? It's really an interesting relationship. We buy a lot of equipment and gear from particularly Eaton and Siemens, not so much ABB and Schneider. At every campaign, specifically in utility, we'll run up against all four of these guys. In the core industrial, we'll typically run up against an ABB.
What we bring to the table is we build our own build, we build our own PCR, we build our own modules, whereas the other four, ex Eaton, because Eaton just bought Fibrebond a couple of years ago, and we'll see how that goes. ABB, Schneider, Siemens, they don't build a building, so they have to go outsource to a building supplier, send their kit to the building supplier, integrate it there. If there's problems, there's a lot of engineering that happens. It's just not an integrated model, and that's our value add.
That's what wins us business, specifically in the core industrial space. What's our strategy? Brett laid this out, our CEO laid this out probably four or five years ago, and we haven't deviated from it, irrespective of this data center craziness. I talked about the automation. We're doubling down on automation. We think that is going to be a real margin accretion machine, and we're starting to see that. We think there's an opportunity in services. I mentioned that GE's breaker joined the Powell family in 2006, and these things are all over the world, and there's a huge installed base.
We think there's good opportunity within the services, the value-add services. Not just wrench turning and buying people to do just general maintenance, but value-added, engineered upgrades, things of that nature. Then adding on to the portfolio. In utility, this is a real key component of our strategy is growing into the utility space. In utility, you have generation, distribution, and transmission. We do the generation and the distribution. We don't do much in transmission at all. Is that a space that's interesting to us?
Yeah, it could be, to broaden our utility scope. From a capital allocation perspective, as we sit here today, our last report out was over $600 million of cash on the balance sheet and zero debt. We've got a very healthy balance sheet, a lot of liquidity. When you think about how that cash is going to be deployed, we are in a highly capital-intensive projects business, so we get a lot of upfront payments. Typically, 12%-15% of every revenue dollar is going to be deployed at some point in the future to work in capital.
That's a big chunk of it, but there's a lot of room for organic growth in R&D. We've really upped our game in R&D. When I joined the company eight years ago, we were maybe half a point of sales on R&D spend. Now we're almost triple that and starting to commercialize some really exciting products. We brought in a new R&D VP, and he's built a good team. In this space, that's what you need to do. You really need to invest in new products and develop IP and patents and the like. Then inorganically, this is something that Brett and I are continuing to knock down doors.
It's tough in the space right now because PE is really gobbling up these smaller electrical companies, whether they're products or whether they're services, bundling them together and either spinning them back out or IPO-ing them. But there are nuggets. Remsdaq was a nice nugget we found last year. So I mentioned of the $600 million+, half of it probably is going to go to working capital at some point in the future. We're spending a lot on facility expansions right now.
We're just finishing up a 335,000 sq f t expansion at our fabrication yard that'll give us the ability to lay down 10 -1 2 additional substations to work on per year. Each substation is $10 million - $15 million, so that'll help us. That's bespoke for large one-piece buildings for the core industrial market, LNG specifically, or refineries. But suffice it to say, we've got a nice nest egg, some dry powder if we find the right opportunity that we can leverage that liquidity. Now getting into the financials.
You can clearly see 2021 coming out of COVID was a really hard year because, again, we're a very long cycle business. When we book an order, that order takes 12- 18 months to execute, and we're all POC, so percentage of completion. So we'll take revenue along the year and a half timeframe. But we were hungry in COVID, took some orders just to feed the factories, and you can see it was about a break-even year. But fast-forward to 2023, when the industrial wave hit us, all this LNG work.
We really filled up our factories with industrial work from the LNG space, predominantly in Louisiana and southern Texas. Then in 2024, 2025, and continuing to 2026, the data center folks actually found us after COVID because they ran out of their typical Schneiders and ABBs and Eatons ran out of capacity. They found Powell. We didn't go hunting for them. But it's been an evolution. We initially worked with the hyperscalers and the colos, and we've found that this new niche now, we booked a large order last quarter with a neoscaler, a new term to us.
But the terms and conditions on some of these hyperscaler colo operations are really tough, so we've found different channels to market. So we're learning our way as we navigate through the data center channel and markets. But we've had some real success going direct with neoscalers and going through distribution predominantly with some of the hyperscalers and colos. So as we sit here today, I mentioned earlier, through nine months, we're about $860 million of revenue. So well on track to meet or exceed what we did last year.
We're about flat on a year-over-year basis on gross profit. We talked about gross profit in our last call. We are seeing some inflationary pressures, but we are offsetting it with price. Gets a little more challenging in a long-cycle, fixed-firm price environment. We've got a lot of levers, such as hedging our copper, we use a lot of copper, and leaning forward on some of our engineer component buy.
We've managed to get a little bit of favorability there, and we hope with some of this configure-to-order as it begins to exit the backlog early to mid-next year, that we can squeeze out some productivity because there'll be less engineering in those types of jobs. Cash flow through the first 9 months is phenomenal, almost $200 million. That's really indicative of the advance payments that we get. When we sign a contract, we'll get 25% of that contract right up front before we lift a pencil.
Last quarter was a record quarter on multiple fronts. We booked last quarter, we announced it. It was a subsequent event to our second fiscal quarter, but it was a $400 million+ data center order that we won in early April. It was formally booked in Q3 this last quarter. You can see in the lower left, it's off the charts from an orders perspective, $934 million of orders in the quarter. Through the first three months, almost $1.9 billion, bring our backlog to $2.4 billion. We also talked about, we get a lot of questions about what's the convertibility of that $2.4 billion, and we do highlight that in our SEC filing every quarter.
Roughly a little better than 50%, roughly 55% of that $2.4 billion is going to exit backlog in the next 12 months. Then you have your book and ship stuff on top of that. This order and the current environment's really helping 2027. 2027 is essentially booked at this point, and we're booking into 2028 backlog. We have intentionally and very thoughtfully been doubling down on utility backlog. You can see from 2021 to current, it's not dramatically different. The only difference is if this weren't dollars, the dollar growth would be phenomenal.
It's maintained about a quarter of our backlog, but our backlog's grown exponentially. As I mentioned earlier, utilities is a core part of our strategic growth plan, both in Canada, the U.S. Those are predominantly the two markets where we're really focusing on. We are not disappointed. Once you get into the utility space, it's almost an annuity. You get on their approved vendor list, and it helps reduce that cyclicality of the oil and gas market up and down, whereas the utility space is pretty static, pretty flat.
You can see the consumer and other industrial, of that 40%, 34 points of that is attributable to data centers. It's almost entirely data center driven. It popped. Clearly, it popped with that $400 million+ contract that we won in the third quarter. You got a third of our backlogs, data centers, a third is roughly core industrial, oil and gas, LNG, refineries, things of that nature, and then the remainder is essentially utilities. This illustrates the differentiation between what I explained earlier, the ANSI and the IEC.
You can see the blue and that orange color. The orange color is countries that have both ANSI and IEC. We play in those countries. We're relevant there. The gray is where we don't have a lot of business because it's IEC, and we're predominantly not in the IEC space today. But very strong in North America. We go toe to toe with the big guy, all four big guys, ABB, Siemens, Schneider, and Eaton, every one of them in every campaign in North America. We hold our own in the medium voltage space.
So U.S. revenue is almost 80%. The remainder is predominantly Canada and a little bit over in the U.K. From a sector segmentation perspective, you can clearly see the growth in commercial and other industrial, not as accentuated as what you see in backlog. But if you play this forward, that 20% likely will grow to 30% over the next 6- 12 months, given where our backlog sits today. I like the utility space. We like to see that at 25% or greater. We want to continue to build that.
Our industrial space will always be between a third and half of the business, because it is a core tenet of where Powell grew up and where we're really strong. We've got a majority share of that market, and we intend on keeping that. We're very strong in the ANSI U.S. core industrial markets, refining, LNG specifically, pipelines to a lesser extent. That's an overview. I'll take any questions. We have a few minutes for questions.
You don't have any long-term debt. But I'm wondering, if you're going to raise additional cash or things going on here, selling stock or long-term debt, what's that all look like?
Yeah. With everything on here, we would not have a need to go raise debt or issue additional equity. From an inorganic standpoint, our inorganic pipeline contains ideas of all shapes and sizes, some very transformational, in which case, if they were to happen, we'd have to do something different, either debt or equity. Some, like the Remsdaq acquisition that I went through a few minutes ago, that we paid cash, and you saw there, we've got about $250 million of dry powder that we could pay cash for something or fund something from our cash reserves without issuing debt.
All our CapEx that we see on the horizon today, there wouldn't be a need for debt. Brett talked about a $75 million - $100 million greenfield facility add-on at some point in probably the next 24 months to address some of this capacity. All things being equal, we could fund that with cash reserves. Yeah.
All right. You shared the new term neoscaler. You said you all didn't know what it was. Could you share with us what it is?
Well, I'm going to give you probably a novice interpretation of my view of what it is. The neoscalers are what we're finding it's real estate folks that grew up in real estate. They've got this idea that, hey, they can put some infrastructure on real estate that they may or may not own, or they're privy to, build it out and lease it back to a Google, a Microsoft. They're the main purveyor of that asset, but they're leasing it for. I heard on CNBC when there was one snippet that Google and Microsoft are paying, like, $1 billion a month for compute.
That's $10 billion - $13 billion a year. But that's what a neoscaler is. And we're seeing a lot more of these neoscalers. They're a little easier to do business with than some of the hyperscalers from a terms and conditions. They're not as arrogant, to put it bluntly. We're not going to leverage the company with crazy terms, consequential damage. The terms from some of these hyperscalers, you'd never see them in a utility contract, you'd never see them in an LNG contract or an ExxonMobil or a Marathon.
Crazy terms like if there's five minutes of downtime and it's due to your equipment, then you have consequential damages for millions and millions of dollars. We're not going to sign up for that.
Yeah. You've displayed exponential growth over the last three to five years. What inning do you think we're in on this growth phase that you're having?
In any specific sector, or?
In the data centers.
Data centers, yeah. That's where I thought you were going. We have this debate quite a lot, and we talk with our peers as well. Brett and I are thinking three to five. We were with a group last night that thought it was five or six. There's a business in Northern Houston, Sterling Infrastructure, and they're heavy into construction and they're riding this data center wave, and they're looking at things five years out now. These data centers, they'll be revamping the data centers in five years. What they're building today are going to be revamped in five years, which goes to our service opportunity.
That's a service opportunity for us. They're going to be pulling out all the gear and putting new gear in with a different electrical configuration. But long and short, the crystal ball that we're using is three to five, but we aren't seeing any letup today, honestly. And where we are in the proposal cycle with all the political noise going on, we're beyond that political toll gate. So once they've gotten local approval, and they've gotten the FID is locked down, and they've gotten the go-ahead from the community, that's when we come in.
These lead times for these projects, like this $400 million project, that happened, it started percolating in late fall last year, and it was done by the end of March. A $400 million LNG project would be 5+ years in the making. This thing happened in six months. So it's very dynamic.
So given that quick turn that you're seeing there, not to try to play the negative, but when the time comes, or if the time comes, the data center turns down, does it potentially turn off like a light switch as opposed to gradually taper?
I don't think you'll see it drop off. Could it soften and instead of going like this, go like this? Yeah, it definitely could. Now, for Powell, these projects are two years in the tooth, so 24 months. So if in the middle of the project, they decided, theoretically, that at one point they were going to do phase two, three, four, they decided, Do you know what? We're going to postpone phase two, you'd still bleed off that. You'd have another year of bleed off. But if that were to happen, your backlog would clearly drop in a circumstance like that. But we're not seeing that today.
These data centers are really consuming so much. Our higher-end medium voltage, 38,000 volts, that's where we stop. We can't build it fast enough. If I was sitting here last year talking specifically about data centers, it was 5,000 volt and 15,000 volt. Now it's 38,000 volt. They need more power. I think the power equation hasn't yet been solved, and that's good for us because these things need so much power, and there's not enough power to feed all of them. It's a good place for us to be, and we're not seeing a slowdown yet. Yeah.
Can you talk about your government contracting?
Yeah. We've launched a government, not an entity, but a group to focus on federal government contracting. We haven't been successful in a true government contract yet. It takes time, but our pipeline is building. We've built that muscle inside the business. It's inside our services business, and we're hopeful. We're hopeful in the next fiscal year, we can land some of these. They're mostly military contracts, whether it's military bases, whether it's ship-to-shore power. Why shouldn't the last American manufacturer of electrical distribution kit supply the U.S. Armed Forces? That's our position. Yeah.
How many more of these mega $400 million projects are in your pipeline?
Brett talked about phase two of the same project, of this $400 million one. We have to execute on phase one. If we don't execute on phase one, you don't get phase two. That's the only line of sight we have to something that large. That's really an anomaly. We won the entire outside electrical kit for a two gigawatt phase one data center. Usually, they'll split it up.
Right.
They will say, Eaton, you do this. Powell, you do this. We won 100% of the outside. We have to execute and then think about phase two. Okay. Thank you.