Eaton Corporation plc (ETN) Earnings Call Transcript & Summary

All right. We're going to get started here. Thanks, everybody, for joining in the room and on the webcast. My name is Amit Mehrotra. I lead the multi-industry franchise here at UBS. Super happy to have Paulo Ruiz, the CEO of Eaton Corp. In the audience, we have Yan Jin. Yan, there you are. Nice to see you. Thanks for joining. I hope this is going to be interactive. There's so much to talk about with Eaton in terms of where the company has been under Paulo's leadership. I think on June 1, where it's going. And obviously, data centers and power and now liquid cooling. Obviously, it's December 2. So we're going to talk hopefully about 2026 a little bit. And I'm trying to make this as interactive as possible. So there's a lot of people in the room today. So please don't feel shy. Raise your hand, we'll get you a mic. But maybe to start, Paulo, thank you for joining. First of all, I really...

Thanks for having me.

Appreciate your time. First and foremost, you've been at Eaton for a while, but obviously taken the top job, the top position mid this year. And it's not lost upon anyone the change in direction, maybe moving into more growth verticals via most recently the acquisition of Boyd Thermal. Maybe just talk about like your vision. So if we look back 5 years from now and look back, what is going to define Eaton and where is the direction that you want to take the company relative to where it was over the last 15, 20 years, which is obviously a very successful direction, but still a little bit different.

No, thanks for the question. Thanks for having me, and thanks for the audience for the interest. I would say this, I shared in March during the Investor Day with the group in New York and you included that we put our strategy around big pillars, 3 big pillars, strategic pillars. And I purposely inserted growth in each one of those pillars, names. So the first pillar of the corporate strategy is lead for growth. It's all about being customer-centric. It's all about being nimble and moving fast and making the decisions. The second pillar is about execution. If you think about execute for growth, I'm an operator. That's my background. That's what I cut my teeth on 20-something years of P&L management. Seizing the opportunity of a growing market requires the company to have strong operators. So it's operating well in our sales teams, our engineering teams, operating well our factories, et cetera. So that supports the growth. And the third pillar is about investments being -- have the conviction to invest in high-margin, high-growth verticals. And we have plenty. We have not only -- everyone mentioned data center, which is pretty obvious growth vector for us. We have the utilities market in electrical. We have the C&I market. We have all the mega project that's just started in the U.S. So it's a long tail of business coming our way. And we have the strongest distribution channels that haven't gone anywhere. So they're a strong asset for the company. So there's a lot of investment in our future growing the company organically. And then inorganically, to your point, I think it's a misconception that we are doing much more. I think we're doing what the company always did well, which is to have tough internal targets where we want to return our investment over 200 or 300 basis points of our capital in M&A, but we are a bit more focused this time. I gave the company clarity on where we wanted to invest, part of the invest for growth mindset. We said we're going to do electrical and aerospace. And within electrical, look at data centers and utilities, and that's exactly what we're doing. So the company has historically done a number of deals. If you think about the position we are in today, the big acquisition of Cooper 13 years ago transformed the company. Since then, many, many, many small deals happened. Some of them didn't even hit the mainline here. But I think the company has that in the DNA. We are disciplined in our approach. We have a very well-oiled machine in terms of diligence and integration. But this year, we were more focused. And I give also credit to Craig when he goes out and makes the announcement on behalf of the Board that I will be elected, he does that in August '24. So that gave me almost a year. So I start working with the team on the new strategy, looking for targets. Don't think as something that we did in the last 6 months. We started 18 months ago. So that helped a lot.

And one thing, obviously, we noticed this most recent quarter is the big sequential step-up in orders of the company within the Electrical Americas business, call it, from $3.4 billion, $3.5 billion to above $4 billion in the quarter. And you were quite bold on the call talking about how you expect this to even improve further in the fourth quarter. Maybe just talk about that because you talked about 100 gigawatts plus, which is just an insane amount if you just think about what that means. Where are we in that kind of build-out or time line? And where is your confidence that you can continue to build on this very, very strong order momentum within EA?

So you talked particularly about data center build-out, right? So think about that in March, when we had Investor Day, we talked about market projections of the data center to go to 70 gigawatts to 130 gigawatt, having 100 gigawatt at the midpoint in '28. Some people were skeptical, but those were market projections, not Eaton projections. We took different projections into consideration. Back then, 2023, the overall U.S. data center capacity was around 20 gigawatts. Approaching the end of the year, we are going to get much closer to 40 already by end of '25. So I would say there is upside and the backlog that our customers have in hands in terms of projects they announced, some started, some have not, is 165 gigawatts. So we moved from 20 2 years ago to close to 40 now in December. And then we have $165 million that's going to be executed. So just follow the CapEx trend. And then if you think about it, this is incredibly powerful for a company like us in terms of winning business. And when I say we need to invest and we are investing, I'm glad we started early because we are capturing that business. When we talk about data center orders in particular, you saw that our year-over-year orders in Q3 were up 70%, 7-0. So there are 2 things here for you to consider. Over time, we improved our lead times. Over time, we have more visibility into our customers' plans. And then over time as well, the data around chips is moving so fast. The new technology is moving so fast. There's a new chip coming out every 18 months. Everything I mentioned here should point to less orders because lead times are coming down. There's more certainty on supply because we're investing in new plants. And with the chips coming -- the changes coming more frequently, you would expect that orders will come down. And orders are coming really up because the market is so strong that is overcoming the lead time reductions overcoming the shift in technology coming so fast. So there's much less multiyear orders now, more like data center by data center build-out because we are ready. So I would argue, of course, the market is really strong. The fact we started to invest early also is helping us to win more than our fair share of this business.

And what's interesting on that point is we talk about this elongation of the backlog. But if I look at your backlog most recently relative to kind of the revenue, I think it was 34%, 35%, which is more than double where it was or double where it was several years ago. Are we actually seeing an elongation of the backlog because of lead times? Or is this really more foreshadowing the revenue growth that's coming kind of in '26?

No, this is the market strength that is pushing that. It's not about lead times allocation. I just told you, we are reducing lead times. So it's actually the strength of the end markets. We talk a lot about the electrical backlog, which is not only bigger is fourfold what we had historically, 4x bigger and growing. In Q3, we said our backlog is 20% higher than last year, being 9% organic and 11% was the acquisition of Fiberbond. So it's growing. It was historical levels. We keep investing. We reduced lead times, and we continue to grow backlogs. But it's also true for aerospace, which we don't talk much about it. Our backlogs are also growing. So I think it's the strength of the end market that's driving that.

And the technology is obviously changing so fast. We were talking about now high-voltage data centers. And when we think about your capacity in solid-state transformers, do you feel like you guys are very well positioned? I assume the answer is yes, but maybe you can offer a little bit more color around how you guys are prepared for that transition and whether your capacity in solid-state transformers will allow you to kind of participate in that growth?

No, I will get to the solid state. I'll give you a broader picture then we get to the solid state. So for data centers, I think we are winning in that space for 4 main reasons. The first one being that we have the broadest portfolio. I'm going to get to the transformer in a minute. We have the broadest portfolio that starts from the utility feeder where the transformers sit all the way down to the chips. This is really important. So portfolio was very large and comprehensive. And with the 3 acquisitions we made in that space, Fiberbond, resilient power and most recently, Boyd, we're just increasing our penetration in that market. So it's an incredible portfolio we have. Two, I think the way we approach that business, which is key to winning across the world, but especially in North America, you need to come with an integrated offer. What I mean by that is about the hardware, it's about the software, it's about service capabilities as well. If you don't have service, you don't win in that space. So we have that. We started, as I said before, talking about capacity getting ready for the growth. That's another area which I think is really important. And the fourth piece, which I think is also important for us to have in mind is, we start to integrate those solutions, not looking at a product sale one by one, but offering full systems to our customers. Getting to a specific point on 800-volt DC architectures, what happened over time, we moved closer to the chips as a company, and we are embedded in the developments of the NVIDIA, the Googles, the Metas, et cetera, on their next generation of chips. So -- and that's what Boyd does extremely well. We made a so interest in that business. When you do that, we start projecting back what the power should be. We got to the discussion about accelerating the development of our solid-state transformer technology. We are doing that organically, and we always do that. We are paranoid about market dynamics. We had a review. And as a team, we said, can we look outside? Is there anyone ahead of us? And the answer was like, we don't know. Let's go check and they found out that Brazilian was actually a couple of years ahead of us. So we're accelerating that development, and we're going to be ready by when those chips that require service without 1 megawatt power, we're going to be ready to supply the whole chain. So we can ramp up -- we know how to do it, and we just don't need to be ahead of the chip development. That's all.

And I think this is an important point because data center exposure is not created equal across all the different players. And if I think about kind of the total data center supply chain from the power generation side to the transmission distribution side to within kind of the gray and white space of the data center, there is a point of differentiation among companies. But when I look at companies like Eaton or Vertiv or obviously Schneider with Motivair, there are questions around kind of what differentiates Eaton versus those kind of soup to nuts players, so to speak, in terms of within the gray and the white space of the data center. How do you think about how you guys differentiate yourself against maybe those specific players?

So I'd rather talk about Eaton than other companies. I agree. But since you asked me I need to address your question. No, I think the big obvious difference between Eaton and those 2 players you mentioned is that we have 3 phase transformers, they don't. And one of the 2 has a very limited power distribution portfolio, which is the biggest item in terms of dollar per megawatt in the power side. So they don't have much of that capacity there and technology and portfolio. And what we also do as a company, as I said before, we go from the utility feeder all the way down to the chip now, and we do that at scale. That's also important. Boyd will give us, it's not part of the company it will be, will give us a scale entry into cooling and we won this because, in my opinion, the cooling and power integration will be won by engineering power. So if you put together the Eaton engineering power, Ingenuity and Boyd, they have 500 of the best engineers together, we're going to continue to win. If you move outside the data center for a moment, another big difference to the 2 companies you mentioned, we are present in utilities. Those guys are not. So we can help our data center operators to have better connection to the grid. And the question you raised before around the current Think about our connection with direct current on energy storage, battery systems or even renewable integration, we can help our customers because we are a utility player as well and the broadest North American player. So we can help them. If you go outside then electrical, we have aerospace as a growth vector that those companies don't have. So my message to you is, yes, of course, we are a formidable player in data center. Those 2 companies also are, no doubt. I believe we have the most complete portfolio, especially now with the recent 3 acquisitions in that space, and we can play outside the data center as well.

And I guess that also just maybe quantifiably, we can move to like content per megawatt as a differentiating factor as well. I mean it was at Investor Day, you talked about maybe a couple of million dollars per megawatt. Now we're at $3 million per megawatt. I mean 100 gigawatts, that's 100,000 megawatts. That's a $300 billion market. I don't think that there's a company out there, maybe one other that has a similar kind of content per megawatt. But maybe talk about that bridge that 2 to 3 pro forma for the Boyd acquisition.

Yes. So we -- today, with the current portfolio we have, we already see our content per megawatt growing in AI data centers close to 2.9 million per megawatt. And the Boyd acquisition would add 500,000 on top of it. So it moved from 2.9 million to $3.4 million, rounding 3.5 million for good math. The way to think about it for us as Eaton is all upside. Why do I say this? First of all, data center is the fastest-growing part of our business, for sure, it is, but still 20% of the company, right? And then within data center, AI, although it's growing really, really fast, it's 30% of the data center. We are talking most of the time around 6% of the company, growing really fast. Now the conversion to AI is to be expected that our leadership will only grow because we can go all the way from the utility feeder down to the chip and the content per megawatt going up favors us. So think about as a very large tailwind of secular growth coming our way.

And just on that point, we talk about brownfield, greenfield, retrofits. Is there something in the order data that we can observe that tells us something about that and what the opportunity shift is if we move more to retrofit versus greenfield?

Yes. It's a very interesting question. We talk a lot about that internally, but also with our customers. I would say today, it's all greenfield, and there is a reason for it. There's a race for in building those -- at the same time, those training data centers for AI, at the same time, need to build the cloud infrastructure for companies to move their data to cloud so they can leverage AI. So that's their race. And the other element, which is important for us to consider here, the demand is higher than the offer of the capacity in the market. So our customers are favoring greenfields. Having said that, it's logical to think when the training models are already developed to a certain degree, you need to have inference data centers that are closer to big urban areas. So I would expect that it's logical to think that more retrofits will happen in those today cloud data centers that are closer to big centers and big cities. And we're already debating and discussing with our customers flexible design architectures where they can have both cloud and inference AI in the same data center. Again, it's something that's coming. It's a tailwind, but it's still very small, and it's going to come.

Can we -- I want to talk about liquid cooling, which is not something we could have talked to Eaton about a year ago as much, but now obviously, it's a different picture. It seems like when you close Boyd, I mean, Boyd is going to grow 70% next year, something like that and good margins. I think there is a question about the defensibility of their business when you think about how much of their business is hardware components? And what is the commoditization risk of hardware components within that even cold plates as we move to two-phase cooling with higher-density racks or whatever. Just talk about like the runway of growth for Boyd. And it seems like Eaton with this acquisition is kind of underwriting another acceleration in orders as you close this deal kind of middle of next year. A lot of questions there, but let's talk first about the commoditization risk of Boyd and how their ability to kind of maintain that market share as we move into higher density racks.

Yes. So I got really comfortable. We asked ourselves those questions over time. We started approaching Boyd 14, 15 months ago. We wanted to get to know the business better. We wanted to acquire the business and was not on the market. So we got closer to them much before the deal was signed. What really attracted us to Boyd is not only the scale that they have in terms of revenue and the deep connectivity they have with customers and the build-out of the chips, but especially because of their engineering power. So in a market that's moving so fast, when chips are launched every 18 months to 2 years, if you don't have a seat on the table today with NVIDIA, with Google, with Meta, with Microsoft, and those people do have that seat on the table with all of them. If you're not embedded in their design of the future chips, when they are launched, you are after that, right? So the race is about engineering and innovation. Why cold plates are important. They are a mirrored image of a chip. So every chip has a different heat map, and you cannot swap a cold plate from one chip to another. So you build a cold plate for a particular chip. And as you're embedding those developments years in advance, when they are launched, you already start producing. It's closer to what we do in businesses like aerospace or even mobility markets where you win a position in a platform and you start getting the orders. That's what these people do. Eventually, other companies can try to copy and emulate that. It's possible. But the time you copy a cold plate and you are able to do that, there's a new generation of chips. So you're always behind. So you need to have that engineering power and you need to be connected to the future development to win. Now what comes out of that connectivity is not only about cold plates. I don't see Boyd as a cold plate company. I see them as a cooling company. So they move up in the chain. They're moving up in the chain in terms of doing CDUs, the cooling distribution units, which are systems that are detached from the rack, but they're really important, and that connectivity is important. So they're already moved up. So think about this, they are moving from the chip upwards, we are moving from the utility down, we are meeting the middle. So it's a real powerful combination between 2 strong engineering cultures, and we believe we're going to win by innovating.

So maybe a few months ago, when you were far along in this process and the headline comes out with Microsoft Microfluidics, I assume that spurred some questions. And obviously, you got very comfortable with that. You've done a lot of work. But just talk about why that's not a concern for you.

It was not a concern. And to be honest, we are also connected with the company that Microsoft was leveraging in Europe. We knew them. We were in connection with them. Those -- my point here only reemphasizes what I just said. Innovation will come out every time, like we just heard about Google's chip launched the other day. It was not a surprise for Boyd. They've been embedded in this development for a long period of time. Those innovations will come out. And I would not -- of course, undeniably today, NVIDIA has an edge over any other chip manufacturer. It's a fact. But I would expect that the major hyperscalers, they're going to come more frequently with their launches. They want to have their own technology. And ultimately, the industry will move to -- from 2 platforms to maybe 5 or 6 platforms around different chips and the companies that will win will be those that can understand this road map, understand the implications on cooling, understand the implications on power and they can be ahead of the game. When the chips are launched, they have the whole technology ready.

And this valuation discussion, I think it was like 20, 22x or something next year's EBITDA coming down to single digits. Is that just more of a function of growth? I mean this is probably not a business where there's a lot of cost opportunity. I would imagine it's just a function of the growth that they're going to endure beyond '26.

Yes. So yes, from the multiple going down to high single digits, the biggest contributor to that is the growth because they're growing at 40% CAGR from '26 to '29. So in 3 years' time, the multiple will be less than double digits. This is the biggest contribution. But also very importantly, when you have a business that is a $1 billion revenue business, they cannot get the same deals we get with suppliers. We have much more purchasing power. And that's an easy, I would say, synergy to execute. So that's the second biggest. The third one, not so big because we normally don't like embedding much of a sales synergy into the model because we are conservative. But there is a third area where we believe we have better connectivity with the multi-tenant, the colos, data center operators than Boyd does. Again, being an engineering company that's focused on chips, they are embedded into the hyperscalers and NVIDIA and AMD's road maps. We believe we can help them with the big colos, winning also more of the cooling business there. But the biggest one is growth.

Helpful. I just want to pivot a little bit more to maybe some observations about what you're seeing right now in the market relative to kind of the guidance and then obviously, sitting at December 2. So I would love to get your perspective on 2026. But first, we're 2 months into the fourth quarter. You guys obviously reiterated the guidance for the full year. Good growth in the fourth quarter. Maybe just talk about kind of how things are trending relative to maybe what you talked about in the third quarter.

Absolutely. So we are on track to deliver on our Q4 and our full year guidance on EPS. We are fully on track. We are fine. I also anticipate in Q3 that our growth will be in the lower range of the guidance, given the vehicle market, the resi market, again, no surprise here. So we are on track for the year as per the Q3 earnings call. In terms of '26...

Before we start to '26, so the margins in EA were better in the third quarter. Growth was a little bit weaker than maybe expected. But obviously, that's reversing in the fourth quarter. Growth is very strong in EA margins, incremental margins implied by our guidance are a little bit weaker. Maybe just -- I know you're going through a big expansion. You're more than halfway through this year. You've got another. Talk about when we kind of get to the point where we can get back to the incremental margins that are more lockstep with what they had been prior to this expansion.

Yes. So we have a lot of activity in Q4. We are doing well, but there's a lot of ramping happening in the Electrical Americas business. That creates a number of inefficiencies. When you have so many plants that are ramping at the same time, training people expediting material, it creates a number of inefficiencies. So that's a reality. We keep investing. So you should think about '26 as a year of continued investment. We should expect towards the end of the year that we're getting most of these inefficiencies away and we start running at full speed and all those expansion projects. And for Q4, specifically, that's what's driving. We had in Q3 some carryover of products we couldn't deliver that we're going to deliver in Q4. I think we are tracking really well. And the comparison point to last year, I said in the earnings call, some people cracked the laugh, some of your colleagues, the comparison point with last year is the easiest, easiest comp because last year was 9% growth only because we had some hurricane and some issues with strikes and things like that. So the comparison point is the easiest one. This business is growing well. We are investing. It's natural. It's a growing platform that you see some inefficiencies. As we get towards the end of next year, we should see those inefficiencies go.

And I think like it's not lost upon me that the margin expansion stories are from companies with significantly lower margins than Eaton as well. So in some ways, you are a little bit of a victim of your own success in terms of where the margin is today and where the incremental margin can be. But do you feel that we can get back this time next year when we're on the same stage, look back and say, hey, we're back to that 35-ish percent type of incremental margin algorithm. Is that the way -- in EA, is that the way you think about it?

We are improving that for sure. I don't want to give you guidance. That's February for next year for sure, but we are working really hard to ramp those factories up. And then more broadly, I think it's also important, the way I talked about execution being one of the pillars of my strategy, we have a lot of opportunity in running our plants better independently on the ramp. What I mean by that, still within the company, if you look at the way we run our vehicle plants or electrical Asian plants versus our Electrical Americas plants, there's a lot of opportunity in Electrical Americas. So of course, the margins are high, close to 30%. We love it. But we are not just relaxing on this big achievement. We are pushing the teams. We are giving them the tools and the talent they continue to improve. There is some juice for us to squeeze here in terms of execution.

Great. And so maybe -- sorry, I interrupted you when you were kind of talking about 2026. So you're exiting this year very good growth, but expectations are also for good growth next year. So maybe just talk about how you think '26 kind of shapes up. I know you're giving guidance in February, but maybe give us some broad brush strokes.

Yes. So guidance comes in February, but we gave you an idea of where our markets will be. So we see our 2026 markets growing around 7%. It's natural to expect we plan to have some growth on top of the market. So make some assumptions there. And then using a decent incremental 30% would be a good proxy. That's the way to think about it.

Got it. And there's obviously you're closing on Boyd, below-the-line stuff can matter. So you gave us the above-the-line variables, but maybe talk about some of the below-the-line moving parts that could maybe either help or hurt '26.

Yes. So I think in terms of '26, below the line, 2 items for me to highlight here. First is pension. We normally don't talk about that. It was a positive guy. It turns into a negative next year. And the other thing, we are acquisitive, so we need to finance and those costs will hit the below-the-line item as well. So we have pressure there on the corporate cost. and we continue to see improvement on the segment margins.

And then just on the acquisitions, obviously, Fiberbond, Boyd Thermal, Ultra PCS, like what's the way to think about the financial impact from those aside from the below-the-line item stuff that you talked about?

Yes. So 2 of those we already closed, right? One is pre-revenue, but then Fiberbond is in full speed and growing really rapidly. So on Fiberbond, I think the assumption could be look at the market growth in data center, assume the same growth and put an increment on top, that's fine. When we close the other 2 deals, when we close Ultra, when we close Boyd, then we're going to give guidance on how we're going to grow those business. But think about market growth as being a good proxy.

And then just obviously, orders such an important KPI for Eaton and a lot of the companies. You kind of were very bold to say, hey, we're going to see another step-up in orders in the fourth quarter. As we think about kind of the runway for this growth and the next-generation platforms and the acquisitions you made, can we have a decent step-up in kind of order activity next year? Is that the right expectation?

So again, it's early to talk about next year. I can talk about Q4. And we -- as you see, we kept growing our orders profile quarter after quarter. So we had a very strong Q3. We expect Q4 to be even stronger. And with large backlogs like that and the value proposition we have, I see no reasons why we should not continue to win business in that pace, right? So that's logical to expect.

Do you still think we can build the backlog next year?

We could, but we're also growing. We are -- we have...

Another way of me to ask you orders, by the way.

I know. I get your point. But for 2 consecutive years, we've been talking about the backlog being 4x, 3x or 4x historical levels would not be a problem the backlog will come down because we're investing capacity. Reality is we kept growing backlog quarter after quarter.

But I guess like the orders, I mean, it's tricky because there are some companies now that are moving away from quarterly order disclosures. And I know we try to get away from that, but we can't really get away from that. And there's always a need to like show sequential momentum in orders, which is difficult because the orders are large and lumpy and timing is important. But you guys did do chunky, chunky orders in the past, and sometimes that's been a headwind as you look prospectively. Do we have $1 billion-plus orders kind of in the pipeline that we could start booking here in your opinion?

So we had -- as an example, we talked about that openly and publicly. Q1 '24, we had a multiyear order, which was above $1 billion from a particular customer. We see less of these multiyear orders because First of all, the customers believe that we can deliver because we invest in capacity. So there is no reason for placing multiyear orders on us. We have a forecast, but we don't have necessarily multiyear orders because they are not needed. The second thing I told you before, with the chips changing every 18 to 2 years, customers would not love to place orders 3, 4 years out when technology might change. Having said all that, we look at how we are performing, and we're substituting this big order with the new bread-and-butter business, which is with high data center, big data center AI data centers is much more common today that we negotiate orders around $500 million, $600 million, $700 million. They're much more frequent. And our negotiation pipeline now is around $8 billion that we are discussing right now. And a big portion of it is data center.

Any questions for Paulo from the audience, raise your hand. We'll bring the mic over to you. I wanted to just -- finally, just in the last couple of minutes that we have, Electrical Global has kind of been -- the spread between Electrical Global margins and Americas margins has been very wide. Boyd, if I'm not mistaken, it's actually going to go sit within global, if I'm not mistaken. So that -- I think that's part of your strategy about moving away from maybe residential machine OEM end markets towards faster-growing, more value-added end markets. So that's great. But that business has actually inflected quite nicely from a growth perspective. And it feels like those 2030 targets seem within reach. But if you could talk about maybe some of the -- what you're seeing in global, particularly in Europe, that's kind of driving that inflection?

Yes. So cooling aside, when you're not a market leader as we are not a market leader in Europe, and the only thing you can do is to win market share. So we -- as a mindset, I talk about lead for growth, the mindset with that team is don't look at PMIs, Don't talk to me about PMIs. Look at the -- all the good things we do in the Americas and try to emulate them in Europe. And we put a new management team together. We actually got someone from the Electrical Americas team to lead Europe with a mission to shift us more towards the growth areas of the market, data center and utilities. Traditionally, our European business has been very strong and it's still strong and will continue to be strong in MOEM, so machine builders, resi distribution business, we value it. We're going to protect it. It's all fine. We're going to continue to invest and grow. But we were not that strong in utilities and data centers, and now we are becoming stronger. So that's driving the top line. And then true to my 3-pillar strategy, we had a strong focus on execution as well. So we said, if we are going to grow in those areas, we need to invest in those areas. We need to restructure other parts of the business, and we are restructuring, and saving money. At the same time, we are improving the way we run our supply chains, the way we run our factories. So it's a combination of a different mindset, going after the right growth vectors, not looking at PMIs and executing better. So that's the formula.

That's great. And just as a final point, I wanted to just maybe tie a bow around how we think about '26. So it feels like another year of underlying market growth in the high single digits, which you should and expect to outperform a little bit. Maybe the incremental margin algorithm is good at 30%, but still kind of being pressured by the first half ramp in capacity. But as we kind of exit the first half into the second half, it kind of might look like a tale of 2 halves, so to speak, as you exit some of those maybe inefficiencies related to capacity growth. Is that a fair characterization of '26?

It's accurate. Yes. Again, guidance is going to come in...

Okay. And then you've got some below-the-line item stuff that's maybe transitionary. Any final questions for Paulo before we end it here? All right. Well, thank you, sir. Appreciate your time.

Thank you. Appreciate it. Thank you so much.