TE Connectivity plc (TEL) Earnings Call Transcript & Summary

All right. Hi, everyone. I think we can get started here. My name is Alex Dwyer. I work with Jefferies. I'm an equity research associate covering machinery and multis, and I work with Steve Volkmann. Today, we're joined by Terrence Curtin, the CEO of TE Connectivity. So -- it's a fireside chat format. We take audience questions. So if anyone has a question, don't be afraid to raise your hand. With that, let's get started. Terrence, thank you for coming.

No. Thanks, Alex, and I appreciate everybody being here to learn more about TE Connectivity. So let's get into it.

Yes. I guess I'll just start out with some Q&A. And so I mean, you've raised your AI sales guidance a couple of times this year, though the growth rate -- the year-on-year growth rate has come down from 70% to that 33% range or something. And I know like the market -- nothing has really changed in the market. I know you're ramping a lot of capacity, and there's been some program timing nuances. Can you just talk about how that kind of unfolds going forward over the next couple of quarters?

Sure. So first off, thank you again. And just to frame where Alex was going, when we look at TE, it's important to understand that we play at the intersection where you need data connectivity as well as power connectivity. And our biggest growth driver, both this year and last year really has to do with where we work with semiconductor companies and hyperscalers on how you connect GPUs and all the connectivity that go into the rack. And it's been a really good growth driver for us. And actually, our expectations are higher than we just thought when we started the year, and the momentum continues to be strong. So yes, we're going to have times when programs change, some programs come in, come out. But what's really good, our position with our hyperscale customers really has set up nice momentum, not only this year but into next year. So our orders are up about 70% this year in our AI space. That's going to set up a really nice tailwind as we go into '27 and as we continue to help them solve the connectivity needs that happen both on the power and data side. The other thing about this year, just to build on maybe outside of our AI business is our AI business has been maybe about 40% of our growth this year. TE is going to grow about $2.5 billion this year. And that's really driven by our industrial segment, which includes what we do for our AI customers. But that breadth of growth that we have is really where we bring our connections, whether it be in the energy infrastructure, across aerospace and defense, which are strong secular growth trends that aren't slowing down anytime soon, which contributed incremental growth on top of what we do in AI as well as our legacy position in transportation, where we have a strong position that really our growth is being driven by content performance that's above market. And I know you're going to click down in all of these, Alex. But the element is the breadth of the growth that we have as data proliferates as well as the power that's needed around that proliferation, I really think makes us a little bit unique as you look at opportunities for investment.

Yes. Okay. Got it. And I know like earlier in the year, there was a lot of commotion about the optics and copper transition. I think you framed it as copper and optics. And I think it's become more apparent that copper is here to stay for longer. But I guess when you talk to customers, have you seen them like try to adopt like more optics racks? Or are you seeing them kind of push against that now? How has that played out as the year has gone on?

So we've always seen our customers obviously look at where optics play from a cost, from a power need and really to get to the data speeds that is needed in AI. So it's not a new discussion, but it also sort of became a little bit of a versus discussion versus an end discussion. And that was earlier in the year was a big discussion. But what we see in our DDN business, which serves our AI customers is and our large customers have said it publicly, Copper is going to be continuing to be the workhorse that you see in the rack. Power connectivity, which is about 1/3 of our DDN business is always going to be copper. When you look at transmitting data and signal, you're going to see areas where optics is used. You're also going to see areas where copper continues to be extended because of its scale, its lower cost as well as lower power consumption. And what's nice about where we play is we get to work with our customers as they look at where do they want to experiment. And in our AI business, it is something that it's not going to be cookie cutter across the architectures. Some customers are going to push the limit. You will see how that comes into the switch in an optics perspective. You have other customers that say, no, I want to be at a lower cost or a lower cost per token. So we see a lot of experimentation across our customer base. But what we get excited about is that the copper TAM will continue to grow, that it is an 'and' discussion, not a versus discussion. And we've also positioned ourselves where we would play sort of optics that comes into the switch because we really play in the scale-up element. We do less in the scale out. And certainly, we're going to be part of that, that will add TAM to us as we look forward. So lots of experimentation going on in any AI architecture, both on the power side and the data side. And they're both things that not only for the growth we've had this year and we expect next year, but also as we look out multiple years and as we work with our customers on those racks and those architectures.

Yes. I guess to stay on the optics piece, you guys made that RAM Photonics acquisition earlier in the year.

Right.

And it didn't seem like there was a lot of revenue attached to that, more like a technology idea. Can you kind of talk about how you see that playing into like the optics adoption? And I know they come with like a fiber array unit. Can you just talk about what that is and how that speeds up the manufacturing issues that the optics supply chain is going through?

So as we look at where optics would play on our road map, we really view from a connectivity perspective in the rack, that happens at the switch and what you would have, whether you have co-packaged optics or near package optics, how do you bring the signal off of that with the density, but also you have to have the scale to manufacture at very high rates. And that's where we've been very focused on our development. To your point, we made a technology acquisition earlier this year called RAM Photonics, which really has some very high density as well as good automation that goes with it. It doesn't have revenue attached with it at all. It's really part of our technology road map from a product perspective. And it really sets us up to where we think we should play in the optical space. We're not somebody who's going to be doing lasers and active optics. There's a lot of companies that have capabilities in it. But when you think about connectivity and the passive connectivity that we do on the electrical side, which is copper, as you call it, we really feel that the fiber attach unit is an area where there needs to be scaling brought to. It ties in with our road map and also the discussions we have with our customers and as we go through engineering qualification. So I view that as part of that TAM expansion that we talked about. It's something that we didn't have at our Investor Day earlier in this year, but it's something that provides a future growth opportunity no matter where optics comes in, whether it's CPO or NPO.

And what do you think about like time line on revenue generating for that asset? Do you think it's kind of a couple of years away or...

Yes. So the question -- I was at another conference yesterday that was a technology conference, this question I got every minute. But the element that you have is it's really going to come down to where our customers see scalability of where this can be that really doesn't change their road maps. And right now, we're in a lot of engineering qualifications. I don't think you get into meaningful revenue into '28, '29, depending upon where our customers have it in as well as where does the broader ecosystem support optics adoptions in the switch.

Okay. And I guess the power connector opportunity, I think is that 25% of your DDN business?

It's actually larger. So when you think about our DDN business, which is about $3 billion of our $20 billion this year, about 1/3 of it is a little bit higher than what you said, Alex. About 1/3 of it is actually around power connections. And one of the things that creates future growth opportunity is as you bring more compute and more optics and other things in, the power consumption that you need increases the content significantly. Certainly, there are some customers that are looking at 800-volt sidecar power racks that support the GPU racks. That's where the power is going. There are types of things that can create a 50% increase in our content on rack. So when you think about the growth we're delivering this year around $1 billion, those types of architectural changes continue to help us look at a road map on growth that's just going to continue to build as AI architecture evolves.

Yes. So I mean, I guess, can you talk about, like what the last 3 years look like from the power connector, like, growth story and how that can differ from, like, as we think about the next 3 years and whether 800-volt comes or doesn't come, like, is it just denser connectors or more volume or higher pricing?

So when you go into there -- typically, when you think about density, you think about density more in the data side. But on the power side, what you're dealing with is you're dealing with voltages that are much greater to bring over. You're dealing with bigger interconnects, not typically smaller interconnects as well as a lot of the materials and electromechanical elements that go with it. So when you think about probably 3 years ago, power connectivity was probably 25% of our DDN business. It's up to 30%, about 1/3. And I think you're going to continue to see that go up as the power architectures evolve versus the standard architectures we've had historically. So when we look at this space, I know you asked about optics and copper. There's elements that relate to what's happening on the data side that we're going to benefit from, both in copper and optics. But then there's the power element. And the power element, we have some customers that are looking at 800-volt. Some are looking at 400-volt steps. You're going to see a lot of different architectures that our customers are experimenting with. And we get to work with them, which is a very important part of our moat with our system architects and our customer system architects. And it goes back to anything that TE does. I know when we're talking right now about our DDN business, but whether it's in our energy business, our automotive business, our factory automation business, it really always starts with you have a semiconductor and you have a power supply and you need connectivity that brings that architecture together. And they're the things that we focus on. And depending upon how hard the technical challenges, that drives the bigger content opportunity. So we always embrace architectural changes. There are things where our customers need our expertise. It's where our engineers excel. And then certainly, we have to make it for a life of any program, but there are things that we get excited about all the time.

Does anyone in the audience have a question or I can keep going if not. All right. I'll keep going. I guess I'm going to move on from DDN segment and go into energy. I guess this -- you've added Richards and you're 70% more -- you're 70% exposed to North America now. The growth rates come up because of that. And I think you even raised it again to mid-teens last quarter. Can you talk about, like, what drove that revision? And, like, what gives you the confidence to think mid-teens organic growth is the right growth rate going forward and talk about that.

Yes. So when we sit there and we talk about our Industrial segment, we talked about our DDN business, which is a chunk of it. But the other big -- other chunk we have, we have a bunch of businesses that are $2 billion. Our DDN business is $3 billion. And once again, it's doing connection, but this is in the power space where we support our utility customers from medium voltage in the grid, all the way down to stepping down to a data center, stepping down to a renewable farm. And it's an area where we spend a lot of time focusing on where do we play from a go-to-market. And we've also done a couple of acquisitions that have helped support where we want to play, including Richards and Harger. But when we look at this space, it's a space that -- these are power connections, 800-volt is low voltage in a utility. I mean, that's low voltage. But when you think about where we're helping our customers here, it is as the energy capacity has increased this market. This is a market 5 years ago that was a 1% to 2% market. We view it's growing 6% to 7% today. And depending where you play, you have growth rates much higher than that. We excel in medium voltage. We also excel in undergrounding. So when you get into grid hardening, you also get into highly dense areas like we are here in Manhattan. You really get into opportunities as the grid needs to be upgraded, certainly, there's deferred maintenance. We feel where we're playing actually gets us into that double-digit rate that you talked about, and it's why we upgraded it. We continue to add capacity, both in our facility here in New Jersey, near here in New Jersey as well as out in Ohio. And 70% of our business is U.S. focused. It is an area where we pick to play stronger in the U.S. Clearly, TE, you typically think about TE being very global. Actually, in energy, it's very much U.S. first, a little bit smaller in Europe is really our position. And as we sit there and think about how you get power distribution, where we serve our utility customers, but just as importantly, how we serve what we call industrial power, the EPCs that help do moving the power from the utility to the step down into the data centers, to step down into semiconductor manufacturing sites, there are other areas that we were very strong at and you use our medium voltage products on. So we get very excited about the growth trends here. I know everybody likes to talk about AI, but there's an element here with where the energy infrastructure is at and the growth rate we see there. We really like where we position ourselves. That's a nice double-digit grower. It's going to be mid-teens this year. I think you can have that expectation as we go forward in the 6% to 7% market. We continue to have opportunities, and it's also an area we'll probably continue to look at across our Industrial segment for places that can have inorganic opportunity.

I was going to ask, are there any -- like as you look at this business over the next 3 to 5 years, are there any like immediate gaps you see in the business that you would like to add organically or inorganically? I assume you probably want to keep building out the North America presence. Are there any regions in the country you're underexposed to -- or -- you mentioned strong in undergrounding. Would you like maybe get like build up capabilities in like the above-the-ground utility? Just how are you thinking about that?

So certainly, we're very focused on medium voltage, and we like that, and it builds on what we've done organically and what we've done inorganically. I think if we could find things that would help in the overhead because a lot of the United States is overhead when you go into the middle of the country. But there'll be opportunities and they have to be on the right returns and growth profiles for us. So I do think the energy space is always a fragmented space. We like the base that we have. We also think there's opportunities to continue to expand it with the relationships we have with our utility customers as well as in the industrial power side. So we'll continue to look at components that could fit into our suite and that we can add value to both for us as well as the customers as well as owners.

Okay. And then I guess on the lower growing parts of the energy business, the traditional Europe business and then the clean energy renewables piece, are you seeing any signs of pickup there? And can you just like kind of refresh on what are the main like renewable clean energy applications you sell into?

So our growth rate, renewables, we typically sell into utility scale solar. So when you look at that here in North America, it's -- we are not as wind exposed. But that's been slower. And even in the growth rate that you talked about in mid-teens, we're absorbing a slower environment and some of that's due to regulation. But that is really where we play in renewables, and that's really here in the United States. In Europe, Europe is traditionally a slower grower than where I talked about from the growth rate. We are actually seeing step-up from an investment from our European customers. So it also gives us confidence that we'll be able to be in that double digit as those investment levels come in, where they're actually putting in LNG facilities actually has a lot of power that needs to come to it. It's very important to our growth profile.

Yes. Okay. Cool. I think I'll stop there on the energy business and then move over to the ACL segment, which is more of like your traditional general engineering -- industrial business. I mean, the double-digit growth started coming before the PMIs started inflecting in January, and it's been 8 months in a row of expansion there. I know you serve a lot of different end markets in that business. Are there any like 1 to 2 like major geographies that have been driving it or any like specific industry? I know factory automation has been strong.

Yes. So when you look at what we call ACL, it's automation is how you should think about it. And this is a business that I just -- when we say the growth rates, I do have to be transparent. It's been a market that's been very slow for a couple of years. So we're finally seeing a cyclical inflection. And to your geography question, Alex, it's broad. We're seeing it in all 3 regions of the world. And to your point, you look at ISM, PMI, you really see that momentum going because it is so broad. Where we play is very much in discrete manufacturing in the automation space in the factory automation space. And we're seeing our customers probably get back to more of a mid-single digit to higher single-digit growth rate in their business. A lot of it comes back to the data that you need on the factory floor. When you look at what we do, and I know we started with AI, the data that's actually needed to make things more efficient actually starts with the productivity that you need to collect the data to make the machine more intelligent. And one of the things, yes, we have a cyclical pickup that's early. I would tell you, even when we look at TE, where we're investing and how we think about how we use AI internally, it's on the engineering side and it's in our factories. And that comes with getting the data off the machines, how do you make the machines more intelligent to have that compute using the models. So it's one of the things that not only have the cyclical pickup, we typically have a content above sort of an industrial CapEx number. That's a couple of hundred basis points ahead. And when I see how we're investing and what we're seeing in the trends, it sort of makes sense that they're dovetailing and it's across our global network. It is not around, hey, growth is good in one region or another. It's really around how we're driving the efficiency that we need to drive the next level of productivity using the tools that we didn't have before that some of the LLMs provide us. So I think we're still early in that cycle, that's just picking up. And it's nice to see the trends our customers are seeing in their businesses, but it once again comes back to data connectivity you need that we connected our machines years ago to really get to this point, and then we're just accelerating it to use the models as they become more relative to drive next level of efficiency.

Yes. And how do you compare this industrial recovery to like different industrial recoveries that have happened over the last decade or so? And I guess there's a debate on how long people think this -- like an industrial recovery could last? And like what would be the biggest risks to kind of derailing this?