Belden Inc. (BDC) Earnings Call Transcript & Summary

Great. Thank you, everybody, for joining. My name is Mark Delaney, and I lead coverage of the U.S. auto and industrial tech sector at Goldman Sachs. I'm very pleased to have Belden joining us for this session. With us from Belden, we have Ashish Chand, the President and CEO of the company; and Jeremy Parks, the CFO. I appreciate you both coming again this year.

Thanks, Mark. Good morning, and it's always exciting to be here at this event.

I thought maybe we could start talking about the evolution of the company's portfolio and there's been a shift in your time as CEO, Ashish, to be more of a solutions provider rather than just individual point products. Maybe talk a little bit about what it means to be a solutions provider and how these solutions are driving value for your customers.

Yes. So if you think about all our customers across verticals, they were using networks with some kind of KPI in mind, and they had an outcome they had to get to. For example, in a power transmission and distribution company, it was about finding the fault faster if they had a fault in the network, right, or to increase productivity. And we realized that as long as we were a component supplier, we would not really address that problem directly. But by combining hardware and software and going across IT and OT, we were able to give this comprehensive solution in that we directly address this KPI. And this obviously came with a number of changes in our own setup and behavior. But it's been going really well, that whole change because now our customers come to us and say, when I move data from point A to point B, this is what I want to do with that data versus help me move the data. And that's really changed the conversation for us.

Maybe you could double-click a little bit on that IT and OT divide. I think a lot of your solutions are helping to address just that situation. Talk a little bit more around how exactly you do help to -- help customers better understand the industrial environment with that IT and OT divide and also leverage insights from the cloud.

Yes. So a lot of people don't realize this, but if you think of a factory or if you think of the back office, for example, of this hotel or a hospital, you have devices there that all speak different languages. In fact, in a typical factory environment, a large factory, you may have between 200 and 250 different protocols because of legacy equipment, different systems, et cetera. So first of all, if anybody wants to get -- make sense out of all the data, it needs to be harmonized. Then you need to take that OT data and move it seamlessly into your IT systems in real time, not through people making notes on paper and then kind of digitizing that, but really from the sensor directly to the cloud. And then third, you want to take the insights you develop on the IT side and take them back to the OT side for some kind of actual movement of something. So this is typically not possible because people have IT networks that are separate from OT networks. So when Belden came in and said, we are able to provide converged networks with a common backbone that have enough granularity for the OT aspect and enough standardization for the IT aspects so that they can be seamless, that solves a big problem. So let me give you one example. Everybody right now is talking about autonomy. They want some kind of AI in manufacturing or services. Now think about that. To do that, you have to first harmonize all those 250 protocols, move them seamlessly to IT, the IT world in that company, where some kind of model has to make a decision and then move it back, and just with legacy systems, that's not possible. So the biggest issue around getting autonomy on the ground is not the data center capacity. It's the lack of a converged network. So I think we're solving some very, very interesting problems right now in health care, in manufacturing, in automated warehousing, in autonomous driving with these converged networks.

Maybe you can give us better understand how Belden enables its customers to achieve this in terms of helping them with the design because you guys have your customer innovation centers, but you also sell predominantly through distribution. So help us reconcile that.

Yes. So the process right now is -- and I think this is an interesting evolution for us. So you might remember this, Mark, 5 years ago, we had 0 consultants. There was no professional kind of job at Belden called consultant. Today, we have over 250 consultants. So it's very interesting. The first stage of that engagement when we know that a customer has some kind of a KPI they're trying to solve is that we have digital automation consultants. This is a new job type at Belden. They go in, they understand that customer's workflow. They understand the KPIs they are targeting. They help them prioritize the right KPIs and then they design a data flow. Through this process, they actually are trained not to use the word Belden. They simply work on the customer's behalf to come up with the best data flow recommendation. If the customer then wants to go to the next stage, they work with our solutions consultants who then design a solution based on that data flow, and these are validated in the CICs. The CICs are pretty powerful. Some people think of them as some kind of a product demonstration center. It's far more than that actually. In fact, we don't have any products hanging off the walls over there. These are validation labs. People come in and say, I'm going to run a nuclear power plant or a train network. And I want to -- for 2 weeks, I want to test how that train network will respond to different kinds of atmospheric changes or environmental attacks or whatever that is. And then I want to make sure that my KPIs are being met, whether they are around automated fare collection or fault identification or whatever those are. And they can change the hardware and software in that process, by the way. So we do this big proof of concept. So really, by the time a customer gets to the CIC with a proof of concept, the chances of us winning are close to 80%. So we have a very, very direct consultant-driven sales approach. At the same time, we appreciate that to deliver at scale, we need to work with our channels because we don't have that kind of logistics capacity. Also, some of our best leads come from channel partners because customers approach them not just for what Belden can provide, but sometimes they get approached, for example, somebody says, I want to buy a fire alarm. And then they say, why? And then it turns out that they're planning some kind of expansion. And then the channel partner gets built and involved and then we talk about the whole design process. So really, the CICs have become our way of selling, but the channel partners remain our way of delivering.

Very interesting. Jeremy, maybe I can go to you for this one. Help us understand how much of the company's revenue currently comes from solutions and any splits between the 2 segments and what your longer-term target is?

Yes. Sure, Mark. So from a solutions standpoint, the majority of our solutions revenue today is coming from our Automation Solutions business. So if you look at us on a global basis, total company, probably 12% to 13% of our revenue is coming from solutions that's 2025. Last year, it was closer to 10%. This year will be 12% to 13% and the majority of that, like I said, is coming from Automation Solutions, which is a little over half of our business. So right now, Automation Solutions is getting closer and closer to 20% of their revenue, which I think is a great number for them. We're still in the early stages with respect to the smart infrastructure business, building out that solution set, validating certain solutions in vertical markets. We're running pilots. We're talking to customers. And so I think we've made a lot of progress on smart infrastructure, but we're still in the early innings there at this point. On a longer-term basis, we would like the total company to be over 20% coming from solutions by 2028. So over the next 3 years, I would expect Automation Solutions to continue to grow. Their solutions business for sure. And at the same time, we expect that smart infrastructure, we can take that business from one that has a very small amount of solutions business today to something more like double digits by 2028.

And I think there's a nice margin uplift on some of these solutions as well, right?

Yes. Yes. So the -- yes, solutions is a great business for us over the long term because when we sell more solutions, we sell more higher-margin products. We sell more switches, we sell more routers, we sell more software because it's a consultative sale, it allows us to differentiate on some of our most profitable products. The other thing it does is when we engage in this, when we go down the road of a solutions engagement, it allows us to create some additional value and do more value-based pricing versus just responding to RFPs and RFQs because when you're doing a solution sale, you're not really just competing on price like you would in a typical product sale. When you sell a solution, you're trying to demonstrate value for a customer to help them achieve a certain KPI, like Ashish said, and that gives you more license to do more value-based pricing, which is a great thing for us.

Maybe we can continue on some of the topics related to solutions. And one of the opportunities has been tied to physical AI and Belden has discussed opportunities in that area, including collaborations with NVIDIA and Accenture. Ashish, maybe talk a bit more around what Belden is doing in this area and what your opportunity is.

Yes. So like I said before, Mark, everybody in the industry is looking for some kind of autonomy in their systems. And for a long time now, people have been trying to figure out what's the best way to get autonomy at the physical layer, right? And so we thought about some use cases that were possible out of this and the one that we piloted with NVIDIA was around a safety fence. So we said there's a large automotive manufacturer. They have a very complex assembly line. It has robots, human beings, all sorts of moving parts working together. Now usually, if they had to create physical safety for people in that environment, they would need thousands of sensors all sorts of connections from those sensors to some kind of cloud infrastructure, some decision-making process out of that. And that would be, a, very expensive. It could still have a lot of latency. So it won't guarantee microsecond decision-making. And it would need to be changed all the time as the line changed in terms of shape and size, right? That was the complexity they were dealing with. So we went in, we took this full assembly line. There were only 3 cameras, no sensors, no data was sent to the cloud. And essentially, it was Belden's Time-Sensitive Networking, which is a very advanced networking protocol versus -- so the regular networking protocols are first come, first served, which don't guarantee that kind of microsecond accuracy in terms of how packets are delivered. So we came in with a very, very complex and high-end switching protocol. That data from the 3 cameras was orchestrated on the Belden Horizon platform, where Accenture had built a bridge to take it to NVIDIA's Omniverse, which is their digital twin application. And basically, Omniverse and the application on that were acting like a human being watching those 3 camera feeds in real time triangulated. And essentially, that system can start, stop, red flag, blow a siren, blow an alarm, all sorts of things that a human can do or a number of humans together would sit and do if they had access to that same kind of video feed, right? So this is a very, very interesting pilot. By the way, all of the compute took place on the edge. Nothing was sent to the cloud. Now this is only video. Now imagine if we added sound, vibration, temperature, pressure, now suddenly, with that multimodal data, the decision-making could expand to many, many other things apart from safety. And all of that is possible on that same network backbone. So first of all, we proved through this pilot that with very high quality of edge compute, which is why we work with NVIDIA, we can create this decision-making system, which is fully autonomous and doesn't need a very complex kind of infrastructure and it's scalable. But second, there are more and more use cases that can be based on the same principle. For example, if you can do safety, you can also do quality inspection. You can also, by the way, have -- there's a shortage of skilled workers in the industry right now. So you could have unskilled workers working on a process that needs skilled workers, but you could have the autonomous system kind of sitting on their shoulder, making them skilled for that process because now you could catch with a camera or with other sources, you could catch all the other data that allows an unskilled worker to become a skilled worker. You could apply this in passenger safety, the same kind of principle. So really, with this pilot, we've laid the foundation for many, many more use cases. I think for physical AI to really happen, 3 things have to happen. First, across different systems, there has to be harmonization. Second, IT and OT have to converge. And third, most of the data, if not all, must be computed on the edge. There is no time for it to go to the cloud and come back. Now you might still have models in the cloud that get trained, and that's fine as a backup. And I think if we can provide these 3 things to multiple customers across different verticals, we will see many, many more such cases.

Maybe we can talk about demand. And on -- in your call, you spoke about some nice growth in your Automation Solutions business, but help us better understand how demand trends in automation are trending.

Yes. So automation is, I think, having a very strong year. On a year-to-date basis, that business is up double digits in the third quarter. Automation Solutions was up organically just about 10%. I think orders were up a little bit more. So things are going in the right direction. I will give one caveat. Last year was a bit of a destocking year, so that helps. But we are seeing good movement in the right direction. Demand is improving. I think every single quarter, which is very positive. The other thing that I would take as a positive is some of these geographies that were drags or vertical markets that have been drags over the past year or so have gotten much better. Like Europe was down last year pretty significantly. Last few quarters, we've been up year-over-year in Europe in the industrial business. Likewise, with China, and when you look at it from a vertical market standpoint, discrete manufacturing, in particular, has been growing double digits the last few quarters on a year-over-year basis. So I think things are going in the right direction. We're not quite where we want to be yet. PMIs, although they've improved in most countries around the world, they're still more or less a little under 50% in most of the developed world. So we've come back a long way. Demand is getting better. Our business is growing. We're selling more solutions, which is very positive because I think we've got some of these secular tailwinds at our back, but we haven't seen the market really take off and inflect yet, but I think everything is going in the right direction.

As you think about the demand environment, do you think tariffs have led to incremental automation investments in the U.S.

I think that they -- I think they are leading to incremental conversations. For sure, what we are seeing is some reshoring of manufacturing. It doesn't have to be building a brand-new factory in the U.S. and closing a factory in China. It could be ramping up production in your U.S. factories or your factories in the U.S., Mexico, Canada and other parts of the Western world. We are seeing that ramp-up in production, and we are seeing some investments happen there. I think on the flip side, tariffs have created a little bit of noise in the demand environment because the tariff environment keeps changing. And so I think that has been something that we've had to work through. But generally speaking, yes, we are seeing investment in the U.S.

And maybe just to add to that, right, I think there are 2 other impacts of tariffs. One is that as there has been more -- and to be fair, even before tariffs, if you think about how the supply chain brittleness became evident post pandemic, right, there was already this feeling that people have to bring manufacturing closer to the point of consumption. But one thing that's happening is that people are trying to bring more sophisticated manufacturing back because otherwise, it won't be cost efficient. But ironically, if you bring back more sophisticated manufacturing, you don't have the skills, so which is leading to a demand for more automation and maybe potentially more autonomy. So I think tariff has this quantitative, but also a qualitative impact on what's -- where the investments are going to be.

Very helpful. Maybe we can shift gears a little bit and talk more on the smart infrastructure part of your business. I wanted to start on the broadband portion of the smart infrastructure segment. Maybe talk a little bit more around how the company is seeing growth rates in broadband and how the BEAD government program may or may not affect that.

So I'll make a couple of comments and then maybe Jeremy can add some numbers. So first of all, we've really invested a lot both organically and inorganically into fiber. And today, our fiber revenue as a portion of total broadband sales is 50%, right, which is a big milestone for us. It used to be, I don't know, like 5% a few years ago, so a big change. We continue to see across the board more adoption of fiber with MSOs, with telcos, so that's great. The one thing we did see in -- and it was reflected a little bit in our Q3 situation, Mark, I'm sure you guys noted this. So different MSOs built different kind of technology stacks on how to get high-speed broadband to different -- to users, right? And so they were not uniform designs. And depending on how they build their stack and how they got interoperability between different vendors, they went faster or slower. So we noticed a little bit of slower offtake on that DOCSIS upgrade in Q3. We noted that, but we also think that's this timing. Overall, what we see is that broadband quality growth is improving. When I say quality growth, speed and bandwidth. And of course, the number of nodes, the number of homes passed is increasing at a certain rate. I think BEAD is going to help. I don't see BEAD as fundamentally changing the economics of the broadband market. I see BEAD as supporting what is already happening, but now they have an additional source of capital. And some of our customers, they won pretty big awards in the BEAD process. So we are looking forward to that coming through in 2026.

Yes. I guess what I would add is from a number standpoint, broadband is roughly flat year-over-year on a year-to-date basis right now for us. I think less an indication of the market, generally speaking, and more an indication of what Ashish talked about, which is that there are some technology upgrades happening and there's been stops and starts in terms of the execution of some of our customers. So they're working through that. I think they're committed to executing on these DOCSIS upgrades. So I think that will happen. I think there's a little bit of some temporary noise going on right now. With respect to 2026, I would see, like Ashish said, that BEAD is good support for the market, right? BEAD is a program that was passed back in, I think, maybe 2021. It does feel like right now, the administration is making it a priority to get this money allocated and hopefully start to get it spent. At this point, the last numbers I've seen is something like 90% of the money has been allocated, $38 billion out of the $43 billion has been allocated. So that money will hit the market and that will get spent over the next call it, 5 or 6 years, I think that will be great support. It will create, I think, a lot of investment in the space. So I view it in general as a nice tailwind in the market, but the overriding story in broadband is what Ashish talked about, which is we've increased the amount of fiber content we can sell into these spaces. We have very deep and intimate customer relationships, and we expect that these customers are going to continue to invest.

How would you characterize inventory levels in the broadband space at customers?

Yes, I think they're fine. They're roughly normal in terms of days of inventory on hand at customers. So there's nothing unusual either high or low.

Maybe we can shift gears to the smart buildings part of the business. And starting with the data center market, I mean we've all seen tremendous growth in the data center CapEx levels. I think this year, the bottom-up aggregation of CapEx estimates from the GS analysts covering those companies points to about 80% growth in 2025. Next year, they're looking for 36% growth at the moment. And just over the course of 3Q earnings season, the CapEx estimates for '26 were revised up by 17%. So the trends there have been pretty positive. Talk a little bit more around what Belden has seen and maybe more importantly, what you're doing to participate in that market.

So let me talk more broadly about smart buildings, and then we'll go to data centers within that. Our goal, Mark, is to be able to provide IT and OT converged solutions to all sorts of verticals, right? And that includes traditional smart buildings markets like health care, education, hospitality, et cetera. And what we figured out is that whenever there is some kind of a KPI, so for example, in health care, they have a KPI around patient health or -- in fact, it's interesting in health care, the KPI they really follow is how soon can they get a patient out because apparently, that has the longest -- that has a correlation with long-term health. Similarly, in hotels, they have KPIs around turning rooms, et cetera. So whenever there's a KPI, which is measurable in an operational sense, that's where we are differentiated when we come in and provide these IT/OT converged solutions. And similarly, in data centers, we participate in the white space plus gray space. So if you were only in the white space, which is inside the data center, the network hardware, it's growing very fast. But to be fair, it's also a little standardized, and it will come under price pressure over time. But if you think of the white space and the gray space together, which means you are now participating in the substation automation, in the cooling, in the security and all of the services that surround the white space, there is an IT/OT convergence there, which is more differentiated and unique. So that's where we are participating. In fact, I think last quarter, we announced, not in Q3, but in Q2, we had announced a big win for a cooling use case in a large hyperscaler AI data center. So we are seeing more and more inquiries and engagements around this converged set of applications. We are also seeing more demand for our fiber products and our connectivity products in the white space. So as a combination, I think data centers are doing pretty well for us. I think in Q3, we had double-digit growth in data centers. It's not the 36% level for us because it's not really our focus to do only that white space. But on a balanced combined approach with that IT/OT converged approach, I think it's far more than it used to be for the old smart buildings business. But again, maybe -- I'm sorry, I'll add one more point. Commercial real estate, plain vanilla office space, is now less than...

Less than 15%, 12%, 13%.

Of the smart buildings portion, so less than 2%, 3% of Belden, which is a big change over the last 5 years.

Yes. The only thing, maybe just 1 or 2 things I would add. First of all, smart buildings is having a good year for us. I think in general, when people looked at this business for us, they would think investors would think nonresidential construction. And nonresidential construction has been weak for several years. Smart buildings is up 5% year-to-date. So it's growing kind of mid-single digits because the strategy here is to focus on the right verticals, health care, hospitality, data centers to some extent, -- and I think that's working for us in differentiating ourselves with customers by trying to sell solutions and pitching solutions, which has been helpful. In terms of data centers, I would say it's a very small portion of our Smart Buildings business, and we've traditionally focused not so much on the hyperscalers. But we do sell, like Ashish said, into the white space and the gray space. And that gray space win that we had in Q2 was a great proof point for us, and I think a lot of opportunity going forward. So I think there's a lot of opportunity here over the next couple of years as we develop and sell more solutions into the space.

One other area of investment in secular growth I wanted to touch on was in the grid, right? There's been a lot of investment around power transmission and distribution, modernizing the grid to support things like data center and other uses of electricity. Maybe talk a little bit more around what Belden can do in that area.

Yes. We love that market, by the way. That is one of our favorite markets. We announced a big win in Q3, $14 million, $15 million worth contract. So if you think about that space, there are 2 interesting things. First, a lot of people don't appreciate this, but there is actually more generation capacity than actually reaches users. A lot of it gets lost during transmission and distribution. In fact, in certain parts of the world, there is a surplus in generation, but still a deficit in usage because of that dynamic. And if you then drill down the biggest pareto around why are there losses is because the grid infrastructure is still largely analog, in fact, to get technical, most grid systems are still based on SONET/ SDH. It's like using the old telephone system, the dial-up system, which we don't use nowadays, right? We all use smartphones. So Belden's role here is to upgrade all those systems from that old push button telephone or dial telephone to a smartphone. So we bring in -- we have this product range called XTran, which essentially is an MPLS-TP or an IP-based backbone solution to connect all these systems in the grid. It basically improves grid efficiency dramatically. It allows for bidirectional communication. And right now, I think only about 10% to 15% of all the grids are like that. So there's like an 85% more kind of white space available for us to go and modernize these grids. I think this is going to be one of the biggest building blocks to more AI usage because right now, one of the biggest bottlenecks is the availability of power, right? And so really, a very exciting market. Belden has a very unique solution. Again, it goes across IT/OT. It's a backbone solution. We feel very well positioned in this space.

Jeremy, I want to touch on the financials and the profitability of the company, if I could, please. There's been so many factors on cost and supply chain. You've got tariffs. We've also seen what's happened with copper pricing. Maybe talk a little bit more around how you're managing the business? And do you think Belden can stay price/cost positive as you look into next year?

Yes. I think to answer your question, yes. I think you're right, there's been a lot of noise over the past really several years in terms of inflation, tariffs, logistics costs, freight, labor inflation, things like that, which I think we've worked through. And I think, in general, we've been fairly successful. We have good processes around pricing. We sell -- we use a lot of copper. Copper is a very volatile asset in terms of pricing. And so we have a lot of practice in passing on copper price changes to customers, either good or bad, that's sort of an expectation in the market we play in. And so we have good processes around that. I think we've been able to extend that into other things like tariffs. I would say we're not the most probably impacted by tariffs of all companies because we do tend to manufacture in region for the most part. So we're not generally manufacturing on the other side of the world for customers over here in the U.S., for instance. So I think that helps us. But to your point, Mark, we've -- I think we've dealt with it. We've not really had major issues up to this point. And I would expect going forward, we've learned a lot on pricing, and we're selling more solutions. And so from my perspective, we've got to keep doing the things we do to manage our own cost productivity going forward, finding opportunities to reduce costs on purchase goods, reduce logistics costs and so forth. On the flip side, there's a lot of opportunity on pricing, especially value-based pricing as we do more solutions and solve more customer problems. I think that allows us to differentiate.

Ashish, maybe for the last question, I can go to you. I mean we talked a lot about the business already. But as you think out over the next, say, 3 to 5 years, what are some of the most exciting things in your mind for Belden? And anything you think the company needs to do to better position to address those?

Sure. I think, first of all, if I look at the macro trends, right, so reindustrialization, reshoring, big plus for us. The idea around physical AI and more autonomy on the shop floor in the back office, it's something we can enable. This needs, as I said, IT/OT and hardware/software convergence. You need to be able to get the data through the hardware, but also orchestrate it so that it's ready for use, and that's something unique. I think we will also continue to see more -- Belden will have to position itself more and more in that solution space. This will need, again, more partnerships with the Accenture and the NVIDIA and other companies in that space so that we can provide a full solution. The one thing I will say that we are working on, which is the fine balance we have to find is how do we simplify enough so that customers can adopt our solutions more easily without losing the complexity that you need per vertical, right? And I think this is where we have to improve. Belden Solutions can still be a little complex to understand and adopt. And as we simplify more, I think we'll see adoption rates going up dramatically.

Great. Well, unfortunately, we are out of time. Jeremy, Ashish, thank you so much for joining.

Thank you very much. Appreciate it.