Texas Instruments Incorporated (TXN) Earnings Call Transcript & Summary

All right. I think people are still filtering in, but we'll go ahead and kick it off. I'm Joe Moore from the Morgan Stanley Semiconductor team, and we're very happy to have back at our conference, CEO of Texas Instruments, Haviv Ilan. Haviv, thanks so much for being here again.

Good morning, Joe. Thanks for having us.

So you just came off of your capital management call. I know that's a sort of focus for you guys every year where you update us on your priorities. Can you maybe give us some of the bigger picture takeaways from that call?

Yes. I think that happened exactly a week ago, and not a lot of change. We are excited about where we are. We are in the last year of a 6-year investment cycle that was executed well. On budget, on schedule, actually ahead, and very pleased to be in that position. But also wanted to go back to what we talk about every year, our competitive advantages, which is manufacturing and technology, a broad product portfolio, the channel advantage and a strong position in the market in the sense of diversity, of longevity of the business. We went into that. Also talked about our market. This year, we have provided a little bit more -- a change in our -- on our market. We've reorganized our market segments. We've added data center simply because the size of it and the importance in our economy is higher.

The economy, we just heard from.

Yes, we just heard about it. And we now have industrial, automotive, data center, personal electronics and communications, gave a little bit of visibility into that. But more important, very happy with our position in these 3 markets that we care about. I think they carry the most growth for the foreseeable future. Industrial, automotive, data center, 75% of our business there. And an improving portfolio, Joe. This is something that we've been working on for many, many years. I am excited about where we are. And to finish with that, we -- remember, our objective is to grow free cash flow per share for the long term. We've given a little bit of an update of how we go closer to the trend line with a very good probability to hit at least $8 of free cash flow per share in 2026 with a lot of opportunity for upside.

Great. And you mentioned a lot of continuity over the years with the way you've approached capital management with different strategies along the way, but a very clear focus on optimization of cash flow. But one thing that is a little different, the M&A, the Silicon Labs deal. Maybe give us an overview of that deal, and then maybe we can delve into what that tells us about your priorities?

From a high level, strategic thinking about M&A has not changed, same framework. And we are doing them very rarely, right? We've done some in the beginning of 2000s with Burr-Brown. We've did another one with National in the beginning of the previous decade and last month, announced a plan to acquire Silicon Labs. We always said the markets that we like are analog, are industrial and automotive. In this case, a very industrial business for Silicon Labs. The focus will be on analog and mixed signal. And I see Silicon Labs as a great mixed-signal asset. I think they also are a great fit or potential fit for our competitive advantages. So if you think about what they bring into the mix in TI, it's a very nice broad portfolio on the wireless connectivity side. People like to call it IoT. That's the way I think about it. And also a great position in terms of diversity and longevity. A very, very broad customer base, very sticky business, long design cycles, which we like. What we want to bring into the mix is, of course, our manufacturing and technology capability. We are now -- we've invested since 2021 in this mixed-signal asset in terms of the Lehi fab we have. We think we can transition the portfolio over there. And also our channel advantage. I don't think we have quantified revenue synergies and actually assumed them at 0. But I think there are also revenue synergies that can come up as we put these products in the hands of our very capable field force. And also ti.com with e-commerce, with the capabilities we have there, I think that can also accelerate revenue growth. So very excited about the opportunity. Of course, we have to execute through the plan, and we are expecting to close that transaction in the first half of 2027.

Great. Well -- and we cover Silicon Labs and like the fit a lot. I like the business a lot. I guess, just a couple of questions though about what it tells us about how you think of your business. We used to write about IoT 6, 7 years ago, and Rich Templeton was sort of saying we're serving that market through our catalog business. And I think a lot of that was when investors kind of focus on individual little product cycles and things like that. But it was very much -- we're not looking at IoT as a vertical. It's much more of a catalog. We're going to have these solutions. SLAB is almost the antithesis of that, right, that they have looked at it as a vertical. And a lot of the kind of core strength of the company is that kind of tying everything together, having multiprotocols back like that. Does that say that TI is less catalog-focused and a little bit more vertical-focused? Or is that more something unique to this asset?

No, I actually agree with Rich. I don't think IoT is a vertical. So the way I think about verticals or it's markets, right? And the way I think about IoT is technology and the product portfolio and a line in the breadth of portfolio of analog, mixed signal and embedded products. It's just a very important line. It has secular growth because as people want to connect, especially with IoT, with AI and data. IoT or wireless connectivity is a great way to add connectivity capabilities to an already existing system, right? You don't have the wires, you add that feature. But it's a very broad business across multiple end equipment, very catalog-ish, in my view. And that's what we like about it. So in that sense, I agree with Rich. We think that the unique opportunity with Silicon Labs, you don't see it many times in a young company that they have the persistence to build their business in medical, in health care, in metering, in enterprise automation, very rare to see early phase companies going after that. You'll see them mainly usually in very fast-growing verticals, to your point, whether it's communication or personal electronics or even data center. To me, that is exactly the opposite in Silicon Labs. Very broad portfolio, very sticky, a great fit to our company, especially as we internalize the portfolio into our own factories.

And how do you think about that dynamic between having the best catalog, having parts that sort of solve these analog problems versus thinking about verticals and thinking, okay, automotive needs these things and they need to interoperate with each other and industrial needs these things. How do you see the balance between those?

I think both are very important. And even in our capital management call last week, we talked about the 3 markets that we put more emphasis on, especially from an R&D perspective, industrial, automotive, data center, is that you have to play the general purpose play. You call it catalog, but a very broad, not application-specific type of socket. But also the application-specific type of socket. Why? Because it's a big part of the TAM. It's not a good idea to ignore it. But I see the foundation of our business as this general purpose business. We love these type of sockets because revenue per socket or revenue concentration per socket is low. You can command high margins. The stickiness of the sockets are very high, simply because not a lot of people are looking at them. So once you've earned your position on the board, it gets reused again and again and again for multiple generations. You still have to have that discipline, Joe, to always make sure that your portfolio is current, meaning best size, best power, best cost structure. And I think TI over the last 15 years have worked very, very -- in a very high level of discipline to create exactly that. On top of it -- and this is an end game -- you add the application-specific sockets. And especially with the markets we care about. So you won't see TI going after huge sockets on the communication market. If you think about base stations, you won't see us going and putting a lot of attention on personal electronics, although we do it, but we are very selective over there. In industrial, in automotive, in data center, you will see us playing in both roles. And this is a muscle that we are continuing to build. And I'm excited about this end game, have a foundation of general purpose business and on top of it, build the verticals that you think present the highest growth for the company.

Yes. I mean, they're both great parts of the business.

Correct.

Yes. Okay. That's helpful. And then with the Silicon Labs deal, you sort of said -- and you talked about it, it's not a fab filler. [indiscernible] But I guess, how do you think about having a lot of capacity, having at the moment, somewhat lower utilization of that, that you can actually add value through manufacturing by buying assets and bringing them into those fabs? Is that part of the benefit of having that capacity that you have? Or are you still much more focused on that as a strategic reserve?

No. To me, the way I think about an acquisition, or in the case of Silicon Labs, is the portfolio and the market position, and we cover that. In the sense of the fab, I think it's very important. And I think we say it, and I think people kind of okay, brush off it, but we say, especially this year, we are prepared for a range of scenarios. And it's related to that strategic position that you just talked about. I think it's very, very important to be prepared. Right now, in the last few years, the market has started to recover, at least in 2025. And I want to be ready for any scenario in 2026. If the market wants to continue at the same rate of recovery, I think that's going to be not a lot of stretch for the company, and we'll be ready for that. But TI wants to be ready of what happens if industrial does want to go back to trend line quickly. What happens if data center continues to grow very, very fast as it has. What happens if the secular growth in automotive are actually getting -- continuing to generate momentum in '26 and beyond. We want to be ready for the opportunity. And in that sense, we've put a lot of discipline on having enough capacity, but also in the short term, have enough inventory, whether in die bank or finished goods, to serve the surge as it comes. So far, look at 2025, we grew double digits at 13%, but it's not been a very, very strong recovery. But '26 is not done yet. I know a lot of people, you guys are spending one time of what it could be, but I want to be ready for each and every opportunity. The excitement I have is that two things are continuing to happen in a meaningful way. One, data center is becoming a bigger part of our business and the growth rate is very, very high. I did not envision that 5 years ago. I think the CapEx investment over there and the need for energy -- for power density, the need for more connectivity, the need for more sensing and more control is growing. So that's been -- it's now -- we left the year at 10% or 11% of our business, but growing at 70% year-over-year. So that's a big part of the business now. And on top of it, industrial is so far away from trend line. Yes, there are some sectors, in our case, it's aerospace and defense that are unfortunately generating new highs every quarter, but that's almost the outlier. The rest of the sectors -- and we have about 10 of them -- are way below trend line. And we are 4 years after the previous peak, right? So that is a market that has a secular growth. That is a market that has content addition from generation to generation, but still trending 25% below its peak. So I think it's due to a correction, and we want to be ready for that opportunity.

Great. So I want to ask you questions about a lot of things that you just said. Maybe start CapEx and inventory. On the CapEx side, you've talked about a business that could have maintenance CapEx of 4% to 5% going forward. So we'll start getting more cash flow relative to the levels that we've had if you sort of stay in that mode. I guess, how do you think about that level of spending now? Do -- if we have a more robust recovery, would you spend more than that? And I guess maybe the question is, how do you optimize long-term utilization? It doesn't seem like maybe in the past, you thought about, let's be 100% at peak and lower trough. You're probably trying to optimize around a lower level than that, which makes a lot of sense given the economics of your business.

Correct, Joe, I think you touched upon it. But first on the maintenance side, look, every company, especially in IDM, that now we are more and more -- we are internalizing our capacity more and more. We'll finish the decade at above 95% of the front end and above 90% on the back end. That's a big change. If you go back, especially to the back end, it used to be about 60% or 55%. So that's a big change. We want to control our destiny across the front end and the back end. And when you run your operations, even if you don't grow rapidly, there are always areas in your portfolio that are growing faster than others, right? So you always have to -- I call it maintenance, but you also want to stay current. If a certain part of the data center market needs a little bit more capacity, you make an investment. The same with certain packaging type. So that is where the 4% is coming from. And that would be at a low level if revenue doesn't want to grow. But I think revenue has an opportunity to grow way beyond that, and this is where we will modulate up our CapEx. Right now -- and I think I mentioned it last week. As long as growth is very similar to last year, either this year or even next, we don't need to add a lot of capacity because we've gotten ahead. And we are internalizing right now, our -- at least our Embedded business from external fabs into the internal one. However, if the opportunity presents itself, we will happily equip the clean rooms and get a little bit of more spare. And you're right, you never want to run at 100%. The factory cannot run at 100% utilization. This is where the efficiency of the factory drops. And you're just building [ WIP ], right? So operating at that somewhere around 80% to 90% is a sweet spot because we also want to have a surge of demand. And this is where inventory comes in. We want to carry more inventory. And this is something that we've learned back in COVID. It serves you very well in the short term. If there is a surge of demand, for whatever reason you can envision, that the inventory is the first aid to serve customers. And we've seen some of it in small cases. Tariff was an example a year ago. There was one company that ran out because of geopolitical tensions, could not ship in Q4. So this is where TI showed up for the game very comfortably and solved customers' problem.

Yes. I think the strategy there makes a lot of sense. I think it's a lot of inventory that you're talking about, 150 to 250 days. But when you think about the life cycle of the products that you have in inventory, it's actually a pretty low risk around that. So just -- how do you view that trade-off? I mean, it was -- I sort of understood why it was a high level. You went to a little bit of a higher level around the Capital Management Day. Just what's the calculus of what's -- I understand the logic of holding more. It makes a lot of sense to me. What gives you to the number, to the 150 to 250?

I think first, tactically, I think Rafael answered that question a week ago, the 150 to 250 is because we are now -- look, we are fully internalizing our capacity. So think about it, when you run foundry wafers, that [ WIP ] doesn't sit on your books, right? The same as on the [ AT ]. So when you go 90%, 95% internal, that's one reason you see a little bit of a higher inventory. The second thing is the more strategic view, and this is where you have to be very careful on what parts you build ahead. So we will build a part that has the diversity and longevity phenomena or traits. You don't want to build a custom part inventory because that's how you get in trouble and scrap that inventory later on. So we are very intentional about looking at our portfolio and choosing the right part, the right wafers to build ahead. And because our business has a high level of general purpose parts, a high level of industrial and automotive, that longevity and diversity phenomena is there. The second point is you see a range. Why do you see a range? If there is a very high surge of demand, your inventory will deplete. Your 250 will go down to 150. You don't want to go lower than that because that means that we are starting to be short and cannot serve the customer well during that up cycle. And you have to model your up cycle. So you have to model the slope of the demand. We have modeled several scenarios, taking our information from previous examples. And we want -- at least if the history repeats itself, we want to be a great supplier through even the upturn. And of course, once you -- the cycle peaks, this is a great opportunity to go and replenish your inventory to prepare to the next one. This is where you see inventory probably climb to the higher level of the range. But that's a theory. Of course, no cycle is perfect, and every one of them behaves differently. But this is where we have to have the discipline to prepare for a range of scenarios, very, very important.

Yes. I think that's a really important point is that you guys -- more than a lot of companies have -- understand the inherent unpredictability of markets, and you're prepared for that range of scenarios. I guess in that context, how do you think about 2026? Most companies at the conference are seeing what you described, strength in data center, strength in aerospace, defense. Everything else getting a little bit better over time, but like maybe the pace of that is a little disappointing. What are the things that might make that change as you move through the year?

Yes. I would say a small -- in general, I agree. So far, it's more or less what you described. Very strong data center, but now it's a more meaningful part of the business. So that's very important to remember. Now it can move the needle on your overall demand. I would be a little bit more -- I would not say optimistic, but we are seeing maybe a different change in industrial versus a year ago because the breadth of the growth in Q4. I look at Q4, we grew close to 20% year-on-year in industrial. Finally, we are starting to see a very broad demand signal, not only aerospace and defense, not only energy infrastructure that are kind of AI-related. You're starting to see the traditional industrial automation, of medical, of building automation starting to show strength again. I think they are due because they are so much below trend line. And we've seen also -- what I was, I would say, encouraged by is that it improved every month. And we are now 2 months into the first quarter. And so far, so good, meaning we are seeing that signal continuing. So let's see how Q1 plays out. There is still March, and we just came out of Lunar New Year a week ago. But I think industrial is due for a correction. Actually, I didn't break it out a week ago during the CAGR of TI growing in industrial, automotive and data center together from 40% of revenue to 75%, but the CAGR is 8%. That's a healthy CAGR even when '25 is not a great year, to your point, slow recovery. But if you look at the industrial market for TI, it was actually below 5%. It's automotive and data center that grew 12%. So I think industrial has more gain. And I would like to see it quarter-by-quarter. We had a couple of head fakes last year. Let's see what it wants to do. But I actually feel that that's the opportunity in 2026. And industrial is a big part of the TAM. It's 35%, 40% of the TAM. So we have to be ready for that opportunity, not only that customers are getting back to their patterns of ordering, but also delivering new equipment with more content. And maybe if they want to build some inventory -- they're not doing it right now, but if they want to build some, we want to support them. We don't want to tell them no. So we want to prepare for that, and let's let '26 play out.

And I guess when you frame that, we're still so far below the peak of several years ago now. And the industrial activity in the economy has grown over that time.

Correct.

But obviously, at the peak, there's always some overshipment as well.

Correct.

So where are we do you think now relative to the level? Are you shipping to end demand? Clearly, there's not much replenishment. Is there still some pockets of reduction? Just -- are you still shipping below end demand, do you think?

No, I think we ship to end demand, but what is happening is that we are 4 years forward. So we are starting to see the new systems coming in. And customers, I don't think they are building inventory. I agree with your assessment. Of course, they don't tell us, but that's my view right now. I don't think they are building inventory. And there is also not anxiety in the market. This is just a healthy broad demand on the industrial side. But who knows? We've seen some other markets where shortages generated a lot of demand. And we don't want to be short this time, we want to be ready for that. So that's what drives our decision, Joe. And again, let's look at it quarter-by-quarter and call it out as we see it.

Okay. Okay. And then automotive, kind of a very different shape of a cycle. You're at peak, but it never really came down that much, and it's grown. But there seems like there's still pretty strong underlying growth. And I think I'm bullish. I don't know if that's playing out or not like the last few months has been a question, but it feels like there's indications that inventories are low. You've seen disruption around some of the geopolitical stuff that you mentioned. Are we ready for automotive to start a more meaningful recovery, do you think? Or is it demand dependent? How are you thinking about the automotive market?

I think high level strategically -- and let's talk about the next 5 to 10 years -- I totally agree with you, Joe. I think content addition is still ongoing. A lot of people get hung out -- hung up about the powertrain. But we see growth in all sectors of the automotive, including ADAS, which is at the early innings of additions, especially across the portfolio of cars. The body, electronics with more comfort inside the car, more infotainment inside the car. And even the safety systems that are being redesigned with brake-by-wire and other actuators that are coming into the car. Chassis control, that is a cool application. And I think early innings of these additions. So at least for the next 5 to 10 years, I think content addition continues to grow. And this is why we've seen a 12% CAGR for TI in automotive for the last 12 years. And you're right, it never corrected down. It kind of plateaued at the peak. I think what you're talking about is more tactical. So China, yes, there was a very strong 2025. Q4 was our peak quarter in automotive in China ever, okay? But the incentives are going away, and maybe there is a small correction. They built some inventory. I think it's a bump in the road. I think there is a lot of opportunity ahead for the foreseeable future. Now you take it decades from now, it will plateau because of saturation of content. But by that time, we have so many other end equipment like robotics. I think data center continues. So in general, these 3 markets have a lot of opportunity ahead.

And within automotive, China is becoming so important. A lot of the innovation is there at this point. And you do have domestic competition that at least, there's a lot of concern that investors have. Can you talk about competing in China in the automotive space?

Yes. First, in China, this is where we have had a little challenge during the COVID cycle because of supply shortage. That's actually driven some of our decisions to divert or to bias our supply into the industrial, automotive, data center market. So if I talk about TI, 75% of our revenue is in these 3 markets. In China, it's actually higher. It's more than 85% in these 3 markets because we were not supporting well, the PE market. But I will say that the competition in China is intensifying, and it's a function of really, the customer base. The customer base is very fast moving. The demand for speed and urgency is high, but also the demand for a broad portfolio and quality and high performance. So we compete. What I'm proud about -- and you can see our results even in 2025, TI has grown share in China. If you look at our growth in China, about 25% in 2025, and automotive was a big part of it. That was faster even than the local competition. And I think the way you compete in China is first with your competitive advantages. You can argue that there is a very competitive price point you have to compete in China, but we have our internal manufacturing where we don't have to share the profit with anyone, not with the foundries and not with the OSATs, right? That's one advantage. The broad portfolio is crucial, especially in the markets I've mentioned. Instead of solving the problem with 15 different companies, which is the case of the local competition, you can solve it with 1 company, TI. And in the markets of industrial, automotive, data center, performance and quality, especially on the automotive market, is very important. You need to have the stomach and the quality discipline to compete in automotive, even in China and especially as they are exporting more and more of their vehicles across the world, if it's in South America, if it's Southeast Asia and Europe. So I think TI held its own, if not gained, and I'm excited about that. Now having said that, we are highly respectful of the competition. We are studying them every quarter as they release their results. We look at their parts they release to market every time there is a new part, and we learn. And we've learned that these competitors are fast moving like their customers, like our customers in China and cannot be ignored.

Okay. That's helpful. And then -- so for a decade, you've been pretty consistent that the focus areas for TI are industrial and automotive. Now you've added data center to that. And I think everybody would agree with that prioritization given everything that's going on. But what -- as a CEO, what does that mean? Like what -- when you say that's a third priority area now, how are you focusing around opportunities in data center?

Yes. I think it's a couple of points. First, it's really allocation of capital. And number one and foremost is R&D. In data center, we have identified the trend of higher power density, higher power consumption that needed us to work on our technology, actually, our process technology and package technology. And that is a journey that we've embarked upon 5, 6 years back. We've identified that as a priority then. And luckily, these investments are coming to fruition, whether if it's in a BCD node for VRM and serving the last inch of feeding or delivering power to that GPU or ASIC or CPU. And also going to higher voltages. So we've invested in GaN 15 years ago, but to support the GaN requirements in data center, we had to update and accelerate actually our R&D, and we've done that. The second thing is the product portfolio, build a portfolio, both general purpose and application-specific. I've mentioned that on general purpose, we've done very well. A predominant footprint of TI right now in 2025 and before is on our general purpose business. We have hundreds of sockets we can fulfill on a board and sometimes thousands, different sockets on a rack. But we want to add to it, the more application-specific that needed this new technology. So this is almost a one-two punch for us. We are we have done that very well. On the industrial and automotive, I remember discussions in TI back in 2010 when we were very heavily based in phones and in base stations and in ASICs. And we said that's not our future, our future is industrial and automotive. This is where the acquisition of National helped us to accelerate into automotive. But moreover, the portfolio has strengthened tremendously. And we've now been at it for 15 years. The broad portfolio we see is, I think, second to none in these markets in our area of Analog and Embedded, and very pleased with where we are right now.

Great. So we have 4 minutes left. Let me see if we have questions from the audience.

[indiscernible].

Yes. You can find it on our website.

Other questions? I think -- maybe if no, I'd love to ask about your microcontroller strategy. I know you haven't been happy with Embedded performance, but there are some elements of pulling away from some of the base stations and things like that. How do you feel now about your direction in microcontrollers and your ability to outgrow the embedded markets from here?

Yes. First, just -- not happy is not the right decision. I would say proud, but not satisfied. So I am very proud of the inroads we've made into the embedded market and specifically to the, call it, the more broad MCU -- catalog MCU market and also more application-specific. What we have said -- and I said it also on the -- someone asked about Silicon Labs before during the acquisition call -- is that I think the right way to do it is build it organically. Because when you look at the pure MCU companies, the number of architectures and portfolio permutations they have is so broad that it's very, very hard to internalize it. I think Silicon Labs was very unique in the sense that they had a very efficient platform discipline, and that was something that we find very rarely. In the case of the MCU market, I think we have to build it ourselves. And I see a lot of momentum here. Our portfolio is broadening quickly. Our portfolio is now entirely running internally. It used to run -- 2 years ago, our [ F65 ] or embedded flash 65-nanometer controllers were all running in the foundries. They are now all internalized at high yields and very high performance. And that is the plan moving forward as we move into 28-nanometer. And this is where we can start to build a more sophisticated MCU with higher memory footprint and higher content per chip. Very excited about that, but a lot of room to go. The opportunity in Embedded is very, very similar in terms of the growth of the Analog one. Our portfolio is the one that has to be built, and we'll stay very persistent in continuing to build it.

Okay. Yes, there's been a lot of good feedback on the product families and the new -- we did some survey work that showed TI at the top of the list.

A lot of work to do.

But when I said unhappy, the market share has come down.

Market share has to -- I mean, this is part of the "difficult." That's what we like. The moat of this market, nothing happens quickly. It's almost like a delayed gratification type of work. You do the work now, you'll see the results -- sometimes we say we are building a business for our kids and grandkids. But we are starting to see results coming in. I think we've started to nudge market share up in the last year, and we have more room to go. I see a question in the [ end ] over there, Joe, if you want to take it.

Yes.

I was curious, just -- obviously, at the Capital Markets Day, you have quite a bit of capacity and ability to gain share. But how are you feeling about that relative to industry capacity, which seems to maybe be reaching peak? Not peak, but high utilization rates like we've seen from GlobalFoundries and some of your peers as well as on the packaging side as we move forward?

Again, I think your question relates to what I described before. And I don't want to call out 2026, but I think the setup is very appealing in the sense of -- yes, you go and some of the foundries in China are reporting utilization, they are up there. We are starting to see hotspots. Not in TI, but areas in the market that we can fulfill. And right now, the way I think about it strategically, let's gain share, okay? Let's use that opportunity to get on the board and to increase our footprint. Later on, market price will do what it needs to do, but I think that's an opportunity. The last point, I invite you to check a little bit what's going on in the biggest foundry in the world in Taiwan and what they have done with their legacy supply or fabs. I mean, there is a lot of conversion into advanced packaging houses, and that's also something that presents an opportunity for TI. So we'll have to wait and see. It hasn't come to fruition yet. I don't see anxiety in the market like we see in other parts of the market in different types of semiconductors. But I can promise you one thing. If the opportunity presents itself, we will be ready, both with capacity, to your point, but also with inventory.

And with that, we'll wrap it up there. Thank you so much, everyone. Thanks, Haviv.

Thank you, Joe.