Okay, good morning and good afternoon. I'm Tim Arcuri. I'm the Semi and Semi-Equipment Analyst here at UBS. Very pleased to have Texas Instruments with us, next, and very pleased to have Haviv Ilan, who is the President and the CEO of TI. So thank you, Haviv.
Thank you, Tim. Good, good to be here. Thanks for having us.
Great. Well, Haviv, let me just start by a question that I'm sure you're getting all day, and, you know, I get a lot too, that your recovery sort of came out of the gates pretty strong. You were above seasonal a few quarters, and things have sort of come back to being more seasonal. So can you just talk about some of the drivers and maybe some of the puts and takes that are occurring in each of your bigger end markets?
Yeah. When I look at 2025, we are in recovery. If I go back, I think we almost, like, had a zero growth in Q4 2024, so finally stopped declining. And then in 2025, every quarter, we saw growth. It's in the double digits. I think company at the midpoint for 2025 is somewhere around 13% growth. But you are right that it was not smooth. So we've seen building momentum in the earlier part of the year, Q1, and especially Q2, with all the tariff noise. It was really hard to decouple what's what is the root cause or what's driving that recovery. But it took a little bit of a step down in Q3. So if I look at it over the last year, we see a recovery. It's a double-digit recovery.
We can also see it on the unit trend when you look at the market, not only TI, but it's kind of a slow recovery. It's a moderate one. You can go back to more or less 2000 to see a similar recovery. And I think there's simply a lot of uncertainty in the market right now. And also, there was a little bit of asynchronous behavior during the upcycle that is now playing in a downcycle. So we are trying to stay away from trying to call the shape of the recovery. From now on, we will tell you how we do once we finish the quarter. But I am pleased to see all markets are recovering right now for us.
In some areas, we are seeing a new peak of revenue established, specifically on data center and also in the automotive market. We got to the same level of 2022, so or 2023 in the case of automotive. So, I'm pleased with where we are. I think there is more room for the market to catch up and get back to trendline. We are still trending below trendline right now.
Can you just talk about some of the end markets as you look in December? You didn't call out any one end market in particular for your guidance, but maybe speak to some of the, you know, relative strength or weakness in industrial, you know, all your different end markets in industrial, and then within, you know, autos, any [GOs] that are stronger than others within auto?
Yeah. I think, look, our main market for TI, about 70% of our business are industrial and automotive, and both of them are showing recovery. I think that the recovery of industrial is more significant. It's in the double digits, but also it fell the most, right? We saw a big drop back in 2024. And if you look at where we are right now, and I think it's not a TI-specific, we've still not established a new revenue peak. I think there is still room to go. And we can talk about some of the macro issues that are creating that, but we do see industrial recovering fast, but still not at the levels where we saw it in 2022. The automotive market is behaving better in two ways.
First, when it dropped, it didn't drop a lot. It was a single-digit drop, and now it's growing single digits, but I do feel like the fact that if I compare to the previous peak, it's at the same level, and I don't forget that the previous peak was probably helped by some of these inventory buildups, so from an end demand perspective, I do like what we're seeing in automotive. I think it's related to the secular growth in automotive. I know that a lot of folks are concerned about how quickly EVs are getting adopted, but I think every vehicle is adding content, from EVs to ICE to hybrids. On the industrial side, I think there is more room to grow. Now, if you specifically about GOs, I can't not think about the China market for automotive.
This is where EVs are gaining momentum. The OEM there are doing well, not only in China but also with their export business. And I think this market still has a lot of potential in terms of content addition. We are seeing a generation-to-generation, we are seeing just more content added to vehicles. I think this is still a growing trend. And I also like the fact that we are competing at a high level across GOs, including China. So in that sense, these markets have been, or this specifically, this automotive market has been doing well. Last but not least, I do wanna add a little bit of a few comments about the data center market because it's becoming more and more significant. It's still a single-digit % of the TAM, but it's moving very quickly.
I call it somewhere between, if for analog and embedded, somewhere around, $5 billion - $ 10 billion, somewhere in between, but moving very, very fast. I think it has a lot of room to grow in the future. So we also see the momentum over there. This is where we see a strong momentum, like all of our peers. Customers are continuing to invest. This is where there is no hesitation whether to put more CapEx into data center, and TI is enjoying that trend, among the industry.
You're gonna start to break out data center starting in fiscal 2026 as a separate category?
Correct. So in January, actually in the earnings call, we provide data center as a standalone market. You'll see TI talking about industrial, automotive, PE, or personal electronics, data centers, and comms. These are gonna be the five markets. We are going to give a little bit of history also, so we'll make it history corrected. So we'll see 2025, and also how fast that market moved over the last years. We'll report this market going forward every quarter. This is how we run the company. We are putting more emphasis on data center internally, and we also wanna share it with the street and with owners.
Yeah. It sounds, I think you said it's about $1.2 billion right now.
Yeah. We are collecting the data. I think last time I checked, it was $1.2 billion. I think it's probably gonna end up a little higher than that. Most exciting is the growth rate. It's well above 50%. And what I like about the market, that I think it has more momentum building, right? As we see architectures changing, at the rack level, we are seeing that, semiconductors are playing a larger and larger role. As long as you believe that data is gonna be important in our, in our life, I think this market is gonna gain momentum in the future and, can one day become a double-digit part of the TAM, if not even 20%. I don't see a reason why it's not, over the long term.
Very important to break it out, very important to monitor it, very important to report it back to the street.
Great. I wanted to ask about factory loadings. One of the main issues coming off of last call was that, you know, gross margin was guided a little bit below what I think some folks thought. I didn't think it was that much of a surprise, but I think some investors thought that it was a surprise. The, you know, guidance implies something like 55% gross margin. It's down about 250 basis points, roughly. That's my numbers, not yours. Can you just talk about how you're managing loadings and inventory? It seems like you finally have maybe reached a pain point on inventory where you don't wanna build any more. So loadings are gonna track demand, and you're sort of looking at Q1, and Q1's not the best quarter seasonally. So, you know, loadings come down a little bit in Q4.
Yeah. I think it kind of answered it, but remember, taking a bigger picture, Tim, and you know, you know us. We don't think about, to be fair, we run the company on free cash flow and free cash flow per share. And the reason we are stopping building inventory at the same rate because we reached the point where we wanna be. So if you think about, even in Q2, we were a little bit anxious about, can we get ahead? Can we build buffers across the entire portfolio? We were in some cases where we were almost like hand-to-mouth on some of the technologies. I'm very pleased that, by the end of Q3, we reached the level that we wanna be at.
And now, not only that, we have capacity ahead of demand, but we also built the inventory level across our portfolio. I think about some areas in analog, especially those that we're building in [Richardson in RFAB2], that were catching up. But you know, you can call it the market was growing not as fast as it could have, and it allowed us to get ahead. Now, when you get to the right level of inventory to serve the short-term surges that you might expect, of course, we manage the company on free cash flow. So you let the cash flow fall through to our owners. So we are gonna see momentum over there. I think we've seen it in Q3. I expect that to continue. That's part of our excitement.
We talked about a six-year process of building capacity and inventory ahead of demand. We finally are getting there. Yes, there is a little bit of leftover in Lehi to complete LFAB too, but we are done on the analog side. We have the inventory level. It's time to go back and grow our free cash flow per share, both from an inventory buildup perspective but also from a CapEx perspective, and we are very excited to be in that finally in that point.
Yeah. I wanted to ask about CapEx and also free cash flow. So CapEx this year is $5 billion. You said that if revenue's between $20 billion and $22 billion, that you'd spend between $2 billion and $3 billion next year. I mean, it seems like revenue probably is on the, if anything, at the, you know, low end of that number. I think you said you can't spend less than $2 billion because of what's going on in LFAB and elsewhere. So you're gonna drop at least $2.5 billion, maybe even $3 billion to cash flow next year because CapEx is coming down. So can you just talk about how you think about CapEx and it and then into 2027? Actually, I get asked this a lot.
It would seem like, given all your inventory and all your capacity, that you could even take CapEx down even in 2027, even if there was a modest recovery.
Yeah. I think, I think let me just elaborate on what you said. You're right. We, we do, we talked about a six-year investment from 2021, 2021 to 2026, and we are five years in now. We have actually executed ahead of schedule on Sherman. So if you look at our analog business, both capacity is built, clean room of Sherman too is built. We've built the right level of inventory. We are ready. On the Lehi side, we are in the process of, of getting the shell and the clean room built in LFAB2 . We wanna be in that position to never repeat the, the issue we had in the previous cycle where we didn't have brick and mortar ready for demand. So we are making progress. I think that we will finish it somewhere in 2026.
That's the main, the largest ticket item, or the big ticket item we have in 2026. This would raise our CapEx somewhere around $2 billion to $3 billion. I can't be talking about precision, but we'll try to be a little bit more, you know, give more information during the capital management call in February. But your estimation is true. There is a $2.5 billion to $3 billion CapEx opportunity over there because we are where we are. If you look at 2027, now it depends on revenue. We are gonna be in this, what we call phase three of modular capacity. Clean room is built, fabs are qualified, customers are taking the new parts. Now it depends on what revenue wants to do. If revenue wants to grow rapidly, we'll be ready.
But if it doesn't need to and we have enough capacity, that capacity comes into kind of what we call maintenance mode. And if people have asked me what does it mean maintenance, but you know, maybe 4% of revenue or something like that is where we used to be when we had the clean room ready at the time when we were kind of ramping RFAB1. I think it's a good estimate. And we can provide some more color in the future, but that's what I have in mind, not to give a precise number. The last point I would make, and we don't mention that many times, but we also care about our assembly and test. We had a mission to internalize our capacity also on the backend side. On the fabs, they're already at about 90%.
By the end of the decade, we'll be at 95%. On the offshore, we started at below 60%. And I think we'll finish the year close to 80%. And again, with the aspiration to finish the decade at above 90%. And this is something that we are still continuing to do. We are ramping our Malacca two factory. I was just there a couple of weeks ago. Very good progress, internalizing part that used to be built in offshore many, many times in China, into our what I call dependable capacity footprint, both in Malaysia and the Philippines. So that's also something that happens in 2026 but can take a step backwards in 2027.
Great. So I'm sure you get asked this also, and this question is more in relation to China, the effort to add a lot of manufacturing capacity in the U.S. and have dependable manufacturing capacity that might in China potentially could cut both ways because some of the Chinese customers are trying to do to the non-Chinese suppliers what the American customers are, you know, not wanting to buy from China given the geopolitical tension. So do you think that having all this U.S. manufacturing capacity does it put you at a disadvantage in China for the products sold domestically in China?
I don't think so because we have options. You know, China is about 20% of our market and also about 20% of our business. And can I supply that 20% with the non-U.S. capacity? Of course I can. TI is not only having manufacturing in the U.S. We have manufacturing in China. We have manufacturing in Europe. We have manufacturing in Japan. And assuming China doesn't wanna take manufacturing only from China, we have the answer. And this is currently what customers are asking us to have. In case something happens, I wanna have a non-U.S. supply, and we are very ready. On top of it, we are also strengthening our muscle of bringing the right technology into China. So in that sense, our China customers are very happy with TI's ability to build for them even locally, in China.
We've been tested over the last couple of quarters. You know, there were periods of tariffs in China, and we've done very well. We could divert our wafers into the right supply chain. Our customers felt very secure, and I'm seeing the design-in momentum continuing to grow in China based on our ability not only to have a dependable, dependable capacity, but also they like our portfolio. They like our level of service. They like the urgency of the company. This is why I see China still as an opportunity rather than a risk. Of course, I don't underestimate the geopolitical tensions that seem to grow over the years. But China will remain an important market not only next year, but I think for the foreseeable future.
If I think about the next five and 10 years, is there gonna be some decoupling between some more decoupling between the U.S. and China? It's probably a good estimate to have. But will we be able to ship, you know, analog and embedded parts, into the China customer's footprint? I think yes. And both things can live in together. So that's where TI is at right now. We see momentum in China. Our business in China is growing nicely this year. It's at a low 30% year to date. And I think there is more opportunity moving forward.
Yeah. I mean, if they were gonna displace you, they've had plenty of opportunity to do that. And.
And that's not new. We've been competing with the local competition since 2018. So this is something that we've talked about way before it was, quote unquote, in the news. We took a decision. I think it's a strategic one, that we wanna be very competitive there. And the hard part was not the capacity of manufacturing. The hard part was changing the culture of the company. Can you keep up on the fast-moving treadmill? Can you react at very fast time to the demand? Can you fulfill the market price? You know, the market price is not set by TI. It's set by the competition. Can you have the right cost structure to make good money when you fulfill these sockets? So I think the answer to all this, we've proven that it's yes.
I think also, to be fair, it's helping us across the globe. China is not the only market. And when you get a team that is fast-moving, that cares about cost, that cares about urgency, that cares about high customer service, it pays dividends across, you know, other geographies. I think that's very important to remember.
Can I just ask about share? So, I ask you this a lot. So if I look at your share of the analog market, it peaked at 19%, almost pushing 20 actually. And you've lost roughly from peak to trough, you lost about 400 basis points. You gained about 150 basis points back this year. So you've gained some of that back, but you haven't even gained half of that back. And the share seems to have stalled at least like last quarter if you look at the SIA data. So, are you focusing on regaining that share? Is that a key metric that you're focusing all the salespeople on? And are you compensating the, you know, salesforce on gaining that share back?
Maybe you, if you can go back and deconstruct, why did you lose that share in the first place?
Yeah. First, a quick answer to your last two questions is yes and yes. Of course, we measure it. Of course, and again, there are many ways to measure it. We just collect all our competitors and see how we do versus them. And on the analog side, as you mentioned, I think we have some momentum, but we have a lot of room to grow. And that's how we measure the team. That's how we compensate the team. So definitely yes. Now, to me, and it's part of the learnings that we've discovered back in 2020 that we were not well prepared for the opportunity. If you think about it, and we had the right part. We were on the board. We had a socket. Customers wanted us to ship more.
And I think everybody struggled during that time, but I think TI was struggling more simply because our clean room got filled up and the lead time to build new capacity was like three years. So that really drove our decision to make this investment cycle in the last five years and to do it. And as I said, we have reached a huge milestone of lots of execution, lots of investment. But when I look at the end of Q3 and where we are, I'm very pleased. Now, in terms of how you gain it back, it's a combination of what the market wants to do and also your execution. So I think the market right now is not testing the suppliers. It's a very gradual recovery.
Every time there is an accident at one, quote unquote, accident at one customer or the other, TI always is able to show up and fulfill some of the gaps. I would love to have more opportunities as the market evolves, but I think they'll come. We see secular growth in our market. We see that TI has invested through the cycle. I think this is a unique position that we have, and I'm convinced that the share of gain will continue over the coming years and, of course, we'll continue to monitor it. Now, the other part of our portfolio is the embedded side. This is where the share loss is not related to the supply. It's related to our portfolio that weakened, and we saw a gradual share loss since 2017. That's more of a strategic issue.
Mm-hmm.
That is not easy to fix. I think we are in the process of fixing it. We are starting to see early results of stabilization of our share, but it's still like we lost almost half of it since 2017. We have the opportunity, in the coming years, and I'm talking about the second half of the decade, to start to make an embedded, I would say, momentum build up, towards the second half of the decade. And the reason I say that we have built a new portfolio, it's starting to get designed in. We see all our internal metrics showing the right momentum. And I think we will continue to accelerate that, in the second half of the decade. The team has to prove it. That's how they are gonna measure them.
We are gonna stay persistent on our mission to rebuild our embedded business. But I think we have. I give the team a good chance to make it happen.
Can you just sort of deconstruct for people why, like, why did you lose that share in Embedded?
Look, I think we talked about it so many times, but I think it's a combination of just where do you wanna spend your R&D or what portfolio you wanna build? I think we had a lot of legacy business coming from our old wireless days that we did big processor, big chips, moving from smartphones, for example, into vehicles. That was not sustainable for TI. So it drove some of the momentum in the previous decade. I think in 2019, we said, look, we want to have a much higher quality type of business in embedded. Let's put our R&D where we care about low power MCUs, wireless connectivity, sensing solutions like radar and others. Instead of big SoCs for processor, maybe focus on low power processors and everything around DSP, drive motors, convert power.
That's kind of where TI's competitive advantages come into play. You can build it internally right now in our new fab that we've acquired in 2021. You can build a broad portfolio so you can solve many problems on the board. Your channel advantage now is coming into play because now you can touch many, many customers to grow your business. And that's how you get to this footprint of longevity and diversity. The good news that we are well into it, you can call it bad news, but I think that's part of the quality of the market. It takes time. This is where these are smaller sockets. You won't see it ripping back up, but it will stay once it gets there. I think it's a very sticky type of business. So, that's what we've done in embedded.
I'm pleased with the execution, but the team has a very high bar to translate into share gains. So far, okay, you've stabilized your share. Thank you. Now you have to grow it. And that's the mission of the team for 2026 and beyond.
And so just back to your point on share on the analog side, which I think the point you make is great, which I hadn't thought of, is that really it's the tension in the marketplace. You actually need the tension in the marketplace for you to actually gain the share.
No, I don't think. I think you will see it quicker when there is tension. I am very convinced that generation to generation we are getting the footprint. But right now, let's take, you know, factory automation. And this is our largest sector in industrial. I know that we are on the boards waiting to be shipped to customers that are going to have automated manufacturing somewhere. But I do see our customers, you know, or the customers of our customers hesitating whether to make an investment right now. And maybe the AI data center is an outlier, but everywhere else in just, you know, our people spending a lot of money building new manufacturing footprint, the grid, solar. We see customers are a little bit in a wait-and-see mode, Tim.
And this is something that we've noticed. I think that's what industrial needs in order to go back to trend and beyond. And when it does, we'll see the momentum. So I'm very optimistic about the long term. I think the market so far has moved slowly, and this is why you see the share gains coming in, but not at a rapid pace. If we need to go in and show the muscle of TI, we'll be there.
I wanna ask you about M&A. Everyone always says to me, every investor says, well, if TI was ever gonna do a deal and buy something, now's the time to do it because you have all this fab capacity. And, and so I, my question is, what are the qualifications when you look at doing deals? Everyone seems to think it would be an embedded, but it seems like because of the IP and the process technology, it probably wouldn't be an embedded if you were gonna do something. So can you just talk about like the, the boxes that any deal, if you were to do one, would have to check?
Yeah. I think the most important part is, does it have a good strategic fit for the company? And we always talk about our competitive advantages not as a slogan. We think that way. So first, can I build that portfolio internally? So of course, when you look at a company that maybe builds at a foundry, you bring it in, you have the COGS synergy that you would not have if you wouldn't build it in TI, right? So that's always part of the criteria. Does it have a broad portfolio? It's very, very important because broad portfolio and a broad customer base is very important to us.
This is where our channel advantage, whether it's ti.com or our sales channel, is very effective because once I already call on a customer, if I can expand my portfolio by expanding it maybe inorganically, that falls into the fall through is very nice. And last but not least, we do like the business to have this, I call it diversity and longevity, which is kind of the stickiness of it. So is it, mainly on industrial and automotive rather than on PE? That would be a criteria.
Now, regarding the portfolio, I think today we now build embedded also inside the company, but I think it has to be a good fit in the sense of it's, it should be more analog mixed signal type of thing other than big MCUs or very high level of concentration on process or that's not where my head is. My head is more on this analog mixed signal side. And we always look at opportunities. And of course, if something comes up, you'll hear us. But we continue to be very focused on organic growth. I think we have an opportunity to continue to expand our portfolio and grow share organically, while looking always at opportunities externally.
Great. And then, I just wanted to ask you about share repo. They've sort of slowed mid-year. They were on a very strong pace in Q4 of 2024 and the first quarter of 2025, and they slowed a bit, recently. However, when I just adjust your CapEx, and I assume you're gonna drop at least $2.5 billion to cash flow, possibly $3 billion as the CapEx comes down. I mean, you're basically annualizing $750 million in cash right now, and the stock is, I mean, it's, it's a, it's at a very low multiple by historic standards at, at, at currently annualized cash flow if you adjust for the CapEx. So it would seem to me that now's the time to lean in and really buy back more stock.
I know your, you know, long-term commitment to, you know, returning 100% of the cash flow, but it seems like now's the time to be maybe a little bit opportunistic.
I look, Tim. You said it yourself. I mean, first, let's build the free cash flow muscle, okay, or rebuild it. And it's really gonna come by executing. And we are executing. So getting through this investment cycle and then whatever revenue wants to win, you think your assessment is right. Probability of below 22 is high next year. So that is a lot of gain on free cash flow, $2 billion-$3 billion, as you mentioned. I will say also that, we've been aggressive, right? We've been returning, at above 100% where we saw an opportunity. We continue to do it even in 2025. I am excited about next year because we are starting to see the fall through. I think we have more opportunity as I think about the trend line of free cash flow per share.
And we showed it over, I think, 20 years in our capital management call. Are we getting there next year? Yes. Do we have a chance to have a new record next year? Yes. And in case we don't, it's because we see such a huge demand that we have to put more CapEx in, right? So very excited about that. It's a unique point. You know, it's been a long journey to run this company in the last five years with Rich through the cycle, keeping a steady hand on our investment. I think it will pay dividends, whatever the markets want to do next year. So in that sense, you're right. Your math is not wrong. And I think a lot of opportunity ahead on capital allocation, whether it's buybacks, whether it's whatever else.
So you think there's a scenario where next year could be a record free cash flow year for the company?
I mean, we've put a target there of eight to 12, right? I think the 12 is a low probability because we needed $26 billion for that. But you know, that eight doesn't sound crazy.
Yeah.
Look at the numbers today. You've done the math of the CapEx. And yeah, we are not gonna force it. I said it, I think in the previous conference. It's not you shall be eight next year, but is there a good probability? Of course there is. Now, I assume there's gonna be some more growth in the market. That thing can change every quarter. I've given up on trying to call the shape of the recovery. We're just gonna recover, report how it went every quarter. But I think this market is still trending below trend line. There is a good chance. And you've heard some of the peers today. I don't know what they've said, but there is good chance for this market to continue to recover, and establish new highs. So we'll see.
Great. Well, we've run out of time. Thank you, Haviv.
Thank you.