Texas Instruments Incorporated (TXN) Earnings Call Transcript & Summary

All right. Welcome back, everybody. I'm Joe Moore, Morgan Stanley semiconductor team. We're very happy to have with us today from the management team of Texas Instruments, Rafael Lizardi, Senior VP and CFO; and Mike Beckman, Director of IR. Thanks, guys, for joining.

Good afternoon. Happy to be here.

Thank you. So your Capital Management Day just happened a couple of weeks ago, and you basically remained on message with the objectives. But can you just give us some of the few key points from that and how you guys are thinking about capital management going forward.

Yes. No, we are holding a steady hand, consistent message, as we've had the last few years. We're going to maintain levels of CapEx to make a significant investment in manufacturing and technology, where we continue ramping our 300-millimeter wafer fabs in Texas and in Utah. And this year, we're going to spend another $5 billion. Next year, it's somewhere between $2 million and $5 million, depending on the opportunity. But in any market scenario, we envision getting back to trend line on our free cash flow per share growth, which is what we think, over the long term, drives value for investors.

Great. Obviously, domestic internal capacity has been a big part of the strategy for the last several years. It seems like that's probably helping you sleep at night a little bit these days, given all of the uncertainty around it. Can you talk generally about the importance of having that U.S. footprint, the benefit that you see in kind of a geopolitically challenging world?

Yes. No, especially in the current environment and geopolitical tension, our investment, we think, is going to become even more valuable. We are building a complex of 4 wafer fabs in Sherman, Texas. That's in addition to the 2 wafer fabs that we have in Richardson. All of those are 300-millimeter wafer fabs. So the largest diameter on the wafer, and that translates into scale where the cost per unit dropped significantly versus older generations. In addition to that, in Utah -- in Lehi, Utah, we bought a factory there from Micron that we turned into an Analog and Embedded factory. And now we're expanding a second factory right next to it, which is going to be a massive 3-level 300-millimeter factory. So that puts us in a really great position to build there. Of course, we're doing that with the benefit of the ITC, the investment tax credit, and some CHIPS grant funds that we were awarded last year. So that, in addition to being 300-millimeter grade cost per unit scalable footprint is being helped -- funded by the ITC and the grant.

Very helpful. Some of your competitors when we've talked about this have sort of talked about a little bit more matching the manufacturing in the region of consumption. So trying to build China for China, Europe for Europe, things like that, whereas you guys have much more of a domestic strategy, as you said. Can you talk about the way you're thinking about that and the reason that you've chosen to do it that way?

Yes. We think that to win in -- winning in China is very important. We're all in as far as selling in China. We -- about 20% of our revenue is based from customers that are headquartered there. And our CEO was just there last week, and he came back super excited of all the wins that we're getting there. At the end of the day, Chinese customers are just like any other customer. They want the best product so that -- to be able to sell the best product to their customers. So in order to do that, they need the best parts. So we have the best parts. We have the broadest portfolio. We did have, as I talked about earlier, the manufacturing and technology footprint that gives us the best cost which then that allows us to go and price in the -- even in the more sensitive end markets, price-sensitive end markets. So we feel that we're in a really good position to continue gaining share in China.

Appreciate that. So a lot of this is framed to me by the shortages that we saw '21, '22, '23, where people realized that these geopolitical strains can put a lot of pressure on the supply chain. But we've now obviously gone from a shortage environment to a little bit of an oversupplied environment. Does that change customer priorities? How aware are your customers of what you're doing? Do your customers still ask you the question, even if they might be drawing inventory down to levels that are lean, they're still focused on where your supply chain is and where their supply is going to be?

Yes. My sense is CEOs of our customers, senior leaders in -- of our customers, they value, they appreciate what we're doing. And that is a key motivator to design us in, in their parts -- in the products. But the designer, the purchasing manager, they care more about the tactical things. So we have to be good at the tactics at having the right part with the right features at the right price and also the strategic view, having that the manufacturing footprint that I was talking about earlier. My sense is that we're in an environment now where many of our competitors are retrenching, are cutting inventory, cutting OpEx, cutting factories, shutting down factories. We're doing the opposite. We're investing, we're building inventory and we're getting ready for the upturn that at some point is going to happen. This -- we've been in this industry long enough that we know that the sun eventually rises again. And we think that, that's going to happen. And when that happens, we're going to be ready.

Yes, that makes sense. As I think about 2026, you've talked about on the Capital Management Day and other times, CapEx of $2 billion to $6 billion -- $2 billion to $5 billion, I think. I think some people were thinking of 2026 as a year where we start to see cash flow go above reported earnings, but it seems like at the low point of that, you're right around your depreciation. So why couldn't 2026 be [ a year ] where CapEx falls further than that and you start to see that free cash flow going above reported?

Yes. So just to recap, this year, '25, we'll do about $5 billion in CapEx. For next year, as Joe alluded to, we have been saying that we'll have a range of $2 billion to $5 billion. And the $2 billion is the floor. And the reason for that is we need to -- that $2 billion is to put a lot of brick and mortar and qualifications in those factories that I talked about earlier, that are not going to give us any revenue on day 1 or even on day 2. But you need to have them so that once you have that, then you can go in a modular -- on a modular incremental capacity addition mode. So that's why we need that there. Penny-wise, pound-foolish. It would be any wise penny-wise, pound-foolish if we were to forego some of those investments just to get the CapEx a little lower.

Yes. But there could come a year where your CapEx starts to fall [ a little lower ].

I'm glad you brought that up. That -- '26, the floor is $2 billion. There's no floor in '27. So if there's a persistently weak market, CapEx in 2027 can go pretty low.

And then you mentioned we're going to have an upturn. We've all been doing this. I've been doing this a long time, too. There will be an upturn. I guess, is part of what you're doing, though, sort of lessen some of that cyclicality? Because you're going to have a lot more capacity. People will know that you'll be able to deliver. You may not have shortages in the way that you did 2 or 3 years ago. And maybe that dampens a little bit some of the cyclicality that we see.

I'm hesitant to say that TI single handle will dampen the industry cyclicality. But to your point, we're going to have plenty of capacity, plenty of inventory, systems in place that will make that inventory more -- allow us to use it more efficiently where it's most needed. So the end result of that should be serving our customers a lot better so that they don't have to get to a position where they're place -- where supply is tight -- demand and supply are tight and then they're having to double and triple order in order to get what they want. So potentially, we could alleviate acyclicality. I'm more on gaining market share. When we're -- when that upturn comes and our competitors are short of inventory, short of capacity and we have plenty of them, then we'll be in a really good position to gain share.

Yes. Okay. That makes sense. So I guess where this all comes to fruition is kind of gross margin. You said at the midyear capital management day last year that under the sort of low-end scenario, you would be at 60% plus gross margin, you're about 55% now. So obviously, right now, you're at 55%, that's a little bit depressed because of the lower utilization. Do you still feel like 60% is kind of the right long term...

Let me correct -- I think what I heard you say. What we said of the 4 scenarios, the top 3 would be 60 plus. The bottom scenario, which was $20 billion of revenue, would be below 60%. Slightly below 60% gross margin. So that was on the 4 scenarios of $20 billion, $22 billion, $24 billion and $26 billion for 2026.

Okay. But that's still -- those are still the right numbers.

Yes, those are still the right numbers. And the way you get there, you don't have to choose our revenue. You can pick whatever revenue you believe, just falling through at 75% to 85%. And for the next few years, just 85%. And then subtract the increase in depreciation, which I've given on the last earnings calls and the last capital management calls, and you can model your own gross margin. That's about what you should get under those revenue scenarios.

Okay. And then Dave and Mike were helping me understand how you manage the underutilization sort of expensing in a quarter. Where like last quarter, you saw lower -- somewhat lower utilization, you take that pain kind of in the quarter where the utilization is low rather than running at FIFO, right? So you..

Correct. That's what underutilization is. It is fixed costs that are not attributable to wafers that you're running, so then you let it fall through in the quarter.

And essentially charged at the quarter where the wafers are in the fab, not when...

Well, you don't have those wafers. So you charge that cost when the wafers are not in the...

But you have over 200 days of inventory. So if you ran it through FIFO, it would take you a year to see the impact of lower utilization, right?

And in fact, it would not be GAAP. So we couldn't do it that way. So the way you do it by -- for GAAP, let's say your -- what you call normal utilization is 90%. If you're running at 90% or higher, everything that you spend you inventory. If you're running at 80%, you take that 10% delta of 90% to 80% and then you apply it to all your fixed costs, depreciation, some of your people costs, some of your electricity. And then you take that cost and you don't put it in inventory, you let it fall through in the P&L that quarter.

Okay. So last quarter, you had lower utilization and your inventory still kind of went up.

We built inventory on top of having lower utilization.

Yes. Yes. Okay. So the implicit -- the idea that you'll be back at that utilization requires a revenue recovery, I guess, obviously.

That's right. Yes. We need revenue. Okay.

Okay. Great. Maybe we could talk about some of the markets. In industrial, one interesting point from your Capital Management Day was you talked about in $700 to $1,000 of contents in industrial robots. How are you thinking about that? Like I feel like TI kind of doesn't talk about those speculative markets maybe as much, but you have a lot of opportunity if those markets take off. And when you make these investments into autonomous driving and things like that. If they take a long time to get into cars, they may show up other places first. Can you just talk about how you guys think about those verticals?

Maybe I'll add and Rafael will jump in. But it is a great opportunity. I think humanoid robots and just robotics in general, the content that you see in those is incredibly rich. I mean, you've got for every joint, if you think of a humanoid robot, there's a motor controller, there's an encoder. There's radar systems, there's camera systems, there's interface systems inside of those. There's just a lot of content opportunity in those. You talked about the size and dollars. Now I'm not going to speculate on how quickly that market will grow or what that's going to look like in 5 years or 10 years, but the idea is we want to make sure we've got the best portfolio to serve it, and it's a great opportunity for TI. So if it does well, we will benefit. But it's not going to be the only place that we have design in. But it's a great opportunity. And I think robotics in general is an area that is small and growing for us. I think in general, in the industry, it is and it has a good potential to be a fast grower over time.

And how do you guys feel about the industrial market at this point? We've been undergoing one of the tougher corrections that I've ever seen in terms of the drawdown. It feels like we have to be shipping below consumption? Do you guys feel that way? And how do you see the market [ pretty much ]?

It is a great market. I mean it is -- it's a diverse market, as we just mentioned, it's a very long-lived market. When you win a socket in industrial, in many cases, these are multi-decade long socket wins. And so it is a good market to be in. Obviously, we're going through an inventory correction in this industry and as Rafael pointed out, we want to be ready because it will recover, and we want to make sure we have the right inventory position, design and momentum, make sure we have the right portfolio to have it. And I'll tell you, having a catalog or general purpose as well as an application-specific standard product or ASSP portfolio, having both of those things make it really a compelling market to be a part of because you really can get a lot of socket opportunity with those 2 types of products.

And so when you talk about a market like robotics or something new, do you have organizations that are framed around that? Do you have vertical exposures with it? I know it's a catalog focus, but do you use steering those catalog focuses into sort of verticals?

We do. So we have -- first of all, we have a product set of teams that are structured and do product development by product family types. Kind of if you go to our website and you look at the product families, that's very similar to how you'll see our product teams that are organized. We also have a team -- a marketing team or a systems engineering and marketing team that knows the applications really well and helps us make sure that as there's emerging opportunities such as robotics or industrial automation or medical opportunities, they are very deeply ingrained in what those applications are going to need and are evolving to. So they help find intersection points and the products we already have, but also help inform the road map to make sure that as we're developing products, we have them. Often a good general purpose or application-specific standard product can serve multiple markets, multiple end equipment. And so they really help make sure that we're plugging those sockets in as many of those as possible and also informing the road maps to make sure that if it needs to have -- in the embedded space, for instance, if it needs to have AI accelerators built into the chips that we have those put in there. So they help inform the road maps.

Great. And then also thinking about automotive in the same context, I mean, there's such a wide spectrum of analog componentry that goes into a car. A lot of the dollars is in more commodity power discretes, which I know you're not participating in. But a lot of the battery management is very expense high and high-capability stuff. How do you think about verticalizing that market and deciding the right way to -- even if it's a catalog approach, the right places to align those resources?

Yes. I think it goes -- it's similar to industrial in making sure that you have a broad portfolio that can address multiple things in a vehicle. And as you may be aware, we look at a vehicle, there's more than just 1 socket in the vehicle. As you know, there's -- we split it into 5 sectors, ADAS, infotainment, we've got body electronics, lighting, powertrain and powertrain EV is the kind of 5 sectors that sit inside of there. And every one of those sectors has growth opportunities within it. And in fact, there's a lot of folks talk about the EV opportunity, the battery electric vehicle opportunity. But even in internal combustion engine vehicles, the ADAS systems are developing. The infotainment systems are developing. The lighting systems are more dense than they were even 5 years ago. So what does that mean we have to do from a road map perspective and a product portfolio perspective? It means you do have to have those -- I use the word general purpose products first because a lot of times when someone is iterating design and doing innovation, they're going to start with using those components to put together a new development. Over time, that sometimes evolves into application-specific standard products. So having both of those becomes really important. So you have to have a portfolio that's in a road map that's structured to head in that direction than we have. I think for gosh, more than a decade, we've had most of our product lines building products for automotive, but also industrial, and we try to do as much as we can to make sure they can serve more markets, not just 1 single end equipment.

Great. And so one of the things that stood out from the Capital Management Day was you mentioned having a cost structure to be competitive in China, and you talked about the importance of China. In automotive, in particular, it seems very clear that a lot of the center of innovation has migrated to China at this point, and you said you're growing there. So how do those 2 things come together? How -- you're going to be in that market. Do you see it as a market that's more price competitive? Can you dissuade local competition with pricing? Just -- how do you think about China as being distinct from the rest of the world?

Yes, we're doing very well in China. As I think I mentioned earlier, our CEO was there last week meeting with customers and came out super excited. We're gaining share. Our manufacturing footprint, 300-millimeter with the ITC benefits and everything is putting us in a good -- in a really good competitive position. We price based on what the market demands. We don't set the price, the market sets the price and we have -- based on our strategy and our manufacturing footprint, we have the capability to get to the price that needs to -- that we need to get to. But price is only one factor that customers care about. Yes, it's not the top 3, right? So there's a lot of other things where our technology comes into play and in the broad portfolio and our quality and our sales force that puts us in a -- and those are our competitive advantages and puts us in a position to do well in China and everywhere else.

So you'll be price competitive where you need to be, but [ it's not ] the first thing you're leading with.

Correct. Correct.

Yes. Okay. And what is the state of Chinese competitors? Do you expect -- Silergy was here the other day talking about making progress in the analog space. I don't think they're really automotive grade the way we would consider it at this point. But what is the status of that? And you obviously have dealt with competition in all markets at all times. But is it different in a market where maybe a TI goes to the local...

My sense is there's clearly some pressure, some incentives for local Chinese companies, our customers to use local competitors. But right now, just to give you the big picture, out of the entire market in China, the consumption in China semiconductors, about 15% of that is coming from local competitors, okay, about 15% is coming from Texas Instruments and the other 70% is coming from other Western suppliers and Japanese suppliers. So we're competing with the local players, but we're also competing with that 70%. And that 70%, they're not all in on China, is my sense. And you've heard some comments from some of them, even here over the last 1.5 days. So we compete well with the local suppliers, but we compete even better with the Western suppliers that are not putting -- they're not as focused in China as we are.

And if you see China being more successful in the export market, whether that's in regions where there are no tariff concerns or even migrating their manufacturing footprint into areas that do, it seems like that's a good thing for TI?

That is a good thing. I'm glad you brought that up, as much -- about 20% of our revenue is for customers that are headquartered in China, but a good chunk of that 20% is for export. So it's not for consuming China, but it's put in some car or in some product that then is exported out of China. And there, the Chinese have to be or just like any customer, globally competitive. So the best product is the one that's going to win. And that's where we feel confident that we have some of the best products.

Okay. Great. I wanted to shift to Microcontrollers and the Embedded business. You guys have made no secret of kind of underperforming a little bit in that business. But some of that is repurposing and reorienting around a catalog MCU. Can you talk about the strategy there and what you see is happening?

Yes, I'll be happy to do that. So if you look at the Embedded business today and over the last, I'd say, 5 years or so since Amichai joined and started leading that team and really making sure that it leverages what TI is good at. The competitive advantage that we have of manufacturing technology, broad portfolio, reach of channel diversity, longevity. If you look at where Embedded was probably a decade ago, couldn't really benefit from those things as much. A lot of it was built externally. The portfolio was smaller, more large, complex SoC-type developments. The reach of channel didn't benefit as much because we're talking about a handful of customers versus a huge number of customers. And so what advantage -- how do those competitive advantages line up to help it in that scenario? And what Amichai has really done is build out -- first of all, bring as much internal as possible. And Lehi helps with that. It runs process technologies that both Embedded and Analog and use, but Embedded certainly benefits probably the most from that. So you have a benefit in internal manufacturing with Lehi. You have a broader portfolio. I encourage anyone here who has a moment to go to the website, look at Microcontrollers or processors or DSPs or wireless connectivity or radar chips on our website, look at the product folder, look at the number of new next to product names. There's just a tremendous amount of releases gone out in the last couple of years. And it's not just 1 device. It's a family of devices. And so to be successful in embedded, you have to have families of products. And so what we're doing is building out portfolio that then wins at many large multitude of diverse sockets versus 1 single large socket. And if you look at what that business starts to look like, it starts to look like a business that benefits from TI's advantage. That's the restructuring that we've really worked on over the last several years. And I think we're seeing really good progress in that, and it makes it a very compelling opportunity for us.

Yes. Great. And that's at margins that are similar to the rest of the business? Gross margin?

I'd say that in terms of gross margin, the point we really think about is, is this a business that's going to help us grow free cash flow per share faster for TI? Embedded does that for us. It is a business that not only now benefits from those advantages, it also strengthens those advantages.

Short term, Embedded is taking a bit of a hit because the new factory that we have in Utah is disproportionately an embedded factory. Because it's underutilized, that's taking a hand -- has given it a hit on gross margins and operating margins. But just like now it's a headwind in the future as we load it, it will be a tailwind. So Embedded, we'll get disproportionate benefit as we move more external load-ins internally.

Okay. Okay. Very helpful. And then I guess, consumer. Consumer has gotten kind of big personal electronics, largely because everything else has been weaker. But that tends to be a business that you guys look at a little bit more opportunistically. Can you just talk about that -- the personal electronics business?

Yes, I think that's the right way to look at it. And it's a market where a lot of the end equipment is in -- handsets and laptops and gaming and headsets and peripherals like that are kind of in a saturation replacement mode. Look you find another planet of people to sell cellphones to, you're kind of now in refresh. And so there's opportunities there, and there's good places for us to participate, but we want to be proportional to the opportunity. And when we look at the markets that we play in, automotive, industrial, the opportunities in data center. Those are places that we want to make sure that we have R&D biased toward because of the opportunity there. But it doesn't mean we don't want to ever participate. It means there are still opportunities we want to have, but we want to have them right-sized to the opportunity that we see.

Got it. Okay. And then just last question for me, I'll open it to the audience. But, do you guys think about market share in Analog? Do you -- because I know there's been some fluctuations and shortages played into it and how much who had inventory who didn't, at various points played into it. I know there's a view that there's a mean reversion kind of share gain coming for you guys. Just do you think about that at all? And just -- or do you look at it more opportunity by opportunity and the market share will kind of fall where it falls?

No, we look at it both ways, and we don't rest on our laurels just expecting some regression. I think what you're referring to on that front, some of that is fair, which is when you hear customers -- competitors talk about long-term service agreements and CSPs and that sort of thing, and there was also pricing. There's a lot of distortions in the last few years that you have to wait until that kind of settles and moves out before you make a real assessment, a more accurate assessment of market share movements. But that is -- those are some kind of macro factors. But in addition to that, we go customer by customer, socket-by-socket and fight those out to make sure that we get our fair share of those.

But if you look at like the growth versus like, let's say, the Semiconductor Industry Association analog segment, there's a lot of stuff in there that where you don't want to participate, right, low-margin commodity stuff. Is high-performance analog as a category, still outgrowing that broader group, do you think?

I don't have the data in front of me to compare the two, but I would say that in terms of where we see the opportunity over time being the greatest for us, it's going to be in -- and when I say high-performance analog, that's a very important piece. But power management, whether it's -- and we like to use the term general purpose and application-specific standard product because that kind of covers the range of the spectrum that we like to cover. And when I say general purpose, that's not what you describe as commodity [ part ] products. It's really these are products that are -- they're everywhere, they're pervasive. They're not interchangeable. But over time, we can definitely design in and gain traction. On the application-specific side, despite what you're thinking of more the high performance -- and we have some of those really, really high-performance devices too in the portfolio. We want to have both of those things. And so having the combination of those, we think, is going to help us be able to grow faster.

Great. Let's open it up to the room. If there's any questions here. There's one in the front here. .

I have 2 questions. So like last year, like NVIDIA announced like Texas Instruments has been the strategic supplier or partnership with TI. So -- but like I say, from your last quarter report for the enterprise system, which related to the data centers, actually, you have low single-digit downside. So I'm used like also, at the same time, Infineon has been a long strategic supplier of NVIDIA, but they still have big business from NVIDIA. So I just say like what's the exposure of TI business to NVIDIA's AI systems? The first question. The second question is like also in your last quarter report, you're saying you have like $1.5 billion direct funding from CHIPS Act. But like this week it was announced like this may be canceled. And also the time line, you mentioned cost is $1.5 billion is like 2026, next year. So that means like where it impacted your plan to get funding?

Maybe I'll cover the first one, and you guys are the second one. So I won't comment on a specific customer or a specific socket, but I will say about data center in general, is it is a good opportunity for us. And it's -- if you look at where the opportunity is, it's in power management is a big part of that, but it's beyond that. If you look at the content opportunity inside the data center, there's quite a bit there. And today, we've got a substantial business in data center. If you look at the size of our enterprise business, it's large. And as we look at where we are allocating our R&D towards, we want to make sure that we are a strong player there. The new process nodes that we're building in SM1 are going to play an important role in that. And so -- it is this area -- if you heard me mention areas where we bias R&D, industrial, automotive and with the focus as well in enterprise and data center.

And on your last question, so last year, in late December, we announced that we were awarded $1.6 billion of direct funding from the CHIPS Act, their so-called grants. So that was an agreement we signed with the government, and we are not aware of any -- we have not been told, have not been informed of any changes to that. There are several milestones that we need to complete, and we are in the process of working towards those milestones in the factories in Sherman and the factory -- in the second factor in Lehi. Let me stress, as we've always said, that the bigger portion of benefit from the CHIPS Act legislation is the ITC, the investment tax credit. That we expect to get benefits of $6 billion to $8 billion over the course of the next 10 years or so. And we've already received cash benefits of about $600 million and expect $700 million to about $1 billion this year.

I guess on the topic of kind of all the geopolitical questions that are happening right now, it's got to be very difficult to run some of these businesses when there's tariffs that are maybe coming, going. We don't know if they're going to be there. Do you see any different behavior? Do you think people are pulling things in to deal with tariffs? Just anything that's different because of all of this.

Yes. It is difficult and uncertain, but fortunately for us at least, we don't expect any meaningful, significant direct impact on tariffs. Of course, indirect, if the world's GDP is affected, then we're going to be affected. But direct tariffs because we have such a diverse footprint and supply chain, particularly with the assembly test operations, we can reroute supply chains so that not to minimize the impact on tariffs and ourselves and our customers. So for example, while we have an assembly test operation in China, what we can do is the products built in China just not ship those to the United States. And instead, we have plenty of IT capacity in Malaysia, in the Philippines, in Mexico and other places where we can then reroute those to customers in the United States to minimize the tariffs there.

Great. Thank you for that question? Mic is coming.

I was wondering if you could help us understand or size the amount of application-specific revenues compared to catalog, either industrial, automotive or both?

It's -- there's not a clear cut line to say, okay, this is an ASSP and this is a general purpose product. They're roughly proportional in size. So it's not like one is 90%, the other one is 10% of the business. I'd say, and again, it's a spectrum, very clear, it's not like it's -- there's devices in the middle where probably 1 person would say it's an ASSP and someone else would call it general purpose. That's why I'm very careful. It's not like there's a clear way you can always say whether one is or not. But they're similar in size, best way to describe it, inside of our total revenue.

Other questions? Okay, we'll wrap it up there. Rafael. Thanks so much.

Thank you.

Thanks so much.