Analog Devices, Inc. (ADI) Earnings Call Transcript & Summary

I guess we'll get started. Good morning, everyone. Thank you for coming. I'm Stacy Rasgon. I cover the U.S. semiconductor sector here at Bernstein. And it's my honor today to introduce our guests, Rich Puccio, the CFO of Analog Devices; and Mike Lucarelli, Head of IR. Before I start, [Operator Instructions]. Now of all companies that I cover, ADI, I think, is one of the most respected. It's got an extremely high-quality franchise in the analog space, particularly as it relates to signal conversion and processing. Over the last decade, so they've embarked on a rationalization program, and they're sort of enjoying the fruits of that now, but they've rationalized their product portfolio, manufacturing footprint. They were smart enough years ago to exit mobile early enough to actually get paid to do it. I think you're one of the only ones that ever managed to do that. Increasingly focused product offerings and high-value applications, whatever the end market and focus on new end markets, automotive, health care, many, many others. They've not been shy to take advantage of inorganic opportunities to boost the franchising growth potential. And they've now embarked on the regime of significantly more aggressive capital return than what we saw in years past. And to talk about all of this, it gives me a great pleasure to welcome our guest today. Thank you so much for coming to today, guys.

Thanks for having us, Stacy.

So this is Rich's first fireside, so I promised to go easy. So maybe just to start, you've been on the job, 3 months, right? I guess...

Actually, almost 4.

Almost 4. Okay. I mean I guess what attracted you to it? And at this point? And I guess, what has surprised you, I guess, either positively or negatively now that you've been in the role for a few months?

So the attraction for me was an opportunity to join a global leader with an incredibly strong franchise that was making what I thought were really interesting strategic moves to move up the stack and to take what is a very powerful Analog franchise, add software and digital and a very strong set of domain expertise to help customers solve their problems. Having come from a very customer-focused place, the opportunity to come to somewhere that will be just as customer-centric and have the customer intimacy that allows us to build ahead to what customers are going to need in the future. And clearly, what they need are higher-level applications and domain expertise and our strong franchise. That's what attracted me. What surprised me -- positive surprises, Stacy, when I got to the company was how far the company had gotten in their resiliency plan, right? If you look across the semi space, if you go back, I don't know, you can go back 5 years ago, most people didn't know what a semiconductor was, right? Until they couldn't get a car during the pandemic, and somebody said, well, you can't get a car because of semiconductors. So now everybody knows what the semiconductor is. So as you think about the supply chain fracture that happened and the challenges, ADI took a pretty aggressive approach. You've seen we've spent a substantially more capital in the last few years than we would have historically to build resilience. And so that resilience has done a couple of things for us. One, it's -- by the end of '25, about 70% of our products will be qualified both for internal and external manufacturing. That was a surprise to me that we've gotten that far.

70%, you said?

70%. The other thing that we've done that is super powerful, obviously, there's lots of geopolitical risk and concern about concentrations in Taiwan. We are on a path to diversify the substantial portion of our product will be available or able to manufacture outside of Taiwan.

How much is in Taiwan now?

So TSMC is our biggest foundry partner. About half of our business is done externally. I would say, over half of that relates to TSM. But one of the things that's going to help even in that regard is -- and we talked about this on the earnings call, is we are going to have capacity in the new fab that TSMC is building in Japan.

In Japan?

In Japan, [ and at the ] finer lithographies than we have, right, all of our internal capacity is 180-nanometer and above. So those -- that was a positive surprise. The other thing that was positive is I've come from a place where you have to think really long term about what you spend your money on and when you're going to see returns and the company is disciplined.

Do you want to remind our audience, by the way -- I should have...

Sorry, yes, I spent the prior 3 years as the CFO at AWS. So when you talk about all the abnormal and irrational purchasing behavior by companies, that was us. But when you hear from TSMC that it's a 2 years to get a product and you can't run a data center without it, even though you've already placed a year's worth of orders, you place another year's worth of orders. So I understand that dynamic from the other side. But one of the things I like here is the discipline and the ability to focus on a long-term development. So one of the things I really liked and it's something we've been highlighting, we just launched or talked about in the earnings call the Sentinel product, which is in our health care space, which is a very powerful innovative tool that opened completely new white space for us.

That was a 6 to 7 years...

6 to 7 years in development. Not a lot of companies are that forward thinking to make an investment that early. So that's a very long answer, a bunch of good reasons why I came to ADI.

Got it. Got it. Maybe to like drill in a little bit to that resiliency point. So you talked about fungible capacity. Maybe, not everyone in the audience and whether this is -- these are not all semiconductor investors. They may not be aware of what that actually means. And just to level set you, ADI in general, tends to have gross margins above 70%. They've been running a little lower than that at 67-ish, but revenues have been cut by, I don't know, 40...

34% to trough.

Not bad...

And to be able to hold 67% versus what we thought when we did our Investor Day in '22, 70% would be the floor. We certainly did not contemplate the size of the inventory correction that could happen. But so I think where you're going is -- so when I talk about this qualifying inside and outside, and you'll hear us use the term, our ability to swing. The best way to describe that is in times of increasing demand at a pace we weren't expecting. We have outside partners with qualified processes that can make our products in their fabs. The flip side, which is why we've been so resilient on a margin perspective is when demand starts to shrink, or do what it's done for the last year, we're able to pull back those products and manufacturing in our fabs. And why that's important is keeping our utilization levels up in our factories is what allows us to keep the margins where they are. absent that swing capacity, we'd have a lot of underutilized capacity and not to bore you, but accounting stuff, the unutilized capacity is a significant drain against your margins...

Like where would your gross margins be if you didn't have the swing capacity?

They'll probably be a couple of points lower, than say, 3 -- probably 3 points lower which would be way off the floor. So that resilience has been important for us and qualifying the -- it also -- as you think about the cost environment, we've incurred a fair amount of capital and expense to be able to build that resilience, but it's super important to our customers because we don't want to be the part that keeps somebody from shipping their million dollar device or their car.

Yes. Yes. I mean maybe to touch on some of that. And again, in this format, I don't like to spend lots of time on the near term, but I do want to...

Okay...

I want to talk a little bit about the short term and get through that. But maybe the -- so when you talked about -- like you guys have sort of suggested Q2 is a bottom. And I know in general, in Analog investors seem to be wanting to buy Analog stocks just in general on that bottoming theme. But maybe talk a little bit about how we got from where we were to where we seem to be in Q2. I guess, what are we seeing by end markets? Because it's been sort of a weird asynchronous cycle where different end markets have sort of been behaving at different points in their cycle, different timing. And I guess, how do we think about the pace and trajectory of that recovery hopefully off of Q2? And then how does that influence some of the decisions that you're making, I guess, as we're going forward from here?

Sure. So do you want me to start with the -- a little bit of the backdrop of what happened?

Yes, maybe. Yes, maybe...

So if you think about the way the trough has happened, obviously, significant decrease in our industrial business, right? And industrial for us is our largest franchise. It's also the most profitable franchise. So we've seen the steepest declines there for a very long period here as the decline was happening, automotive held up better than most of our other segments. Clearly, the wireless, in our comms world, the wireless piece has been soft. The 5G rollouts aren't happening...

That is infrastructure. You guys don't really do a ton of like smartphones and stuff, right?

Smartphones is in our consumer bucket.

Yes. It's not that big, though, right now. Yes.

So this is in our -- in that space, it's the wireline and wireless. And as it turns out, the wireline is doing a bit better, and part of that is where we're helping with the data center connectivity, et cetera. Consumer has been a challenge as well. So -- but what we're starting to see is improvements, and we've had 3 consecutive quarters of increased bookings, cancellations have declined. So we're feeling much better about the trend where we're headed. And in fact, even if we look at Q2, and if you look at what happens geographically for us, Asia Pac actually grew sequentially. And if I look at China, in particular, China has been our weakest market for a number of years, and China actually grew sequentially. And based on the new wins -- and we talked about some of the new wins on our earnings calls, based on the new wins, we feel like we're getting some pretty good traction, and we'll see growth -- continue to see growth in China despite some of the macro headwinds and some of the indigenous China pressure. So we're seeing those signals. If you think about the Industrial piece, which is our biggest business, our business is highly correlated with the PMI index. So PMIs have been above 50 for every month, every month in '24. So that gives us some comfort that we're starting to see the return. What we've guided to at the midpoint is about 5% sequential, we haven't. Did guide out to Q4. There's still enough macro uncertainty. And if you think about geopolitical interest rates, any number of factors that could cause a continued dampening, I'm not -- I don't have enough -- we don't have enough and I don't have enough visibility to say that this is going to be a steep recovery. So I actually -- I think we're reasonably comfortable that 4Q can be sort of low to mid-single digit sequential coming off of Q3. I do think that we will see faster growth and acceleration into '25 as we get fully past the inventory digestion issues. And look, the -- if you look at the curve and Mike been using this as an easy way to explain this, if you look at the historical growth curve for the company, even if you take 5%, right, because the historical business has been pretty close to a GDP growth kind of business. If you look at where we've been relative to that kind of growth rate, we've clearly -- at this point, have undershipped pretty significantly end demand, both in our end customers and in the channel. So -- and we think we've been under shipping for at least a year. So at this point, we expect that we will start to see sequential growth, and then if we stay on this path in Q3, the signals we've started to see, it's likely we will start to ship to end demand in the fourth quarter.

Got it. So no refill though, at that point. Okay. So, yes.

And literally, we've spent a substantial amount of time and effort getting our inventory back down into our target levels so...

You're back in the 7% to 8% range now?

Today, probably at the top of the range. Closer to the 8%, but we will expect to take more inventory out of the channel in 3Q to get us solidly in the middle of the 7% to 8% range, which is where we'd like to be. And I said, if we stay on that track and are seeing ourselves achieve that, then we'll be able to ship to end demand in the fourth quarter.

Okay. Got it. That's all in the short term. [indiscernible]. What are some of like the broader like -- I always say, like for Analog's side, for me, it's sometimes I feel like it's hard to build a differentiated perspective on growth because I mean it's -- you always think about it -- if you sell 100,000 different things to 100,000 different customers, and how do you get a handle on that? I was like, ADI though, it's maybe a little more easier, at least to articulated growth narrative. You do seem to be investing and involved in a lot of the other like higher growth areas, you mentioned some health care. Vince has talked about [ aero ] and defense and space and now certainly EVs and AI, I'd be remiss if we don't talk about AI. But I guess, what are some of the opportunities that you guys are most excited about and where you're putting more of your investment dollars these days. And then I think we can drill into some of them.

Sure, I'll start, and then I'm sure Mike will add some details for me. Certainly, one of the ones I'm super excited about is the health care opportunity because I think...

You mentioned Sentinel, what is that?

So Sentinel is actually a wearable patch that monitors vital signs, and it's -- doctors will use it to help patients with congestive heart failure. And so it's super powerful because it does -- it's not a surgical implant, which most of the devices that do that. And it is tied to a monitoring solution. You put the patch on for some number of minutes every day to get measurements and then the algorithms that we've worked with doctors and hospitals to build give feedback to the doctors because ultimately, congestive heart failure is the single most expensive chronic disease, particularly in the United States. The goal here is to keep patients out of the hospital, because that's where the costs incurred and to give better care and this device helps do both. And it is a great [ melding ] of the sensing, the software algorithms and the scientific knowledge. That one is super powerful. It's a $5 billion TAM. So that's what I'm super excited about. Obviously, I like what we're doing in aerospace and defense clearly, with all of the things going on in the world, that's an area where not just in the U.S., everywhere else in the world, they're [ buffing up ] their spending on military and aerospace. So that's a great opportunity. You mentioned AI, so I'll go there. There's sort of -- we think about it in 2 phase -- sort of 2 stages. This first stage is the infrastructure stage, which all of you are hearing about, hyperscalers are spending hundreds of billions of dollars building data centers. There's huge opportunities there for us. And 2 big opportunities to highlight. The first is if you talk to anybody in the space, one of the broader concerns for the hyperscalers and probably for everybody at this point is power. Power is going to be a constraint. And so anything that can be done to help make the hyperscalers and their data centers more power efficient is a big win. So we've got a win in with one of the hyperscalers on a vertical power solution. And that vertical power solution can reduce about 30 -- up to 35% of the power loss in running those chips in the data center. So that's a super powerful win, and we're getting traction with all of the hyperscalers. That's one. The other thing is all of these things are being built, the high-performance compute and the high-bandwidth memory require much more intense testing. And we are shipping high-value solutions into the tester space to allow them to test faster test more channels. So 2 really powerful opportunities. Then if you think out longer term, and this is where the compounding effect happens for us is as you think about AI at the edge. So getting compute and AI out into the physical edge is going to be super powerful because -- I use an example, I like this one. I'm a simple guy. This is a simple example. When you think about noise cancellation. And I don't just mean the Bose that you wear on your headphones when you're in plane, it could be hearing aids. Sound is very variable. We're shipping product today and developing products that have neural nets built into the devices that allows the algorithms to do the work to sort the sound to do the noise cancellation, super powerful, super effective. Another good example I like is you think about factory automation, which is one of the big trends, it's going to be a big driver for us on the industrial side. When you think about the robotics, the robotics that are in automated factories. Well, if you can get compute at the edge and everybody today is talking about large language models. [indiscernible] ChatGPT, it's all the large language models. I think more about it as small language models with compute at the edge. So you think about that robot arm that is sensing and actuating, that data has to go back and be computed to determine the next move to follow its algorithm. Now you can have that compute power in the device. So you think about the benefits, less latency. It's closer to the actual source of the data when you do the compute, less power consumption and more secure because you're not taking all of that data, physical data, converting it to digital, sending it to the cloud, you're doing that at the edge. So that the mid- to long term is a huge opportunities for us, because it builds on our analog expertise, our ability to convert the analog and then to do the compute. So that -- I think that's one of the really long-term things that I'm excited about.

Have you guys ever sized like put a number behind like the AI opportunities for you guys? Or...

So the AI -- exposed AI today for us is sort of low to mid-single digits, say, $400 million-ish of our revenue today that's growing double digits and some of these things we've talked about as we get more design on the power side. The other piece, I did miss one, I apologize. The other piece that's exciting is particularly when you think about the AI data center and the GPUs and the connectivity, the speed and so we have optical devices that are part of enhancing that speed. Everything is moving up, right? We're going from like 1.2 terabit speeds for communication, and we play a role there with our devices. So that's another really strong area. And those could contribute another 5% or 10% growth to our comms business, right?

Got it. Got it. Stacy, that was a great question on the earnings call, but you guys chose to focus on 2Q and revenue, whole time.

What, say that again?

I'm just giving you a hard time, just give me a dig at the sell side a little bit.

We can take the chair away actually. We could do that.

I can stand.

I want to go to auto. This has been a pretty good -- I remember like when I first started covering ADI, they still had auto like lumped into the Industrial business because it was still small. Healthcare is still in the Industrial. I always ask, like when are you going to break healthcare out? When is it going to be big enough...

Do you want to ask the question -- is that...

No, I'll just say auto was -- so auto is not in there anymore. Auto is -- I can't remember, 20% or 25% of revenues now, something like that?

30%.

30%, okay. So closely, there's an EV portion and we tend to look at EV just as strictly as BMS, but I kind of maybe get the [indiscernible] maybe it's broader than that. Maybe you could talk a little bit about I guess, the broader auto opportunity and then as well as like what do EVs mean for you?

Sure. And I guess you can think about it across the 3 categories of vehicles, right? So you've got the ICE vehicles, you've got your hybrid vehicles, and you have your fully electric, right? And if you think about what has happened in the car, you drive the in-cabin experience has changed dramatically. There are more speakers, there's more microphones. There's more sensors. There's more compute. There's more screens inside. You have to transport all of that data, right? The video to your screens. So in addition to the battery management, which we'll get to, we have our GMSL products, which are helping facilitate all that activity inside the cabin. And that is across all of the vehicles, right? Because you're going to get -- you're going to have the luxuries whether it's in an ICE vehicle or a hybrid or an EV. Then we get an incremental step up in our concentration in auto when you move to the EV and the hybrid, obviously, there's more battery management in the full EV than there is in a hybrid, but so you get a little bit more concentration and a little bit more dollar value in each of those cars. And the things that have been happening for us that have been fantastic. We talked about this in the call. We've gotten designed into 2 more of the largest 20 OEMs. In addition, we've got design wins in one of the largest Chinese players. So we feel really good about that opportunity. And the concentration continues to grow, right, as they add features and functions. And we start to think about more and more of the automated driving and ADAS and where it's going to go. So that is a good opportunity for us. And actually, in the most recent period, auto continued to perform best for us and we expect that it will be -- it has proven to be a shallower trough than we've seen in industrial, for instance. Now look, are we certain, 100% certain it's the trough on the auto side? No. But do we see signs given that we have got signals that we're going to see growth there, that we're coming back out of it? Yes.

Got it. I guess within the BMS business, you guys talked a lot about like wireless BMS versus wire. Can you talk a little bit like what that is, what the advantages are. And I'd love to get some feeling of like the actual like content that you guys are seeing in the cars right now.

I'm going to [indiscernible] pass the...

Sure. Yes. So you're right. So our legacy has been wired BMS. We're the leader in wired BMS. That's not going to go away. BMS is battery management systems, by the way, for -- basically, we manage the battery, and we imagine more accurate than our peers, so you get more miles per charge. We've moved to the next generation with wireless. Wireless basically is it's all the features of wired, and we can do it wirelessly. Why does that matter? It's better R&D efficiency for the customer. They can design a few platforms as modular across different vehicle models. They can manufacture that through robotics versus human, faster, lower cost to manufacture batteries. Which are very expensive. So those are 2 big features. Now what do we get? We get 2x the content. If you use a wireless system, you get the wired plus all the wireless connectivity in that. And what we've seen is we have 4 or 5 OEMs adopting it. Today, we're shipping one for revenue, 2 more in the next, call it, 6 to 12 months, another couple beyond that. So you're seeing a proliferation of wireless. And really, the reason why they're doing it is driving down the cost to develop these batteries, which is super expensive and hard to do. So I think you'll see both over time. Wired won't go away, but wireless will increase in penetration over time.

What kind of content are we talking about? Like is it hundreds or thousands, like what, I wish thousands...

Hundreds. Multiple hundreds.

Okay. Got it. And I want to ask a little bit about industrial. So it is your biggest market. I always joke like it's not an end market, it's like hundreds of end markets, right? I mean what are the biggest pieces of industrial for you? How are those growing?

So as I break it down, you're right. Hundreds of markets, but there are a few big ones. The biggest piece of industrial is instrumentation and test close to 30% of sales. Think of those like automated test equipment, electronic test and measurement, scientific instruments. That's a big part of our business, 30% are probably the most exposure than anyone in the analog industry to those markets. Number two is factory automation, close to 25%. That's been weaker recently. It was a little higher now it's below 25% because of just overall inventory digestion, but a very important market for us. Those are the 2 biggest pieces represent over 50% of industrial. Then number two -- sorry, 3 or 4 would be aerospace and defense, and health care, both between 15% and 20% of sales. And then sustainable energy is probably close to 10%, maybe high single digits. That's about 90%, 95% of Industrial. The remaining is what we call other as a grab-all bucket.

Got it. Got it. And back to the comment I made earlier, right, if you think about the industrial space, it's the testers for the high bandwidth memory and the compute. And then -- sorry.

It's working. That's...

Okay. So we're catching some really good secular tailwinds as you think about it. And that's why we're optimistic.

Got it. Yes, that automated test business can be very -- you had some quarters back in the day where it was like 0, I think.

0. But you're not wrong. So inherently, [ ATE ], automated test has been very lumpy. What you've seen is these testers need more and more semiconductor content. It's much harder to test high bandwidth memory than NAND, general NAND. So we ship more content per tester than we used to ship and there's more demand out there everything needs to be tested. Everything -- every growth we talk about, whether it's EVs, AI, wireless systems, factory automation. I mean you could go through the whole list. There's more growth drivers of our business. Therefore, we need more tests. And we haven't seen these drastic lumps of recent history just because it's a more diverse business with more content per tester.

That makes sense. It makes sense. I want to shift gears to China. So you talked a little bit about -- you said it looked like things are starting to improve somewhat. I guess maybe can you talk a little bit like which of your businesses have the most China exposure? Like what do you see, and in general, I would say your overall China exposure just in terms of percentage of revenue seems to be lower than many of businesses. I don't know 20%-ish?

It's less. So I think in our 10-K, we had about 18% exposure in China today. It's about 15%. And if you think about it, 95% of that is B2B. So it's 80-ish percent is between industrial and auto, and there's 10% in comms and roughly 5% in consumer. And what we started to see is order rates increasing. I mentioned we had the design win in at one of the larger Chinese automakers. And look, the factory automation tailwind is real. And so we're starting to see that. Look, there is continuing pressure from the domestic made in China for China, but one of our advantages is we're not playing in the low-end commodity space. We're in the high-performance space. And my prior comment of how much of the China business is industrial. It makes it very hard to attack because you have to have breadth and depth and performance in that space.