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

Everyone, I'm Chris Danely, your friendly neighborhood semiconductor analyst here. It's our pleasure next to have one of our topics, Analog Devices. The normally scheduled program of Vivek, the VP of Ops, is not here today. So being as it's baseball season, we've got the best pinch hitter in semis up here, Mike Lucarelli. VP of Finance, IR, FP&A, probably he has more alphabet soup behind the name. Something about Vivek's brakes went out unexpectedly today, and you were the only one that could step in. I don't know how that happened, but any rate.

Desperate times, desperate measures.

Yes. Here is Mike. And as it always is, this is meant to be interactive. So if anyone has a question, feel free to raise your hand or throw something at us, somehow get our attention.

I guess just to kick it off, Mike, we've had a few analog companies here so far. You guys are one of the most recent companies to report. You guided up, I think, for the second consecutive quarter, sequentially. Other analog companies are still seeing the revenue down. What do you think is separating the basic business? Because I know you study all of your competitors. What do you think is sort of separating ADI from the rest of the folks that are not seeing as good a business conditions as you are?

That's a good question. I think what you're seeing is a million dollar business about a year ago. We really worked hard to normalize our lead times, getting back to 13 weeks. So we got our lead times back at 13 weeks about a year ago, which started the process of inventory digestion at our customers, about a year ago. I think if we started that process sooner, could have normalized lead times. Two, we've been very actively undershipping the channel for the past 2, 3 quarters to bring down our channel inventory, which were too high. They're above 8 weeks, now it's back to 7. We've normalized that. And three, if you look at our business, our business is not the same. I mean, we're an analog company and there are other analog companies, but we have different exposure than a lot of our peers, especially in the industrial market. If you look at our industrial business, we have a big instrumentation business, which is about 30% of revenue, and aerospace and defense like 20%. And part of instrumentation, that relates to AI like ATE, is seeing demand and Aerospace and Defense is also seeing good demand. So there's some demand, there's some timing to it. And basically, let's get through that inventory digestion, I think, faster than our peers and then layer on top different exposures of our business.

Yes, the different mix. How would you characterize your overall visibility now versus, say, 3 months ago? Better or worse, somewhere in between?

Same. I'd say visibility -- so visibility is dictated by lead times. Our lead times haven't moved. They're about 13 weeks. 13 weeks give us about 1/4 of visibility, no different than a quarter ago. Now way back when the lead times were long, we had "better visibility." But when lead times are long, that visibility, you never know if it's good or bad when you get there. So I will call our visibility is good, but short, meaning 13 weeks visibility.

Yes. Okay. I had a number of people telling me to start off with the automotive question. So we'll just get that out of the way. Maybe start with your sort of views of the underlying demand trends in the automotive industry.

So I think -- so what's going better or worse for us versus the peers. If you look at our auto business, auto definitely got a bit weaker for us than we thought a quarter ago. And you saw it in our outlook, we actually guided auto down. I think 90 days ago, most thought it would be flat to up. So what really changed the automotive market was, halfway through the quarter, demand just got really weak broadly across automotive. It wasn't one geography or one sector of auto is weaker. We did see it get weaker over the summer. Now the good news is it did pick back up from that low, I would say, from an order standpoint, but that's the good news. The bad news is, it's still below where it was over a quarter ago. So overall, auto, I think, is a mixed bag. I don't think it's falling off a cliff. It's definitely softer than it was 90 days ago, and it's kind of broad-based across all applications of auto. And then if you look at our business in auto, I think it's important to point out, about half of our business will be called as a production base. Things that are really more volatile with SAR or production. And the other half of the business is content growth drivers like BMS, GMSL. GMSL connects radars and cameras in cars, power management around those radar systems and A2B, which makes microphones. That other half, that 50% is still growing year-over-year. The production part is down over 20% year-over-year. So you're seeing different dynamics in our business. So the peak to trough of auto is a lot shallower because you have this content stuff that's still going on and some inventory digested on the other side. So net-net, yes, auto got softer. I don't think it's that bad. It probably takes a quarter or 2 to kind of filter through all that digestion. But I think as you enter '25, I think we're in a good position both in auto and the rest of our markets to have a good growth year.

And would you say either of those 2 parts of the business correspond more with ICE or EVs? Or is it basically everything?

It's every -- so a good question. BMS is all EV. So BMS manages the electric vehicle battery. That business has been weaker this year, not surprised. I think you've seen a lot of EV announcements, not a lot of purchasing of EVs, and also a mix around hybrids to EVs. We are the leader in BMS. We have the highest market share out there. So you've seen that business modulate for now 2, 3 quarters. And our view was it could get better in the second half of this year on the BMS side, it didn't, except China. So China actually improved for us sequentially, but the rest of the world was not. And that's not a share comment. That's just overall inventory digestion of EVs. So that BMS business really correlates to EVs. The other parts of our business related to high-end vehicles. High-end vehicle has more radar, more cameras, more microphones, more speakers. Those doesn't matter if it's an EV or an ICE car. If it's a high content vehicle, like ones you drive, Chris, that's what it correlates to.

I'm all about content, my Toyota Tacoma. And so I guess let's start from the beginning of the year. Because I know Vince talked about this on the call that you saw bookings increase, I guess, in auto, and then they fell off and they stabilized and they fall off again and now they're better. Is that how it's gone? Or I'm just trying to hear how the year has gone in auto.

Auto? Yes, auto, let's talk about auto. Start of the year, every market, including auto, up into the right, bookings good, demand good. Even the start last quarter for us, so kind of the May time frame, auto good. In the summer, you saw bookings fall in auto a bit, summer time frame, call it, June, July. Now since that time, they fell and now that they came back up, just as they bounced off the bottom, but they're still below where they were starting the quarter. So positive side is they're not still falling, and they're not flatlining at a low rate. They actually picked up. On the negative side, you could say they're below where they were to start the quarter. But that's why I think net-net, it's not that bad. It is softer than it was 90 days ago in the automotive market. So start of the year, auto, the first 6 months, [indiscernible] like the rest of our business up and to the right, saw some choppiness over the summer. Now some investors point out, it could be seasonality. I mean, summer is usually weaker. So that could be some of it. But I do say that it did fall a bit more than we thought in the summer. So we'll see how it goes from here on to see -- that will dictate what 1Q and 2Q is. But from a positive standpoint, it has gotten better. So we'll see if that trend continues.

And you think the reason that the bookings from the auto space fell off again, was that just because demand took a downtick? Or was there some excess inventory out there, a little bit of both? Or...

I'll stick to the reference. I think demand took a downtick and therefore there's so much of inventory. And it only fell once. So you keep in again, it fell in the summer time frame, in June, and it bounced back since then. And I would call it more demand-driven, resulting in inventory reduction.

Got it. And then what are the customers telling you guys from the automotive side? Do they feel like their demand is stabilized? Are they nervous? They feel good?

Mixed. It's a very mixed bag. And even by OEMs versus Tier 1s, some say inventory is good, some say it's too high. It's a very mixed bag across the landscape.

Any particular geo or area or anything that's better or worse? Or is it just [indiscernible] customer specific...

Europe and North America are weaker than the rest of the world.

Got it. And you said you expect 1 or 2 quarters for the inventory to normalize?

That's what it feels like. I mean, we'll see, it could be sooner. And when I say 1 or 2 is, 1 quarter would be the fourth quarter. And our first quarter, which would be quarter 2 for that correction. This is an unusual time to see strong demand in auto. So it feels like you probably have a down quarter, and then we have no visibility to 1Q, but usually it's down, maybe it's down again. We're talking low single-digit decline in share. So not a big drop off. I think a lot of the worries is, you saw industrial went way up and it fell like 35% peak to trough. Right now, I think auto is down, call it, 15%, maybe 20% peak to trough. That feels kind of like where it's going to end up, and then we start going the growth cadence again for the automotive market.

And do you think the issue over the summer was more of an EV problem or an ICE problem or both?

Both. I think it's both. EVs did not help. But I think overall -- I mean, if you look at IHS, they reported out, they would reduce production. If you look at our customers in the Tier 1 space with OEMs, all their earnings reports were bad, they all lowered outlooks. It was a broad-based demand impact, whether it's a traditional combustion engine or an EV. I mean, I don't know for sure, but rates are high, people finance cars. It's expensive to buy a car today, therefore is less demand.

That makes sense. And then last question on the inventory. Do you think that the excess inventory or where they're burning it off, do you think it's at the lots? Is it in the supply chain? Is it somewhere else? Is it more components or cars or both or? Any idea?

Supply chain and components, I would say, more than lots.

Got it.

Lots were bare. I mean, if you go buy -- I mean, lots were bare for a while. I think there's more cars on lot, but I don't think it's too high. I'll call them more at the component Tier 1 level than the actual OEM level. But like I said, it's very mixed, honestly. So it's tough to call out is precisely where that inventory is and at what makes this semi industry fun.

And then our auto analyst has been sort of hammering this point, you think China is going to continue to gain share in EVs, is that good, bad or indifferent to ADI?

Depends, and what they're replacing. So you're saying trying to gain share globally [indiscernible] market?

[indiscernible] EV makers, yes.

That would be a net positive. We have a very good share of the China OEMs, both on a BMS standpoint for the export market and also on the connectivity side for GMSL. So why I say it could be indifferent is if they're shipping the car, BYD car into Europe and it's taking share from another company we have share at, it could be net neutral. But we do have good share there. So if they're taking share from someone we don't have good content, and then it's net positive. So we'll see. But it is fair to say, we have a very good presence at those OEMs in China that are moving more to a global production model.

Okay. And then some time ago, some crazy analyst out there, I forgot his name, rhymes with [indiscernible] or something like that, talked about you guys losing a design win internally at a Chinese automotive production company, but we haven't heard about anything since. So maybe just address that. Do you expect this to continue? Did it stop somehow? I get [indiscernible] questions on that.

Yes. So there was the biggest guy in China look to in-source more of their own chips. It wasn't an ADI issue. They chose their own chip versus any other supplier, whether it's ADI, TI, NXP. They want to be more China indigenous for those cars. We saw that with -- that's in our numbers. Last year, we got hurt by some of that. I can't say we didn't. Why? Because we had high share and it went down. That stabilized, and we actually have gained back share, added OEM as well. So it's hard to do what we do. And if you want the best performance, whether it's BMS, GMSL or converter, you name it, you choose ADI. If that customer solving for something else, like geopolitics, they might choose someone else. I think that's what you're seeing in that market, to your auto analyst question and why would that customer go internal. They're not choosing performance then, they're choosing something else. Our first BMS chip was launched back in linear in 2010. We're in year 14. Competition is not year 14 in that market. So inherently, they don't have as good a performance as we have.

How do you see your China business growing long term versus other geos? And has that growth estimate changed at all over the last few years?

Good question. So China, historically speaking, was one of our fastest growth markets, I would say that. Our view going forward is, it is still a growth market, but we're unsure if it's the fastest growth market going forward. Now there's 2 dynamics at play. The GDP of China used to be 10%, now it's 5%. So that's slower growth. And two, over time, China will try to do more stuff locally, which probably trims some growth. But our view is if China wants the best performance, they'll choose ADI. If they're choosing something else, then they'll go for someone else. But it doesn't mean we can't grow in China. And we will grow as you look forward to this year, and China was one of our best growth markets last quarter. So it doesn't -- there is a lot of nerves around the China indigenization. So far, it's been more worries than actual impact in our revenue. Yes, it was an inventory correction, but that was an inventory correction, not a share loss. And if you look at our China business, what we track is design win. Design win momentum in China is up the past 3 years. That portends to a good -- at least growth in the next few years as well. So yes, there's an existential threat out there that at some point, they'll take some more local. But in the near midterm, it doesn't feel like that's happening. So I would say, yes, we're cognizant of it. We see it. Our goal is to design the best chips, and if they want the best performers, they will chose ADI.

Yes. Great. All right. Better news. Let's talk about Industrial. It sounds like we're well past the bottom -- past the bottom. Maybe just give us some signs on what you're seeing there. And I know you gave some nice delineation on the various flavors or verticals of Industrial that are bouncing back quicker than others. Can you just give us the spreadsheet, the matrix there?

Yes, sure. It's hard. I talk to my hand. Now I have a broken hand, so it's hard to talk to my hand.

You want to borrow mine?

No, we're good. So if you look at Industrial, the biggest part of Industrial is instrumentation and test, 30%. Automation is about 25%; Aerospace and Defense about 20%; Healthcare about 15%; and Energy, high single digit 7%, 8% and then there's probably remainder of just other. That's the breakdown, the spreadsheet of Industrial. If you look at the pieces and what we're seeing in each market today is your question, Automation is the weak spot. That's easy. Automation is the only segment that was down sequentially in 3Q, and we'll see what happens in 4Q. How about that? I don't think -- it could be up, let's call it, flattish, maybe up. Automation is definitely lagging in this recovery. And the question is why? I think what you're seeing is companies have a lot of plans to build new factories and these new factories have a lot of more automation in it, more content for ADI. But they're not deploying that CapEx to build the factory yet because demand is weaker.

Less than some of the others?

No. Automation is down a bunch. So automation for ADI will be a long-term growth driver. It's just in this low period right now, because if a company wants to build a new factory and they want to make a sovereign supply chain in U.S., Europe, it's expensive to build those factories. So what do you do? Robotics. Robotics means more content for ADI. Fixed arm robot about $100; a mobile robot, $300, $500; a humanoid -- one day, there will be humanoids, $1,000. So what you're going to see is more content per factory. But right now, you're in this inventory digestion and no one's really building factories with not much CapEx outlay, that's impacting the automation business. That's the weakest part for us. So that's the bad -- sort of the bad. Let's go to the good. On the instrumentation and test side, there's 3 pieces of that. Automated test equipment, ATE. There is electronic test and measurement and scientific instruments. The last 2, ETM and scientific instruments, I'll call it, have normalized, meaning they're not burning inventory, but there is not much demand in those areas. Think of customers like Keysight, Agilent, Thermo Fisher in those areas. The ATE portion, think of like Avantas, Teradyne, those customers, that demand is strong. Why? We ship into the guys who test high-bandwidth memory and also test SOCs and GPUs. So good demand in that ATE vertical within Instrumentation. Aerospace and Defense, I think we're only going to in detail why that's good demand. The world is, unfortunately, a more unsafe place. Demand is quite strong in Aerospace and Defense for us. Energy actually came down. Energy is starting to grow again. Now it's a small part of our Industrial business. But I think if you had to ask me one area I'm most excited about in the Industrial business for us is Energy. Why? Energy consumption has not grown in decades. Now you have everyone wants an EV, electric cars, AI systems. It's all driving energy use up, means you need a new grid. A new grid means monitor, control and storage; all things ADI does. So I think you'll hear us more talk about energy going forward than we used to because there's a new demand driver for that market. Now it is small. So we move the needle on growth, but it is a new growth driver for our business than 2 years ago when we outlined at our Investor Day. And the last one is Healthcare. Healthcare actually is a bit weaker than I thought it would be at this point. It's bottomed out. It's starting to grow. But we haven't seen much demand in the healthcare market. I think there it is. It's more a broad market business, and we're just seeing digestive inventory at our customers.

Okay. Take us through the last couple of quarters, this quarter and then the previous quarter of Industrial sequential growth, and do you see it continuing to accelerate? Do you think this is inventory replenishment? Do you think that demand is better, like what's driving the better Industrial business?

A couple of things. So on the Industrial side, what we've been doing is reducing channel inventory, and we're done reducing channel inventory. So because we're not reducing anymore, that's helping the growth rate, because you're normalizing the shipments into the channel and out of the channel to net zero. Before a quarter ago, it was down tens of millions and 2 quarters ago, it was down over $100 million. So that's helping the growth status of Industrial. I think what you're seeing is, things have normalized now and you're kind of handing off from inventory digestion and channel reduction into more normalized demand. I don't think we're inventory replenishment yet. And I think we talked about it on the earnings call, 1Q for Industrial is usually seasonally down. We don't have any visibility yet today, but it feels like it's probably where it's going to be. So [ bombs ] in, feel good about that. We had 2 growth quarters. 1Q is normally down. I think it's a normal quarter, and then we'll see how it goes from there. But at some point, that industrial business is going to have a snapback. And one thing [indiscernible] add is if you look at our Industrial business, ADI plus Maxim pre-COVID. So get rid of the COVID noise for a second. Look at '18 and '19, we did about $1 billion a quarter or $4 billion a year in Industrial. This year, so 5 years later, we're going to do just over $4 billion. Industrial is not a flat market for us. It should grow at least 5%. So if you think about that $4 billion and you just draw a line and grow at 5%, you're shipping $250 million a quarter or $1 billion a year below trend line. So at some point, once that the digestion stops, which is, I think, has stopped, and demand picks up, you're going to start seeing us grow faster than the industrial market. Now the multibillion-dollar question is when will we see that?

Right now, all right?

I wish. I'll say sometime in '25.

Okay. And then you guys guided the Industrial to grow, I think, high singles, sequentially this quarter, and it grew mid-singles last quarter. So if it's above seasonality for these 2 quarters, why would it go back to normal seasonality for next quarter?

Because the channel helped that growth. The reduction of -- the less reduction in channel helps some of that growth.

Got it. Okay. And then let's skip over to the comm market. So most folks out there seem to be calling for a bottom, seeing their business bottom, bottoming now, picking up, definitely not getting worse. What are your thoughts there?

The comms you said? Yes, I agree. So in communications for us, it used to be 55% wireless, 45% wired. Over the past 2 years, flipped. It's like 55% wired, 45% wireless now for us. On the wireless side, it has been -- I mean everyone here, I think, has heard 5G a million times, but now no one wants to talk about 5G. It's kind of washed out. The good news is it's not going down. It actually increased for the first time for us in probably 2 years, the wireless piece. Going forward, it probably -- there's still not much demand on the 5G side, but I feel like the inventory burn is done. At some point, it probably picks up a little bit in '25, but the demand for 5G is still very muted going into next year. The wireline or wired side we call it, there grew a lot more for us last quarter, and it should grow again for us in 4Q. And even in '25, there's a lot of inventory out there. I mean you got some customers go way up and way down in that wired. That also, we've been shipping well below demand on the wired side for probably 3 or 4 quarters now. And at the same time, you're seeing more demand on the AI side. So we ship into data centers for connectivity and power and there is more demand in those areas. And now that the inventory correction is done, you're starting to see that demand actually drive revenue growth. And that started last quarter, will be again in 4Q and through next year. So comms next year, I think, has a decent year. Wireless okayish, but wires have a good growth year for our business.

And if you look at the overall comm space, I think you guys guided it up slightly sequentially this quarter.

Flattish.

Yes, flattish?

Giving room on both sides.

Got it. Equal opportunity. And how was it last quarter?

It was up a bunch last quarter. So what we saw last quarter, it was up close to 10% sequentially. But the reason why -- if you look at our comm business, we have a lot of pull for a lot of stuff that goes into AI systems. And I don't know if it repeats again. So I took the conservative approach saying, I think it's not going to repeat again, it's about flattish, but it could. So comms could surprise the upside, but I'm not going to build in our outlook, something I saw those kind of lumpy last quarter. But the good news is it lumped up for the first time in a long time, and that really related more on the wireline side.

And then maybe take us through the linearity of bookings was the book-to-bill. I think the book bill is above 1.

Auto brought it down.

I mean for comm.

For comm, yes, it was -- fair.

Great. Yes. And then just to quickly touch on the consumer side, that seems like we're back to normal seasonality?

Consumer side, I'll say it's normal seasonality, and we actually did a little better than that. I mean, usually we are up about 15%. We talked on the earnings call about new design wins ramping up, call it, key customers in the consumer space. It's not one customer, it's multiple customers and multiple applications that's helping drive above seasonal growth here in 3Q and actually to our double-digit growth sequentially in 4Q for us as well.

Okay. So in terms of the broader overall business, getting back to some -- commentary by your competitors, some are saying pricing is still strong, some are saying pricing is going back to normal, some are saying pricing is terrible. Where does ADI fit on the pricing spectrum?

Pricing is more or less normal.

And can you define normal?

Normal for us is -- our goal at our Investor Day is at flattish pricing. So you can consider this year close to flat for pricing for us. I mean we manage our business -- different parts of our business have different pricing dynamics, but our goal is to make that pricing minimal impact on our P&L or our gross margin. And that's kind of what we're back to. I think others had bigger price declines than us. I think what we do is we play a value game. We want to drive innovation. So pricing is less important to our customers in what we deliver.

And your pricing is probably a little more stable, up or down versus the [indiscernible]...

Correct.

And then how would you expect pricing to trend next year?

We'll see.

Same as normal or [indiscernible] go down?

I think it would be similar to what it was in '24. Again, those conversations are just starting now. [indiscernible] trying to get too ahead of it. But our goal, and the way we drive our business, is to deliver value, sell the value. Pricing is the last thing we discussed, the goal to keep it about flat.

Okay. And then how does that correspond with what's going on at the disties? You mentioned inventory going up, going down. How do you see your inventory positioning at disties right now or disti inventory overall...

Disti inventory is actually very lean. So if you go back to beginning of the year, we probably got -- just go back a second. Our target model of disti is 7 to 8 weeks. We were north of 8. Not north, and I would call it, 8.5 weeks. As of this quarter, we're down to 7, the low end.

Yes. North of 8.5 was coming into the year, right?

Yes. So we brought it down quickly, and we like to stay in that 7 to 8 weeks. So as you go forward from here, our goal is to match supply/demand. What we ship in, what shipped out, net zero. If demand picks back up, we want to keep it in 7 weeks, so we'll have to start shipping more in if demand gets better. But that's an if statement that won't impact our 4Q, maybe [indiscernible] we'll see how it transpires. But our North Star is to keep inventory in that 7 to 8 weeks. 7 to 8 weeks allows them to meet demand and not have a big cycle on our side. Yes, it got above it, but I mean demand fell so fast in Industrial, the fact that it only got up to 8.5 weeks is kudos to the team to really managing that to [indiscernible] up to 10 weeks.

Okay. And then in terms of your own outlook or the overall ADI outlook on 2025 for a calendar year, conceptually, generically, however, obtuse or specific you want to get? What is ADI, like, most excited about, most worried about?

So I don't really do calendar year as good as fiscal, but a similar answer. I think if you look at next year, early view, knowing I'll be wrong, maybe high, maybe low, don't know. Industrial and consumer probably the 2 strongest markets for us and auto and comms still grow, but to a less degree. Now that's an early look here in, what, September. I'm not sure I did it last year, but I think I blame Prashanth if we were wrong. I can't blame anyone except myself. I'm wrong...

And then circling back to your own inventory and utilization rates, what's inventory now in terms of days? What's the goal? And then what are we thinking on utilization rates these days?

So we brought our inventory down by about $300 million from the peak. So it peaked out, I think fourth quarter last year. Since then, we brought it down about $300 million. Days are still elevated, but they're about 170, 175, which is a bit higher than we like, but revenue is very low. Our view is now inventories normalize in our balance sheet, we'll grow into those days, meaning that revenue grows, days will start falling. We don't have a specific target, but 170 is too high, and I would say 120, 130, the old mile is too low. So split those 2, you're probably not too far off. So that means utilizations have bottomed last quarter and are starting to pick up here a little bit as you go into the back half of this year. So gross margins at bottom, revenues bottom and operating margins at bottom. So we've turned that corner. Now it's the shape of that corner and the upturn, that's still unknown.

Would you say that your utilization rates will reflect demand? So if a quarter is up, then utilization would go up. If the quarter is down, it would go down? Or will you have more of a smoothing effect?

Is more or less smoother than that. So typically, 1Q utilization rate usually down, because we do maintenance around the holidays. That's just normal maintenance we do. We didn't do it during COVID because you need all utilization in all the products that you could get. So usually, utilization is flat to down in 1Q because of that week of maintenance. But you're right to think, utilization is more or less to track demand now. Because we've done burning inventory. But we do have some -- a lot of inventory still, so what's going to happen is, even though demand picked back up, utilizations won't ramp as fast as you might think because inventory is still 170 days. But you're right to think it will go up as demand comes back up. And the question I get from investors is, when can we get that 70% floor that you blue passed on the way down? To think about that 70% floor, we need at least $2.7 billion of revenue. I think we get there some time in '25, but it feels like the first half of the year might not be above that, we'll see. But just how to think about it, what is the fall-through of gross margins and when you get back to 70%, $2.7 billion plus is probably a good number to kind of put in the back of your head.

All right. And then does mix or pricing have anything to do in terms of the...

Good question. On pricing, no. I mean pricing is, I'll call it -- we make pricing gross margin neutral, is not accretive or dilutive to margins. So mix well, if Industrial doesn't come back, Industrial drives our gross margin, drives internal factory utilization, and it's also a good gross margin business for us. So if Industrial lags, I don't think it will, but lets say it does lag, that $2.7 billion might be a hard threshold to get there. But I think Industrial will be one of the leaders coming out of this, given it fell over 30% this year.

Maybe rank the gross margin drivers going forward generically, top 3, top 4. How would you rank them?

So product portfolio drives a lot of it. So the innovation we bring to the market and ASPs of those products drive gross margins. That's number one. But from here, it's mix and utilization.

Utilization more than mix or mix more than utilization?

They're very similar.

Okay.

They're very similar. And if you look at peak to trough for us, we're at 74%, trough 67%, half of that decline was utilization and half of that decline is mix. So it will be similar in the way up if the ramp is similar to the decline. So really, pricing is not a gross margin drag for us or accretive. We don't want pricing to be gross margin accretive. We want our ASPs and innovation we bring and the value we charge our customers to drive gross margins, and then it modulates around that based on mix and utilizations.

Okay. And then you mentioned something earlier about your inventory goal is to be 120, 130. Now it's 170. It's going to be higher than 120, 130. We've heard other analog companies talk about their inventory goals being a little higher than they used to be. Why is that?

So good question. So what we learned in the downturn is, we can hold more on our balance sheet, less in the channel and keep that inventory on our balance sheet at die bank. It's cheap to hold. You don't reserve it very often either. And it's very quick to turn that die bank, go through the back-end cycle time and turn it into revenue. So we'll hold more small lots of die bank on our balance sheet to respond to fluctuations in demand quicker. I think that's for ADI. I don't know others, but that's probably some sentiment around across most companies, yes.

The same thing. And no increased risk associated with the higher inventory?

So no, minimal. I'll say very minimally. Because when you keep it at die bank, it's fungible across different markets and different customers. Where you're getting trouble in inventory is, if you move that die bank into finished goods, then there's more risk associated with it. So keep it at die bank, such that you can send it to different customers, different markets and where demand is. So that's why we keep it there, because of the very low risk association with that die bank.

Okay. And then just one more on gross margin, a bit of a longer-term question. Can your gross margins get back to the previous peak? And if so, what would it take to get there?

To get to 74% again, it's possible, but it will be challenging. And why will it be challenging? If we looked at when we hit 74% last time, Industrial was a record mix, 53%, and utilizations were 95% plus. Normal utilization for us is 85%, 90%. So mix utilization dependent, making that 74% challenging. I'd not say it's impossible, but to think [ 72 ] -- [indiscernible] we should get back to 72%, 73% gross margin. And I do think we can get operating margins back to the upper 40s, 50% again from a leverage standpoint, as you grow back into the model.

So if we did get that back to that mix of industrial and that 95% utilization, would you get back to 74% gross margin or would be something...

You asked the question, but we need a higher revenue base. Could we have more capacity today than we had when we hit it last time. So you need more revenue than you did then. Because if you think about it, we did -- I think we were at $12 billion of revenue. We added capacity since then, so we need more revenue to fill that utilization to get up there. But to your point, if utilization gets back to where it was and mix goes back to where it was, there's no reason why we can't get there again.

Great. And then how does it work with foundry? Couldn't you have less reliance on foundry, take more internal and increase utilization rates that way? Maybe just give us a refresher course on what the split between foundry and internal...

Good question. So we're about 50% internal and 50% external from that standpoint. And we have swing capacity, meaning we can build this product at a foundry or internal ADI. And what we've been doing during the downturn is bring more in-house. Our utilization will be lower than they are if we didn't have that swing capability. But really, we like that a hybrid model because it gives us flexibility of where we build our products, gives us flexibility around what process technologies we can use and it gives our customer flexibility to -- if they -- what kind of -- do they want supply from U.S., Europe, Asia, we have that ability to. So customer and resiliency, optionality in the financial model is why we like the hybrid model, and we'll invest in the hybrid model, we are. We put a lot of more capital in the ground in the past couple of years on the CapEx side to enhance that model, and we're working more with our foundry partners to increase exposure to other areas outside of Taiwan. We announced a quarter or 2 ago about using TSMC Japan. So you're seeing us really reinforce our hybrid model over the past couple of years, and we'll continue to do so as we go forward.

Okay. I have a couple more questions on that, but getting towards the end, I would be remiss if I didn't throw it out to the audience in case anyone has a question out there. Going once, going twice, probably, we'll feel free to keep flapping my gums up here. Usual. Okay. So why not have a 300-millimeter fab, especially given how big you are and you're about to buy [ microchip ].

Untrue. Chris...

You think I'm kidding.

Chris, it's a thesis. Yes. You're done. Nice knowing you, Chris. Good conference. Next year, I'll see someone else here.

I'll be washing dishes in the back.

There you go. Make more money that way now. The -- lost my train of thought. Oh, 300 millimeter. So why? Because we pick up the phone and call Texas and use one of theirs. No. Just kidding. So why would -- we don't have the volume. We are value-based. We don't have the volume to fill those fabs. If you look at our business, we're a 90-nanometer and larger node company, 90, 135, 180, as we do internally. That's like 90% of revenue is those nodes. Sub that, 300 millimeter makes more sense, we're talking 10% of our revenue. To fill a fab and make it productive, we need more revenue than that. So from the standpoint of, can we do it? Sure. We don't need to do it. And everybody thought if we need to do it, we would. But we have the capacity in place for very strong revenue growth in the next 5 years. And we look at it every year to see does it make logical sense to do it. And I think it really comes down to, our goal is not to ship more chips, it's to ship more value per chip. So we're not counting chips. We're recounting dollars per chip, and that's how we make our business.

Do you see your foundry mix continuing to increase over the next 5 or 10 years?

No. I think it stays stable around 50% plus/minus. I think maybe if you look at the next internal go up a little bit, but I'm talking we're like 45%, 50%, maybe go to 50% to 55% internal. And then really -- we don't say this is internal, this is external. We say, 180 and above, we have capacity for that, and we'll fill that out. If we have too much demand, it goes external. Below that, it goes external all the time. So we don't really say we want this. We don't target a percent, we target the nodes, and the mix is dictated by the nodes. That makes sense?

Yes. Does your foundry business in general carry higher or lower margins than the internal business? Or is there pretty much node [indiscernible]?

It's market -- it's minimal difference between the 2. And basically comes down to, if you have the sunk capital, then filling up of internal fab is more productive. But if you add a fab or add capital, it's very cost-intensive until you hit a certain threshold. So for us, like-for-like, we don't really care if we make it internal or external as long as it drives good margin for us, and the margin really comes down to what business is it selling into more than anything.

Got it. Okay. I did get a couple of people wanting me to ask you about CHIPS Act. So when do you think you will start to get the money, talk about the tax credit, when that impacts? What are we going to see some...

I'll back to ITC credit, [indiscernible] everything we spent in the past year or so. It's on our balance sheet. I want to say we've accrued $300 million for ITC credit. That's already impacting the P&L minimally, meaning when you put the equipment in place, the cost of that equipment is 25% less than it would have been, therefore, the depreciation over 10 years is less. And again, that's a minimal impact you're seeing there. We'll keep accruing more ITC given our CapEx spend. On the grant side, we apply to Europe and U.S., we're still waiting to hear. It's a long process, and I can't do a time line on when do you think we'll hear.

Got it.

But I would say, our numbers won't be as big as others. We didn't spend as much as others. We went in, our goal is to reinforce our hybrid model, not build out a big footprint of internal capacity with the CHIPS Act money.

Okay. And then last question, you guys continue to generate a ton of cash. Nice dividend, goes up every year, knock on wood. Your buyback has been pretty steady. As this cash continues to build, what is the use of that? Just continue to go dividend, buyback equal, something else out there?

That's pretty much it, dividend, buyback equal. And you're right. If you look at our CapEx, our CapEx is coming down. I mean, we did $1.1 billion a year ago. This year, $700 million, and you look to next year, dollars in CapEx will be down again next year in percent of sales. So that will help free cash flow. And you're right. We look at our free cash flow as -- our goal is to grow our dividend 10% through cycle and everything after that, less free cash flow go to buyback. So really, our target is 100% free cash flow return, take it with the dividend, and the rest goes to buyback with a goal to keep share count flat. And usually, our goal more or less is to take it down over time.

Great. With that, we're out of time. Thanks, Mike.

Thank you.

Thanks, everyone.