Allegro MicroSystems, Inc. (ALGM) Earnings Call Transcript & Summary

Good morning, and welcome, everybody, to the 25th Annual Needham Growth Conference. My name is Quinn Bolton. I am the semiconductor and semiconductor equipment analyst for Needham. It's my pleasure to host this fireside chat with Allegro MicroSystems. Allegro is a leading supplier of magnetic sensors and power ICs, targeting the automotive and industrial markets. The company holds #1 share position in mag sensors worldwide as well as the leading position in current sensors. We believe Allegro is well positioned for a recovery in automotive production and for growing content per vehicle driven by the trends towards safety and ADAS as well as electrification. Joining me from the company are President and CEO, Vineet Nargolwala; and CFO, Derek D'Antilio. We also have [indiscernible] Hoover from Investor Relations in the audience. Vineet, Derek, [indiscernible] , thank you for joining us.

Good morning. Thanks for having us.

I wanted to start just sort of to sort of frame the automotive market with a question on automotive production. It looks like automobile production was up by a mid-single-digit percentage roughly in 2022. Your revenue was up 21% approximately. We've seen this outperformance through many other automotive semiconductor suppliers. Could you give us some sense -- how much of that revenue growth came from content gains versus potential inventory build at your customers?

Yes. Quinn, thanks for the question. For us, we're really not seeing any inventory build. So even though automotive production grew single digit, what we're really focused on is the transition to electrified vehicles and the increased adoption of ADAS feature sets. And that's really driving our content. And so when we look at our performance versus market, we feel really good about our exposure to the higher content applications. And we think that, that's a secular growth trend that's going to play out over the next decade. The transition to e-mobility, which is the combination of electrification as well as adoption of higher-end safety features of ADAS feature sets.

Got it. And do the auto OEMs give you pretty good insight into the amount of all components they're holding? I mean do you have a pretty good look into what they're either holding and/or what the sell-through is through the automotive channels?

Yes. So we have, over the course -- we've been in automotive for close to 3 decades. And really through the pandemic, we built deeper supplier relationships with the OEMs. We've always had great design relationships and engineering relationships, but we become much closer to them from a supply chain standpoint. And we are very careful because we've been in a supply-constrained environment, not to ship it to inventory. And so we have daily, weekly calls with our major OEM partners, where we're getting checks on their bare demand. And we're still in a supply-constrained environment. It's getting better, but we are still supply constrained. And so we are very careful to make sure that there's no inventory being built up, and we don't see any evidence of that happening.

In your conversations with the automotive OEMs or your view of third-party market research, what's your outlook for automobile production in calendar '23, '24? I mean, I think we're sort of in the low 80 million range in '22. Do you see that flat, up or how does that trend over the next few years.

Yes. So we haven't really given any guidance and we're still sort of calibrating our views on what is going to be our fiscal 2024. I'll make some comments just sort of at a high level. As you pointed out, we finished in 81 million, 82 million units of production last year. That's a far cry from the 94 million units. We were -- the world was at not too long ago, right? And we think there is still room to run in terms of just overall production levels. From our perspective, equilibrium is probably somewhere around 90 million units of global auto production. This is where supply and demand are more balanced. You have more inventory on dealer lots, pricing is more reasonable. OEM lines are running at a high utilization. And so the industry is still out of whack from a balance standpoint. And so we think that production will continue to run upwards until we get to that sort of 90 million units, right? And do we get there in a couple of years or 3 years, who knows? But that's sort of the trend that we would expect.

In the last couple of years, we production was constrained, it feels like the auto OEMs sort of mix shifted up to higher-end vehicles just to maximize their dollars. As we get back to that more normalized production level of 90 million vehicles, do you see any sort of mix shift back to more low-end or mid-priced vehicles? And could that be a headwind to your content gains that you've seen over the last couple of years?

Yes. So it's difficult to predict where the OEMs are going to drive their overall portfolio mix. I think one can expect it to be -- to have more sort of broad availability of models across the spectrum. I'll tell you, for us, our ICs are not driving nice-to-have feature sets in the car, right? We are going into powertrain. We're going into safety feature sets, which is essential for the safe and efficient operation of the automobile. And so regardless of what the portfolio that our OEMs put out, we feel really confident about our ability to grow above market because of the applications we serve, whether it's in electrified powertrain, whether it's an adoption of feature sets around ADAS or even in the, what we call safety, comfort and convenience, the content that we are part of is going to continue to increase.

I wanted to move to a few questions on backlog and orders. On your last earnings call, you stated that your backlog extends over a year and that automotive backlog actually grew quarter-on-quarter in September. So a few questions first. You've previously stated that you've opened up your order book to customers to allow for cancellations to try to scrub the order books to see what real demand is. Can you say how customers responded to that opening of the order books?

Sure. So about 70% of our business is automotive, another 20% is industrial. And as a result of that, over the last few years, we've entered into like many people, preferred supply of programs, long-term agreements. And we have built up over a year's worth of backlog, and we've had that for some time now. But we are operating and have been operating in a supply chain constrained environment, particularly to wafers, so 200-millimeter wafers. And so we went through and looked at our backlog and said, where do we really want to ship these wafers, where do customers need them now and speak to the question on inventory between OEMs and distribution. And distribution sort of bottomed in Q1. And we went through in the summer time and opened up the largely noncancellable, nonreschedule backlog and asked long-standing customers, particularly with orders that we were past due on, do they want to reschedule those or cancel those? And we had some uptake. When I look at that backlog, backlog declined 2% from Q1 to Q2 as a result of that exercise. And it was really skewed towards what we'll call the other portion of our business, which is the last 10% to 12% that includes some interesting consumer and smart home applications, which we expected, and some of the long tail industrial. That allowed us to actually direct some of these scarce wafers at the wafer level downstream into applications where there's immediate demand.

I was going to ask is, were you providing incentives to cancel orders that may not be needed by saying "Hey, if you cancel the parts you don't need, we can pull forward perhaps the critical parts that you are most short on." Or are there working capital requirements for the customers are saying, "Hey, cost of financing inventory has gone up, and so we will actually try to take orders?"

It's actually a little bit of both. So the incentives really work for the customers to make sure that they're getting the right parts at the right time, so there were active discussions within customers about which parts do you really need, help us give us some more visibility so we can give you the right parts. So as Vineet mentioned, that's given us a lot more supply chain connections with our customers. And on the other side, there were orders that were just past due that we've asked to reschedule. So I would say there was a higher proportion of those orders that were rescheduled and canceled, and it was really to help us balance our own supply and demand internally as well as our customers.

With backlog at roughly a year or better, can you talk about your lead times? How are they trending? Are you -- have you been extended with lead times, and when do you think they might normalize?

Yes, I can take that one. So we have been in extended lead times for better part of this fiscal year for us. We do see things improving, and we are working with -- in targeted applications to significantly reduce the lead time. We are confident that by, call it, middle of this calendar year, we will be able to get back to normal lead times or close to normal lead times.

Okay. With lead times compressing, we've heard from other customers that tends to affect the order book. You know if I don't need to place orders out 12 months, maybe I'll hold back. Have you seen any changes in order patterns more recently, whether it's because they're shrinking lead times or just because of the global uncertainty, threat of recession?

Yes. I would say that it depends on the market. In our automotive business, which is, as Derek pointed out, 70% of our business, we have not seen any change in order patterns, right? And there we have a lot of visibility based on the design win cycle. So our leading indicator is really the funnel that we work, the design wins that we score that then turn into backlog a few years down the road. So we have a lot of visibility into where our OEMs are going and what to expect directionally from an auto patent standpoint. I would say that in some of our industrial businesses, we have seen some shifts right? I think we were -- we talked about data center in our last earnings call, where we believe we are in a period of digestion of some of the inventory that was taken earlier. Our design cycle and momentum in data center and really overall in our industrial segments across clean energy and automation continues to be very robust. So we feel really good about the long term. There might be a little bit of a digestion period here as things settle out in data center.

Just a question on sort of the e-mobility design wins. You mentioned last quarter, they were up 68% year-on-year, sort of well ahead of target. Can you help us understand what the design win to revenue cycle looks like on the e-mobility side? And what are the e-mobility design win targets? I don't know if you sort of specified what those targets are?

Yes, we have not specified what our targets are. I will tell you that e-mobility is now shaping up to be the majority of our sales activity in automotive and certainly represents an unfair share within our funnel as well as our recorded design wins. Typically, the design win to revenue cycle can be anywhere from a year to 3 years, depending on the application, depending on the customer. In cases where we are winning with something that we already have on the shelf, so it's an existing part, just testing is required, the design to revenue cycle might be shorter. Where we are starting from scratch, we're designing a brand-new product, there the design cycle is going to be more typical of automotive, which is 3-plus years.

I wanted to move to some supply chain questions now. To what extent did wafer or foundry wafer supply constrain your revenue in 2023? And how are you feeling about wafer commitments from your foundry partners as you look into calendar '23?

Yes. We've been in a wafer constrained environment for the past almost 2 years at this point. We have 3 major suppliers for wafers and the latest of which is TSMC, which we've engaged with for the past 3 years. TSMC has committed to giving us approximately 15% of our overall wafers coming from TSMC. And they're largely there right now. So they're about a quarter ahead. Still in a very constrained wafer environment for this year. And so when we started this year, we talked about we grew 30% growth -- revenue growth in fiscal '22 and we talked about having a growth year just below mid-teens here in 2023 fiscal. And as we move throughout the year and got incremental supply from each of those suppliers a little bit ahead of the TSMC ramp, a little bit more from our partner, Polar, and a little bit more from UMC, we were able to continue to increase that expectation of revenue growth in the year from below mid-teens to about 20% in the last earnings call of 24%. And so that expectation of revenue growth is really all calibrated to our comfort with the supply on the wafer side, given that we have over a year's worth of backlog, which really all supply constraints still at this point.

And as you look to calendar '23, have you completed the negotiations for wafer supply for this calendar year? Or are those negotiations still ongoing?

So it's actively ongoing. So our fiscal year starts April 1. It will be our fiscal '24. So they're active ongoing discussions with our wafer suppliers. And what we found typically is TSMC is really good at doing what they'll commit to do, and they're largely where they said they would be. And Polar has been a great internal partner for 18 years, and we have a great relationship with UMC. So those are actively ongoing, and we're balancing all the factors of when we want capacity; two, the right technology; three, of course, cost; and then the fourth one is geopolitical balancing having a wafer supplier in North America and the other 2 in Taiwan.

Probably up until about a year ago, I think you mentioned that the TSMC, UMC, the Taiwanese foundries were able to get you lower wafer costs, and there might be gross margin advantage to shifting more of the wafer supply to those Taiwanese foundries. I think in the last year because the Taiwanese have increased wafer pricing, and I think Polar has become a little bit more competitive, the gross margins may have balanced out. But is that still the case? Are you pretty agnostic at this point from a margin perspective where you source the wafers or do you -- is there still a preferred vendor from a margin perspective?

Yes. It's not completely agnostic. Cost is certainly a factor and they're different in prices. But over the last 2 years, TSMC has been very aggressive in the market in terms of pricing and 200-millimeter is not their sort of leading edge. So they've been very -- even more aggressive there. So I think there's been some normalization of pricing between sort of the high end and the North American suppliers, particularly at Polar, and Polar has really come up the ramp in terms of scale and their facility to what they can do right now, so they've made some cost improvements. So I think it's become less material in terms of the mix. There's still some mix balance we can do there, but it will relate to technology, cost per wafer and gross margin per wafer as well.

You mentioned TSMC now up to 15% of wafer supply exiting fiscal '23. Does that have room to go higher? I know you haven't given a new target yet for '24, but should we think over the next 2, 3 years that the TSMC could become a bigger partner? Or are you pretty comfortable with the mix as it exits '23?

It will really depend on the fact as I talked about. There's a technology factor. There's a cost factor, of course. There's a regionalization factor. So we haven't really -- we're in that mix discussion right now where we'd like to get our wafers from. So I don't think we've committed to it being any higher. About a year ago, we said we'd get to 15%. We're there about a quarter earlier than we would be. Still working through some yield issues, quite frankly, and that happens as you take up a new supplier. So it's a really long cycle time to bring on a new supplier in the wafer world to begin with, but then when you put on the automotive grade, it's even longer.

Last question on sort of the supply side, wafer pricing has increased over the past couple of years given how constrained the environment is. As wafer foundry utilization start to loosen up, do you expect to see wafer pricing still moving higher for you through calendar '23 or fiscal '24? Or do you think it starts to stabilize?

Those are active discussions, as Derek said, obviously, our hope is not, but that continues to be an active dialogue for us.

And when you look at the headline numbers of utilization of like UMC or TSMC it's a bit more nuanced because we're in 1 particular fab with certain technologies on 200-millimeter, 0.18 micron. So that's automotive grade. So I think there's a little bit of nuance even in their own pricing.

Okay. I want to move to the magnetic sensor market. Maybe just to level set everybody, can you talk about who the primary competition is in current sensors and magnetic sensors? And sort of how do you distinguish what all sets Allegro apart from the competition in that market?

So magnetic sensing is obviously a very large market. There's a lot of players. We are the leader by any stretch of the imagination with over 20% share. We've been at it for the longest, I would say, and we've been really focused on that, and we've built leadership both through IP, but very specifically in automotive, we're the leader by far. And that comes with decades of year and of experience, customer intimacy, understanding of applications. So it goes beyond just the product set. And obviously, and I said this before, we've built up a really strong IP portfolio that serves as a barrier to entry. I would say what sets us apart is not just the product, but how we understand the applications within those products or that the products get applied into. And so everything we do is automotive grade starting from the substrate on. We understand how end customers engineer the products, how they source the products. Our engineering teams work very closely with our customers' engineering teams to solve problems as they're emerging. And so that deep customer intimacy and application expertise really lends itself to innovation that is very targeted on those applications. And so that takes years and years to overcome. So when I think then about current sensing specifically, we've translated that years of expertise in magnetic sensing into really leadership in current sensing. Today, we have probably the most accurate, most robust, most reliable current sensor on the market. And we continue to innovate as we move the industry from Hall-basis to xMR based technology, which will bring more accuracy and more robustness to these applications. There are over 40 current-sensing applications or sockets, if you will, on an electric car. And they are continuing to increase in terms of the demanding nature of those applications. So higher voltages, higher power ratios, more efficiency being required, and we are up to the task with a big chunk of our R&D going towards innovating in those areas.

Are those applications mostly in the traction inverter and the onboard chargers? Or is it a pretty diverse set of applications?

It's a very diverse set of applications, right? When you think about an EV, you've got a 400- or 800-volt power source, and then everything on the EV is operating a 12-volt, 24-volt, 48-volt. And then you -- so you've got a pretty big power conversion challenge. You've got to do it safely, and you got to do it efficiently. And every time you do power conversion, you need to measure the current before and after to make sure we've done that power conversion safely and efficiently. And then you've got the traction motor, which is operating at AC, and so you've got a DC to AC power conversion there as well. And so you definitely need current sensing all across the EV. But then there are applications like steering columns, where you wouldn't think you need a current sensor, but there are motor drivers, more importantly, our motor drivers that are operating most of the steering columns in the world, and you need current sensors to make sure that your -- the motor driver is operating efficiently. And so there are multiple applications across EV, which lend itself to our current sensors and the type of robustness we offer. And so we feel really good about our leadership position there.

You've said your strategic areas of EV and ADAS increased to 41% of the automotive...

Sales.