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

Thank you for joining us at our 50th Annual TMT Conference. I'm Josh Buchalter, semiconductor analyst at Cowen. Very pleased to be joined by Ravi Vig and Derek D'Antilio of Allegro MicroSystems. As Ravi just informed me that this will be his last conference as CEO. So I guess for those who haven't read our initiation report from last week, would you mind introducing yourself and the company?

Thank you, Josh. Yes, it is my last conference. I've had a good run. Allegro is a semiconductor company and located in Manchester, New Hampshire, focused on automotive. Great secular growth factors: revenue about $768 million last year, great profitability. We came out public in 2020, in the fall of 2020 with an operating model that we had expected to get to about this revenue and gross margins in the 55% range, establish growth vectors in xEV, ADAS, industrial. We've achieved a lot of those milestones and -- a little bit earlier than we had expected. So I thought it would be time to hand over the reins to my -- to a successor. So that's a quick summary.

I guess all it took was us initiating coverage for you to quit. But would you mind sharing some background on your successor, Vineet? Why he was chosen? And without pre-announcing anything that he's going to do, what you think he'll bring to the table?

Well, we've got great tailwinds as a company. We're coming into this with a high double-digits sales growth rate for the year. We are looking at gross margins becoming sustainable in the mid-50s towards the end of this year. We've got a solid company that he's coming into, design wins, secular growth vectors in automotive xEV. So I think the challenge for the company now is -- we've got line of sight towards $1 billion. The challenge for the company now is to start scaling, scaling every aspect of the company, whether it is internal infrastructure or whether it is through M&A in terms of driving additional growth. And I know that when he comes in, he'll sit with the team and go through some of the long-term strategic planning that we've already done and validate that or make his mark on it.

Got it. Before we dig into the questions I'd actually prepped, you guys announced an acquisition yesterday after close of -- Heyday Integrated Circuits. I believe it's gate drivers for targeting compound semiconductors. Can you discuss -- give some background information on the company? What it brings to the portfolio? And how it fits with what you're doing in power today?

So our power business pretty much is very heavily focused on gate drivers. So we currently in automotive, especially power steering, braking, et cetera, have great expertise in driving the lower voltage end of the gate drivers, MOSFETs, et cetera, for 3-phase motion control. But along with that comes a lot of redundancy, aero detection capabilities, ISO 26262, which is a safety protocol that you have to be able to design to -- a lot of deep expertise there. And as we look at the industry, we see that electrification is happening. Current sensing is what we do in electrification. We've got great market share in current sensing at this point. And there's a parallel space there in the inverters, for example, where we do the current sensing to be able to drive these compound semiconductors. And it is a technically difficult problem. It is a difficult problem to build the compound semiconductors, but it's also a technically difficult problem to accurately drive them, level shift them and bring some signals coming off a 12-volt rail. So Heyday has made that task easier. They have a power supply -- mini power supply module that they do, which takes low voltage signals in, convert them into -- isolates them and converts them into a high-voltage driver, optimized for the speeds of SiC and optimize for the speeds of GaN. So we think that this is a solution that has great legs as these compound semiconductors start getting adopted. Now you will hear -- I'm sure during this conference, you will hear other companies talk about the compound semiconductor growth trends. And we can tell that it's a tremendous market. It adds about $2.9 billion to Allegro's SAM, which is a very big number. It's almost a 50% increase in our SAM. So we're excited about what this offers.

And so these are CMOS-based gate drivers that would go in a module alongside a silicon carbide or gallium nitride MOSFET. Is that correct?

That's correct. So it is a CMOS-based multichip with isolation in between to allow the driving of these -- high side driving of these high-voltage MOSFETs. And the problem is quite difficult because these compound semiconductors switch at very high speeds. They need the current very well controlled, otherwise they get damaged. They radiate a lot of noise. So having a very well-designed module makes the design problem much easier for the customer.

Got it. And a last one on the deal, maybe for Derek. Anything you can share on expected financial contributions or -- on the…

Yes. The deal came with virtually no revenue at this point, no material revenue. We expect material revenue to be in about 3 years, so call it fiscal year '25 for us. We will be incurring about $5 million in OpEx this year and we'll have some incremental OpEx the next couple of years. We expect the gross margins to be at or above our corporate margin here given the fact that the value that these bring to us.

Got it. I guess shifting to the auto market. You've been a supplier for over 30 years. Any perspective you can give for both your positioning and the broader industry just given all the noise we're hearing in semiconductor growth that's growing so much faster than SAR. Where do you -- are you comfortable -- I know most of your sales are direct, but what can you tell us about what you're seeing out there?

Yes, we do have -- we have been in the auto industry for quite a while and we do see these cycles. What's different at this point is that we are going through these megatrends in the automotive industry. And there's 2 basic trends that are happening at the same time and they intersect. So one part of the trend is the electrification of vehicles. And for example, last year, electrified vehicles grew 40% year-over-year even though car production only grew between 5% to 7%. So a tremendous adoption rate. What electrified vehicles do is they increase for Allegro content per vehicle by over 50%. So it drives growth. And the other part of it is ADAS. So ADAS, we look at autonomy as sense, think and act. The sense being the camera modules, camera LiDAR, et cetera. Think being the CPU. And then act being the steering, braking systems. And as you get to that level 1 plus, level 2 kind of LiDAR -- I mean, ADAS, rather, you start ending up with steering systems that are self-controlling the car. For example, today, a lot of cars have lane keep assist. What that means is it senses the lane -- that the camera senses the lane. It then adjusts the steering wheel. Steering wheel control technology has become enormously more complex. Semi content has doubled or tripled within the steering system, for example, in these kind of vehicles. So there is a -- so when you start looking at the semi content, you've got the -- in automotive, because of these megatrends and because of the infotainment that's occurring, lighting systems that are being upgraded to handle the cameras, et cetera, the semi content has been growing quite dramatically. So we believe that at least from an Allegro perspective, the demand is real. The backlog is real. We are running at -- our demand really shows the trend to these new businesses, these new -- and the content acceleration that's occurring here. So we don't see pockets of inventory at this point at least to our visibility.

Got it. By the way, this is meant to be interactive. So if any investors in the audience want to ask questions, feel free to jump in, wave your arms, throw something at me, whatever you want. I guess if I could ask that question in another way. You mentioned you have, I think, 60% of content bump in electrification and I think $10 to $12 in autonomy. Any metrics you can give on how much of that is your magnetic sensing portfolio versus power?

Yes. So it's really -- magnetics is a predominant part of our business. It's 70% of our -- 65% of our sales. Power being the balance, power being 50-50 in automotive -- maybe it's a little more than 50% now in automotive versus the rest of the business. From an Allegro perspective, our -- we see the growth in both areas. Power is actually having a much faster growth pressures just given the motion control aspect of power. So motors in the vehicle are moving to 3-phase motion control -- motor control. And it's right in our sweet spot, whether it's the gate drivers that I spoke about earlier or whether it's embedded fan drivers, et cetera. So -- and that is rapidly expanding as vehicles electrify. You eliminate all the hydraulics. You eliminate the vacuums. All of that goes away. It's all going to electric motor control. So -- and it fits right in our sweet spot.

Got it. And then you mentioned in a lot of your materials about moving from, I think, L0 to L1 adds $10 of content. I mean does it adds into -- there -- does it continue to layer in as we move into higher levels of autonomy? And then, bigger picture, any update -- I know you guys a few years ago acquired a LiDAR asset. Any update on that and when you would expect that to contribute to revenue?

So what we see with the acting piece of vehicles is that they're actually being reimagined. So you've got the steering system that is increasing in complexity. More and more content will go up there to that $12, and more vehicles will increase in their content. But as we also look at it, there is the whole skateboard design of battery electric vehicles, and the skateboard design eventually starts talking about moving steering systems from within the vehicle to per wheel. So you start getting a multiplicity of motion control systems to control steering. And it's a similar concept on braking. So braking systems also, which today are electrically assisted, are moving to full motion control where you start driving the caliper with an electric motor and retracting the caliper with an electric motor. So -- and it also gives you regenerative braking. These are all brand-new systems that are being reimagined, but they work very well with the skateboard platform because then the stalling of the vehicle versus the [ steering ] of the vehicle can be separated from the chassis itself and it makes the battery electric vehicles much more scalable. So we are very bullish on motion control in automotive.

I think on the last earnings call, you mentioned, I think, around 36% of revenue -- of auto revenue was tied to ADAS and electrification. I mean how much faster is that growing? What do you expect that portion to be a few years from now?

We continue to see that outgrow the market. So I don't have the data on what that -- when we expect that to be more than a majority of our business, but it certainly would be -- within 2 to 3 years, we'd be heading in that direction.

And moving on from autos. I was surprised here on the last earnings call that you expect the data center to be your largest growth driver on both dollars and a percentage basis. Can you talk about what you're playing there? Is it predominantly 3-phase cooling fans? And how much is it power delivery today?

Yes. So we -- a couple of quarters ago, we had announced that we had signed major agreements with multiple data center cooling systems providers, and we were anticipating that the backlog would really fill in nicely, that the projects would fill in nicely. And they have. And last year -- our fiscal year ends at the end of March. So last fiscal year, we had -- it was about 3.5% of our sales. This fiscal year, we anticipate it to be 7.5% sales, but of a higher sales number. So we are guiding the market today towards the high teens kind of a year-over-year growth rate. So data center is really a major piece of it. What Allegro offers there is a fully integrated cooling 3-phase motion control solution. So again, all the IP of how you drive a cooling fan and how do you communicate with the CPU is all embedded into the IC, which replaces a multichip solution, a complex PC board. That is the alternative, and it's been very well accepted. So we're pretty happy about the progress in this.

Any idea -- any handlebars you can give us on what's the content per server opportunity that you're addressing here? And…

You could have 3 to 5 cooling systems per server. For example, you could have smaller power ones sitting on the board. You could have some at the back panel. And then you could also have one much larger blower for the entire server. So it depends on the architecture depending on the manufacturer. But in general, cooling systems are a great area of focus because the power levels are increasing in the server -- in the server farms, but also energy consumption of the cooling systems has become a big problem. So 3-phase systems really increase the efficiency of the cooling system.

And we cover the compute companies and they had some well-publicized server refreshes coming in the back half of the year. Do we think about that as a catalyst for you? Or is it just more slow and steady consistent growth and execution?

We're almost agnostic, right? Because there's a conversion rate happening in terms of the current technology because even the current technologies need to become much more efficient. And so you've got existing technologies being converted. But the newer technologies start looking at even more cooling required, because the horsepower, if you want to call it that, of the processors is increasing, the heat generated is increasing and you need better cooling.

I guess moving more broadly within Industrial. I think we as investors struggle with how to -- what are the big chunky areas of growth? What should we be monitoring there other than PMI? Can you talk about where you fit within electrification in a broader set of industries? What are the sort of bigger, chunkier wins and verticals that you could be addressing?

And it's a really interesting segment, because as you said, industrial is really a big giant catch-all, right? And it -- but it's made up of small areas of energy, if you want to call it that. And for us, electrification is a piece of it. So when we talk about electrification industrial, you talk about solar, for example. That's a piece that is -- goes through cycles, right? And right now, you could anticipate that solar is -- with energy prices going up, solar is going to increase again in terms of production capacity. But we also talk about the electrification for infrastructure. So when you talk about electric vehicles, the whole charging systems and charging stations, that's a big area for Allegro that we are -- I think last quarter, we announced that we were in -- we had 10 major design wins in Europe in terms of infrastructure -- charging infrastructure electrification. A couple of quarters -- it bounces around. It moves around. A few quarters ago, our largest opportunity ended up becoming the current measurement for factories that were building batteries. So it bounces around. But the general theme is still the same. Power measurement is a big issue and we've invested heavily in power measurement. But it bounces between these. But we also get robotics as being an area for our motion control products, both our angle sensors as well as our motor drivers. We also get factory automation, similar kinds of things that we see a lot of growth in.

I mean you're well entrenched already in autos and the industrial businesses -- like do you spend as much time on industrial and charging infrastructure as you do in autos? Or like how big is the long-term opportunity if you think about it in relation to what you're already doing in the automotive business?

Yes. So I'll answer it in a slightly different way. So we have our industrial and other businesses, server store, distribution channel. And we love this channel, because what it does is it connects smaller customers to our products and -- over 10,000 customers to our products. And these customers pay us over 1,000 basis points higher pricing than our OEM customers. So this is a major margin driver for the company. And so we've grown to love these small customers. And so we market them. We set up a dedicated distribution team that goes and reaches out to industrial customers and makes sure that they get serviced appropriately in all kinds of applications. So yes, the company is pretty committed to this particular space.

And how much of your business is through distribution today? I believe it's almost entirely the industrial segment, right?

It's 37%. And it's not entirely industrial, but 37% of our sales goes through distribution. Some automotive goes through distribution. That might be -- some of it might just be -- simply be fulfillment. But it also might be, for example, in China, a lot of Tier 2 and 3 car manufacturers buy through distribution. Interestingly enough, we had actually some business through to Tesla that actually went through a distributor that we had no idea that it was going through. So you find that even large OEMs are -- sometimes their supply chain ends up buying through distribution. But in general, most of our industrial and other business goes through this.

Is there a major difference in your mix of sensing versus power between auto and industrial? And then I guess within the power bucket, is there much of a mix shift between higher voltage applications in industrial versus autos?

We work -- as a company, we invested all our technology working in that 12- to 60-volt kind of space. So -- and when you say 60 volts, it really means about 100-volt wafer fab processes. So both auto and industrial line up very well. Previously, auto was very heavy in that 12 to 40, 45 volt range, and we did really well with the fab processes. Industrial has been moving in that direction in terms of much more precise motion control, et cetera. So it's lining up very nicely. And then we aligned our processes to the 48-volt auto bus, which lines up very nicely with the industrial 48-volt trends that you're seeing. So there's a lot of commonality, which is one of the beauty of our operating model.

I guess stepping back, you made a strategic shift a few years ago to move to a fabless model and divested your -- the wafer fab or semiconductor and partnered with UMC and now TSMC for your foundry partners. Can you talk about the rationale for that strategic shift? And what that's -- how that's allowed your business to change and be more nimble?

So when you are an IBM, you tend to be able -- you tend to be limited by the capabilities of your internal manufacturing. For those of you who've had any experience with wafer fabs, the problem is that you have to keep them -- well, you have to keep them utilized. So you tend to align your product road map to what your fab can do as opposed to get your manufacturing capability to what your product road map wants. It's a little bit backwards. And for a small company like Allegro -- I know it's $750 million. But in the world of semis, it's a small company. For a small company like Allegro, we didn't have the gas to be investing in advanced -- internally investing in advanced wafer technologies as well as investing in R&D to develop products. So we felt that we are better suited to allocate our capital to developing these -- co-developing these processes with our foundry partners who have really great technology. And in addition to them be -- investing in -- and invest our internal capital into R&D.

When you say advanced foundry technologies -- I believe the vast majority of your portfolio is on a 180-nanometer. Is it on geometry or is it other parts of the process?

Well, it's -- 180-nanometer technology for the foundries is old stuff. So they've got great experience in it. 180-nanometers for a small U.S. fab internally in the U.S. was actually very, very advanced. So it's a question of perspective, right? So for us, the 180-nanometer is actually -- add a 110-volt, being able to drive 3-phase motors. There are very few technologies in the market with that particular capability. And for us, it is a -- it was an existential technology that allows us to do things like we're doing for the data center products, et cetera, that allow us to embed these high algorithmic motion control products or sensor motion control products, smart sensor products. And that particular digital density is sufficient.

I mean the whole trailing edge industry has been constrained basically since COVID. Have you felt it? How much has that held you back? And then looking forward, you mentioned bringing on more capacity at TSMC in the latter half of this year. Can you talk about the magnitude that, that incremental capacity could bring? And what's going on there?

Yes. So we've worked with TSM for probably 3 to 4 years to transfer Allegro's wafer processes to them. So it's been a long road, and they've been working great -- very well with us as a partner. They've been ramping the customized tools that we needed to ramp our volume. Those are being installed as we speak. And that volume has started coming out of TSM. So we expect that TSM's shipments for us -- actually, our shipments from TSM wafers will start leveling out in our Q4. So we will continue to ramp all the way through Q3, and then we'll start leveling out in Q4 at a high rate. So our -- so we're guiding to high teens. It's rationalized the wafer capacity that's been available that we've been committed to by both TSM and UMC. But having said that, we expect to get additional wafers from UMC next year. That's long-term agreements that we have. So -- they are a wonderful partner for us also. And we're fortunate to have 2 of the best foundries out there working with us at this point.

And how severely have you been constrained in the last few years -- go ahead.

So what -- right now, we're fabs constrained, wafer constrained at this point. So we're not increasing the Polar amount because Polar is also at capacity, but we're not decreasing it either. So -- but net-net -- as TSM comes on, as UMC continues to ramp, Polar will continue to be a lower and lower percentage of our overall sales. As our foundry capacity continues to increase, we'll look at the percentage that we buy from Polar, but only in the event that supply exceeds demand. At this point, demand is so far ahead of supply that reducing Polar is not contemplated. Yes. So we have guided next year to mid -- let me answer it in a different way, because I don't have the exact wafer numbers off the top of my head. But what we have guided next year is the -- this year to the high teens and next year to the low to mid-teens. And have capacity rationalized with that. So what that means is that we've looked at all our wafer supply and we are able to support that in volume. Now it's not a simple math of saying, "Oh, that means you get 15% more wafers because die sizes and all are different." But we rationalize that all the way through to our product mix. But it's a rough number.

And then I think given the move in margins after you divested Polar, is there material margin difference between the capacity ramping at UMC and TSMC versus Polar?

Yes, I mean-- I mean -- yes, maybe you can take it.

Yes. I mean, we'll definitely see a benefit in the gross margins from having an incremental increase from TSMC and UMC. But in the near term, you're seeing both -- capacity increases from all of our suppliers in the near term. And then as you said, from our gross margin model, we're at the 55% where we stated to be publicly 1.5 years ago. There's still work to be done there in both leveraging our back end, there's work to be done on efficiencies and then die-optimization and those things…

I guess a question we've asked most of our companies the last couple of quarters. Any metrics you can give as we decompose the growth as to units versus ASPs over the last several quarters?

Yes, it's been pretty inconsequential. In 20 -- when fiscal year '22 just ended, about 3% overall of the growth -- the 30% growth came from price. And then moving forward into 2023, right here, we're expecting high teens growth, and that's rationalized to capacity. I'd say less than 20% of that is, what we have stated publicly, is going to be related to price. The rest will be units.

Got it. And financially, you keep exceeding your growth target and bumping up against the profitability metrics. And it's hard. What should we think about the long-term profile of the company?

Sure. When the company set out to be public, the targets were mid-teens growth rate, 55% non-GAAP gross margins and about 30% OpEx targets of revenue. And we're bumping up -- we're exceeding all of those. Revenue in FY '22 was 30% growth. We talked about high teens for '23. Gross margins hit that 55% or just north of that 55% here in Q4. And then OpEx, you'll see in Q1 at the midpoint of our guidance about 30%. So we expect to incrementally keep improving that gross margin. The next step up is, what I'd call, sustainably be at 55%. And then look for opportunities to head north of that, the next large step up there. And then on the OpEx, you'll continue to see leverage on the OpEx as we'll continue to invest in R&D. So that will be about 14% or 15% of overall revenue. And you'll see the leverage on the SG&A dip below 30% as we start to get towards that high teens revenue growth this year.

And I guess on the R&D line, what are the areas that you're prioritizing for investment? Is it for power delivery? Is it magnet resistance sensing? What are the primary areas that you're investing in for growth?

I mean the investments are really targeting our growth vectors, right, the megatrends. So electric vehicle electrification, which means it's magnetic sensing in terms of current sensing, but also in terms of xMR, GMR in silicon, TMR in silicon. This is a big area. We believe that our leadership position in TMR and GMR in silicon is going to drive market share in the future. It's a long-term play. We're doing great. We started shipments already. We're expanding that particular line. So -- and we've got some great wins going on. In addition to that, what we are looking at from a power perspective, it's -- motion control is continuing to increase. The braking systems are now getting much more evolved in the motion control area. So we see that. We've talked about the Heyday activity. That's really something that we will invest in. And photonics, I think you had asked that a little bit earlier, Josh. That we have our first prototype that -- from a platform perspective that we sample some customers. We'll continue to see the photonics product portfolio develop. And hopefully, it becomes a platform for 15 to 50-nanometers for our customers, where we have the various APDs on top of readout ICs that makes that whole system design a lot easier for them.

I guess -- well, go ahead.

Yes. So publicly -- Sanken has made public statements repeatedly, including recently in Japan, that they intend to hold more than 50% of the share -- 50% of the ownership. They own 52%. They consolidate our financial results. One equity owns 11.89% of our shares. And they've been pretty disciplined about when they have had opportunities to sell down the shares. And that's up to them, of course, when they want to do that. Our shareholder agreement is a fairly shareholder-friendly -- all shareholder-friendly and robust shareholder agreement in the sense that either one of them can take unilateral actions. It requires a full board vote.

All right. With that, I believe you're out of time. Ravi, Derek, thank you very much for your time. And Ravi, especially, thank you and congratulations on a great run.

Thank you, Josh.

Thank you, Josh.