Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Let's get started. I'm Vivek Arya from the BofA Securities Semiconductor semicap equipment team. And welcome to all of you, and I'm absolutely delighted and honored to have our guests from Microchip, President and CEO, Ganesh Moorthy; and CFO, Eric Bjornholt. What we will do is maybe start with some opening comments from Ganesh and then we will get into Q&A. But during the Q&A, please feel free to raise your hand if you would like to bring up a question as well. But with that, welcome, Ganesh, welcome Eric, and Ganesh over to you.

Great. Thank you, Vivek. Let me start with the obligatory -- during the course of this discussion, we will be making some forward-looking statements regarding future events and future financial performance of the company, and we wish to caution you that such statements are predictions and the actual events may differ materially and refer you to our SEC filings. Our June quarter business remains steady and on track to the guidance we provided a month ago at our earnings call, and we continue to believe that it's unlikely that the September quarter is going to be a sequentially down quarter for Microchip. We do recognize we operate in a cyclical industry. We're not immune to business cycles. But if you also review Microchip's peak to trough performance through the business cycles over the last 15-plus years, you'll observe how robust and consistent that cash generation, gross margin and operating margin have been throughout the cycles. And although we don't foresee a significant decline in our business, if we were to experience any kind of an inventory correction and/or other adjustment, we expect that our performance on these metrics of gross margin, operating margin and cash flow will remain robust and resilient as we have consistently over the many years. Microchip 3.0 is the strategy that we have been working on now for just over 18 months. Now it's firing on all cylinders. We're delivering premium results exemplified by a 5-year compounded annual growth rate, which we had shown as 10% to 15%. We're ahead of that track at this point in time using fiscal '21 as our baseline. In fact, over the last 30 years, our compounded annual growth rate is 16.4%. Our trailing 12-month non-GAAP gross margin is 67.8% of gross margin percentage, our trailing 12-month non-GAAP operating margin is at 46.9%. So all high and on track to the long-term guidance we provide. Our trailing 12-month EBITDA is $4.3 billion, just over 50% of revenue. Our leverage is down under 1.5% or 1.45% at the end of March and continue to head down into this quarter. Free cash flow as a percent of revenue is in the 88th percentile of all S&P 500 companies and the capital return to shareholders increased 76% last fiscal year and is on track to be 100% returned on free cash flow by the March quarter of 2025. And lastly, our valuation based on an enterprise value to EBITDA as compared to S&P 500 SOX index, some of the high-performance peers, et cetera, we believe presents a compelling opportunity for long-term investors to buy premium results at a discount. And on that, I'll hand it back to you.

Excellent. Couldn't have said it better myself. Thank you.

So maybe let's start with the macro in Asia and then work our way into the Micro. If we look at your results by end market, right, versus what the assumptions were at the start of the year. Could you help us contrast how you see the demand trends in the big end markets, industrial, automotive, data center, consumer, comps? What do you see now versus what you thought would be the case the start of the year?

So -- when we look at it over the last 6 months, incrementally, customers have become more cautious. Some of the -- their own markets have become weaker. And I think there is a great deal of uncertainty that everyone is operating under. So in that sense, what we have seen in our own results and guidance and all that reflects where our customers are seeing business at. There is still strength inside of that in certain segments. But I would say in any given segment, there are pockets of strength and pockets of weaknesses. It's just some are much many more strengths than weaknesses. So you take automotive, you take industrial, yes, they're in the aggregate, still strong for us. But within it, we've got some customers who are feeling more stress and -- but on the average, more customers feeling so confident about what they're going off and building. The segment that has been the weakest for us has been consumer. It is dominated by consumer appliances, but as many other consumer type of products that go into it as well. And then I would say data center is taking a pause in terms of where it's been at. But even within data center, things like AI servers and all that have continued to be strong for us. So net result is 6 end markets. There's no -- or 5 end markets. There's no uniform, it's all strong or all weak. It's a degree of weakness or strength among customers in each segment.

I see. Is there any certain geographical area, which you think is getting weaker or is stable at the margins?

So of all the regions, China has been the weakest. It has not bounced back the way we expected it to post the shutdowns and post the Chinese New Year. And so it is consistent with what we had expected when we came into the quarterly guidance and all of that. But it is definitely weaker than what a normal June quarter in China should be and what the bounce back most of us expected would be coming out of the 2 quarters of December and March with its own unique weaknesses.

Got it. Now normally, Ganesh this area of diversified chipmakers. It's not really a share bin-loss area or a small period of time. But when I just look at the difference in the growth rates, right, yourselves versus the peers. So your trends suggest that Microchip could grow sales almost double digit, right, this year, whereas one of your largest peers is expected to have sales that are declining. Why do you think there is such a contrast in terms of relative growth rate? Is it just mix or is it something that Microchip is doing that's helping you outgrow your peer group?

So there's probably 2 or 3 things, right? One is our relative end market exposures can be different. And so depending on the strength of that end market, whatever exposure you have, you may have a weighted average that helps you there or not. I don't know each person's exact end market exposure. But there are 2 things unique to us. One is what we have done for the last 4 or 5 years to focus on total system solutions, which is how do we harness 10-plus years of both organic and inorganic acquisitions and be able to take it as a complete solution to the market and how we make it sticky with hardware, software and services, all that put together to make these solutions grow and grow disproportionately. And second, our attachment to the 6 megatrends that we have picked. And I reported at the conference call that those megatrends in the last 2 years for us grew at 2x the rate that Microchip grew. So Microchip numbers were about 55% growth over 2 years. The megatrends grew close to 110% over those 2 years. So that is giving us both a high content from TSS and a higher growth rate from the application segments and megatrend markets where we have the most amount of growth taking place.

Got it. One unique thing that I think Microchip kind of started and then many of your peers followed, was this notion of, right, having what you would describe as the preferred supplier program. So give us a sense for how would you grade that? Like if you had to do the preferred supplier program. So maybe first explain what it is, right, where you are in terms of how it represents your backlog. But then, let's say, if you had to do it all over again, what would you change about that program?

Sure. So preferred supply program was something we started in February of 2021. So it's like almost 2.5 years old. It was done at a time when there was a tremendous imbalance between supply and demand, and it was done at the behest of customers who were looking for ways in which they could have a better sense of having priority in the constrained supply that was there. The bargain was that a customer would be willing to pay -- would be willing to place orders that are 12 months of noncancelable backlog on a continuous basis. In exchange for it, they got priority over everybody else in what they did, and it was driven by when they placed orders and all that. The program has been phenomenally successful, right? When we kicked it off, we thought at best, maybe we'll see 25%, 30% of the backlog that would take that type of program. And it's been well, well in excess of 50% of what it's done. And it's been pretty consistent as a percentage of our total backlog for the last 18 months of time. So this is not something that customers have taken their foot off of in terms of the desire for PSP. Overall backlog has come down and with it the PSP backlog in absolute has come down, but as a percentage is staying pretty consistent with that. I would rate the experience of the customers who give us feedback on it as an A to an A+ and what it has done for them. It did exactly what they wanted. They want a peace of mind. Those have signed up for it, they'd extremely well with their end businesses because they got the supply that they needed from us without having to come through the battles that many others that came in late and went forward. And how did we redo it at each point in time, we make small adjustments as we go along, But at no point, what I have said with what we knew at the time we launched it and as we went along, would we have made a change. It's been a phenomenally successful program for us. We do it all over again, and it's a tool, right? It's one that when it outlives its use, when customers no longer see value in it, we'll sunset it.

So what percentage of your backlog are these PSP orders now? And how has that mix trended over time?

So we haven't broken out the precise percentage because it got frighteningly high. But as we said, it's over 50%, well over 50%. And as a percentage of total backlog, it has stayed remarkably consistent within 1% or 2% of each other for the last 18 months.

Got it. Now one concern that investors have is right, which is kind of a broad industry concern is that many semiconductor companies have their versions of the PSP program, but how enforceable will they be, right, if they get into a moderate or severe downturn. So how enforceable will these orders be? Or will you give flexibility to customers to push out? Like are you going to be flexible on units or pricing or both in case the macro environment changes?

So PSP is not the first time we have had entered into noncancelable orders, right? So it's been a piece of our business historically as well on certain types of -- it's a custom product or something that is a unique requirement that went into it. And we've always managed that business under those terms. PSP just increased the amount of the business that began to go through a noncancelable process. The bargain in this is that we went off and made investments based on the customers signing up for non cancelability as well. And we think, in general, customers don't sign up for noncancelable backlog unless they've done their diligence, they've done their buyoff from inside of the company, et cetera, just as we -- nothing gets -- we don't place a noncanceled order on a supplier into that. Eric or Eric's senior team looking at approving at it. So it does go to a degree of scrutiny that is much more than normal backlog. That said, we know conditions change. We know customers are not prescient for everything that's in front of them. And as those change, we will work with them, and we have worked with them, and we have over the last 6-plus months been indicating that we are taking and as much as we can, pushing our product to help them out with it. What we are not very flexible on is the noncancelability. This is not going to be a customer who says, heads I win, and tails you lose. That is not a recipe for long-term mutual benefit in this kind of a program. What we are willing to be flexible on is how does the rescheduling, which is typically nonreschedulable. How can we work on that and what can we do to help each other in the process of doing that. And we're a reasonable business people on that.

Got it.

The last part of your question, I think had to do with pricing and PSP was never a pricing program. It was a unit volume program. And if pricing changes, the customers are subject to that.

Okay. That's a good distinction to know. Have there been examples where customers have come back and said, "Look, I was under the PSP program, but I needed rescheduled for, I don't know, 6 months, 9 months. And in that kind of scenario, how has Microchip reacted?

Yes. I mean there's -- with the amount of customers we have and how high the PSP percentage is, there's every flavor of a customer needing help that comes about. And sometimes it's a month or 3 months or 6 months and probably in some cases, they've been 6 to 9 months as well. It all depends, right? It's not a digital answer. We might be able to do a portion of it for a set period of time, a portion more for future out in time. So our objective is to help the customers out, smooth out any overhang there may be in the inventory that is in the channel or in the customers hand itself. I mean, in the end, this is helpful to us as well. We don't want a hard landing at some point in time as well. And so in that sense, and we want our customers over the long run to be successful customers, and we want to continue to work with them on future stuff as well.

All right. Makes sense. One more near term and then some longer-term questions. You mentioned that September, you think is unlikely to decline sequentially. If you could give any more color around that? And how much of that is based on the assumption of normal seasonality, right? You mentioned consumer is weak? Like is it just the consumer coming back because it's a seasonally stronger period? Or do you think that, that view is based on your core industrial auto data center businesses continuing, right, to be resilient during that time frame?

It's largely a function of how do we read the tea leaves for all the information we have by customer, by product, by backlog and where we think things are going to be, et cetera. We're not relying on a consumer comes back trend or something else that's out there on it. And so we haven't put any numeric thing about what it is. But I don't think when I look at the trends, we feel comfortable that it's unlikely to decline. I don't know if you want to add more on it?

No. I think it's -- a lot of business is driven on both backlog and our ability to provide a baseline capacity that's in place.

Right. And just conceptually, do you think we should think about September kind of the way we think about June, right, in terms of that -- is it kind of sustaining the kind of trends you're seeing in June persisting through September? Or is there something very different that can happen in September?

What we're unsure of is where does the macro go as time goes on, right? So we can only see -- it's a bit like driving in fog. You can only see a certain distance, but there's not complete clarity. And that's what we don't know is, does China have any improvement that happens in the second half of the year. Is there something else with interest rates that might affect. What happened in the U.S. and what happens in Europe, et cetera. So -- it's a time of uncertainty. And I think we're trying to provide the best clarity we can within reason for where we see our business.

Got it. On the automotive end market, for the most part, we have seen companies sound very confident, right, and optimistic. But very recently, one of your peers kind of suggested that they have started to see some softening of demand. How do you see your automotive business?

That's exactly what I said at the beginning, which is, I think, in any segment, automotive included, there are pockets of weaknesses and there are pockets of strength, and it's just the pockets of strength outweigh the pockets of weakness. There is no segment that is uniformly strong with perhaps the exception of aerospace and defense, where everything is strong at this point in time. Aerospace and defense is 9% of our revenue. So it was not trivial in terms of where it's at. But with that set aside, everything else is a mix of strength and weakness. And just a matter of does the points of strength outweigh the points of weaknesses, certainly, it does in automotive.

Got it. But have you seen any recent -- any increase in cancellation activity, any change in customer behavior in automotive?

As I said, in automotive, we have customers who are looking for help, and we have many other customers who have shortages. They're trying -- we still have shortages by the way. So we're not done with shortages in some of the products. So we have that too going on in it. The balance is more in the areas of strengths, than it is in the areas of weaknesses within automotive.

Got it. Then on a longer-term basis, Ganesh, we have a situation where all the semiconductor companies exposed to more mature nodes, right? They went through a period of shortage, but now there are investments that are coming up, right? And one large player is planning to build a tremendous amount of capacity. So how do you see that in the context of the industry? Like when should investors start to get nervous that, oh, there is too much capacity, right, that could come on? And so how do you deal with maybe a competitor who has not actually added too much embedded capacity, now planning to do so? Or how do you see that? And how big of a competitive threat can that be?

The strategies that each company pursues is defined by their view of the world. Our view of the world is that the capacity we have split internal versus external, is exactly where we want it to be. The additions we're making in measured steps, and we will pause and accelerate from time to time in where we want to go, is exactly where we want it to be. And it is to give us the growth and profitability that we see is an important part of how we try to manage the company. So we're trying to manage the company towards this 10% to 15% compounded annual growth rate on a 5-year basis using '21 as the -- fiscal '21 as a baseline. To have the gross and operating margins and the cash flow that we want to generate, deliver the results that we have promised to create the free cash flow, and particularly the free cash flow per share that we want to have available to come out of this whole thing and what we can return to shareholders as well. So our actions are governed by what we call it Microchip 3.0, but what we're trying to do in there. I'm sure there are other options, other actions that can be taken in it. There's nothing in the market today that suggests that, "Oh my God, I'm somebody is going to come in and use up all their capacities, take all the market share away." These are very, very complex products and applications with very long design-in cycles and very sticky sockets that we're in. And this isn't subject to a variation in market share or something else or price. And by the way, it's possible that these competitors, whoever they may be, are also subject to it. Larger portions of their business being in commodity spaces. If you're a commodity player, you should be afraid about where the capacity is at. If you're a proprietary player, you sell on a different basis than price.

Got it. From a -- at some point, I believe in the past, you had considered, right, exploring 300-millimeter fab option. So what made you consider that then what made you change that view? And don't you think it will be an advantage, right, to have 300-millimeter capability?

Great question. So when we went down that path, there was an uncertainty whether our partners -- and by the way, all out 300-millimeter is done with outside foundry partners. Whether they would be willing and they have the bandwidth to invest in the types of technologies that we needed, and most of what we need are trailing at 300-millimeter technologies, 28, 40, 65, 90 and those are primary nodes and where they were at. And so when we couldn't get that and we had a growth agenda, and we had a business objective that couldn't be achieved through that. We said, okay, we're going to look at other options. And then the CHIPS Act helped us to derisk some of those options on where we went off. So we did a complete review headed thing on it. But what we came to the conclusion was we could do it. Financially, we could do it. And had there been no other option, we probably wouldn't have done it because the revenue and gross margins we could generate were still good. But for a period of time, when you have a factory, a factory is a large fixed cost. And so if you don't have loading in that factory, your underloading costs are going to be quite significant. And you will have to absorb that for any number of years until you bring enough loading into that factory. We didn't see that as a time line that made sense and we thought from a return on invested capital standpoint. Once our foundry partners began to see the light of what needed to happen and they were willing to make the investments and they were willing to do it at a scale that makes sense, that was a far better ROIC answer for us and it gave us the same requirements. I mean the second thing we needed was geographic diversity, which we were also, at that point, getting a lot of questions about, and we have absolute commitments from our foundry partners and what they will do to help us achieve that as well. So that outside of the China, Taiwan theater is where we'll see some of the incremental capacity come online over the next 5, 6, 7 years of time. So it met all our business objectives by going the route of staying with our foundry partners and in fact, created significant risk for us and in cash, consumption and all of that and doing it ourselves. We had absolute confidence in doing it. We just didn't see the business benefit of doing it.

The other thing that became clear to our foundry partners is that the market was willing to pay because the capital costs and the capital intensity had gone up significantly, and they saw through the cycle that we just went through that their customers, Microchip and whatnot, and then our customers are willing to pay those higher capital costs that required to make those investments and justify that to them.

Got it. That makes sense. On gross margins, if somebody had told you that Microchip would be doing 68% gross margin like 5 years ago, would you have believed that? And like what has led to these kind of gross margins?

I'll let Eric talk about. I'm usually the optimist in this thing. So...

I'm usually the pessimist in the room. So I might not have believed it, but the bottom line is what's happened over the last 5 years as we have integrated Microsemi. There's been a lot of improvements from that. 5 years ago, we had some underutilization charges in our factories, too. Those have gone away. The pricing environment has been good. And we haven't used that for gross margin percentage improvement. But we've proven over the last really 8 to 10 years now that there was not really a need for the proprietary products that we sell to offer ASP discounts to customers every year. It's a design in business, sticky products as Ganesh talked about, and these prices are sticky and will stay. So lots of moving parts, but utilization and product mix are a huge impact related to gross margin, and we're in a very good spot today with that.

Got it. So maybe on the positive side, how much is pricing a factor in terms of growth this year. So even if it's neutral to gross margin, still adding to growth. So what is the assumption on pricing for this year? And then kind of the other side of it is that what needs to happen to take gross margins down? Like are these kind of gross margins sustainable? Let's say, hypothetically, we get into a year where sales go down 5% or 10%, what is the leverage on the gross margin side?

So with -- we don't really break out pricing versus units. We've said the last 2 fiscal years that volume has been the larger driver of sales improvement than pricing, and I'd expect that to be the case going forward, too. But our margins are very consistent. And if you look at the last 15 years of history through cycles, our gross margins have fallen 200 to 300 basis points during a downturn. And I wouldn't really expect it to be any different this go around. It might even be better. We are a little bit less dependent than we were in some of those earlier periods on internal manufacturing. It's a little less than 40% was fab in-house last year. Our internal inventory -- so inventory that's manufactured internally is still quite low. Our die banks are lower than we'd like them to be. And we're a little bit higher on the foundry side where we took advantage of capacity freeing up in the December and March quarter and took some of that inventory in. So I think that's going to allow us to continue to run our factories at a high level and maintain very high gross margins throughout the cycle.

And in fact, some of the historical numbers even included dilutive acquisitions before we've made the improvements. So it was not all our gross margin decline only, but just the additive number that came.

Got it. And then finally on use of cash. Where does Microchip in the journey? I know your plans are to get towards that 100% free cash flow returns by March '25, if I remember correctly. So where are you in that journey? And then at what point does M&A also come back on the table?

So we're bang on schedule with what we said back in November of 2021. We've been very consistent with the programs of how we have deployed the free cash flow that we wanted to return, a combination of both dividend growth as well as the share acquisitions. Those are progressing to where 100% of the free cash flow by the March '25 quarter is returned to shareholders. Ideally, in about a 50-50 year split, obviously, the dividend will be a fixed component. The share buyback will be the variable component of that. And that's very much where we're headed to. M&A for us is not a top-of-mind issue. It is not part of our strategy. We have a pretty rich portfolio. We have the scale. We will do small pinpoint acquisitions of specific technologies or markets or other things that we would be interested in and you will see some of that. We've done about half a dozen of them in the last few years, and we'll do more of those as well going forward. But there's no large M&A that we see as required or imminent in our future.

Got it. Actually, I wanted to ask one other question, which is we don't normally think of Microchip as a big player on the data center side. But I think you have a very interesting portfolio there. So I can't let the presentation go by without mentioning the 2 good words of A and I. So maybe this is the chance to...

So I'll speak to AI, but more specifically around data center, right? Nobody knows what we do in the data center. And it's about 17.5% of Microchip's revenue at this point in time. And it is split between the compute and storage infrastructure, the memory infrastructure, the solid state drive control that we provide, the data center interconnects, the timing and precise timing that you're going to need, the root of trust, the power management, on and on and on. There's a whole bunch of things that we are deeply embedded into. And within that, there's a subset of things that are specifically around AI servers themselves. And this is not in the last 2 weeks, and this has been there for 3, 4 years of time that as AI servers have been growing significantly. It's been a big part of what has driven our data center growth as well. But specifically, in an AI server, where you have multiple GPUs and maybe 1 or 2 CPUs, switching of data is one of the most critical things that takes place there. And then we have the PCIe switches that are in many, many of these data center platforms. With every GPU, you're going to need a precise clock. We're often the clock that goes next to it. Many of the -- and with every GPU, not just one per system, and then you need route of trust requirements. You need some of the memory infrastructure to help it be moving stuff faster. So there's many, many elements of data center in general, both compute and storage and then AI specifically on the parts of the market that are more AI driven.

Great. So have you seen this big upsurge in activity in the last few weeks or it's...

You know, we've been designed into many, many of the data center platforms and AI platforms themselves. And with it, whatever happens in that market, we will get our unfair share.

Understood. Terrific. Thank you so much, Ganesh. Thank you, Eric. Really appreciate your time today.

Thank you. Thank you, Vivek.