Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Good afternoon, everyone. Thank you for joining us. Really delighted to have the management team from Microchip join us this afternoon. Ganesh Moorthy, the President and CEO; and Eric Bjornholt, the CFO. So thanks to everyone for joining. And what we will do, format-wise, is I'll turn it over to Ganesh to make a few introductory remarks, and then we will get into a fireside Q&A. With that, Ganesh and Eric, welcome. And Ganesh, let me turn it over to you.

Great. Thank you, Vivek. Before we begin today, I wish to remind you that in today's discussion, we will be making some projections and other forward-looking statements regarding the future financial performance of Microchip. And these statements do involve predictions, and the actual results may differ materially. I refer you to Microchip's filings with the SEC regarding some important risk factors about the company. With that, let me start with a few brief statements about our business. The March 2021 quarter that we reported back in May, almost a month ago, was a record in many respects, including revenue, non-GAAP gross margins, operating margins and EPS. And we have guided the June 2021 quarter to even higher levels as we continue to execute our long-range plans focusing on providing total system solutions to our customers. Business conditions continue to be exceptionally strong. Backlog remains at record levels of product to be shipped over multiple quarters. The demand increase continues to outpace the capacity improvements we're able to make and hence, the gap between supply and demand continues to widen. With the strong demand and low levels of inventory both at Microchip as well as at our distribution channel partners, we are experiencing constraints in all our internal and external factories and their related manufacturing supply chains. We are continuing to ramp production in all of our internal factories as well as investing in capital additions to expand our internal factory capacity based on the strength of our backlog, especially our noncancelable Preferred Supply Program, or PSP backlog. We're also working closely with our supply chain partners who provide wafer foundry assembly, test and materials to secure capacity through the combination of both our internal and external capacity actions that we have taken so far. We expect our overall capacity will continue to grow every quarter in calendar year 2021. And while our capacity will continue to grow every quarter, we also believe that wafer fab as well as assembly test constraints are here to stay with us through the middle of calendar year 2022 and possibly beyond that. About 4 months ago, we launched our Preferred Supply Program, or PSP, to provide customers with supply priority beginning 6 months after their order in exchange for at least 12 months of continuous noncancelable orders. The customer response to the program has exceeded our expectations with direct customers and distributors alike. And today, about 50% of our backlog is in the PSP category. Although it is almost 100% of our backlog in some of the most constrained capacity corridors as well as a high percentage in some of the end markets where this has been more favorably received. Additionally, the additional PSP backlog continues to come in every week. And all of this gives us a solid foundation to enable us to prudently acquire constrained raw materials, invest in expanding further capacity, help factory capacity further and to hire employees to support our factory ramp. Finally, our Board of Directors is systematically moving towards a higher shareholder returns model, and we are rapidly deleveraging our balance sheet with a laser focus on becoming an investment grade-rated company. That concludes our update. And let me now pass it back to Vivek for the Q&A session.

Excellent. Thank you, Ganesh. Thank you for that introduction. So let's actually pick up from that one very important feature that you mentioned, the Preferred Supply Program or the noncancelable order program. I'm curious, is this the first time Microchip has established this and what led to the establishment of such a program? And I ask this because we have seen a wide range of views in the industry when we talk with, for example, the companies with a higher analog exposure. They have similar programs, but not to the same extent. So what drove your decision? What has been the experience in the past? And is there any downside of having a preferred supplier program?

Sure. So to answer the first part of your question. We have never taken a noncancelable window out to 12 months on standard products. We have done it on custom products and that type of activity. If you go back a year ago, our noncancelable window was 30 days for just about all of our products. We extended that first to 45. And then at the beginning of this year, we took it to 90 days. So that's what all products are, with or without PSP. PSP came through many conversations with customers, who were trying to grapple with how do they get some degree of certainty in a world of complete uncertainty about capacity. And in that discussion, the genesis of the idea came about, how about they give us commitments that gives us the ability to prioritize capacity for them. So PSP was born through that conversation. It is not the right answer for all customers or all end markets. But in those end markets, where a customer has a much more durable demand, a end product which has a significant multiple on the semiconductor content that they are buying, this became a solution. So that's why industries like automotive and industrial, et cetera, communications infrastructure, they're all seeing this as very critical to enable them to plan better and give us that so that they get -- they don't get guarantees. They get a priority. And by and large, at this point, we're able to fit most people's demand in in most of the capacity corridor. We find that it is very effective. We get high marks from our customers for having done something to enable them to be able to plan their business. And so that's what we're doing.

Got it. So you said the PSP covers about 50% of the backlog. What is the backlog? And how much of revenue coverage does it provide for you if we look at the next 12 months as an example?

So we haven't broken that out, but backlog is running at record, record levels, right? It's almost -- and it's because lead times are longer, and people are wanting to provide more backlog visibility in a constrained environment. I think another way to look at backlog intensity would be how much backlog, in excess of what we can ship, is being requested in a current quarter. Currently, we're running over 50%. In excess of what we can ship is the unsupported backlog in the June quarter. And that number, if I go back to December, was in the 30% number. So the gap between what we can support and what is unsupported continues to grow despite adding capacity. And that gives you a sense of the demand intensity and add to that the visibility over the next 12-plus months. And actually, we're getting some PSP backlog beyond 12 months. We don't require, but for some customers, they want to have that peace of mind well beyond 12 months as well.

Got it. And you mentioned, Ganesh, that there is some differences between different application areas or products where you have more or less coverage versus the 50%. Could you give us some kind of ranking where you are kind of fully supported by backlog versus you have less kind of -- fully supported by PSP versus where you have less PSP coverage?

Sure. So I'll give you one example. So from an end market perspective, automotive backlog is 80% in PSP at this point in time. So it's 50% for all of Microchip. There are others like industrial, which will probably be north of average, although I don't have the data with me. And then there'll be still others like consumer, et cetera, which will be south of average in where they're at. But I think industrial and automotive and comm infrastructure and defense and aerospace, these tend to be areas where there's durable demand, long-term requirements, a lot better long-term understanding of the market and therefore, a lower risk for them in being able to participate. And so automotive is the one I have a number for today, which is at 80%. Even within that, there could be specific capacity corridors, a given process node, a given package type, where it could be closer to 100% in some cases as well.

Got it. And final question, which I know is probably not a fair question. It's really a question for other -- your peers in the industry. Why have they not implemented as much of this program? For example, I'm talking about some of your more analog peers, right? One mentioned that their noncancelable windows only go up to like 90 days, right? So why is there a difference? Is it because of the kind of products you make? Like what's so special about your product mix that is giving you the opportunity to put this program in place?

I can't speak for them, but I think our program works for our customers, right? We took the insight for what business problem are our customers trying to solve and designed the program to give them a solution for what they needed. We think ours fits for what our customers' businesses and business model are. Perhaps those are not what our competitive competitors have in terms of their mind. So I don't know. I think the other part of the noncancelable that is also done is, to some extent, it shakes out demand that is uncertain or in excess of what is really required, right? Because if you've got to commit to a noncancelable window, you're going to be much more thoughtful about what backlog you're placing. And I think that is an added benefit in the current environment, we think, when there are constrained capacity demand.

Got it. Let me ask this final question about the PSP program. So when you look at the kind of customers and the amount of orders that are being placed with you, does that look like normal demand? So for example, I'm sure you have some visibility into what 2022 is shaking out, right, could shake out to be. Does that look like kind of trend line kind of demand, but with very good firm coverage? Or does it look like above trend kind of demand with firm coverage?

It looks normal. There's obviously growth assumptions that everybody is putting in place and growth assumptions from both '20 to '21, which are quite large, but '21 to '22, which is more normal. So I see nothing in the noncancelable or the PSP type of backlog which would indicate that it's in some way in excess of trend line. It's normal, which is why I think it is half of our backlog. It's not 100% of our backlog. And people are placing the backlog with the knowledge that they're financially committed for 12 months. And I think that's not an easy decision for them to make because their companies are now financially on the hook for it as well. So I think people have been thoughtful of placing the backlog they need and not trying to place excess backlog that would not make sense for them to put in place.

Right. So maybe one point that I'd like to make. If we didn't have this program, and we had a 90-day noncancelable window, we would have a huge amount of orders from customers that would be sitting between 90 and 120 days that would have a lot of churn in it, right, because it could be pushed out, canceled by the customer at any time. And Ganesh is absolutely right, that this makes customers very thoughtful in the backlog that they're placing on us because they've got a firm commitment.

Right. Absolutely. The next aspect I wanted to get into is, given the situation we are running into, right, which is kind of these unprecedented supply shortages, in the past, when we have gone through these situations, the industry has responded with increasing, right, capital spending. And we are starting to see like you're also increasing capital spending, I believe, by almost 75%, right, looking at the next fiscal year, right, like some big amount. How are you ensuring that you don't run into an excess of supply from your side, right? We don't tip into the other part, right, of this cycle where, yes, demand is there but like you said, demand is trend line, but your supply is above trend line. How are you making sure that doesn't happen?

Sure. So first, maybe to correct the perception about our capital intensity. Our long-term capital intensity has been between 3% and 4% of revenue is where it's at. The last 2 years, we've been well under that, right? We've been -- it's been slower business. And so while relative to the last 2 years, this year looks high, it's actually on a normalized trend right about where it should be, 3% to 4% of revenue. So what are we doing to ensure that we don't overshoot? That's your question, right, on the capacity side of things. In our internal capacity, we are working so that the capacity comes online from, one, a little bit at a time, in part because the equipment only comes in a little bit at a time. We're also making sure that as the environment affords, we would actually have enough time to both -- there's going to be a need to replenish our channels. Our channels are running on fumes at 22 days and heading lower, we think, by the end of June. And so in whatever time, when things begin to normalize, you would have a buffer on what you have to do to replenish the channels. We would need to buffer ourselves because we're running on fumes with respect to inventory. That would need to happen. Three years, the PSP capacity or the PSP demand and backlog would give us about a 9- to 12-month head start on where is that demand going, which would allow us to have a way to thoughtfully begin tapering some of our capacity programs if need be, so that we don't overshoot because that's a pretty good leading indicator of where things are going to be. And so those are all different elements of the -- on the capacity side. Clearly, we have a component of our capacity, which is outside and outsourced, and that's not directly what we are driving for in terms of our capital spending. But all in all, we are doing the things in our control to try to achieve a soft landing. And you can look at many things we did in 2017, 2018 with similar objectives of a soft landing, right? And we are not trying to get all the capacity in 1 quarter. We're trying to do it over multiple quarters. We did the same thing at that point in time. And we had a fairly successful soft landing without the benefit of longer noncancelable windows at that time. And I think many of those things, plus some additional things we're doing this cycle, is what we expect will help give us a soft landing.

Got it. And is it fair to say that having this very good visibility around the shape of demand is perhaps actually helping you put CapEx in the right places, right, internal versus external, front end, back end, et cetera? So is this actually almost helping you from an operational perspective also?

Absolutely, not only in terms of the CapEx but even in terms of the -- how are we using that capacity, what wafers are we building, what packages are we building and all that, right? The stronger your backlog is, the more effectively your capacity gets utilized because you're not building things speculatively. You're building things largely based on backlog not for unforeseen turns, which may or may not come in in the shape as you're thinking. So yes, strong backlog helps us to be able to build in the best mix to serve demand.

Got it. And Ganesh or Eric, if you kind of polish your crystal ball and we look out a year plus from now, which end markets do you think, right, come into balance sooner rather than later? And which end markets would kind of trail that kind of process of normalization?

It's hard to tell because right now, there is strength -- although automotive gets the most airtime as to all the issues because of the plant shutdowns and all that stuff on it, the constraints are really in just about all the end markets. And many of the underlying capacity elements, right, the wafer fabs in which we run these product, the processes they run on, the packages, the tests, they're not unique to a given end market. We build products and with the underlying capacity, it can go to any of these end markets. And so we don't see a softening in one end market or we don't anticipate one that we're smart enough to tell you today is the one that's going to be the first one that comes off. And as that happens, what will really happen is we'll just -- it'll just move around to other parts of the constraints we have because the underlying capacity is really fungible across end markets.

Got it. What is the risk, Ganesh, of customers? I know you have stressed many times that, look, what you're building is a very proprietary product. So it's hard to kind of dual source them, right, from somewhere else. But at some point, I imagine customers who are going through this experience and having to go through these commitments, et cetera, at some point, they will say, you know what, if we have the same issue 3, 4 years down the road, right, what steps do we need to take in place to take those steps today to make sure we don't go through this kind of experience? So do you think you are running the risks of perhaps a shift in market share towards some of your competitors? Have you seen any kind of design out activity to -- even if it's to a small extent?

On the contrary, what we see is lots of requests to rescue people from capacity that is not available. So for some competitors, the proportion of large customers or a given end market like automotive is quite significant. And often support for that comes at the expense of other markets and customers that they have. So we actually have more inbound requests for people who are asking for an immediate help because of whatever has happened on the end of supply. We don't see it, and they could be on the fringes, somebody who's going in the reverse side of that is. But largely, we don't think that's the risk we have.

Got it. And I want to ask this in the right way, which is one of your competitors, right, has a lot of 300-millimeter capacity, right, in the U.S. And they have highlighted that a number of times as kind of a competitive differentiator, right, in a way. Whereas, I think, from whether it's the balance sheet or leverage or just kind of the tight control that you've had on the spending side and not having a lot of inventory, do you think that creates a little bit of a competitive risk down the road? That if somebody has a better balance sheet, that they can afford to build up a lot of inventory and that maybe helps them grab more market share in this kind of an environment?

Well, if there was ever a time to test that, it would have been in the last year, right. Because we were constrained as we went through the second half of last year in our ability to build inventory because we were trying to balance what we needed to do in debt reduction, our leverage ratios and all of that. And, in fact, we started this cycle at probably the lowest inventory point at which you are starting a brand-new cycle, right? Most cycles we start with days of inventory that are 135, 140, 145 days. We were at close to 125 as we entered the subcycle. So we were not as well positioned as we would have normally been because we were constrained in what we could do. But despite that, you can see what we have done in performance through the first -- basically the first 3 quarters of this upcycle. And you can benchmark it against any competitor in terms of what their capacity is. And maybe we have not had all of the possibilities covered but we've done a fair amount of what we needed to do to cover. I can't speak to the whys of where it's at. But it definitely does help to have some of our internal capacity, our packaging and testing, having been in-house has helped us a lot. Our ability to grow our internal fabs has helped us a lot. But we still have 60% of our capacity that we have in the foundries and of our fab capacity that we run in the foundries. And theoretically, I can understand why there should be an advantage. But practically, in the last 9 months, we have not seen a disadvantage.

Got it. And I know historically, Microchip has not wanted to kind of break out, right, the microcontroller market in terms of, right, the bit lens and 8 versus 16 versus 32. But because that is one of the externally available metrics to us, that's why I'll still ask the question. What is the difference between these 3 markets? How do you allocate R&D resources? And the fact that you have been very strong in 8 and 16 but the 32 market is perhaps a lot more competitive, how does that guide what your blended market share will be over time?

So I think it's probably older news that our 32-bit is not large or competitive, right? Our 32-bit is pretty big as a business. We broke some of that out at the last conference call. It's well over $1 billion, I think it is $1.3 billion, somewhere in that neighborhood. It's #4 or #5 in the rankings. It's growing -- it grew faster than the 32-bit market did last year. It grew faster than our 8- and 16-bit businesses grew as well. So we have a pretty dynamite 32-bit business in terms of where it's at. But the way we look at the market is not in terms of 8s, 16s and 32s. We're organized that way to be able to execute products. But we really think about it in terms of what is the customer's application need. And we lead the customer to what the best solution for them is. And if that's in 8-bit, so be it. If it's in 16-bit, so be it. And if it's in 32-bit, so be it. Rather than say, hey, you've all got to move to 32-bit because that's what I have, or that's what I believe, right? So we don't have any dogma about 8, 16 or 32. We have a belief that we help customer -- we help understand customers' needs and lead them to the best solution that they need to have. And we're a powerhouse in 8-bit. We're a powerhouse in the parts of 16-bit that we play in. And we're clearly, in 32-bit, a much, much stronger player today than we were, say, 10 years ago.

Got it. Ganesh, does the choice of architecture matter in that the spec or ADR, right, or ARM, is it a given that everything will become ARM at some point 5, 10 years down the road and that will really increase a lot more -- that will increase the competition in the microcontroller market? Or do you think these are actually separate markets. There is a very loyal group of developers for each architecture. And like you said, it's really about the solution and what it can do as opposed to, right, having some religious attachment to one architecture, right?

Yes. There are probably camps in customers who say, hey, everything has to be ARM or whatever. But I think there's some history you can go back. You can go back and look at the '90s, right? 8051 in the '90s was the ARM of today. And everybody thought 8051 is just where we're all going to go and that becomes a standard architecture, and it didn't. And it's because I think choice gives people opportunity to employ different capabilities and all that. We today have vibrant 8-bit architectures, one of which is PIC, the other one is AVR. We have a vibrant 16-bit and 32-bit. We have a MIPs-based core and we have an ARM-based core, and they're both doing quite well in where they're at. So we have always said the core really is not as critical as the peripherals. The customer really interacts with the peripherals. And what does the outside world see? How does the customer write software? Are there a group of customers for whom the core is super critical? Sure, there is. And we have a solution for them, too. And if the core is not the only thing that they're interested, we have a broader set of solutions for them as well. So there's no monolithic description of a customer that is you got to be ARM or you got to be this or that. There are pockets. Some people have just a familiarity. A lot of people use PIC microcontrollers because they have a familiarity with the product, the tools, the peripherals, the software they've written. They may have an installed base of what they have developed that they can reuse, et cetera. So many inputs come in to that decision-making process.

Got it. Got it. One last question on the product side. How do you forecast industrial market growth? Because again, whether it's auto, right, it's kind of conceptually easy to think about units and content. Industrial is the toughest market to forecast. So how do you forecast industrial market growth?

So we obviously don't pay that much attention to growth by different market segments, et cetera. And we're honestly not smart enough to say, hey, this is what auto and this isn't. I think we can see what does the demand in the aggregate look like. What are we doing from a demand creation standpoint? How are we overcoming whatever is the macro growth rates to doing things that drive growth rate that we would like to have. And in that, we have an aggregate growth that we're driving for internally. And then within that, we say our products, by and large, most of them, not all of them, but most of them can be sold into multiple end markets. We don't have very many products that are only sold in one market. We do have in data center and a few other places. But many of them can -- so if the demand happens to be stronger in industrial than perhaps in automotive or than something else, the products just move around to where the demand is rather than having to forecast by end market growth and products that we have to plan for.

Got it. Two questions on the financial side. First, on gross margin and kind of use of cash. So on the gross margin side, right, just exceptional job in expanding gross margins on a consistent basis, right, record 64%. You're getting close to your longer-term target of 65%. What role is pricing -- so first of all, just what have been the drivers? And is the PSP program a positive or negative from a pricing perspective? Like, are you having to give up the ability to change pricing dynamically by having somebody locked into a PSP program? So just talk to us gross margin drivers and whether the PSP program is a positive, negative or neutral factor when it comes to gross margins.

I'll answer the second part quickly, and then I want Eric to really speak to the rest of it. So PSP program has nothing to do with pricing. Prices can go up if they need to, prices can come down if they need to. PSP is really about a supply priority and has no effect on pricing or gross margins, other than perhaps it continues to have kind of good loading for what we're doing. Eric, why don't you speak to the drivers for gross and operating margins?

Sure. So gross margin has absolutely been a highlight for us over the last several quarters. We were underutilizing our capacity back in last calendar year. And as those underutilization charges have gone away and they were completely eliminated in the March quarter, we saw pretty significant increases in gross margin, a record last quarter. And at the midpoint of the guidance for this quarter, guiding towards another record of 64.3%. Our long-term model is 65%. So if we come in at the midpoint, we'd be 70 basis points away from that long-term model. And there's a number of drivers that are going to help us get there. We've talked about many times externally about increasing the percentage of assembly and test that we do internally. So we can do more packaging and testing on -- in our facilities in both Philippines and Thailand, we do about 55% of that production today. We're also growing into our clean room space as we add capacity. With that, we're spreading our fixed cost over a larger volume of product, which is all incrementally positive to gross margin. And then we've got some longer-term projects that we're working on for shutting down some of our older factories that we've acquired through acquisition and transitioning those to our more efficient factories. And those are multiyear programs but all progressing in the right direction. And then the other thing. Outside of where we're seeing input costs into the system increase today and then passing those on as price increases to customers, which is really margin-neutral, pricing is because of consolidation in the industry has been quite stable, and we're holding prices flat outside of the current environment. So all those things combined together are really going to help us drive to that 65% target.

Got it. And does bringing on new capacity, does that impact gross margins even if it's on a temporary basis, or it doesn't? Like does it change utilization and hence, impact gross margins?

So it does. When we're not adding to our factory footprint in terms of clean room space, we're adding equipment lines into the existing factory footprint. That makes all the production that we're doing in that factory efficient, more efficient as that production comes online.

Got it. And then finally, the company has gone through a very good deleveraging process over the last few years. Where are you in terms of the deleveraging? Can you just remind us what the current leverage is, what the target leverage is? And once you get there, how are you prioritizing your use of cash?

Go ahead, Eric.

Okay. So we ended last quarter with net leverage of about 3.71. And we've said we're very much laser-focused, and I think Ganesh even said in our introductory comments that we're laser-focused on becoming an investment grade-rated company. We believe that we can do that over the course of the next 12 months. We had a very positive update from Moody's a couple of weeks ago, changing us to a positive outlook in that respect. So moving in the right direction. We haven't set a specific leverage target. Clearly, we want to even once we become investment grade to continue to delever the balance sheet. But as we get there, the Board is signaling that we're going to increase our capital returns to shareholders. We've increased the dividend pretty significantly on a sequential basis, both in February and in May, and that's the direction that we are heading to continue on as we move towards investment grade. And then once we become investment grade, it gives us more optionality not to necessarily return 100% of free cash flow to shareholders, continue to use some of the cash flow to delever the balance sheet. But we can take the dividend higher. We could introduce a share buyback program. Those are decisions that the Board will make as we get closer. But the cash flow from this business, with the operating margins that we have and the relatively low capital intensity, it produces a lot of cash and the leverage is coming down very rapidly.

So if all goes according to plan, and I know, right, it's a dynamic situation, could you be at that point of decision-making in the next year or late this year? Like is there a certain kind of rough time frame? And the one word I did not hear in that use of cash is M&A, which has historically been a very, right, important attribute of your strategy.

Yes. So we absolutely believe that we can get there in the next year. I'll let Ganesh comment on M&A.

So we have not found a reason we have to do M&A or valuations that would make it attractive to us. We had bandwidth issues, both management and financial, for the last 3 years. We're getting to the point where those are all largely going to be behind us. But we're not particularly attracted by where M&A valuations are today. And we are perfectly okay not pursuing M&A for as long as it takes. If it's 10 years, it's 10 years and -- and be focused on organic growth. So indeed, in the capital allocation options, for us, paying down the debt, increasing the dividend and looking at share buybacks are where the priorities are. M&A could enter the equation somewhere down the line. We don't know if and when that will make sense. But right now, it does not make sense. And so it is not on the table.

And we are very fortunate that with all the acquisitions that we've done and how we've grown organically over the last 10 years, we have a very full, robust product portfolio. And when we look at it, there's not a gaping hole in the portfolio saying, hey, you need to do M&A. We have what we need to be able to supply the full system solution approach that we have with our customers.

Excellent. So I'm glad you're not trying to buy software companies. So on that positive note, we can wrap up the call. Thank you so much, Ganesh. Thank you so much, Eric, for joining us this afternoon, sharing your insights with the group. And thanks to everyone for joining the call. We can close the call here. Thank you so much.

Thank you. Thank you, Vivek. Great questions. Appreciate it.

Okay. Appreciate it.