Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

All right. We'd like to get started. Good afternoon, everyone. My name is Toshiya Hari, I cover the semiconductor space at Goldman Sachs. Very excited to have the team from Microchip with us this afternoon. We have Eric Bjornholt in the center, SVP and CFO of the company; we also have Sajid Daudi, Head of Investor Relations. Gentlemen, thank you so much for coming.

Thanks for having us. Hi, everybody.

All right. I'll kick off with a bunch of questions here. In terms of the near term, probably one for you, Eric. You cited a moderation in customer bookings. You guided September quarter revenue down 7% sequentially at the midpoint. How has the quarter trended so far? Any standouts, any surprises to the upside or the downside?

Okay. So let me start by referring you to our filings with the SEC that identify important risk factors about the company. We will be making certain projections and forward-looking statements, and I refer you to those documents to identify risks. So the quarter, we announced earnings in early August, and we talked about a couple of things that we were seeing in the business. One is that we are seeing fewer requests for pushouts and cancellations. That has continued. We are seeing more requests for orders that have been placed out in time to be pulled in. And the other thing that we were seeing, although bookings are at a lower level than we would like them to be, that the percentage of our bookings that are coming in with shorter delivery dates more in line with our lead times is improving. So those things are all continuing as expected. I think from an end market perspective, we aren't really seeing anything different than what we talked about back in the earnings call. Industrial is weak, automotive is weak, going through inventory correction. We are seeing some strength in data center this quarter, and we expect data center overall for us to grow in both the September and the December quarters. And before it was just kind of the AI piece of data center, which for us, is a little less than 5% of revenue, but now we're seeing some momentum in the overall data center business, which makes up about 18% of our revenue. So all that is really unchanged.

Okay. Got it. So again, September quarter guidance is down 7% sequentially. I think many of your peers are guiding the quarter flat to maybe up a little bit, down a little bit, somewhere in that range. I realize one quarter doesn't make a trend, certainly. But what is driving the delta there in your view? Is it inventory dynamics? Is it share? How should we think about the relative performance in the near term?

So it's not share. I think if -- you have to look over the history of this last cycle. And Microchip definitely overperformed at the upside of the cycle or the top of the cycle and now we are underperforming on the down cycle. And I think it's just a matter of -- there's always going to be customer and end market differentiation, but I think a lot of it is just based on inventory. And we entered the cycle later, and so our recovery is coming a bit later than most.

Got it. And on that point on customer inventory, I realize you guys don't have perfect visibility into customer inventory. But how would you characterize where inventories are today relative to 3, 6 months ago or what you would consider to be more healthy?

So the inventory position is improving when we look at customers and distributors. So we give -- we get real-time data from our distributors, so we can tell you exactly what's happened over a period of time. And distribution inventory came down in dollars by $125 million in the March quarter and about $85 million last quarter. We expect sell-through activity from distribution to be higher again this quarter than what we were expecting in terms of sell-in activity. So distribution is correcting. And the thing that we're pointing out to investors on that is that, that is on a lower level of sell-through activity because distributors' customers are also adjusting inventory. And right now, customers, end customers, distributors, contract manufacturers, whatever it might be, all know that Microchip and the rest of the industry have ample levels of inventory. We've got a lot of inventory. Lead times are extremely short. And with that, they are managing their working capital. We've got interest rates that are still stubbornly high. I'm going to hope for better news on that next week. We'll see how it goes. But with that, they are comfortable taking inventory down to lower levels. And the surprising piece of it, to me, is if you had asked me 18 months ago, are customers going to return to pre-COVID levels of inventory? I would say no way, right? I mean, we've just gone through this crazy up cycle where lead times pushed out longer than what we've ever seen in our company's history. And customers couldn't ship an $80,000 automobile or a $30,000 piece of metal equipment because they couldn't get a $2 microcontroller, right? It makes no sense to run inventory down that low. But customers appear to be very comfortable at least in the short term in taking down inventory. So we'll see how all this plays out. What we would really love for our customers to do is at least give us better backlog visibility, right? And that positions us to make sure that we're building the right things for them. And they have very little risk in giving us that sort of visibility because they can change, push out, cancel that order any time prior to 45 days of delivery to them. So that would be the most helpful thing to us to make sure we are positioned appropriately to respond, because inevitably, customers overcorrect on the upside and they overcorrect on the downside.

Right. Right. Got it. On the PSP program, now that you've had multiple quarters to look back on the pandemic and the PSP program, I'm curious, if a sharp cyclical upturn were to occur over the next, say, couple of quarters and your customers came to you asking for reassurance of supply, long-term supply, would you reintroduce the PSP program? And what aspects of the latest iteration or version would you inherit? And which aspects would you tweak or change?

Okay. So the PSP program was a good program for us. We introduced that program in response to customers wanting a higher level of assurance that they could get product. And this is when lead times went from 8 weeks to 52 weeks, right? And so we had never been in that situation before. Now was it a perfect program? No. We thought that when we were getting these noncancelable and nonreschedulable orders, that it was a true reflection of what the customer is going to need. But the challenge is when you're going out that far in time, they're guessing in terms of what end market consumption is going to be. And you might have an industrial customer that thought their business was going to grow 10% this year, and now it's down 5%, right? And that automatically puts them in a situation where they're over inventory. So I think there are several things that we would do differently next go around. Now I am hopeful we don't have to introduce a PSP type program. I've been at Microchip for 29 years and had never seen lead times stretch out like they did this last cycle. That's not saying that it couldn't happen, but it's happened once in my pretty long career at Microchip. So I hope we don't get there again. But we take every opportunity that we have, where we've executed a program, we've done an M&A transaction, whatever it might be and take learnings from that and figure out what could we do better next time to make the program better. And there are several things that we're looking at. We want to look at customer run rate activity, make sure that something flags it in our systems. The customer has been buying 10,000 units a month for the last 3 years and now they need 40,000 units a month, right? That should be questioned, right? Because we don't want a single customer to consume the capacity of a certain device. We could also say, hey, we have a PSP or NCNR program, but we only let 50% or 60% of the capacity for that given product to go on the program, and that gives us flexibility to manage the rest of the business and not have it completely be consumed by PSP programs. So we'll see. Again, we're hopeful that we don't have a situation where lead times stretched to the same way they did this last cycle. But we'll be better the next time and implement a program if needed that gives us better indicators of what's happening.

Microchip sells -- you guys sell a broad range of products to a broad range of end markets and applications. But I am curious, if you had to call out -- you talked a little bit about data center earlier. But if you had to call out 3 or plus or minus drivers that could really turbocharge the business forward over the medium to long term, what would they be?

I mean the things we talk about, and this is more than 3, but we talk about 6 megatrends that our business is focused on. And these are areas that we think have legs over the next 10 years to provide above-market growth. And those are IoT, data center is a piece of that, right, whether it's AI or not, data center, the amount of data that the world is creating is growing at an exponential rate. Within automotive, you have ADAS and you have EV. We've got 5G as a piece of that and then sustainability, which covers a wide variety of things that we do to help our customers be more efficient from a power perspective or a consumption perspective. So those are the areas of focus from an end market perspective. And then we look at it from a product perspective, how do we position our products together to be able to take advantage of what we call TSS or selling Total System Solutions. And all those teams are working together to make sure we've got reference designs, customer collateral material to make our salespeople successful with the customer and ultimately provide a benefit to our customer with lower engineering costs, faster time to market and more complete solutions.

All right. At the most recent Analyst Day, I feel like it was a long time ago, but the most recent one, I think the long-term revenue CAGR target you provided was in the 10% to 15% range. Obviously, you guys operate in a cyclical business. But when you think about the through-cycle growth rate, is that still the relevant appropriate range in your view? A lot has happened.

Yes, a lot has happened. And I think you have to go back to that November of 2021 meeting, and we were using fiscal '21 as a baseline. And you say long term, it was 5-year, that seems long term, but we didn't say this was a continuous 10% to 15% CAGR. Really, we want to position the portfolio, the products and what we do with our customers to drive market share gains. We target internally to try to grow at twice the rate of the market. And obviously, that's going to be different year by year. I think if you go back to fiscal '21 to now, we are clearly well below that 10% to 15% growth, but we were well above it for a few years during the cycle also. So over time, we just want to position ourselves to gain market share. And internally, we probably target about 2x the market growth. And we think the megatrends, combined with these selling more complete Total System Solutions will help us get there.

Got it. Again, going back to data center, I think it was 16% of fiscal '24.

18%.

18%. I apologize, 18% of fiscal '24 revenue. Not all data centers are created equal. I'm curious, as AI comes into the picture if not already, how should we think about that impacting and influencing your business going forward, total data center?

So it's important. Again, data center is a megatrend for us. AI is a piece of that. But we have a lot of products that sell around these more complex chips that you hear about from some of the big guys, the people that talked this morning early on at the conference, right? We have PCIe switches, we've got retimers, we've got SSD controllers, we've got standard microcontroller and analog products, timing solutions, security, root of trust products. So there's a lot that goes around that, and we try to get on the reference designs and then be complementary to these other products that are being sold in. It's been a great business for us. A large piece of this came to us through the Microsemi acquisition, which happened back in 2018, and the business has grown nicely. We've got great margin structure, and we're excited about the opportunities in the future. Sajid, would you highlight anything else on the data center?

No, that's about it. And I think additionally, just pre-COVID, it was a pretty robust grower too, going into this environment.

Got it. Eric, you talked a little bit about your Total System Solution strategy. Maybe expand on that a little bit? How is it different? How does it allow you to differentiate vis-a-vis your competition?

So we are fortunate in the product portfolio that we have that we are selling the brains of the system in many of these embedded applications, right? We've got anywhere from 8-bit to high end of 32-bit microcontrollers. We have microprocessors, both on 32-bit and now 64-bit. We have FPGAs. We have some ASICs in the portfolio. So that is typically the first chip that the embedded design engineer uses when they're building out their system. And then they lay out the analog, the connectivity, the timing, all these other products around it. And so by doing that and approaching it that way, we have a lot of the anchor products and these solutions and then can position the rest of the portfolio around that. And it works well when we go into a customer, we've got a working reference design, hey, we've done this 25 other times for customers, we can speed their time to market. Almost all of our customers are engineering resource limited, right? So if we can do anything to help them speed their time to market, reduce their investment in R&D, make them more cost effective and, at the same time, make us more sticky with the customer, the benefits are on both sides. So we're really excited about the opportunity. We can see it and every business unit review that is given internally, they talk about, hey, well, this is what we've won for their particular product division, and this is what's sold around it. And also identifying, well, if we didn't win, what are the areas from a product perspective that would have helped us get there -- what were we lacking that didn't allow us to get there? And that's what the whole focus is within Microchip: megatrends and TSS.

Got it. I guess somewhat related to that, your FPGA business has done really well over the past couple of years. You've outperformed the industry, I believe, over the past 12 months. I think many of your peers have gone through a pretty significant correction. Which end markets or applications have driven your success relative to peers? And what are your long-term aspirations in FPGA?

So in FPGA, again, this is a business that we acquired in the Microsemi acquisition 6 years ago. And when we acquired Microsemi, the business was very much focused on the aerospace and defense market. That's been a great market for us and has continued to grow, and it's been quite steady through the cycle. We've called it out as being a high performer from an end market perspective throughout the cycle. So that's been good. But we have then taken those products and expanded into other areas in industrial, in automotive, in communications. And the products play very well there. And the product line is continuing to introduce new products all the time also. Our kind of brand name for our recent products there is PolarFire. We have PolarFire 2 coming out soon, and we think the specs on that are fantastic, are going to do extremely well in the marketplace and allow us to expand. We're kind of in the midrange of the FPGA space today, and we have aspirations to encroach on the lower end of the high end and the higher end of the low end also with the product portfolio.

Okay. Great. Maybe on pricing, I realize you guys don't operate a commodity business, and the design cycles are long, and the product cycles are long. But curious how you're thinking about pricing going forward. You've gone through -- not you specifically, but the industry has gone through a fairly inflationary period. I would say a lot of capacity has been added both at the IDMs as well as the foundries. So I guess there is a concern out there that pricing could decline more so than historical patterns into '25. But how are you all thinking about overall pricing for your overall business?

Yes. So from our perspective, nothing has changed on pricing. Pricing is always extremely competitive at the point of design, right? We tend to be leading into new designs with our newest products that are on the most -- more advanced process technologies or more cost effective, have a better feature set and can position those products with aggressive pricing. Because we see that over time, the margins on our product improve. And in many cases, we don't even peak in revenue on a product until 10 or 12 years after it's introduced, right? So we have a lot of time. And we look at where is the margin going to be over the life of the product. And typically, new products have lower margins in the early stages. That's planned for, and then they improve over the course of time. So our position with customers is that, yes, we are competitive at point of design. But once we're designed in, the work has been done by Microchip, the work has been done by the engineering side of the customer. And that pricing should be fixed over the life of the product. Now we did, as others did during the up cycle and the inflationary pressures of higher capital cost, higher wage costs that were outpacing the rate of chains that we could get from a cost reduction standpoint, we did increase prices. We were fair with customers on that. We were transparent. And those prices are staying firm at those levels.

Okay. Okay. Great. On the geopolitical backdrop and how you all think about your manufacturing footprint. One of your competitors has been pretty aggressive in expanding capacity here in the U.S. They've also been very vocal as to how having access to geopolitically dependable capacity is a competitive edge. How do you guys think about that dynamic?

So we are very comfortable with what we are doing from investing in capital. Now the competitor that you are mentioning has a much different position than we do, right? We do about 40% of our manufacturing from a wafer fab perspective, in-house. Those are in 3 large fabs in the U.S., one in Arizona, one in Colorado and one in Oregon. And we have continued to expand in those locations. But we are process technology limited. Anything that is 90-nanometer or below or on a 12-inch wafer, we are dependent on the professional foundries for. And we evaluated the possibilities of doing a 12-inch factory, thinking about chip stack, money, investments, partnerships, whatever it could be. And we could not make it cost effective for us. It's not saying what the competitor is doing is wrong, it's just a different business model. But we are quite comfortable with the strategy that we have. We feel that our foundry partners understand what our long-term road map is and will be there for us and investing in new capacity. They have done that. In some cases, they are investing in capacity outside of Taiwan, as an example. And again, that can provide some benefit to customers if they feel that being not in Taiwan or China is important to them. I will say customers just want reliable supply. And over time, our foundry partners have been very reliable for us in terms of delivering what we need to service our customer base. So really no change on the manufacturing front. We think that we'll stay about at this 40% level for internal fab. We'll probably increase the amount of assembly and tests that we do. Today, it's roughly 60% on the assembly side and high 60 percentage for assembly, will probably -- excuse me, for test. Assembly will probably take to about 70% and tests will take to probably 80% over the course of time.

Okay. The medium term?

Yes. And I probably should emphasize with this. Our capital that we have to invest in the business is pretty minimal. In the up cycles, we might invest 6% of sales in capital. In the down cycle, would be about 3%. But the capital intensity is much different for us than the competitor that you compared us to. And again, reasons for that.

Got it. That's helpful. Chipset funding. You guys announced an agreement with the CHIPS office back in January. What are some of the key milestones that you need to hit for you to actually see the funds come through?

So just to clarify that, we entered into a what's called a PMT or preliminary memorandum of terms of the CHIPS office back in January. And we supported the CHIPS office through diligence, through March and got through that. And we have not signed a final memorandum of terms for -- or a final agreement at this point in time. We've got $162 million that they have kind of offered up to us and we are working through the details of the contract with them. We're still quite hopeful that we get there. But it's not a make-or-die situation for Microchip. If we get the money, it's great. If not, there'll be business reasons why it didn't make sense to do it. And again, the CHIPS office has been supportive, their understanding of what our issues are, and we're working through those things.

Shifting gears a little bit. Competition with local Chinese suppliers is a big topic, and I'm sure you guys get questions all the time. Can you remind us what percentage of your business today comes from China? What percentage of that is China for China? What are you seeing on the ground from a competitive standpoint? Is it more intense? What are you seeing from your perspective?

So we sell based on where we ship the product, about 18% of our revenue into China. We feel that about half of that is designed outside of China for consumption outside of China. So domestic China exposure is somewhere around 9% to 10% of revenue. And of that, we think that half of it is highly proprietary, unique stuff to Microchip that competition in China can't touch. And the other 5% of revenue is more standard microcontroller analog products that will be subject to more competition over time. Now there's a reason these customers have chosen to design with Microchip. It's because we have a huge portfolio of products. We provide great support. And it's very difficult for a China competitor to come in and compete with that, because customers want options. Most of the time, a customer starts their design, what they go to market with is something different, right? And when you've got a wide portfolio, let's take microcontrollers as an example, from everything from the low end of 8-bit to the high end of 32-bit, with thousands of products, customers have choices. If they go with one of these local competitors and they change, right, do they have to start over from scratch? Do they have to go to another competitor? So they don't run that risk with Microchip. Also, we're selling these more complete solutions, right? And so that is value add to the customer. So we aren't saying that competition isn't real in China. Obviously, they're investing a bunch. But we think any bleed we have on this 5% of revenue that we have there that is really more standard product that's subject to that competition will be slow. And again, you have to remember that we design products in into applications that in some cases, sell for 20 years. So we're focused on it. We're continuing to support customers and winning our fair share of business.

Yes. I'm going to pause here and see if we have any questions from the audience. If not, I can keep going. If you can raise your hand. Maybe going into the model, gross margins, puts and takes, maybe starting off of factory utilization rates. I know you guys have been moderating production, if you will. What's the plan going forward? You talked about pricing being stable. Any other kind of pluses and minuses that we should be thinking about as it pertains to gross margins?

Okay. So we are underutilizing our factories today. Obviously, we had ramped our facilities and our foundries have ramped up to be able to support a quarterly revenue run rate of about $2.3 billion. And we're you're roughly half of that today. So there's lots of capacity. We are underutilizing our factories. I think last quarter, we took an underutilization charge of about $36 million. And we don't start taking that charge until utilization gets back to kind of normalized levels. And we were running factories at like 100% utilization for an extended period of time. So today, we're getting that charge. On top of that, we are taking some inventory reserve charges, which are accounting charges that are required based on our accounting policies. Those will probably be the first thing that go away to help gross margin. As once the business environment starts to improve, bookings activity picks up, business confidence from our business unit leaders picks up, those reserve charges will come down. And quite honestly, we're not building products that we don't think is going to sell. Eventually -- and I don't know if it's going to sell a year from now, 2 years from now, 3 years from now, but eventually, those reserves should provide some tailwind to gross margin.

Got it. Okay. So to your point, for you to meaningfully take up utilization rates, it's bookings, sustainability, its business confidence, those kind of things.

Really, we need revenue to return to a higher level, and that will happen. It's just a matter of the timing.

Right. Okay. Makes sense. OpEx is down about 25% peak to trough. Is it entirely due to the cycle? Is there a permanent change? How should we think about OpEx on the way up as business recovers?

Yes. So our peak OpEx from a non-GAAP basis was about $464 million a quarter. This quarter, we're guiding to about $351 million, which is about this 25% reduction that you mentioned. And there's various pieces of that, right? I mean we've got everybody on the company on a salary cut today. That's 10% for most of the population, it's 20% for executives and others, believe it or not, have volunteered for more. We've got a pretty unique company culture where in these down cycles, we don't do layoffs, right? We lay people off for performance issues, but we don't do broad-based layoffs. And what that does is it gets everybody bought into a shared sacrifice, shared reward system. And the business was doing great. Last year, a year before that, bonuses were paid at a very high rate. And we reminded people when we were paying those bonuses, this is a cyclical industry. This is temporary, don't spend it all. And most people listened to that. But it's difficult to have employees on a pay cut. Ganesh has made the commitment to the broader base of employees that after they've been on that pay cut for 9 months, which will be kind of middle of November to early December for the employee population, we're going to return them to full salary. So that's happening partially through next quarter. Anybody at a director or above level in the company, so the executive team will stay on the pay cut for longer and as long as we need. So that is a temporary reduction. We were paying bonuses at 200% to 250% of target back in the up cycle when we were producing 48% non-GAAP operating margins. Today, those bonuses have gone to 0, and they can stay at 0 for as long as they need to until the business improves, and we start showing significant improvement in margin. So there hasn't been a lot of what I would call permanent reduction. We have had -- through this cycle, we have had some turnover. We've had some retirements at higher ranks. And we do, as Microchip always does, is we give people opportunities from within and then replace people with new college grads or new people to train through the system. And so the cost structure is lower because of that. We've taken out a lot of discretionary expenses for travel, advertising, things like that, that will come back into the P&L as it can afford it. But you can view that, June quarter of 2023 OpEx is where it hit its peak that, that was inflated in dollars because bonuses were so high. So you shouldn't expect it to bounce back to that level.

Got it. Okay. It sounds like a lot of the changes, whether it be the gross margin or the OpEx is very cyclical in nature. The long-term margin targets that you guys threw out at the Analyst Day, those seem fairly reasonable still. Is that the right way to think about...

They do. They do. So -- and those were 68% gross margins at the midpoint and 45% operating margins on a non-GAAP basis. We well exceeded that in the up cycle. I think our peak operating margin was 48.4%, and we see no reason to change the model. The cost structure is in great shape. We just need revenue to get back on track and it will.

Capital allocation and maybe M&A, you've been really methodical and transparent with your approach to capital return. I believe by the March quarter of next year, you'll be returning 100% of free cash flow to shareholders. Post that, is M&A still a consideration? Do you feel like you have the right pieces in place to execute your strategies and therefore, you're done? How should we think about the priorities from a capital allocation standpoint?

So we don't view that we need to go out and do a large-scale M&A transaction. We have the pieces of the puzzle that we need to be successful and support our customers. It doesn't mean that every business we have, they're all working on new product introductions. But we're not -- we don't have this gaping hole that we need to fill. And our acquisition strategy between 2010 and 2018,helped us build up this portfolio along with our organic efforts. So we'll do small tuck-in acquisitions. And I would never sit here and tell you that there'll never be a large M&A for Microchip. But our cash flow will be at a point where we're generating a lot of cash. And if we were to come to market with a deal, and again, I'm not signaling that, that's not the intention today. I'm saying longer term that we could back off from some of the buyback activity and justify to investors, hey, this is why this is going to give you a better return than buyback. But today, we're really not focused on any large-scale M&A. We've done a couple of small tuck-in acquisitions. We did 2 in the June quarter, and we've just reduced the amount of stock buyback for the amount that we're paying for those deals.

Got it. Makes sense.

So we want to keep leverage in terms of dollars, pretty stable. Net debt to EBITDA is obviously decreasing right now, increasing when EBITDA is falling, but again, that's temporary in nature.

Very clear. AI is obviously a very hot topic these days and certainly at this conference. At Microchip, do you guys leverage AI whether it be chip design or day-to-day operations vis-a-vis a couple of...

I can start on that. Yes, so at Microchip, we've been leveraging AI for almost 6 years now, starting with machine learning and deep learning in a lot of our operational activities. And then today, it's becoming a more and more integral part of our business today. Roughly 87% of our orders are auto scheduled through a deep-learning engine. And then more recently, we've been leveraging generative AI, which is helping a lot of our teams translate technical documents, software libraries and other areas and translation and stuff. So yes, so we've been deeply using AI in many different facets, and we feel that sales efficiency, operational efficiency and product development in all 3 areas is where we can leverage that.

And I would say that Ganesh and Richard Simoncic, who's our COO, challenged all of us in terms of how are we going to use these tools to make ourselves more efficient. So every business unit, every administrative function is looking at it. And I'd say it's early stages, but we're having some success.

Okay. Great. I guess in the last 2 minutes, Eric, I wanted to give you the opportunity to kind of speak to anything that we may have missed. Or I know we're keeping you busy with meetings all day. Anything as a collective unit, we underappreciate about Microchip or misunderstand about the cycle? Anything you want to highlight?

I mean, obviously, there's a lot of short-term focus from investors at this point in time, and we fully understand that. But we want to continue to emphasize what we have done in terms of investing in product development, new product introductions, how we are supporting our customers with these more complete solutions. And those are really the key to the long-term success and market share gains of Microchip. And we think we're doing all the right things. Design activity is very high today. If you go back 18, 24 months ago, it wasn't because customers were just scrambling to figure out how to get their products out the door if they couldn't get the exact product from Microchip or one of our competitors, could we tweak the product to get them something in a shorter lead time. But now it's full focus on new product introductions. We're seeing that with how the seating activity is happening through our catalog distributors, activity that our salespeople are having with customers. Design activity is high, and that bodes extremely well for the future. Outside of that, I think we covered most of the important topics in your Q&A.

Okay. Awesome. Thank you so much. Really appreciate the time.

Thanks. Thank you everybody.

Thank you.