Great, we're gonna get started. Good morning, everyone. Welcome to day two of the Stifel 2024 Cross-Sector Insight Conference. My name is Tore Svanberg. I'm a senior semiconductor analyst that cover analog, connectivity, and processor companies, and it's my great pleasure to introduce Microchip this morning. With us from the company, to my right here is Ganesh Moorthy, who's the company's President and Chief Executive Officer. We also have Sajid Daudi here in the front, who is head of investor relations. The format for this particular session is a fireside chat, so with that, we'll just get going. Thank you. So good morning, Ganesh, and thanks for coming to our conference. Really appreciate it.
Good morning, Tore. Good to be here.
We usually start these sessions off with a very soft question. If you could just give a very general introduction to, you know, those investors that perhaps may not be that familiar with Microchip.
Sure. So Microchip is a semiconductor company that's been around almost 35 years. We play predominantly in embedded control and embedded computing, which means we're deep inside customer systems. These are in a broad range of markets, ranging from industrial, medical, automotive, aerospace and defense, data center, some amount of home appliances. So just about any end market you can you know think about and which touches your lives, we're in, and very fragmented across a 100,000-plus customers and tons of applications in each of these end markets.
Great, thank you for that introduction. So maybe we could start with sort of the near-term dynamics, and as I've been asking a lot of other semiconductor companies this week, this last cycle was quite the cycle. You know, you and I have been through a lot of cycles. They're all different, but this last one, I think, was exceptionally unique. So, I was hoping you could talk a little bit, just take a step back and think about or talk about the last couple of years. What are some of the things the company did really right to manage the cycle, and what are some of the things that you probably would have done differently?
Sure. So, you know, we've been through probably the most challenging cycle of my career, and I've been in this industry 40-43 years at this point in time, just in terms of what the cycle went through, first on the upcycle and how extraordinary that was, and then how this cycle reversed and where we are at the current point, which is on the downcycle and how extraordinary that is. So it is truly a cycle we haven't seen before. Lots of factors that go into it, but you know, primarily coming out of the COVID, you know, dislocations that took place in 2020.
There was a lot of spending that took place in many of the applications and markets that we are in, and because the industry was not prepared for the rate at which demand was growing, the supply-demand imbalance very rapidly got to where, you know, demand was, in many cases, you know, 2x and 3x of supply, at least perceived demand as we could see it. And we went through a process of how would we rationalize that? How would we put brakes on it? How would we get it to try to shake out what the real demand was? And to some extent, we did.
But also, anytime you have a supply-demand imbalance, demand solves on time, and so you deliver product at some point later in time, and the farther and farther that goes out, the larger the error bands become, because no one knows what the future is 52 weeks from now, which is about where lead times were when I go back 18 months ago. And of course, the environment changes, and people's environment was very strong in 2021 and 2022, and going into 2023, they began to have just a normal, you know, cycle of the end markets that they're in. Plus, we saw some extraordinary weakness in China, which is a big part of the overall semiconductor market, plus interest rates began to go up dramatically.
And so many of the businesses felt both cost of their own inventory and other things that they were investing in, plus headwinds they were seeing in their own business. So all of that has resulted in an overhang that the industry is correcting through at this point. And, you know, we think we're getting to the bottom of this cycle and where it's at, but I don't think there's much visibility into what is the shape of the recovery, going to look like.
Great. And to follow up on that, Ganesh, you know, to your point, I think a lot of companies, including Microchip, have talked about, you know, finding a bottom. And now the big question is just how the shape of that recovery is gonna look like. And it's tricky, right? Because, you know, arguably, the revenue run rate right now is quite a bit below consumption, at least, you know, that's what I believe. So what are some of the, you know, milestones and things that, you know, you're looking at, you know, to try and determine, how this recovery is gonna look like?
Oh, great question. And so the indicators that we are paying attention to, and that would give us the basis for knowing how strong is the recovery going to be, the early indicators are the ones I spoke about on the last conference call, which are, you know, what are bookings doing? And are we getting to the point where... You know, there was, if I go back two years ago, I mean, we had bookings that were four or five times what they should have been, and then we went into bookings well below where they should have been. And when do we get to the point where bookings begin to restart a growth path?
We were on a growth path for, you know, February, March, April, and, you know, May just finished up, and May is just about where, slightly ahead of where April finished up at. And this won't be every month isn't gonna look, you know, different. Sometimes it'll slow down, sometimes it'll speed up. But bookings are the first thing, because that indicates customers are seeing their inventory coming down, and they're seeing their demand environment having sufficient confidence that we will place those. Second is, in what time frame are they placing bookings for? You know, with lead times down, we expect, and we are seeing, that more demand is being placed nearer term.
Mm-hmm.
That tells us their visibility is providing them the ability to say, "Okay, I do need to ship the product." We are seeing demand as the bookings are placed, much better for the nearer term than it was in the preceding months and quarters that we were at. The third thing we're seeing is we had a consistent push over prior quarters for more push-outs and cancellation requests. People who had gotten ahead of themselves, whose demand was not as good as they thought, and we, as we would work and accommodate that. Well, that's stabilized, and it's bottomed out in terms of where it's at.
We always have, in good times and bad times, some number of, you know, of cancellations, push-outs, so it never gets to zero, but at least it's not continuing to go down. And then the final piece is customers who, either because they waited till the last minute or because they were surprised by the demand, but coming in and saying, "Hey, I'm placing the order now, and I want you to ship it a lot faster than what your normal lead times are." Or, "I've placed orders with you, and I placed it for delivery in August and September, but I really need it in May and June-
Mm-hmm
... because my demand has changed or my market has changed," whatever has changed in it. Those are all good signs. So we're seeing many, many more expedite requests of people coming in. So these are early signs, and that's why we call them the green shoots, and we've got to see many, many more months, of this. And, but I think that is all, these are all the ways in which a bottom forms, from which, you begin to take the next leg.
Very good. And I'll get off the cyclical topic here soon, but another thing that's pretty unique about Microchip, the company does a phenomenal job managing downturns. The company has a, you know, pretty impressive variable cost structure. I mean, revenue's down 40% peak to trough, but you're still able to generate some very, very strong margins. I know, you know, you're still having shutdowns every quarter. What are some of the signs that you're looking for in order to, you know, stop those shutdowns and sort of get back to normal?
I think it's really how does the revenue into the September and December quarters look like, and what confidence do we have in it?
Right.
-that begin to tell us what actions we should take. And the actions are a little bit different between, factory versus non-factory, right? So in factory, we have an inventory challenge to work through.
Mm-hmm.
And so the only way to bring that inventory down is, in fact, to take some of the capacity out, where it's at. We have shutdown days, both in our wafer fabs as well as in our packaging and testing factories, et cetera. And what we will do is, as we see demand strengthen and inventory begin to come down, is you can unwind the shutdown days at some level in terms of where you go. In the non-factory areas, really, the work hasn't gone away.
Mm-hmm.
I mean, we're still... The innovation of new products and technologies we need to bring, the customers we need to support, the new designs we need to win. So that work, none of it has gone away.
Right.
But we need to bring costs down on the operating expense side. There, we have gone through both some amount of discretionary cost reduction that we can take, but also, the entire company is on a pay cut at this point in time, starting at the senior-most levels, so it's at the highest levels, and then smaller levels at levels beyond that. And that's part of our variable comp. Some of it is in bonus that we, you know, don't have during these times. And in extraordinary times like we're in today, we do a shared sacrifice-
Mm-hmm
... that we have. You know, over time, as business improves, that shared sacrifice, as we have great quarters, as we did in 2021 and 2022, et cetera, will pay off significant shared rewards as well on it. So that's what we try to do, is on the factory side and the non-factory side, to manage expenses through a cyclical industry.
Great. To maybe start taking this a step further and thinking about the recovery, now Microchip is one of the most diversified companies in the entire semiconductor industry. Are there any end markets that have been structurally hampered the last few years? You know, I'm thinking about, you know, some areas that are just exceptionally weak, you know, like 5G or... But based on what you can tell, are there any areas that are kind of structurally have sort of changed, and you're now, you know, actually don't think that those markets will come back to consumption levels?
I don't know of any that won't come back to consumption level, but there are some which are weaker, for a longer period of time. You mentioned 5G as a great example of that. I still think there is a 5G infrastructure upgrade that is gonna take place.
Mm-hmm.
It may be more back-end weighted. You know, they go in decade changes, so 5G began at the beginning of this decade. 4G was 10 years before that. And maybe there's more to come as you go into the second half of this decade, more exact. But which is why our diversity helps us, right? We're not dependent on any one megatrend or any one end application or any small set of customers.
Right.
You know, we're gonna always find that there's somewhere in the market opportunities for us, even as others may not be as good. And you can go back a few years ago when people were saying, "EVs are not going," then EVs became great. Now, EVs have slowed down. So it doesn't matter to us because we're not smart enough to know exactly which megatrend is gonna go at which, at what rate, in which year. But we wanna be present in the places that have the best chance of ongoing above-normal growth, and that's where the megatrends have been, for us.
Great. Thank you. So moving on to some more important topics, especially from an R&D and product perspective. So when I talk to investors, I mean, 90% of the time, everyone talks about the microcontroller franchise for obvious reasons, right? Because Microchip has always been in that market, done really well in that market, steadily gained share every year and so on and so forth. But I don't hear a lot of questions on analog, on FPGA... this is obviously part of the company's total system solution strategy. So can you give investors a bit of an update on how that is going?
Sure.
Because I think it's an important part of the story, but it doesn't get discussed very often with investors.
Sure. So as we think of total system solutions, right, we build around what we call anchor solutions. Anchor solutions are like microcontrollers that you mentioned, but, you know, we've gone from just having general purpose microcontrollers, which is kind of the core of how we got started in this, to many specialized microcontrollers, to networking microcontrollers, to microprocessors, to higher-end microprocessors, including we kind of talked about next month bringing out a 64-bit processor to DSP, digital signal processors, and then to FPGA. So these are all computationally or control-oriented centers of the customer systems, and those are usually early decisions a customer makes. But you hear about microcontrollers only because that's kind of how, what people know us about. But these other parts of the company have all become important.
Just to speak to FPGA for a minute, you know, we break it out once a year, and we talked about our FPGA business in fiscal 2024, which just finished at the end of March, being a $670 million-plus revenue business. Grew, I think, 20% or so in the last year, and, you know, continues to do extremely well in the space it plays in, with gross and operating margins that are north of corporate average and more. Signal chain, power, and clocks, which are an essential element. Every system needs power. Every system needs clocks of some sort. So when you look at signal chain, clocks, power and all that, those are all important parts of what we bring together. analog now has become a $2 billion-plus revenue for us.
You know, I think, and if you look at the merchant analog folks , we're back in the number 8 position and, you know, starting to move into the top 10 and where we're at. And it's a nice complementary piece to what we-
I talked a lot about... Obviously, today, a lot of the excitement is about AI infrastructure, but eventually AI is gonna move to the edge. And I even highlighted Microchip as a potential beneficiary of that. So, could you talk a little bit about the company's role in AI, especially at the edge, and how do you see this market evolving, right? Because I'm sure today there's still bottlenecks, right? There's, you know, bandwidth issues, there's power issues, but I'm sure you're having a lot of discussions with your end customers about this trend, so,
Yeah. So AI is an important market for us. In today's environment, the AI subsegment of the data center. If you take our data center business, you know, the AI component of it is just under one-third of that revenue. So overall, for the company, AI is about, the AI and the server is just under 5% of our revenue. It's growing year-over-year. We are in all the major AI platforms with a range of products from our switches, our products that go into security, specialized clocks that we have in them. We have brought out more products now that are doing some of the Ethernet retimers. We will be working on things which are PCIe retimers, and some of the solid-state drives that go into these devices, the solid-state drive controllers, power supplies.
Longer term, we also see the opportunity for AI at the edge and at the far edge. So at the edge could be in a factory, could be in a hospital, could be any one of these things where you're trying to aggregate and go, you know, make decisions faster without necessarily all the training needing to be done at the edge, although some of that could take place there as well. And now you get into a smaller class of what kind of processes you need, what kind of work you need. So Smart Embedded Vision is one example of a solution that we have that is aimed at, you know, factory vision, medical vision, et cetera, where AI can be applied.
And the processing power you need is more akin to our high-end processor, in which there's a motor that is trying to do something, there, right? How do you bring a level of intelligence to that where it is agile, it's aware, it's able to do, recognize what needs to be done, what needs to be changed, et cetera. So we think AI will pervade all of these solutions, but today, a lot of it is in the data center infrastructure. And we have a, you know, meaningful play there and then a growing play over time at the edge itself.
Right. I also wanted to talk about another big trend, of course, which has been automotive/EV. As you mentioned before, clearly some fits and starts the last couple of years with that market. How do you think about that market longer term? And the reason I wanna ask you this is because it's a pretty concentrated market, right? I mean, in industrial, there's, you know, tens of thousands of customers here. Maybe there's top 15. You know, even a lot of them are in China, where, you know, perhaps the opportunities are not as important because of insourcing and things like that. So, what is the company's strategy in automotive as far as, you know, making sure that the barrier to entries are there, you know, making sure there's no concentration of customers and so on and so forth?
In some sense, that's what automotive has been for us for the last 30 years, right?
Mm-hmm.
I mean, you could say the concentration of how many, OEMs and Tier 1s were there, has always been there. Now, that number of players has grown over time-
Right
... as, you know, both newer companies as well as EVs and all that have come about. So, we don't feel all these things don't care. So those are gonna be there no matter what.
Mm-hmm.
And then EV has a number of incremental opportunities, which replaces what would be the, the ICE drivetrain and, some of the other motors and battery management and all that. And those all present opportunities for incremental silicon, that you can go in them. I don't quite know the rate at which EVs are going to penetrate or not. Those numbers keep changing over time. But-
You know, a hybrid or a mixed model when it comes to in-house fabs and, and outsourcing. Yesterday, we had a peer of yours here, Texas Instruments. They talked a lot about inflation. In fact, they, the reason why they wanna do a lot of the manufacturing internally is because they're concerned, not just about geopolitics, but also concerned about, you know, foundries raising prices every year. I mean, that was clearly an issue the last few years during the pandemic. So how, how do you—and among, you know, this is, of course, is a pricing question, but it's not a, it's not a pricing question, you know, because of competition, it's a pricing question because of foundry inflation. So, how do you, how do you respond to that based on Microchip's mixed model?
Rational actors in marketplace and do respond to what they need to do from a supply-demand standpoint, as well as being market competitive. I don't think foundries have a philosophy that they're gonna keep raising prices every year.
Mm-hmm.
You know, we went for 20+ years before that with a foundry model that was reasonable, rational. I think they all recognize that over time, the market grows and the pie expands if we're able to provide cost-effective solutions for the market to absorb more of these things. So for semiconductors, as well as for foundries, I think the last 2 years were all an anomaly of where we were all at as we were trying to deal with this extreme imbalance of where we were at and what we needed to do to respond to it. Whatever that price at the end becomes is how they determine value.
Mm-hmm
... for what you're doing, not just the silicon that you're delivering.
Right. Related to that, you know, this is something that I've learned more recently, and I mean, I've covered analog semiconductors for 25 years, but it seems that the 28 nanometer and below, the capital intensity has changed. I mean, I don't look-
I don't know if it changes the profitability necessarily. I think what it says is, as that CapEx becomes substantially more expensive, if you're planning to build a factory, you better have the way in which you can fill the factory and, and amortize the cost, right? So we don't have anything that is in 12-inch as an internal factory.
Right.
We do all the, or a good amount of the 200-millimeter inside. But 300-millimeter, for us, as fragmented as our products are, it makes far more sense, to be able to work with partners, with different specialties and capabilities, whether that's at 90 or 28 or something smaller than that. And we're in all the nodes, from 90 down to 3-nanometer, in terms of products that we make-
Mm-hmm
... and go into. It's more cost-effective at the scale we have and the scale we see to be able to do it. We ran an exercise during, I think, 2022, on should we build a 12-inch fab, as Microchip? It was an internal exercise. It got out into the press, so we then talked to people about it. Even with CHIPS funding, even with, you know, other ways in which to tackle this thing, it just did not make financial sense, when we think about, you know, what is the loading that is needed? What are the foundry partners able to do? We think it's a far more stable way. Maybe there is a size and scale at which it makes sense. For us, it didn't make sense, and we decided not to.
Right. So we only have a few minutes left, but we wanna make sure I address the audience to see if you have any questions.
During the course of this meeting, we are gonna be making forward-looking statements and other projections. All of those come into your own risk, and so please refer to our filings with the SEC. Thank you.
All right, any other questions beyond the safe harbor? All right, but I have a couple more, Ganesh. The one thing that is pretty interesting to me is, you know, Microchip has been on this journey the last 10 years, where the company made a lot of acquisitions. There was quite a bit of leverage on the balance sheet. You fixed that. You went through a cycle. It's gone up, it's gone down. But, you know, for some reason... And by the way, you're now close to returning 100% of your cash flow back to shareholders, yet the stock is still trading at a big discount to so many of your other peers. What do you hear from your conversations with investors, what the issue is?
Is it a perception of the past, or is there anything else that, you know, really bothers them? 'Cause I'm a little bit surprised after this 10-year journey, that by now, investors wouldn't be, you know, a little bit more upbeat about what you've accomplished.
So firstly, the valuation gap is closing. If you look at on a, you know, 12 months ago, where was it? To now, where is it at? It is closing, and it is closing, not at a rate we would like entirely, but at least it is making meaningful changes in that, and I think that is people recognizing that. Second is, some of these things are journeys, as you said, and journeys take time.
Right.
People need to see some consistency in the strategy, the results, and test it over cycles, right? And so whether that is growth, growth in operating margins, cash flow, free cash flow that we're generating from it and what we're doing to return it, all that needs some time to play out. And I think as it does play out, I fully expect that the valuations will continue to head towards where they should be. So it's that, that's the biggest thing we hear is, you know, "Just stay at it.
Give the market time to absorb the changes you've made." And in some cases, there may be a rotation of some investors, as well, that, are investors who maybe had a different view, a view from a historical perspective, to investors who will see you in terms of where you're at and what better aligns. And as you better align who you are and what you do with what investors are interested in, that, you know, are, are all synced up, your valuation gives, gets the appropriate, you know, value that comes from that.
Great. So that's the opportunity. I wanted to finish up with a risk factor, which is on everyone's mind these days, which is China. So you could just talk a little bit about, first of all, the exposure to China and how you see that changing over the next 3-5 years? But then, more importantly, from a competition perspective, I still don't think investors quite fully understand what's going on in China from an analog and microcontroller perspective. I think the FPGAs, they're way behind, but what's your view, and is China indeed as big of a threat, you know, as we always hear?
So competition in China is not new. It's been there for 30+ years of time on microcontrollers, on analog, et cetera, you know. And competition from China is not meaningfully different from competition in other places in terms of what do they do to come after the market, et cetera. And so we've always had to be competitive in whatever markets we're in through innovation, through cost reduction, through the things that we have to do. I think what is different in China has been a renewed focus on putting more money into both companies as well as capacity, as well. Not all of that is completely in place, and not all of it is gonna be productive. We see lots of companies that start up, and 2 years later, you know, end up, you know, not necessarily making it.
Capacity is still going into place. Our business in China is about 18% or so of our overall revenue. That's mainland China. About a half of that is designed outside of China, so the design point of influence is US or Europe or somewhere else. The other half of it is designed indigenous in China. Of the half of it that is in China, again, about approximately another half of it is on the more complex end of our product lines, where it's got substantial end market application, software, architecture, knowledge that is very hard to duplicate in terms of where we're at.
Mm-hmm.
Then the other half of it is what I would call more of the standard products and microcontrollers and analog and all that. So if you take that, you know, quarter of our business in China, about less than 5% of our total revenue, that's standard microcontrollers. Yes, theory holds in analog to go after, but there is other factors, you know, geopolitical factors that cause them to say, "Well, should I be putting all my eggs in this basket?" And so there'll be some of that headwind-
Right.
-that comes, where people have to decide what's in their best interest, and do they wanna have a domestic source, or maybe there's some pressure, if they're in automotive, to say, "You've got to use X% of domestic content," et cetera, that goes with it. But there are lots of markets outside of China that are also developing and a lot of players in China who are now developing products, manufacturing products outside of China, too.
Right.
As some opportunities close, there's gonna be other opportunities that open.
Right. No, that's great. Great perspective. So with that, we've run out of time. Thank you, everyone, for joining the Microchip session. Ganesh, Sajid, thank you so much for coming to our conference. Enjoy the rest of your day. Thank you.
Great. Thank you very much, everyone.