Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

All right. Welcome back, everyone. We'd like to get started. My name is Toshiya Hari. I cover the semiconductor and semi cap equipment space here at Goldman in the U.S. Very excited to have the team from Microchip with us today. We've got Eric Bjornholt, Senior Vice President and CFO; and Rich Simoncic, SVP Analog, Power and Interface business unit. And I will kick off with a bunch of questions, but I will open it up to you all in the crowd. So please be prepared. So Eric, thanks for doing this. Really appreciate it.

You're welcome. Thanks for having us.

I wanted to start off with a near-term question and then sort of transition longer-term as we progress. In terms of the near-term environment, you guys had guided the current quarter in terms of revenue to grow 2% to 3% at the midpoint. I think on the earnings call, you talked about September also unlikely to be down, I think it was the language that you guys used. I realize it's only been a couple of weeks, several weeks, but any change in your thought process there? Any standouts from an end market perspective?

So no change. What's giving us confidence in the current quarter and next quarter is our backlog visibility that we have. And we're in a good position to grow midpoint of guidance this quarter is 2.5%, and we still believe that it's very unlikely that it will be down next quarter.

Got it. And then, I guess, applications, end markets, any standouts not necessarily potential deviation from your guidance, but relative to the past quarter?

Yes. Really no change from what we guided to. I mean, clearly, China is our weakest geography. That hasn't -- we really haven't seen any sort of significant bounce back since Chinese New Year. But outside of that, there's really nothing to call out industrial, automotive and data centers still our strongest markets and consumer appliance, which is our lowest kind of percentage of revenue. It's about 12% is the weakest area.

Got it. Okay. And I guess I was hoping you could -- excuse me, if you could talk a little bit about customer inventory and channel inventory, the fear that investors have in terms of the questions that I'm getting is that you've been over shipping relative to true consumption. And at some point, you would see a decline in your business. Maybe talk about your ability to gauge and monitor where inventory levels are? And kind of what gives you the confidence to sort of, I guess, is the backlog, but what gives you the confidence to look out beyond the quarter.

Yes. So I mean distribution is just a little less than 50% of our revenue. And so that's pretty easy for us where our distributors provide us with inventory and sell-through reporting at least on a monthly basis. So we've got a good handle on that and the distribution inventory has grown a little bit from its lows, the low is we had about 17 days of inventory. Last quarter, we ended with about [ 24 ] in distribution, which is still kind of on the lower end of what we've seen historically. So I think distribution inventory is okay. For us, customer inventory is a little bit harder for us to get real-time information on that really comes through discussions with customers. And we have had a, what I would call an increasing level over the last quarter of requests for pushout activity from our direct customers, and we get a lot of the same questions from investors that you get related to our PSP program. But ultimately, that preferred supply program, we think that the backlog that we've had under that program is better if we didn't have it because it's very well thought out that -- those orders in the PSP program are noncancelable and nonreschedulable, and we have given customers a little bit more flexibility on the potential to push out orders but not cancel orders under PSP. But that -- those orders are well thought out because customers have had to go to higher levels within their organizations to get approval to even place those orders. And PSP is still strong and well over 50% of our backlog.

Got it. And I guess that was sort of my next question on PSP. Maybe take us back to Feb of '21 when you introduced a program and you sent that [indiscernible] letter out to your customers. What drove -- or what led to the creation of PSP? And has it sort of served its purpose, if you will?

Actually -- it's actually 2 things that led to the creation of PSP. One is to try and prevent double ordering or triple ordering. And one of the biggest factors that always plagues the semiconductor industry is the broker industry, where brokers come in and start consuming up electronics and then reselling them on the marketplace. So PSP was a method to curb broker activity so that you don't have this huge buildup of inventory, not just in inhibitors, but within a broker network. So it's meant the 2 biggest reasons. One, give -- 3, give us visibility; two, ensure that we didn't have a lot of double ordering or low-quality backlog; and the last one is to prevent bad behavior that plagues the semiconductor industry of broker activity. And it really worked on all 3 of those dynamics.

Yes. I mean customers who signed up for the PSP program early have been serviced very well and then most of those customers continue to be PSP participants today. And -- some of those customers are now entering into long-term supply agreements with us, which cover typically a 5-year period. And those customers have been served very well through the PSP program and are now looking for further assurance that they will have priority of our capacity into the future.

Okay. I guess on that point, Eric, the long-term supply agreements, have you ever disclosed how big that is as a percentage of your backlog or -- that would be a subset of your PSP, correct?

So every customer that has entered into a long-term supply agreement with us has been a PSP customer. So that's really just an extension of that program. they typically don't give us more direct coverage in terms of order visibility beyond 12 months. It's kind of functions as PSP, but then they have the highest level of priority going out in time and allows us to put capacity in place for them, either whether it's in our directly in our factories or through our partners.

Got it. Okay. I guess just on PSP. When you introduced it, I think, collectively as a market, we were under the impression that, again, there was little room for negotiation and rescheduling of delivery schedules and things like that. But to your prior point and the point you made on the call, you're being a little bit more flexible. But I just want to clarify, pushouts are okay, but cancellations, you're still not taking...

That's correct. I mean, in some cases, we will offer to a customer that, hey, you could pay a pretty significant penalty as a cancellation fee and customers just aren't taking advantage of that, right? I mean, they -- this is backlog that will need eventually, I mean, typically, its customers that have very durable end markets that are entering into these PSP programs or long-term agreements with us. And with that, if they don't need the product next quarter, they likely need a quarter after that. And as they've seen the economy weakening, they've been able to make adjustments because it's a 12-month backlog that we're getting as when they're placing that 13th month or next month of backlog, they modify it and place 50% of what they had the month before as an example.

Right. Okay. Got it. The one question that we get quite often from investors is on pricing. It's accretive to you guys. It's been accretive to the broader industry. Can you speak to sort of the magnitude of price increases that you've experienced over the past couple of years? And now that demand appears to be moderating should we be expecting a potential decline in pricing? Or do you expect pricing to be relatively sticky given the nature...

I'll start with an answer, and Rich can chime in if he wants to. But pricing has been very stable for us historically. Really 8 to 10 years ago, we stopped giving kind of annual price declines to customers. But as we felt inflationary pressures on our business, whether that was in increased capital and labor cost of our own or what our foundry partners or other service providers were passing on to us. We we're inclined to pass that on to customers, not to gouge them from a gross margin percentage perspective, but to pass those costs on to them. and we were consistent with that with our customer base. I think they were appreciative in terms of the transparency that we have through that process. And really, it was relatively easy to do that, and customers were seeing that in other areas of their supply chain also. In terms of the stability in pricing going forward, we've seen some of these costs not decline, but the pressure go down a little bit. So we aren't seeing as significant increases from our foundry partners as an example. We are -- we've already made the investments that we've had to in capacity, and we have more capacity coming on next year. So I think it will moderate, and we did say for both fiscal '22 and fiscal '23 that the majority of our revenue increase actually came from volume increases, not from pricing. Anything else on pricing you want to add, Rich?

I think most customers have realized that the trickle down of equipment, depreciated equipment coming down into trailing edge technology doesn't really exist anymore. That model is dead. The equipment to use on 3 or 6, it does not trickle down. And in fact, the growth of 3- and 6-nanometer devices that want a [indiscernible] that occurred over the last 2 years, is that you can't sell that device without all of these trailing edge technologies. And so the entire industry has learned a big lesson in the last 2.5 years, that you have to actually not depend on depreciated equipment on those older nodes. They have to invest in new equipment. And so that depreciation cycle will take 5 to 7 years depending on whatever companies in there. And I think most companies know that or most customers know that today.

Got it. That makes sense. In terms of profitability, in your business since the beginning of the pandemic, I think you've expanded gross margins by nearly 700 basis points operating margins by 13 percentage points. You have a couple of charts in your presentation that are extremely helpful in sort of framing the cycles, if you will. But how should we think about the resiliency in your business model, you're a bigger company, you've got a bigger manufacturing footprint? How should we think about margins going forward, assuming sort of a soft landing, if you will, in your business?

Yes. So we've made an outright statement that we don't see any scenario where operating margins could fall below 40% on a non-GAAP basis. They're about 48% today. I think we will do much, much better than that. And there's a couple of reasons for that. So let's start on the gross margin line. So we are less dependent on internal manufacturing today than we have been historically. I think our last fiscal year, we did 37% or 38% of our wafer fab in-house. The rest is all outsourced. So if we had to cut production in our factories, the impact as a percentage of cost of sales would be much lower than it's been historically. Also in those charts that you referenced that are on our website, we were integrating acquisitions through much of that timeframe and acquiring companies that had a much lower gross margin. So that was impacting the gross and operating margins during that timeframe. And that's not the case today. We haven't done an acquisition in the last 5 years and don't intend to do another large-scale acquisition. So that should really help on the gross margin side of things. And on the operating margin side, we are well below our operating model today in terms of operating expenses. And because we've been limited on resources, we've been paying our employees very high variable compensation over the last couple of years. And it did again this last or these are quarterly programs. So if we were to hit a softer spots, those are easy things to pull back on and take these high percentages down to a lower percentage and moderate any impact on OpEx. So operating margins will stay high throughout the cycle.

Okay.

And I think some investors don't quite understand the impact of the acquisitions in terms of margin synergy. I mean, Microsemi was essentially operating as 22 separate companies. And when you integrate all of those entities into one operating system, and unleashes a tremendous amount of synergy, buying power, leveraging technology across business units, how we go to market with different solutions, that was something that we realized with Atmel on a larger acquisition where they were over a dozen separate companies within one company. Microsemi was 22 companies within the umbrella of a company but working extremely inefficiently. And we finally finished the integration of those 22 different companies in the last year. And that have leased a great deal of that margin that you see today.

Yes. So in integrations, he's referring to integrating into common ERP systems. So now we're functioning as a single company.

Got it. Okay. That makes sense. Eric, you talked about how you rely less on internal capacity today vis-a-vis past cycles and you're dependent on external foundry. How easy or not easy is it to cut wafers at your foundry suppliers? Because oftentimes, they talk about commitments from their customers, people like yourselves. So how smooth is that adjustment?

I mean we have partnerships with all of our large suppliers, including our foundry partners. So you need to be very balanced in that and we can take a foundry from 10,000 wafers a month to 2,000 wafers a month overnight. That's not how a partnership would work. But we do moderate those purchases. And actually, our foundry inventory is much higher today than our internal factory inventory. So we'll moderate over time based on the environment, and we are comfortable holding higher levels of inventory than we have historically. Obviously, since the Microsemi acquisition, the net leverage we have on our balance sheet has come down tremendously from almost 5x to under [ 1.5x ] today. So we have balance sheet capacity to hold higher levels of inventory, and our inventory is very long life. So we don't really have to worry about obsolescence, but it's a balance. And today, we are focusing on taking a little bit of that foundry inventory down and continuing to run our internal factories pretty hard.

Okay. Got it. I think a couple of quarters ago, you guys had floated the idea of building a 300-millimeter fab in the U.S. And ultimately, you decided against it. What were some of the pros and cons as you were debating that potential move inside the company?

There was a number of factors that led to our final decision. But first of all, during the up cycle, our foundry partners weren't really committing to make the investments that were needed for us on the what I would call the more trailing edge technologies because they saw the benefit on the leading edge, and we're allocating most of their capital dollars in that fashion. But what the foundry partners and the industry found out is that our customers, Microchip and other companies like Microchip and then the customers for Microchip and our competitors were willing to pay higher prices and then to support the capital investments that it would take for our foundry partners to make those investments. And so through that process, we were looking at would it make sense for us to build our own 300-millimeter factory with chipset funding other government funding versus our foundry partners doing that. And ultimately, we got to the right decision that our foundry partners were going to make those investments and not require us to do that.

Okay. All right. It's very clear. Maybe transitioning to more of the business side, if you will, and get Rich, to make a couple of comments on the analog business specifically. It's been a great business. You grew north of -- well north of 20% the past 2 calendar years, taking share. What's been sort of the driving force there in terms of whether it be end markets or device types or key strategies inside the company?

So broadly, the key strategies that we employed the same thing with the TSS. We identify anchor devices. So whether the anchor device is a PCIe bus switch device from Microsemi and FPGA, 32-bit MCU or MPU. And then we target a particular megatrend or application segment and that we build out that reference design or solution. And so we've become quite proficient at bringing 28 different business units together and then serving up those solutions to clients. The other thing that has really happened from a TSS standpoint is we've really gone in providing additional tools to that. So now we've automated a lot of the serving up of those devices and solutions. We actually have created -- we started investing over 2 years ago in an AIML engine that actually serves up the solutions when our sales force puts in an anchor device and an application it populates that solution with the appropriate devices based on all of our design history from our databases. And so it not only looks at past history, but it looks at customer journey that's taking place on our web and then a number of other devices that we have within the company and then serves that up to customers. So we really -- we see a dramatic improvement in overall productivity of finding the right device that goes with the right anchor device at the right time during that design cycle. And that is when -- we really did a great job during COVID period. And since we aren't doing any large acquisitions. We got to take a whole bunch of our internal resources and just work on overall productivity. TSS, megatrends, reference designs. And that focus is what's really helped catapult a lot of the attached growth that we're seeing now with the analog.

What differentiates you from other companies that perhaps also have analog and microcontroller technology under one roof? I know you've got more than that, but...

I think the biggest thing is we don't we don't reward different business units for outgrowing the other, right? And so the competition is not internal for resources. The competition is all directed external. And so from an employee equity or bonus system, it all is how well all 28 business units work, not individually how each business unit performs. And so when you remove that internal competitive element for natural for resources, everyone just works together to maximize the revenue externally. And then the second part of that is we are still the only semiconductor company and thankfully so that doesn't pay sales commissions to our [ CEMs ] and so our [ CEMs ] are not competing against each other for various sockets or design wins at different locations around the world. So the reason why it works is essentially those 2 things. And -- and those are the 2 hardest things for any company to put together is how do you compete for resources internally, R&D resources specifically? And then how do you eliminate a commission-based sales force? And -- and most companies haven't figured that out.

Yes. I think one thing that Rich didn't mention that I think is really important is that we've got such a strong foothold in microcontrollers and MPUs and FPGAs, which tend to be the first product that is selected by an engineer when they're building out their embedded system. And then that gives us an early view into what else the customer needs to be successful with the analog and the timing and connectivity and security products that go around that. But I think that, that is a strategic advantage that Microchip has over much of our competition.

Got it. Okay. Interesting. You talked about megatrends. And I think one of the beauties of your company or your business model is it's very diversified. So it's hard to pinpoint to one thing and say that's driving the company. That said, what are some of the things that you guys are focused on, excited about investing in when you're thinking about resource allocation?

So right now, we're seeing conversation that happens in every customer is security and sustainability, right, efficiency? How do I draw use less resources? Everybody is targeting net zero at 2030, 2040, 2050, those programs don't exist without semiconductors, right? And so there is a huge amount of discussion or focus on how do I reduce the overall energy usage in every application. So that is probably one of the largest or fastest-growing areas for us, and that's why we made that a megatrend last year. So we folded in AIML, like essentially was emerging in almost every megatrend. So rather than have that as a separate one, that just becomes a function within all the megatrends and then sustainability or preservation of natural resources as just paramount and almost every customer conversation.

Got it. And in terms of that AIML sort of exposure, if you will, what kind of technology, what kind of device type is most sort of geared to that cycle?