mic working yet? Okay. Good afternoon. Welcome to the second to last session of the TD Cowen's 51st annual TMT conference. You're almost done having to look at me and Matt. Very pleased to be joined by Eric Bjornholt and Sajid Daudi of Microchip Technology. Thank you both for joining us.
Thanks for having us.
Let's just try to get the short-term stuff out of the way as much as we can. It's a very dynamic business environment. I can't recall, and you've been doing this longer than me, a cycle where the end markets moved at such different paces and different times. Can you maybe walk us through, remind us what you said on the last earnings call of what you're seeing across the different end markets, and how auto and industrial are doing compared to consumer computing and whatnot?
Sure. I mean, we've said the last couple of quarters that, you know, consumer has been the weakest part of our business. For us, that's consumer appliance. It's only about 12% of our business. You know, our larger end markets are industrial, which includes Aerospace and Defense, data center, and automotive, and they've all been hanging in there pretty well. You know, we were very supply constrained for a long time, and that is obviously getting better. Foundry capacity is more readily available. Our own capacity has come online, but, you know, we're still got a lot of backlog that is unsupported.
We're shipping to those customers. You know, what's changed recently is we've got a few more requests for pushouts from customers, and we're trying to accommodate those, even if they're part of our PSP program. We can talk about that more if you guys want to in a bit.
For sure. Let's, let's talk about it now, I guess. You, you mentioned in the last call there were the more pushouts. Are those primarily focused on the consumer markets? Also, can you help tie that together with your commentary on inventories? I think you were at 169 days, which came in higher than your expected and are expected to trend down in the back half of the year. Can you walk through, you know, your comfort with current levels? Is it a soft, trying to take down? Are you cutting utilizations? What are you doing to manage levels now?
I'll address the inventory question. As you said, we ended our fiscal year in March with 169 days of inventory. There's a few components of that. We had about seven days that relates to last time buy inventory from some of our foundry partners that are end-of-lifing process technologies that, you know, we see 10, 15 years of opportunity for sales to customers at very high margins. We've taken that inventory in, and that will stick with us for a period of time. Although it's not permanent, it will be with us for some time.
The other piece is our internally manufactured product from our own fabs is still pretty low. Our die banks are lower than we'd like them to be, so we're continuing to run our internal factories hard. Foundry capacity, as it started to free up in the December quarter and the March quarter, you know, our business units and our operations teams and our customers were really starved for product. We took advantage of that capacity coming online and took quite a bit of inventory on, and that was really the driver of the growth. In the current quarter, we've actually guided the street that inventory will decline.
In dollars and in days, we're targeting about a five to 10 day reduction in inventory. Foundry will be a piece of that, as the supply chain has loosened up, we were carrying higher levels of raw materials, and we'll start to bring that down. The other piece that impacted inventory last quarter is we were starting to accept some pushout activity from customers. We had product that was further along in the manufacturing process, some even in finished goods, that we ended up holding and will ship to customers at a later date.
I think, can you spend a minute talking about the PSP program. I think there's some confusion between the PSP program, non-cancellable orders, and enforcing those. When you have a customer in the PSP program that comes to you and says, we'd like to push out our orders, how does that conversation go? Then, you know, perhaps most importantly, are the pricing levels firm within the PSP program?
Okay. All right, PSP, we put it into place in February of 2021. It was based on customers requesting a program to give them more surety of supply. The program was developed for customers to provide us with 12 months of non-cancellable, non-reschedule backlog. Customers who have participated in that program have been serviced very well. It's been a very good program for customers. Customers do not always have a crystal ball of what demand is going to be, and obviously, the demand environment has weakened in some end markets and some geographies.
With that, we have received some requests for cancellation and/or pushout. You know, we are not really accepting cancellation requests. We might offer up a cancellation fee that would be punitive to the customer. You know, ultimately, the customer is going to need this inventory.
We're working with customers who are self-identifying these inventory positions and allowing some pushout of that product as it makes sense for us. In some cases, there's another customer that's desperately waiting for that product, so that's an easy thing for us to do to satisfy the one that's in need of the product and the one that wants to push it out a little bit. In other cases, we're just giving customers some flexibility.
Okay.
Eric, would you characterize maybe you can walk us through the end markets where you're seeing those activities, right? I wouldn't imagine that that would be uniform across the business, where people are coming back and asking for pushouts, or there's cancellations. Maybe what types of products, what end markets? Are there anything unique geographically? Just, maybe we could start there, then we can dive in further after that.
Okay. All right. You know, it actually has been really across all the end markets that we've had some of these requests. It's not that, you know, we've been completely constrained in data center, and now no customer is requesting a pushout activity. We've seen it across all. Again, these are selective. You know, this isn't the vast majority. What we've always said about PSP backlog is it is better backlog than if we hadn't had the program in place.
Because if a customer doesn't have skin in the game, they can book an order, schedule it right outside the cancellation window, which for us would normally be 90 days, and then just continuously push that out or cancel it at some point in time. Microchip then doesn't know what we need to go and invest in from a capital perspective or people or make commitments to our foundries. With PSP, a customer generally needs to go to a higher level within their organization and ask for an approval or sign off on that type of order. That's what would happen at Microchip.
I'd need to sign off on it, Ganesh or a senior VP. I think customers naturally kind of hedge themselves a little bit with that program. We've always seen that the first six months of PSP backlog is fuller than months seven through 12. You know, with that, I think that the backlog is still good under PSP, but in some cases, customers overforecast it a little bit, and we're giving them some relief.
Got it. I'm gonna ask one more on this. I just wanted to say to the audience that we do want to make this somewhat interactive. If you guys have questions, just wave your arms around, and we'll sort that out. One of the follow-up that I guess wanted to dig into a little bit more is when you guys signed up customers for PSP originally, and then they wanted to sign up, do you, do you guys get any view into customer inventory levels as part of the program? Is, and just speaking more broadly, like, you can monitor your distribution inventory. I'm sure you do very, very closely, as most of your peers do.
Different ones of your peer companies have different levels and different programs to get on at customer inventory level visibility. I just wondered, like, the closeness of the partnership with yourself and your customers through PSP, do you get any of that data back? How has the visibility changed over the last, I don't know, 24 months?
We don't, we don't get any sort of reporting from our direct customers on their inventory levels. Distributors all are required to provide us with inventory and sell-through reports every month. We reconcile their inventory and do all that. On the direct customer side, it's more anecdotal in terms of having discussions. PSP does allow a higher level of conversation to be had between senior management at Microchip and the customer. I think there's better interaction than there used to be, but we do not get any sort of real-time, regular reporting in terms of inventory levels at our customers.
Remember, you know, our customer base is very diversified. We service 125,000 customers. Our top 10 customers are, you know, around 10-12% of our revenue. Many of those larger customers than that are contract manufacturers that are servicing, you know, dozens and hundreds of customers.
Any just anecdotal view on levels of customer inventory? Y ou've talked pretty openly about where things are with the distribution inventory, but I don't know, just gut feel about customer inventory now.
Our gut feel still is that we were so constrained for so long and still are constrained. We have a lot of what we call unsupported backlog that customers have requested, that we are delinquent to their request date. You know, because we are manufactured on these trailing edge technologies, where a lot of these supply constraints were, you know, we don't think there's been enough time for significant inventory to be built up. With the diversified, large customer base that we have, there's no doubt there's pockets of inventory out there that have built up, but we can't quantify that.
A lot of your peers have talked about de-emphasizing the channel and going more direct. I would be curious to hear your views on that. Is that a short-term thing that, I guess, companies are doing in light of the current environment? Is it a structural change because of the difference in exposure of trailing edge semiconductors now? I'd be curious how your strategic view of the channel long term.
The channel is super important to us. You know, we probably have 120, 130 distributors that we work with, that's a variety of the global distributors, like the Arrow and Avnet of the world that you're all familiar with, DigiKey and Mouser, that kind of seed the market with new design activity. Then the larger piece of our distribution revenue comes through regional distributors that are really dedicated to the Microchip product line, get trained just like our sales force, and do a good job of actually going out and creating demand for us, winning designs.
The way our distribution program works, if a distributor is providing logistics-type services, fulfillment services, they earn a fulfillment lower margin. If they're involved in the design activities, they earn a higher margin, and we monitor or manage that through the way we sell the distributors at a uniform higher price. Once the distributor goes out and identifies an opportunity, they register that design with us, and then that distributor is the only distributor for that product going to that customer that can get a reduction in that original purchase price. It's worked well for us for a long time.
We are not de-emphasizing the channel. The percentage of our revenue that goes through distribution has been pretty steady over the last three or four years. It's declined just modestly. We saw maybe about two years ago, maybe it's a year and a half ago, that distribution took a little bit of a sharper drop-down in a particular quarter, but that was more based on distribution being later to join the PSP program than many of our direct customers.
Because our distributors then had to go on, if they were going to be on PSP, had to then go work that with their customers to negotiate NCNR of their own, because that's not our relationship, that's their relationship to manage. It's kind of bounced back. It's about 48% of our business today.
I'm going to try to make this one the last short-term one. There's widespread concern of pricing that's been up objectively in the semi- trailing and semiconductor industry for the last 2 years, but starting to soften as your end market soften. Your results and your peers have objectively shown that that's not happened. When things soften, or if and when things soften in auto industrial further, do you anticipate a sharp change in pricing? Like, have your conversations with your auto industrial customers changed versus the past, of are they willing to, I guess, work with you more when demand softens?
We are not anticipating price declines. You know, we took the stance with our customers probably 8 to 10 years ago, of not granting annual price declines, right? You know, the semiconductor was giving away value with that. We made that change and, you know, gradually got better and better, and better at enforcing that. Pre-COVID, you know, there was really no annual price declines that were happening with customers. Now, the supply and demand situation, and the increase in inflationary pressures on our business, increased capital costs, increased labor costs in our factories.
That impacted Microchip directly with our cost structure. Our foundry partners and other suppliers were experiencing the same thing. We have passed those cost increases on to customers through pricing increases, not to be gross margin percentage accretive, but we are earning a Microchip-like margin on those increases, and we had price increases the last two fiscal years. What we've said is the majority of our revenue growth was driven by unit volume, not by pricing increases. Now, we think that those prices are very sticky and here to stay.
The two largest components of those cost increases were capital, which once you buy the capital, it's fixed and gonna be depreciated over five years, seven years, 10 years, whatever that schedule is. Labor costs aren't going down, and we think that our supplier base is seeing the same thing. There's still inflationary pressures on the business today. It is not our intention to pass on additional price increases to customers, but if inflation continues the way that it has, we would need to consider that. We have no intention of decreasing prices, and I don't think our customer base expects that either.
Unless there's a question in the audience, maybe we'll shift gears a little bit. Sajid, Microchip's identified, I think it's six megatrends, and the last we heard were about 45% of revenue. Can you walk through, you know, what are the growth compositions of this versus your legacy business, and what are some of the more, I guess, meaningful growth drivers underneath those six megatrends that the Microchip team's excited about?
Yeah. I'll highlight a couple of them in there, a couple of megatrends, right, which make up a bigger proportion, automotive and industrials, largely. You kind of think about automotive, it's about 17% of our total revenue today. And there's a lot of areas that we see opportunities, growth opportunities and current and new opportunities as well, right? In, you know, electric motors, a lot of our components go in there. These are motor control, controllers, positioning controllers, and the DC-DC converters, we play a big role in. We also have in EV chargers and a bunch of other areas.
Automotive is a nice growth opportunity for us. Supplementing automotive or complementing it is on the ADAS side as well. Lots of opportunities there with sensors, controllers. You know, we're seeing a kind of shifting, and we talked about that earlier, too, on the kind of the automotive architecture as well, and that's gonna create, I think, a ton of opportunities for us, largely on the Ethernet side and the PCI Express side as well.
Because as these ECUs get replaced by a central compute, you're gonna have a lot more sensors, a lot more data that's getting processed at the point of collection, and that creates opportunities for us. You know, lots and lots of areas where we think that's good. On the industrial side with we are in. T he design activity is longer on the industrial side, but once, you know, it starts, once it's designed in, the ramp is 12 to eight to 10 to 12 years. That's a very meaningful ramp there. A lot of the areas that we're in is your kind of core industrial application.
Think of some of the newer stuff would be the smart city infrastructure that's going through, smart lighting on the streets, traditional, you get factory automation. That's a big, evolving process there that's happening. Automation, access, building access points, a lot of the, smoke detectors, and medical is a big piece of industrial as well. Lots of different areas of focus. There's not really one specific core area that we are dominating in.
It's just really our products are used across the board. Same thing with, you know, on the 5G side. As that connectivity goes through, we're still probably in the early stages of that 5G transition as well. What else? e-mobility, right? We talked about e-mobility a little bit, and that's on the automotive side. Industrials, I would just point out one thing that's currently about 41% of our revenue that moved up from 40%-41% last year. Within that, about 8%-10% is Aerospace and Defense, and that, of the total revenue, so 8%-10% of the total revenue, representing about 25% of the industrial component.
The Aerospace and Defense segment, you can probably think about it in three partitions. You have the commercial aviation, space, and defense side of things. All three have been doing really well. Commercial aviation recently has come out. Yesterday we were at a different conference and we put up, you know, some slide that showed our content in a large airplane, which is greater than 100,000. I forget the model of it, but it was one of the larger Boeing planes. So, you know, really a broad. That, I think, is what drives some of that resilience and differentiation that you're seeing.
Like Eric pointed out, a lot of our consumer stuff is separated from that. T hat would be the kind of areas that we're excited about there. Sustainability, sorry, is another one that we've recently added as a megatrend. We're just seeing a lot of applications come through. These are energy preserving or greenhouse reduction type of tools or better and more energy efficient tools. Megatrends are expected to grow 2x our core revenue. We're pretty excited about, you know, what we can bring. That's where a lot of the TSS activity happens as well.
Yeah, we showed at our Analyst Day back in November of 2021, kind of a baseline of where those megatrends were at about a third of our revenue. You know, leaving this last year, just two years forward, fiscal 2021 to fiscal 2023, they represent 45% of our revenue. Over that two-year period, Microchip grew about 55%, and the megatrends grew at double that rate.
Thank you. I guess within auto specifically, a lot of attention gets paid to high-voltage power discretes and ADAS processing, 'cause they're easier to model with an attach rate. There's a ton of other semiconductor components that are getting pulled into the vehicle, both because of those and in addition to those trends of EVs and ADAS. You know, one of your larger peers recently took up their long-term target model because, explicitly because of auto and industrial growth. They're not exposed to those high-voltage power semi verticals either. You know, what's a reasonable rate?
In general, auto semis have grown mid-to-high single digits. It feels like we've hit a knee in the curve here. We're coming off two very strong years. How should we think about the longer term growth rates of auto and industrial growth for your, quote, core exposure?
You know, we don't break any of our end market out in term to, you know, what they are gonna grow in the future. I mean, we've consistently outgrown the market year after year. I mean, we've been organic growers for the last five years since we've acquired the Microsemi acquisition. You know, EV and ADAS are a significant growth driver for us, but it's really the whole electronification of the car or the data center on wheels that's gonna help continue to push that forward, as well as then the consumer's expectation of what's in a high-end vehicle today, moving down into lower-level models over time.
We're excited about automotive. Again, it's not the largest piece of our business. It's at 17%, but it grew kind of in line with the corporate average last year. It's not really all about the number of cars produced, it's about the content in those vehicles. In our slide deck that we've got up on our website, you'll see an example of a vehicle, the new Iconic Six. It's got 88 different parts from Microchip, controlling all sorts of features throughout the vehicle, to something very simple, like a power window, power door lock, to more sophisticated systems with an arm, we'll be able to touch screens and things like that.
Eric, I'm gonna switch gears just a little bit. You guys are talking about megatrends, but one of the things that needs to happen for you to serve that kind of growth is capital investment and capacity, right? You guys have a hybrid model. You get a bunch of capacity from outside the company. You obviously have your own fabs, and you're investing. There's been a lot of back and forth over the last, I don't know, 12 months on the strategy there.
Maybe we could revisit how that conversation's gone with yourself and Ganesh and the board, and there was thoughts maybe to do a 300mm fab yourselves, then that got shelved, and you're not gonna pursue that anymore. The market's been dynamic. For a while, you couldn't get tools. You see some of your peers in trailing edge semi spending gangbusters over the next five years to build out capacity. Some have leaned more into external foundry.
There's a lot going on around your company, and different competitors of yours are taking extraordinarily different tacks as to how they're gonna pursue the next five years. Maybe just revisit the last 24 months, leading up to this and just how you're thinking about that now.
Sure. Sure. F irst, I'm gonna say we're very comfortable in our own skin, and, you know, what our competitors are doing from a CapEx perspective doesn't really drive decision-making within Microchip. Maybe start with the 300mm fab that we explored. We do not own intellectual property to run the process technologies that our foundries have in our own factory for 12-inch, because we don't, we don't make 12-inch products today. We have six-inch and 8-inch factories. That was going to be a challenge for us.
What we found is when foundries are not looking to expand in a certain area, they are willing to grant us a license. We took a license from one of our major foundry suppliers a couple years ago now, to take inside an 8-inch process technology that they were no longer gonna invest in. They could turn it into a long-term revenue stream for them, made sense for them, made sense for us to bring it in-house because we saw 10, 15, 20 years' growth on these products. You know, that's kind of how we evaluate these things over time.
With CHIPS Act funding, we thought it was appropriate to evaluate the possibility of a 300mm fab. We were talking with state and local governments, and we would have liked to done this outside of the public domain, but we were having these discussions, and there was a very high likelihood that word was gonna leak, that Microchip was talking to Arizona, Oregon, other municipalities. We said, we're evaluating it, and this is why. There was a period during the upcycle where foundries were not making commitments to invest on the trailing edge.
T hey were seeing a higher benefit for them to invest on the leading edge. What's happened is that the foundries and the industry has seen that with these higher capital costs, right, there's no 8-inch used equipment on the marketplace, as an example, 12-inch equipment is very expensive, that customers are willing and able to pay a higher price for these products to support those capital investments. As we were going through the analysis of should we do this ourselves, should we do it with a partner, or should we allow the foundries to be able to support this for us? That's the analysis we went through.
Ganesh had lots of detailed discussions with upper management at our larger foundry partners and got comfortable through that process that they were going to make the appropriate investments in appropriate locations that made sense for us to drive our capacity needs for the next five to 10 years. I'm not gonna sit up here and tell you that we'll never have a 12-inch fab in our future. It's possible, there are challenges from a loading perspective and from a process technology perspective where we don't own the IP rights. Capital, in general, the capital costs are significant.
We spent near the high end of our range of our long-term model last fiscal year. Our long-term model is 3%-6% of revenue. We laid that out at our Analyst Day a year and a half ago, two years ago, and, you know, we feel very comfortable staying within those bounds. This fiscal year will be lower. We've guided to $300 million-$400 million, or $350 million at the midpoint of guidance versus last year, which was about $500 million. Capital equipment lead times have been long, we made a lot of commitments over the last 2 years, and some of those are gonna continue to come in in the current fiscal year.
If we need it from a capacity perspective, we'll install it, place it in service. If we don't need it, we'll wrap it for a period of time and use it in the future. I don't know if I answered all your questions, but I think I got most of them.
No, I think so. No, thank you for that. There's a lot of moving parts there. The mix of the business, I guess, to follow up to that then, is the mix of the business that you expect to be external versus internal, has that, through this whole evaluation of the 300mm facility and the decision you made, has that long-term balance changed?
It really hasn't. We finished this last fiscal year at 37% or 38% internal fab production. The rest is outsourced professional foundries. At our Analyst Day, which I mentioned before, Analyst and Investor Day, we talked about potentially getting to about 45% internal over time. I still think that's possible, but I think it's kind of in that range where we are today, to up to 45%. On the back end, assembly and test side, we do more of that in-house. Today, we do about 60% of our assembly, maybe 67% of our tests, and those percentages will likely go higher over time.
I guess since we're on this sort of evaluation of the business kind of tone, Microchip, for a long, long, long time under Steve's leadership, was a very M&A-intensive organization, then that strategy completely changed, right, to be one much more focused on capital return as the portfolio matured. Valuations have changed up, down, and sideways. And I don't think we've ever seen the last five trading days in semis like there's just been. We can revisit this conversation hour by hour. How are you still on this firm commitment to capital return side of the pendulum there?
Have we swung back toward the middle at all in consideration of smaller M&A, larger M&A? T he big picture from your perspective, was there anything large enough to be material to Microchip that you think you could actually get done, given the regulatory environment?
Yeah, all good questions. On the M&A side, you know, we started a journey in doing M&A kind of in the 2010 timeframe. At that point in time, we were about $1 billion in revenue. We were primarily a microcontroller company. We had a nice portion of analog, but it was smaller. What we just wanted to do over the next multi years was to expand the portfolio, not just through our organic efforts, but use the cash flow of the business, the business model that we had, to go out and acquire various pieces of the semiconductor pie that made sense for us in providing more complete solutions to our customers.
Post the Microsemi acquisition, and, you know, there was a lot of acquisitions between that 2010 date and 2018, when we acquired Microsemi, so we've got a very robust analog portfolio today. You know, it's about $2.4 billion. We've got everything from the low end of 8-bit to the high end of 32-bit in microcontrollers. It's not just that. We've got FPGAs that came from Microsemi. We've got all sorts of connectivity products within the portfolio. We have timing, we have security, we have memory.
That allows us today to go to our customers and present them with more complete solutions, where we can come in with a working reference design, speed their time to market, have them avoid the hassle of dealing with parts from various different companies interfacing with each other, reduce their efforts in R&D, and become more valuable, sticky suppliers for them.
That was the goal all along, and post Microsemi, we don't look at the portfolio and think that we have a gaping hole that we need to fill through a large-scale, transformational acquisition, like a Microsemi was, like Atmel was for us at the time. Today, we're very comfortable with the portfolio that we have. We think that we're providing our customers with what we need.
Now, we will likely do some small acquisitions, tuck-in, IP, R&D teams, things like that, and we've actually done some of those post Microsemi, but they're a blip, from you guys' perspective. You know, it doesn't take a lot of cash. These are kind of sub-$20 million deals that, you know, just help us propel some of our roadmaps further. Post Microsemi, we've been focused on the megatrends in the industry, growing organically, selling TSS, or total system solutions, to our customers, and through that time period, reducing the leverage on our balance sheet.
We've taken it from about 5 x, down to last quarter was about 1.4 or 1.5 times. Now we've laid out a very clear path for investors that we want to get to 100% free cash flow return by the March 2025 quarter, so that's 7 quarters from now. You know, we're just gonna execute on that. Today, we're doing 67.5%. We'll increase it by 5% each quarter until we get to the 100%.
When we get there, I can't look at our business and see it's much different from some of our high-quality competitors in the analog space, when you look at gross margins, operating margins, free cash flow, and capital return, there's a huge valuation discount that we have today. All of you can look at that and make your own assessment, that valuation gap should close as we move towards 100% free cash flow return.
Well, well done. Right on time. Thank you, Eric. Thank you, Sajid. Really enjoyed the conversation, and thank you for participating.
Thanks, everybody.
Thanks, everyone.