All right, welcome back, everybody. Joe Moore from Morgan Stanley, Semiconductor Research. Very happy to have with us today the CEO of Microchip, Ganesh Moorthy. I think, Ganesh, I think you want to make a few opening comments, and then we'll go into Q&A.
All right, thank you, Joe. So let me remind you, during the course of this fireside chat, we're going to be making some projections and forward-looking statements regarding future events and financial performance of the company. And we wish to caution you on these statements and refer you to our filings with the SEC. Today we're not updating our guidance, which we provided at our earnings call in early February. That guidance, just as a reminder, was $1.325 billion in revenue, ±$100 million as the range for it. And we believe that at that revenue level, we are significantly under-shipping to what the end market consumption is. Thought I'd give you a little bit of color on the business before we go into the questions.
Our bookings improved modestly here in February from what we have seen over the last two quarters, but they still remain below what we would consider to be normal levels. The cancellations and push-outs are stable and manageable. In the March quarter, we're expecting that distribution sell-through will be significantly higher than sell-in, even as some of that sell-through will be lower than prior quarters. So we are draining inventory both with customers directly as well as through our channels as well. With the short lead time, we are seeing some early signs of green shoots, very early, so I don't want to get everybody too excited by it. But it's not at a point where we can call it a trend. We have modeled using the Street modeling of revenue at flat-ish for next quarter versus this quarter.
We still have a lot of turns we've got to take in the June quarter as well, so we've got a while to go. But we are seeing some places where our expedite requests are starting to come in; there are bookings starting to come in which are aging nicely into the near term as well. And we stay focused on the long term, which is around, number one, our own innovation that we're delivering to customers. This is all of our new products. Two, our design activity with our customers to enable and empower their innovation. And three, working on the things that are improving Microchip so that as the upcycle comes, we are more fighting fit and ready to go as that part of the cycle comes in as well.
All of that is based on, focused on the key Megatrends and then selling a complete solution with Total System Solutions we have. And with that, I'll turn it back to you.
Hey, great. So it might not be time to rent a hall and hire a band, but you definitely see signs that we're getting near a bottom in all of this.
I think some of the initial signs are there. What we'd like to see is multiple data points trended over time, to call it a trend. So we have a first data point, and let's see how it all trends in time.
Great. Well, maybe we could focus on the shorter-term dynamics as different than the script we had, but just because you brought it up to start. This revenue decline has actually been quite severe, and we've looked at it not just for you but for the peer group. Revenues have come down from peak to trough only two other times more than this, and that was 2001, 2009, which were periods of real demand destruction, where now I feel like demand not great, maybe, but this is primarily inventory. So it's kind of maybe the worst inventory correction of all time. I guess, would you agree with that? And do you does that surprise you? Obviously, it was also a very good environment that led up to this. There was inventory accumulation. We all recognize that. How surprising is that to you to have such a severe decline?
Yeah, I think the peak-to-trough correction has been quite severe. However, if you look at it on a full-year basis, fiscal year basis, right, we had some stronger quarters, we had some weaker quarters. On a year-over-year basis, using our midpoint of guidance for March, down 9.5%. So it's not out of the ordinary with respect to how an inventory correction would go. But certainly peak-to-trough. And I think in the early quarters, there was shipment ahead of what was required now in retrospect that we can see. And people were placing orders farther out in time. People were growing inventory for their own needs, thinking their business would grow and/or to protect against whatever concerns they had on supply chain.
Great. And you talked about bookings improving at the margin. Can you give us a sense where those bookings are relative to the billings?
So book-to-bill remains under one, and that's not unusual because book-to-bill usually will reflect lead times, and lead times are very short. And so people are going to give us very low visibility because they don't need to. And where they have the uncertainty in their own business and no concern about supply, there's no need for them to give us that. But all that said, at least bookings are for a month firming up as compared to where the prior six months were at. And it has to go quite a while before we get to where we would say bookings are coming in at the same rate as billings.
OK, great. I mean, I just as a philosophical matter, I like to buy semiconductor stocks when they say stuff like this. You're under-shipping demand. You have these tailwinds that are going to kick in. But again, the inventory build that preceded it was quite large. You spoke to the inventory reduction at distribution. Do you have any sense for how big inventories are beyond distribution at the end customer and where we are in terms of burning those inventories off?
Yeah, no, it's a great question. It's harder for us to see because we have transparency on what distribution revenue distributions or inventory looks like. What we didn't quite appreciate was how much inventory exists with customers off distribution themselves. And so they are slowing down purchases from distribution. So we've seen sell-through revenue is weaker than we would have expected, even as we are shipping less into distribution over this period of time. So it's hard to know. I think it will clear itself out as time goes on. But we don't have a clean insight into how much inventory are people carrying beyond what distribution customers are telling distributors in terms of what their demand looks like.
OK, great. As you look at these trends, is there any difference by end market or geography? Or is it a problem to isolate which markets are better or worse than others?
So I would say mostly the end markets are all in different forms of inventory correction. The two places where we have strength in the current environment, one is our aerospace and defense business. It's composed of defense, space, and commercial aviation. And there is strength in all of those. And it's only about 9% of our revenue, so to put that in perspective. The second place we see strength is in the AI subsegment of our data center business. And so with the growth in the AI servers, there's content from Microchip that is in on these, and they are growing extremely well. But in the grand scheme of the Microchip scale, it's not enough to move the negatives that we see in other areas of inventory correction.
Yeah, OK. I assume the other parts of the data center business that aren't AI, there's some offsetting weakness?
Yeah, so data center overall is down. The AI piece of it is up. The other pieces of it are down. And we don't know how as the year goes. There's been an underinvestment in the normal data center capital. And if some of that will come up, anecdotally we hear perhaps that will. But it's clearly AI servers are where a lot of the capital is being invested in.
Yeah. I feel like people are surprised when we talk about this for you guys because they forget you've acquired PMC-Sierra at all these businesses that have all these high-quality data center exposures.
Yeah, and our data center exposure is not limited to what PMC-Sierra had that came through Microsemi, right? I mean, that's clearly one piece of it. But if you look at Microchip products that go into the clocks in data centers, the root of trust in the data centers, just this morning we announced our next generation of products for backplane monitoring. We have products about a couple of months ago we announced which are the active electrical cables. So we do have multiple plays in the data center. And then we've always had a historical position in the power supplies for energy-efficient power supplies that are needed in data centers.
I guess just since you mentioned that stuff, you guys have so many interesting businesses within Microchip, and you tend to not talk a lot about it. You talk at the kind of broad segment level. I understand they're oligopoly businesses. There's probably a limited amount. But some of your peers talk a lot. Talk about like 18 growth drivers for the next three years and things like that. What has led you to strike that balance? Do you think because I feel like people look at your businesses maybe monolithically, a lot of microcontrollers and things like that, and there's a lot more complexity to it.
Sorry, ask your question again.
Yeah, so the question is, what leads to the decision to sort of not give us more color on some of the smaller exposures that you may have?
Because really, the power of the business is around the whole being greater than the sum of the parts. Yeah, it's yes, we have a great FPGA business, but its value is not just on FPGA. It's how does it pull through a whole bunch of our other product lines that are in all of these applications. The same is true in we talked about data center. Yes, we have a wonderful storage networking story, but it's much more than just that alone. So what we try to do is really sell to as complete a solution. We call it Total System Solutions that bring all of the strengths from the acquired companies with the strengths of Microchip pre-acquisitions into a more complete solution. And that, I think, at the end is what drives the overall performance of growth, profitability, et cetera, that we have shown for the business.
Yeah, OK. By the way, I'm not pushing you one direction or another. I actually appreciate that you're not super promotional around a lot of those things. You bring up a good example in FPGAs. The EV/sales of the standalone FPGA company is quite high. So there's a lot of things that maybe you don't get full value for. Maybe we could step back and talk about the strategy. I mean, you've built the company through inorganic growth for a period of time where you bought some really interesting assets. I think you grew the core business at the same time, but you were definitely very focused on building out scale and scope. Then you sort of came to the point where you were comfortable with where you were, and you focused more on growing organically. Can you talk about that shift?
What the assets that you have got you to a certain point, what was it that got you to say, OK, at this point we're focused more on our internal growth?
For about 8 years or 9 years, starting about 2010 through 2018, we did a number of acquisitions, about 17 of them in all. 7 of them were public. The rest were private companies. It was intended to build out the portfolio into many dimensions that we saw as being needed and which we didn't have. It gave us both scale in the portfolio but also scale in the size of the company. In that process, we were able to dramatically improve the overall performance characteristics of the company, gross margins and operating margins over that period of time, dramatically improved. We also bought at a time when the valuations of the companies that we acquired were significantly more within a realm of what we felt were good, good for the acquired entity and good for us.
And then the scale of the last two acquisitions, Atmel and then Microsemi, were quite substantial. Each was about 40% of the size of the predecessor Microchip standalone. And the work involved in trying to bring out the synergies and both synergies on the cost side but synergies on the revenue side were tremendous. So by the time we finished the Microsemi acquisition in 2018, we had a powerful set of assets that addressed many, many parts of the market that were important to us. But we also had debt. By the time between Atmel and Microsemi, we had a net leverage of close to 5. And we needed to start paying down the debt. We needed management bandwidth, which needed was all being applied towards the synergies. We didn't have management bandwidth to go off and do more acquisitions.
And so we focused inwardly on take advantage of the portfolio, improve the profitability, generate cash, pay down the debt. We got to a point about 2021 where we looked around and we said, OK, we have the bandwidth. We have started to bring the debt down to a level where we could do more acquisitions, but we didn't like the valuations that were there. So we put the focus back on organic growth, get the best out of what you can in the portfolio. We launched the Total System Solutions combined with Megatrends as a growth strategy. Then we decided to take the free cash flow and begin to return it to shareholders.
Incrementally, we have taken it where today we're at 82.5% of free cash flow being returned. Goes to 100% over the next 3 quarter, 4 quarters by next March.
OK, that's very helpful. Thank you. So then when you talk about extracting these synergies and the total systems solutions, can you give us a sense for the scope of that effort? How much investment are you putting into that? And when do you think we might see that pay off?
So I think it is already paying off today. There is not a design at Microchip that we approach a customer with where the approach, the selling process, doesn't comprehend an understanding of what does that customer's board look like, what is the system requirements, and how do we position ourselves to have a more complete part of what they're trying to go do. So it's much more than just saying we want to go do it. It's got to have all the systems from at the point we touch the customer to how are we going to accept the orders, how are we going to plan for them, how are we going to fulfill them, how are we going to next generation of product that we're going to develop. So that's all become part of the DNA, and it takes time to put it all together.
But the growth rates that we're expecting for the company that we expressed back in 2021 where we said, hey, we expect to grow 10%-15% on a compounded annual growth rate using fiscal 2021 as a baseline is based on that assumption that we'll get a multiplier. And we have seen it in terms of how we go to market, how much of a customer's board we're able to get. And then the second piece was to then attach it to the fastest growing areas, which is the Megatrends.
That's helpful. So I guess we've been quite positive on your stock. And I hear a number of objections, concerns that people have. So maybe I'll just run through some of those, maybe starting with the PSP program, which to me was an effort to try to sort of build some trust between supplier and customer, but maybe a perception that you sort of forced more inventory on customers' balance sheets than they would have liked. Can you just talk to that concern ?
Sure. So PSP was launched three years ago, and it was in like January, February of 2021. It was in response to a supply-demand imbalance that was just enormous. And we needed to sort out what demand was real and what was not. And so part of this was we said, hey, if you want us to go put the capacity in place and hire the people and build the factories and all of that, we need to make sure your demand is in fact real. And one of the ways in which we'll judge it to be real is if you're willing to stand behind it as noncancelable. It worked extremely well through 2021 and through 2022 and the early parts of 2023 where people got priority for placing those orders. Now, in that period of time, lead times also began to stretch out farther and farther.
There came a point in time where that supply-demand was starting to shift, where supply was improving. Then customers' view of demand was inflated and began to change as they saw business conditions for them saying, hey, I don't have the same growth rate. But they had long lead times. So they were placing orders non-cancellable as it might be, one year out in time without sufficient understanding or visibility into their own business. And that created the overhang. And certainly, in some of the early phases, we were trying to hold people accountable to orders that they had placed. And we had, for even all the way through the second calendar quarter of next year of last year, people who were screaming for, you got to ship me product. You got to ship me more. You got to ship it yesterday.
Then within 2 months , 3 months , 4 months, people began to say, no, no, slow down, push it out, et cetera. The program was designed for what the problem we were going to solve. And the problem began to get solved by supply-demand began to balance. And then we began to evolve the program. We cut the amount of time that they had to give us. We finally got to the point where we no longer have the program because it's not useful to do it. But was that a period of time where customers were taking product that they felt some remorse over having to take? Sure. And by the way, we made investments that we have remorse over as well, right? I'm sitting on $350 million of capital that we have purchased and is sitting on our balance sheet unused.
But we all have to take risks, and we have to have some sense of a shared risk and shared reward.
Yeah. I mean, I feel like every company that I cover in the analog microcontroller space has seen some variant of this play out. And I think I would just say when we saw checks that your business was starting to weaken in a meaningful way, within a month, you were on an earnings call saying, "We're seeing early signs of weakness in all of our businesses." So I feel like there's a disconnect between the perception and maybe the reality that you guys have actually done a very good job of characterizing to us when the business started to soften. The other couple of anxieties I hear are China and then pricing, which maybe go hand in hand. But China obviously has built a lot of capacity.
With an uncertainty as to what they're going to build in that capacity, I feel like when I talk to investors, it's like, oh, microcontrollers. That's what they're going to build. Which it seems like the microcontroller business is actually not like that. It's quite fragmented. You need a lot of products to build factories. I know you have provided some data in terms of how vulnerable you might be to Chinese incumbent competition. But just broadly speaking, I feel like you have a business that's pretty insulated from all of this. How do you feel?
Sure. So mainland China forms about 20% of Microchip's revenue. About half of it, so let's say 10%-ish, 10% is designed outside. So it's designed in the U.S. or Europe or somewhere else and gets produced there by the contract manufacturers. That part of the business, the point of design is outside of China. In the remaining part that's inside mainland China, about another half of that is of the type of products that are not at a degree of complexity that are not really what the local manufacturers are going to be able to work on. So there's about 5% of Microchip's overall revenue, which fits within what I would call broad-based microcontrollers and analog and power and all of that that you could say trailing-edge capacity could be applied to.
The beauty of that 5% is, sure, you can have a product, but the surface area of attack is quite narrow because every design is a small volume and hundreds of customers and designs that you have to win. So that's a fragmentation. And to win it requires a large investment in products and technology, a long patience for design- in cycles, and then a customer design- in effort that has to be over thousands of customers to be able to win it. So are we going to lose some of it? We probably will. But there's always been competition. And we may not have lost it to the Chinese company. We may have lost it to some other competition as well. We have to continue to provide value to customers.
That value comes in innovation of our products, not only the silicon, but the software and the tools and the support that comes with it has come in other forms that a customer views our solution as a better solution. If we don't differentiate along those lines, then we will lose some of that. But this has been 30+ years of time where the competition has been different, different names, different regions of the world. And we have had to work through it all.
Thank you for that. Then finally, pricing. We've had long periods of underutilization of these kind of analog nodes where pricing was pretty stable for years, like 10 years ago. I guess one difference now, we did see like-for-like pricing increase quite a bit during the period of shortages. There's some anxiety that maybe this is different this time, and we'll see more like-for-like pricing pressure. I know there's some of that at the margin always, but I'm talking about different than what we've seen in normal cycles. Can you speak to that?
Yeah. No, it's a great question. So firstly, our product lines, like 99% of the product lines that we sell, are proprietary products, and they are sole source from Microchip. So price has never been a weapon we want to use in the upcycle, nor do we want it to be something we cave on in the downcycle. So price is a point at the point of design in that a customer needs to have confidence that they can design in with a proprietary product, sole source from Microchip, and not have any capriciousness with how pricing is applied. The only time when we've had to apply a price adjustment is when the cost input cost increases was so much higher than the normal efficiencies that we were able to do and that we can absorb.
And so we and everybody else in the industry did apply price increases in 2021 and 2022 to that effect. Now, those are structural cost increases, whether they were pricing of what we were buying from the foundries or what equipment that we were buying or people and materials and all of that. So those prices are pretty much where they're at. It's not the place where tomorrow, if we were to magically reduce our prices, wouldn't increase our business in terms of what we did. And we wouldn't go raise the price because we could. That's not, again, the way in which we do business. Price is, however, one element of what is important in new designs. Now, that is the way it has always been, even if I look back pre-2019. At the point of design, you have to look at, do you have the right solutions?
Do you have the right roadmaps? Do you have the right technical support? All the other pieces that a customer uses to decide, what is my cost of ownership with a Microchip solution versus something else's solution? And so in that respect, we have to be competitive at the point of design. Now, after the point of design, the price competition is gone. But we need to continue to work on improving our costs. And if we took a piece of business that perhaps is not as good a margin at the point in time, over the years, the margin will improve as we work on the cost reduction work that needs to go with it. And we fight with the most cost-effective solutions on a new design rather than trying to fight with an older design that may not have the same cost structure.
Great. That's helpful. So since your CFO is presenting in Orlando at another conference, maybe I'll ask some of the financial questions. Starting with inventory, your inventories are up at 185 days, still climbing a little bit this quarter per guidance. How high do you let those go? Is there a risk that you have to take utilization down further to clear those inventories going forward?
Inventory measured in days is a backward-looking indicator. We are kind of at a significant under-shipment to where consumption is at. By definition, as the revenue comes down and the COGS associated with it come down, the days will balloon up. We expect, I think our guidance was in the 225-ish days of inventory, is what exiting March we would be at. What happens in June is a function of what the revenue in June is going to be. We have to think of days of inventory not just in the rearview mirror. We have to be thinking of, where does the business go as it even heads towards consumption? Because at some point here, inventory, that's in the customers, and all that will get to normal. People need to be buying at least at consumption.
So we have to think through the lowest point quarters in terms of where we want to be. We're comfortable with the actions we have taken between shutdown days or shutdown weeks this quarter or next, plus what we've done in underutilization, that we've taken the right actions to start bending the curve on that and keeping enough optionality that as business conditions change, we are able to address the requirements and not have the other flip side of now shortages starting to develop with it.
With more of a hybrid manufacturing model where you have internal fabs and foundry, if we do have a period of extended underutilization, do you have the ability to bring in some of that foundry product and improve your utilizations?
Yeah. So we were doing some of that even during the period of the shortages because we couldn't get investment from foundries on trailing-edge capacity. And so we were qualifying more and more. Somewhere along the way, I think I even had talked about we had licensed one of the process technologies. And that's getting. It's already qualified at this point in time. So yes. And then in the back end, it's easier to do. We've always done more assembly and test in-house during the downcycle. And so we keep a percentage of in-house and outside in the back end as well. But yes, one of the ways in which we're addressing internal utilization in the fabs is by producing more internally, where the option exists to do a combination.
We never want to go all in one direction or all in the other direction, but have a degree of flexibility that allows us to flex in and out over a cycle.
I mean, one of your big competitors in TI is obviously going to lead with domestic manufacturing kind of certainty of supply. Do you find customers mentioning that, wanting you guys are they directing you to build certain products in certain locations, or is that not something that's happening?
So I think the real customer need is not just built in the U.S.. It may be for a defense application, that kind of thing, right? But for the vast majority of our thing, it's really, do you have sufficient resiliency for what might happen under different scenarios? And geopolitics obviously is one people talk about. But besides that, you need resiliency for natural disasters. You can have floods, earthquakes, fires. You can have disruptions of shipment based on pandemics, based on, and I think that's a process that we have well thought out over multiple years on where can we build cost-effective resiliency. So resiliency isn't free. And it's like insurance. And those are the discussions we have with our customers. Now, in today's environment, everybody's fear is about China and Taiwan. But I don't know where five years from now the concern is going to be.
Is it some other part of the world? So you cannot have infinite options. But yes, we want to have resiliency not just for geopolitics, but for natural disasters and a normal way of how we want to have fewer and fewer single points of failure for any kind of eventuality.
Great. And then finally, I'll open up the audience after this. But you mentioned the debt reduction. You've paid off over $7 billion of debt in the last, I think, 22 quarters. And you're kind of approaching this point where you're going to be able to start to return all of the cash. How do you think about that? How do you think about the balance sheet going forward? And do you want to maintain flexibility in case maybe assets do come back to being at an interesting price again?
So with where we are, we don't have any large M&A planned. So with that, when we look at the use of cash, use of free cash flow, largely at this point, the debt is at a level where it's going to be structural debt that we maintain at the $5.5-$6 billion kind of range. So the balance is what we are at this point returning to shareholders. Long term, it'll be half in dividends, half in share buybacks. If there was to be an opportunity at any point in time, we could always adjust the share buyback piece. The dividend piece is always a fixed piece. The share buyback could be adjusted. We don't have anything planned that's going to go there.
And so we're quite comfortable with how structural debt and then the free cash flow being returned to shareholders in the way we have articulated works.
Great. Well, thank you for the overview. Let me open it up to see if we have questions from the audience. All right. Well, if not, we'll wrap it up there. Ganesh, thank you very much for tonight.
Thank you very much, Joe.
Thanks.