Welcome back to this session. I'm really delighted to have Eric Bjornholt, CFO of Microchip Technology, join us. I'm Vivek Arya from BofA's Semiconductor team. What I plan to do is go through my questions, but please feel free to raise your hand if you have a question, and we'll be sure to get a mic to you. But very warm welcome to you, Eric. Really appreciate you joining us.
Thank you. Thanks for having me.
We'll go through the business, but I just wanted to kind of touch on the most recent, right, thing with Microchip. You recently issued a convertible, and I was hoping you could describe that, you know, what is the rationale behind that? And I think you had a covered call as part of that, you know, versus a standard bond. So maybe just give us some more color around that.
Sure. So maybe before we start, I'm gonna do the standard disclosure that during the course of this discussion, I'll be making certain forward-looking statements about the financial performance of Microchip, and I refer you to our filings with the SEC that identify important risk factors about the company. So we were in the market last week with a convertible debt transaction. It's a 6-year instrument. It has a couple of relatively unique features to it. We purchased a capped call up 75%, and that cost us about $104 million to do that. This is a $1.25 billion convert.
The other unique feature of this convert is it has an investor put feature, that the instrument can be put back to us, and really, that at a, at one day in time, 3 years out, June first, 2027, that gave us a lower coupon, but does give the investors the ability to put it back to us. And really, the only reason they would do that is if the stock price is kind of around the levels that it's at today or lower, which obviously the hope is that, that is not gonna happen. But if it does, you know, we have cheap money for 3 years, and then we would go and reinvest that. So it, we think it's gonna be a good transaction for us.
We think the cost of capital over the six-year period would be about 2.25% if the stock does not exceed that 75% up, and it would be cheaper than investment-grade bond issuance over the same period, which would probably be at about a 5.5% rate, all the way up to a stock price of about $194. So anyway, that's the transaction we executed last week.
Got it. At a high level, Eric, what does it mean to kind of the P&L math for this year?
So, I mean, it has a positive impact on interest expense-
Right.
in the short term. Now, we do have a couple of maturities coming up that are at lower rates. So we have a investment-grade bond that is coming due in September at about a 1% interest rate. So you know, the, the coupon on that is, is similar to where we are at the convert, but it was not a convertible transaction, right? The, the interest rate environment three years ago was very different than it was today. And then we've got a $665 million convert that matures in mid-November, and, you know, essentially, we're using these proceeds to be able to address those. In the short term, the proceeds, we've, we've paid down our commercial paper balances, and that has about a 5.6% interest rate, replacing that with the 0.75%.
So clearly, a cash cost savings there for us, and then we have the ability to use our line of credit or issue additional commercial paper to take out these maturities that come up later in the year.
Makes sense. Let's go to the business. Maybe you could just start with the state of the union, right? Every company is going through its own timeline of inventory, you know, adjustments. Where does Microchip sit today? And if you could maybe walk us through whether it's by product type, whether it's by end market type or geography, whichever way you think is the right way to kind of segment your business, where you are in that inventory adjustment process.
Yeah. So, you know, clearly, we're in a pretty significant inventory adjustment phase at this point in time. You know, our last quarter was down 25%. Sequentially, we're guiding the current quarter down another 6.5%, and if you look at our business year-over-year, midpoint of June guidance to June of last year, it's down almost 46%. So we are clearly shipping way below consumption today. Now, last June, we were obviously feeding inventory into the channel unknowingly at the time, but, you know, customers had gone through a significant supply-demand correction and were fearful on getting product, and were ordering product out in time that ultimately they did not need at the run rates that they needed. So distribution's correcting inventory.
We made a comment related to our March quarter results that distribution drained about $125 million of inventory last quarter, and that is on what I perceive to be depressed level of sell-through because distribution customers are also draining inventory. Our direct business in the March quarter was down 30% sequentially, and, you know, clearly, we think that they are going through an inventory correction also. So yeah, we, we're gonna have two back-to-back quarters that are down pretty significantly, and, you know, we know that the business will bounce back to much higher levels, but it's a matter of how each customer has managed the cycle and their inventory position in terms of when that correction comes. Now, we are seeing what, our CEO, Ganesh, calls green shoots in the business, right?
We are seeing bookings improve on a monthly basis, and I'll just walk through that. February was the highest month of bookings that we had had in 8 months. March was the highest level of bookings that we had in all of fiscal 2024, which ended in March. April was higher than that, and then May, which just completed, was higher again than April. Now, it was higher by a small amount, so it's not like bookings are going through the roof, but it is improving. Bookings are still well below where we need them to be to get back to kind of a normalized revenue run rate as customers work through the inventory situation. The lead times are short, and with that, we are seeing a change in behavior of customers. We are seeing much fewer requests for push out and cancellation.
We are seeing more on the other side of that, where customers might have an order that is scheduled out in time in August, and now they need it in July. So we'd call that a pull-in. It's not a new booking, it's an order that's already outstanding, but they're pulling it in for an earlier delivery date. And then the bookings that are coming in are coming in more in line with where our lead times are, which are really short today. You know, the vast majority of the products have lead times that are eight weeks or less.
Got it. Given, you know, Eric, you have been through a few of these cycles right before. Obviously, cycles are different every time, but do you think the recovery, the shape of the recovery this time around is, you know, is it a V-shaped recovery? Is it something that, you know, you need to have a couple of flat quarters before you start to see the recovery on the other side? Like, when will you know that now you can trust that the recovery is happening?
So, you know, you kind of have to make assessments based on what we've seen in other cycles, and these green shoots, as we refer to them, are kind of the early signs that the recovery is coming.
Mm-hmm.
And, you know, they're-- You know, we still need turns to have the September quarter be up from the June quarter. We have significant turns that we would need to take, right? So there's no guarantees on this, but these are classic signs that customers are starting to get through their inventory challenges, and with 125,000 customers that we service, that all doesn't happen at once. But, you know, to see the signs that that's starting to happen for a portion of the customer base, and it's not geographic specific, it's not end-market specific.
Mm-hmm.
We can have a very similar customer in industrial-
Right
that one of them is pulling product in, and the other is saying: "You know what? I don't need to book anything for the next three months 'cause I've got inventory." So each customer, as I said, handled the supply and demand environment and what they did from an inventory perspective differently during the last upcycle.
Got it. On the industrial side, I think what I remember from the last call was that, look, certain customers are getting better, certain customers are worse. Help us understand, what are the big moving pieces of your industrial business, right? Is it more exposed to aerospace or defense or factory automation or test and measurement? Like, what are those big moving pieces, and are you seeing a distinction between what is getting better versus what is not improved yet?
Yeah, so to address the second piece of that first, it is very customer specific, and we can't say that just because somebody is in test and measurement, that they're through the inventory correction, or if they're in safety and security, they still have inventory that they're working through. So our industrial business is very broad. It's about 43% of our overall revenue. The strongest piece of that over the last year has been our aerospace and defense business, and that's 11% of overall revenue. So it's about a quarter of our industrial business in aerospace and defense, and that continues to perform very well.
Outside of that, and really the AI piece of data center, which I think you have a question on later, you know, those are the stronger areas of the market, and everything else is going through this pretty significant inventory correction.
All right.
So the industrial business is our broadest business. You know, you mentioned a few things, so A and D, we talked about testing and measurement, safety and security, factory automation, smart cities, all these things as part of that, as well as medical.
I see. And outside of aerospace and defense, you think that we should be prepared for, you know, more downside in the next couple of quarters? Do you think that we are getting to a flatlining level in those other non-aerospace, non-defense markets within industrial?
Yeah, I mean, the visibility is low right now, but I'd say we are seeing the formation of a bottom, for sure. You know, these signs that we're seeing with these green shoots tell us that the bottom is forming.
Okay. On the automotive side, what is Microchip's positioning on the automotive side? And I am asking these questions because Microchip has never described itself as a play on, you know, one thing or another, right? As a broadly diversified company. Many of your peers describe themselves as a play on one thing or another. So what are Microchip's kind of key focus areas in a car, and how is what's happening in the automotive market, right? Flat production this year, EVs may be growing. How does that inform us about, you know, how you can do in automotive?
Yeah, so, so automotive in total is about 18% of Microchip's business. I think that increased by 1% from fiscal 2023 to fiscal 2024. We have a very broad set of products that go into the automotive market, and, you know, we show a couple of different sets of pictures to investors, but, you know, they're. It's not uncommon for a high-end automobile to have 80 of our parts doing various different functions within an automobile. It could be something very simple, like a, you know, power window, door lock, seat, it could be lighting, it could be the infotainment system in the vehicle, it could be a touch screen, it could be USB, could be camera system. So it's very broad-based. There's not one thing that is driving the growth.
Now, we have the mega trends that we're focused on, and two of them are automotive specific, and one, that's ADAS and EV. And, you know, EV is obviously had a few short-term challenges, but, you know, longer term, we think that's an area that there's gonna continue to be growth and an area that we're focused on.
Mm-hmm. What's your average content in an EV versus a traditional car?
So, I mean, a lot of the same functions go into an EV versus a combustion vehicle. You know, generally, we think that the opportunity is probably about twice the size in an EV from a dollar content perspective. And EV, this is in the mega trends that we break out, is about 3% of our revenue. So of the 18% that's automotive-
Six, right.
3% of the total revenue is in EV today.
Got it. In line with how EVs are as a percentage-
Exactly.
of the automotive
Yeah
market. In the automotive market, is that a market you still think that's, you know, that there is a sense that maybe it's still the last shoe to drop in terms of, you know, whether it's excess inventory at Tier Ones or at OEMs? Do you sense that also, or you think you have a much better handle on the automotive business?
I don't think we necessarily have a better handle on it. I think that customers across the end markets, including automotive, have inventory and are working through that. You know, one thing that we thought would change based on this last crazy upcycle that we went through and the supply and demand shortages is that customers who were producing high-value end equipment, and could be automotive, could be data center, could be medical, whatever it could be, would learn their lesson that, you know, semiconductors is a complex supply chain. These aren't commodities. There's lead time involved.
And we still hope some of that sticks, but, you know, it seems like these lessons are kind of short-lived, and the purchasing managers are back to, you know, really managing working capital, taking inventory down, and again, I hope we don't find ourselves in a situation like we did last time, where automotive stopped buying during COVID, and then it set off one of these, you know, very huge supply-demand imbalances that we've had, the largest that I've seen in my career at Microchip, and that was very painful for our customers and obviously very difficult for a company like Microchip to respond to.
Right. Based on this, bookings, right, improvement, I know you said it's not or the, the normal level, right, May better than April, better than, you know, March and Feb and so forth. Is automotive recovering faster, or is industrial recovering faster in, in your bookings?
You know, this sounds a little funny, but I really don't have a way to break that out, right? We do not track our business by end market on a quarterly basis. We summarize it for investors once a year, and, you know, we just don't have reporting within the company that tells us to that level of detail, what's happening. You know, but it is true that there is more of the business that is direct in automotive than in some of the other end markets, but, you know, general feedback is, again, that, automotive is still correcting inventory and, you can just see the difference that's happened in terms of the vehicles that are on the lots today versus what it was 18 months ago.
I think the Tier Ones are sitting on quite a bit of inventory and, you know, I think some of that needs to get rationalized still.
I see. And, you know, you mentioned last quarter, direct sales, you know, went down more than distribution, right, sales. Conceptually, why wouldn't Microchip have better visibility into your direct customers, right, versus distribution? Or you think that direct customers just don't provide you with that level of feedback?
So, you know, we do not get any sort of inventory reporting from our direct customers. At least distribution, at least once a month, provides us an inventory report.
Right.
They provide us a sell-through report. We can see if inventory is going up or down. We don't get that from our direct customers. Now, there are interactions and discussions that are had-
Right
On a, you know, Microchip salesperson to the direct customer, but you do not get any sort of reporting that shows you that inventory has gone up or down.
I see,
So it's less clear. Distribution is easy for us to quantify.
Right. Over time, do you think that Microchip will have... So how much is distribution right now? Is it what? About 60%?
Distribution is 47%.
Forty-seven percent.
Yeah.
So you think over time, does it get to be a bigger part of your sales, or you think direct distribution remains... Does distribution become a smaller part of your sales, or you think it stays at this level?
No, I think it stays in this general range. I mean, we're pretty agnostic in terms of, you know, how the customer chooses to go to market. You know, they can come to our website that we call Microchip Direct and order product and have relatively little interaction with a salesperson from either distributor or Microchip. They can pick and choose the type of support-
Right
That they want. Tends to be that the larger customers are direct customers, but, you know, some of those might choose to purchase through distribution because distribution is providing a service that they value, that might be kitting, it might be programming, it might be logistics, it might be payment terms, right? So, you know, definitely distribution has a role, and we have multiple types of distributors that we work with. We have catalog distributors that tend to see the market. That could be like the Digi-Keys and the Mousers of the world. We have the global distributors that you're all familiar with, the Arrows and the Avnets, and then we have regional distributors that tend to be a little bit better at demand creation activities. They're fully trained on our product set.
They tend to only carry, as an example, one microcontroller line, and they can go out and be proponents for our products in selling them to, to our customers and really showing them the benefits of what Microchip brings to the table. And if a distributor is involved in a demand creation activity, we are willing to pay them a higher margin.
Got it. On the data center side, Eric, it's you know, we don't think of Microchip as a data center, right, company-
Yeah
But it's almost as big as your automotive business.
Yeah.
So maybe help us unpack what is in your data center. You know, what do you see as the big growth drivers for that?
So, you know, there's probably a point that I should make here 'cause Ganesh made it at the Stifel conference earlier today, is that with this roll-up that we did of our end market data that we posted on the website today for fiscal 2024, that the data center business was relatively flat year-over-year in terms of percentage of revenue. It's about 18%.
Eighteen percent.
But of that business, about 5%, just under 5% of revenue is AI related. So that was new information that we shared with the street today, and so it's a little bit less than a third of our data center business that is focused on the AI portion of the market. Now, we provide a lot of different products to data center. Your data center-specific business unit came to us through the Microsemi acquisition, and they do things like PCIe switches.
Mm-hmm.
They do SSD controllers, as an example. But also that goes into the AI portion of data center. You know, you're going to have things like microcontrollers. There can be FPGA opportunities, there's timing opportunities, there's clocks, there's power management, all these things. And Microchip tries to position those things around an NVIDIA GPU or an AMD product, in a reference design. And our business grew very well last year in AI. So the AI portion of that business grew significantly last year. Data center overall was down. You know, CapEx spending and data center was down, and that impacted our business, but the AI piece of it was up nicely.
Got it. So that is correlated to what we are seeing in the accelerator.
That's right.
Okay. On the CapEx side, you know, your large competitor is building out a really large fab footprint, right in the U.S. How do you think that changes the competitive dynamics over time?
So it definitely doesn't change anything that we are doing. You know, you'd have to ask them in terms of their strategic rationale behind that, and I'm sure there is rationale behind it. But, you know, we essentially have a different model. We do a little less than 40% of our wafer fab internally, and the rest we were reliant on third-party foundries for us, TSMC, UMC, GlobalFoundries, and others. And, you know, we are comfortable with that model. We do not have a 12-inch factory. We evaluated during COVID, during kind of discussion of CHIPS Act, would it make sense to make that investment? And it didn't for us, from a volume perspective, we just couldn't make it cost-effective.
And then any process technology that is 90 nanometer or below, we use the foundries for, and we're comfortable with the model, we're comfortable that the foundries are investing appropriately for us. For the capacity that we own, that's roughly 40% of the business, we have made significant investments. We ramped significantly to be able to drive us to $2.3 billion revenue in a quarter. You know, that was what June of 2023 was, and equipment lead times were very long at that point in time, so we've actually received quite a bit of equipment. I think it's about $350 million that is sitting, not depreciating, not being used for manufacturing today. That is paid for and ready to go when the next upcycle comes.
I think our capital intensity over the next few years will be quite low, and that's not saying that we don't expect good things from the market, but we've got the capacity in place, and we're comfortable that we've got what we need.
Got it. So you don't think having a lot of US-based capacity can give you a competitive edge?
You know, I think for some customers, that might be important, but it's not necessarily if it is US, it is if it's outside of China, or maybe for some customer, if it's out of China and Taiwan. And so our foundry partners are working on investing outside of those areas, to help drive the growth that our business needs over time, and we're comfortable that they're committed to make those investments.
On pricing, is the thinking that pricing is relatively stable, right this year? So does it mean that there are products where pricing is still going up and products where pricing is still going down? What are the kind of the puts and takes? Because, you know, stable sounds like such a clean and efficient, right, assumption, and I'm sure there is more to it.
Well, there's a little bit more to it than that, but in general, you know, once we are designed into an application, that product can hold that average selling price or that selling price for that particular customer in that socket.
Mm-hmm
Through its life. You know, it doesn't mean that purchasing managers might not come and ask for it, but pricing is always competitive at the point of design. And you win based not only on specs, but you have to have the right price that makes it cost-effective for the customer. But once you're designed in, that product can hold that pricing for its life. We've taken this tack with customers for the last 10 years, and we've been very successful with it, and customers are accepting of it. And again, they choose Microchip for a reason, because we provide them the right specs, we provide them a cost advantage, we speed their time to market. Whatever the reason they might choose-
Right
To design with Microchip, that holds through the life of the product. But we do have to be competitive on price at the point of design, and we've always done that, and we'll continue to do that.
But, Eric, historically, you know, we have had that same situation. Still, industry pricing came down before, so why can't the same thing happen again?
So we don't run our business based on what the industry is doing, right? We run it based on what makes sense for us. Obviously, we want to support our customers appropriately. Like I said, we made this change about a decade ago-
Mm-hmm
And have been very firm on that. And the only time that we've had any significant movement in prices over that time frame is either an acquisition that changed something or, when we had the high inflationary costs that happened over the last few years, we were passing those costs on to our customers.
Got it. On gross margins, so about 60% or so in June, they are, you know, 8+ points away from the high. Can we assume 60 is kind of the line in the sand that, you know, Microchip should hold this level of gross margins? Or if not, you know, what are the kind of the puts and takes around whether it's mix, whether it's, you know, further factory utilization actions that you might take?
Yeah, I believe around 60% is the floor for us. You know, that's the guide for this quarter. There hasn't been much of a change in what we're doing from a factory perspective, quarter-over-quarter. In the March quarter, I think it was 60.3%. Yeah, we're comfortable at this level. You know, we're taking underutilization charges in the P&L today, where we have two-week shutdowns in all of our wafer fabs. We're taking days off in our assembly and test factories. All that is factored in to the gross margin. Utilization is significantly lower than what it was at the peak. So you know, the costs that are being capitalized to inventory today are higher, so our cost per unit is higher sitting in inventory than what it was a year ago.
Mm-hmm.
But, you know, we're taking the underutilization charges. We've taken some pretty significant inventory reserve charges based on our accounting policies and, you know, as the business improves, you know, those things will go away, and we're very confident in our long-term model of 68% non-GAAP gross margins.
How much is underutilization as part of the 8-point delta from the peak that you were?
So it was $32 million last quarter, and I'm not expecting it to be materially different than that in the June quarter.
Got it. And then on the OpEx side, you know, you have taken some pretty drastic actions, right, to control, to control spending. So as the market normalizes, should we be prepared for some sudden jump in OpEx as a catch-up? You know, because I think your quarterly OpEx is almost $100 million off, right, over the last several quarters. So should we be expecting a quarter where, as soon as you get to, let's say, seasonal sequential growth, right, whenever that is, that we suddenly see a $100 million jump in OpEx also in that same time?
So, no, that's not what you should expect. So, June quarter of last year was our peak revenue, and it was our peak OpEx. The OpEx was about $465 million on a non-GAAP basis. This quarter's guide is somewhere around $353-$354 million, so over $100 million of OpEx reduction. So that's come from a couple of different areas. One is just general expense control from the employee population, but the two biggest pieces are we have quarterly bonus programs, and those bonuses we're paying out at roughly 300% of target last year at this time. Those bonuses are zero today, so that's a big change. We also have the entire employee population on a 10% pay cut.
Executives are on a 20% pay cut, and, you know, you should view that as temporary, right? We can't have our employees on a pay cut forever, but we will start to bring those back as the P&L can afford it, as revenue grows. And then, you know, we would expect OpEx as a percentage of sales to come down, the dollars to come back. But that full amount of $110 million that we have taken out, you should not expect that to come back in the short term because bonuses were at exceptionally high levels-
Right.
-when revenue and profitability was so high.
Got it. On free cash flow, margins, so they have stayed pretty strong, right? 30%, which is remarkable given how much the business has fallen off. What are your thoughts about free cash flow returns as the business starts to get to, to normal? I know there is a very specific formula, right, that, that you're following. So when do we get to the point at which, you know, dividends start to grow at a more kind of normalized pace, and you're able to devote a lot more to whether it's buybacks or, or M&A?
Okay. So, so as you said, we've got a pretty specific laid-out program of how we're moving from where we are today to 100% adjusted free cash flow return. And so the plan is by March of 2025, so just a few quarters from now, we'll be at 100% free cash flow return. We're at 87.5% this quarter, and that's 87.5% of last quarter's adjusted free cash flow. Next quarter, we'll move to 92.5, the quarter after that, 97.5, and then we'll be at 100 come March. Now, our, our desire, or the board's desire over time, is for that to be roughly a 50/50 split between dividend and buyback, and that's never gonna be kind of perfect math that happens. You know, we're, we're never gonna cut the dividend.
The board definitely has decreased the amount of the quarterly increases in dividend. You know, all of last year, the buyback was higher than the dividend, and that is not the case in this quarter as an example. So we need, we need revenue and profits to return to a higher level to generate the cash flow to do that, but the board is fully committed to this program. I think a piece of your question was kind of M&A related.
Right.
You know, we have no plans to do large-scale M&A. We did a couple of small tuck-in acquisitions in April, and we reduced the amount of the buyback that we did this quarter for the dollar amounts that were paid for those acquisitions. But again, they're small, and those are really kind of IP, R&D team-type acquisitions to help accelerate a particular business unit in a certain area. But large-scale M&A is challenging. First, we don't have any sort of large hole in the portfolio like we had historically when we were doing acquisitions that we think we need to fill through doing an M&A transaction. Secondly, you know, when we did our past two large acquisitions of Atmel and Microsemi, from the signing of the definitive agreement to close was about 90 days.
Today, you know, you'd be lucky to get a deal done in a year, maybe two years. If you take two years to do a transaction, the company that you're acquiring is totally different. You know, there's just a lot of change, there's a lot of churn, a lot of uncertainty, a lot of turnover. So it's just not something that we are actively pursuing today, and we're comfortable with the assets that we have, the investments that we're making, and you should expect that there'll be more of these kind of small tuck-in acquisitions that don't move the needle financially, but position us well for growth.
That's terrific. Thank you so much, Eric.
Okay.
Really appreciate your time.
Thank you.
Thanks, everyone.