Morning. Welcome. Let's get started. I'm Vivek Arya from BofA's Semiconductor Semicap Equipment Research Team. Really delighted to have Eric Bjornholt, the CFO of Microchip, join us this morning for a fireside session. I'll go through my questions, but please feel free to raise your hand if you'd like to bring up anything in between. A very warm welcome to you, Eric. Really appreciate you joining us. Maybe as a start, I know that Microchip recently updated the guidance upright. Thank you for doing that a few days ahead of our conference. I think you mentioned that Steve also gave some incremental details about your inventory write-ups. Maybe just give us a start with state of the union, how you are seeing demand, and then what led to the raise in guidance.
OK. All right. Thanks for hosting me, by the way. Appreciate that. Thanks, everybody, for attending. During the course of this discussion, I'll be making certain forward-looking statements about the future financial performance of Microchip. I refer you to our filings with the SEC that identify important risk factors about the company. We did update our guidance last week. Essentially, we removed kind of the low end of the range. Things are trending well in the quarter. We noted that bookings for the month of May were the highest level of bookings that we've seen in monthly bookings for a couple of years. That's all good. Turns into the quarter have been a little bit stronger than anticipated. Because of that, we updated the guidance around the previous midpoint and the high end of the previous guidance. That's going well.
Clearly, we have been executing on Steve Sanghi's nine-point plan since he returned to the company. Kind of the big focus areas have been inventory reduction. We did some fab consolidation. I think we'll get into some of those questions as we go through this. Incrementally, what Steve shared with the street yesterday is we had not been as specific about breaking out what our inventory write-off charge impacts were on gross margin. That was over $90 million. I think it was $90.4 or $90.6 million in the March quarter. Combine that with our underutilization charges, which were about $54 million. Combine that is almost $145 million in charges that, as the factories ramp, as things improve in the business environment, those charges will be reduced. Now, inventory reserve charges never go to zero.
We have a clear line of sight to getting to our 65% non-GAAP gross margin targets. If you just add those charges that I just went through back into the March quarter non-GAAP gross margins, the underlying product gross margins were almost 67%. That is the good news in the business. Obviously, we have a ways to go to get there. What happens with these inventory reserve charges is it is an accounting charge, right? We are building inventory that we believe will eventually sell. Whether it sells in nine months or four years from now is hard to predict. Our products have very long useful lives, 10, 15, 20-plus years. With that, we are building inventory that we think eventually will sell. Anyway, with that, let's open it up to your questions.
Yeah, absolutely. Thank you, Eric. The rebound the industry is seeing right now, is that because you think the end demand is improving? And/or do you think that inventory has got so low and so depleted that it is a matter of just getting them back to a trend line?
Yeah. Microchip peaked out in revenue at close to $2.3 billion per quarter during the upside cycle. Clearly, there was distribution inventory build, customer inventory build, contract manufacturer inventory build reflected in those numbers. We fell as low as $970.4 million in revenue last quarter. Clearly, on the upside, we were overshipping end consumption. On the downside, we're significantly undershipping. I think this is a matter right now of inventory replenishment that is needing to happen after a couple of years of inventory being drained off. As our distributors and end customers are getting pockets of their inventory that is now getting too low, they're having to place orders. We're seeing that reflected in bookings. This just isn't bookings that they need immediate delivery on. It's in line with our lead time.
We can see our backlog building into the September quarter, where our September quarter backlog is significantly higher than where the June quarter backlog was at the same point in time.
Oh, excellent. OK. Is this recovery being driven by distribution or by direct customers? What does the difference tell us about the state of this recovery?
Yeah. So it's a combination of both those things. Inventory was built up at the direct customers. It was built up at distribution and at our distribution customers. The June quarter is the first quarter in many quarters that we expect our distribution sell-through to increase. That is a reflection of the inventory at the distributor's customers starting to work its way through and improve. They will need to start placing orders onto distributors that will pass it on to us. Distribution has been draining inventory for quite some time for us. I think if you look over the last five quarters, our distribution sell-through, which we get reports from distributors every month that reflect that, compared to what our reported revenue based on sell-in was, it's like a $530 million difference. Last quarter alone was a $103 million difference between distribution sell-through and sell-in.
I expect that during this fiscal year, which will end in March, that that gap is going to converge and get close to zero. Because distribution is starting to place orders at a higher level. We talked about that back in Steve's March 3 update to the street that we were seeing strength in distribution. Now we're also seeing strength from direct customers today. And I think that inventory replenishment is happening across the board with customers.
I see. Any certain applications or geographies that stand out in terms of who is driving this recovery?
It's really broad-based. Obviously, when lead time is pushed out, it took no prisoners. It wasn't just one end market. It's really all end markets, all geographies are seeing these same trends.
Got it. I know you look at industrial and automotive on an annual basis. Is there some relative comparison between the two where you are seeing a better recovery on the industrial side versus the automotive side, et cetera? Where do you think Microchip is feeling more confident about this pace of recovery and where the goodness is still to come?
Yeah. We can't really make a large distinction between industrial and automotive, I think, depending on the customer's appetite for holding higher levels of inventory during the upcycle and making those purchases. Inventory was built across the board. We are seeing recovery across all of our end markets, including industrial and automotive. Industrial is our largest end market. Excluding aerospace and defense, it's about 30% of our business. The A&D portion is about 18%. By far, that's our largest end market.
I see. Eric, if you look at the $103 million gap relative to we can either take total sales, which are in the zip code a billion, right, billion plus, or we can take just the distributor sales, right, that are obviously, I think, about half of that. Is that the right way to quantify how much you are "undershipping" versus where distribution? How do we get a sense for where you are relative to what that trend line should be?
That's a question that everybody would like answered. And quite honestly, we don't know the specific answer to that. But the $103 million difference between distribution sell-in and sell-out only represents what's happening at the distributor. What it doesn't represent is what's happening at their end customer. As I indicated, the sell-through had been falling for many quarters. Now that's starting to turn, where this is the first quarter we think that sell-through is going to increase. I think there's multiple levels of inventory. The $103 million difference in distribution sell-in versus sell-out is just a piece of that.
I see. Do you have any similar metrics or anything representative for your direct customers also? I appreciate that it's harder to get a good sense for where you are. If you were to look at historical trends, I imagine that I don't want to use AI or something that gives you a sense of where inventory should be and where it is right now.
It's really impossible to do. Our direct customers don't give us an inventory report of their holdings of Microchip stock, right? There's obviously conversations. We know that if a customer is buying 100 products from us, maybe 25 of those have bottomed out and they're having to buy those again. Those other 75 will come over time. That's just a hypothetical example. Distribution is easy, right? Because our partnerships with distributors require them to provide us detailed inventory reports at the end of each month, as well as sell-through, so we can see part by part how inventory is trending and make assumptions from there where it's heading.
I see. From this data, is there a way that the investment community can get a sense for we equate undershipment as the potential for above-seasonal quarters, right? That is kind of the simplistic correlation that is made. From this data, is there a way to say that we have x more quarters of above-seasonal opportunity?
We guide one quarter at a time. It's hard to tell. I think there's multiple quarters in front of us. How this thing returns to a more normal level of shipments is hard to tell. Again, with 110,000 or 120,000 customers, it's very difficult for us to know that x% of the customers have corrected or are now back to normal and how many still have a long ways to go. It's a mix.
Right. If we zoom out, Eric, and look at where Microchip is relative to kind of the pre-COVID levels, right? It is still below even that run rate, even though the industry has experienced a lot of pricing gains in between. Do you think the nine-point plan that Steve has proposed, that is enough to kind of get you back on trend? I do not want to revisit too much of the history. What are the two or three things that you think are happening to help Microchip kind of get back to a positive slope versus that baseline?
Yeah. The nine-point plan encompassed a lot of different things, right? A heavy focus early on was what we needed to do to correct inventory. One of the major pieces there was assessing where our customers are at and how they're feeling about Microchip. Obviously, when lead time stretched out, non-cancelable, non-returnable programs were in place, there was some fear by the investment community and even at Microchip that, hey, were customer relationships damaged? We've been through a pretty detailed process in meeting with customers, surveying them, getting feedback. Actually, that survey results, which we shared back on March 3rd, showed that 24% of the customer relationships of those 7,000 or 8,000 customers that we hold, that the relationship had improved because they were serviced well by our PSP program, right?
There was about 12% of customers where the relationships were damaged and we had been de-emphasized by them. Those are the customers that we really focused on. At this point, we have met with the vast majority of those customers, fell on our sword, apologized, made sure that our product roadmap is there for them, asking them how we can support them. There is a very, very small percentage of those customers that have not moved us back to kind of a regular routine with Microchip in terms of engaging with them. That is well. We think our customer relationships are in good shape. Over time, we are very confident in our ability to return to higher levels of revenue, gain market share, and be quite competitive in these competitive markets that we participate in.
Got it. The skeptics would say, not for Microchip specifically, but for the industry, that, look, the industry was not very good at spotting when it was going into overshipment. Why should we believe the industry when they say we are very good at spotting the undershipment, right? What specific kind of checks do you have in the system to give you a sense that what you're shipping is not for some kind of pull-in or trade or tariff-related situations? That this is just a two or three-quarter kind of inventory build, and then things will maybe trend in a different direction.
Yeah. I think I'll start by saying from where we peaked in revenue to where we troughed was very significant. Clearly, I think it's clear to investors that Microchip is significantly undershipping what end consumption is. In the customer discussions, obviously, we've seen a very nice trend of improving bookings. The bookings activity from January through May has been very healthy. Our backlog has been building. We had a positive book-to-bill last quarter. I mentioned that backlog for the September quarter is building nicely. That, and then you add in customer discussions where all of our large customers that have come in, and we've talked to them, or we've met with them in their offices, we ask them the specific question, are you doing anything differently because of tariffs? The vast majority of the responses to that is no, right?
The tariffing situation is uncertain. The vast majority of our products are exempt from tariffs today. That does not mean that a customer could not be building their systems because they are trying to get out in front of it. The feedback we are getting is no. I specifically asked the person that runs our global sales support team, is she getting any feedback from customers that they are doing anything abnormal because of tariffs? The answer has come back as a unanimous no. It is customer discussions. Now, with 110,000 customers, can I say that there is no customer out there that has not done something different because of tariffs? I cannot. I think it is very small if it has happened. I think there are as many customers that have frozen and say, hey, let us wait to see where this settles rather than pulling in.
If we were getting massive amounts of orders because of the tariff situation, you would think that those orders would be placed today and deliver immediately or deliver as fast as you can. That is not what we are seeing with the orders that are coming in.
There is no specific geographic or concentration or application-type concentration to these? These are broad-based, exactly like that.
That's correct.
Got it. I am also curious. You mentioned there was a certain proportion of customers who did not have the best experience, right, during the PSP program. Does this broad-based recovery apply to, are you seeing re-engagement with those customers also, or are they still kind of sitting on the sidelines?
We absolutely are seeing re-engagement. These customer discussions that we've been having have been quite positive. All of our competitors had challenges during that period too where lead time stretched out significantly. Many had these non-cancelable programs. Microchip clearly waited too long to take our foot off the gas pedal on that program. We've capitulated on that, obviously, quarters ago now. Customer relationships are in a good position.
I see. One other explanation that sometimes people have given for PSP is that Microchip was almost kind of required to make that priority call because you do not have access to too much capacity. People will always draw the contrast with you and your other U.S. peers who are building out a lot of capacity. Do you think, first of all, is that a valid explanation for why you were almost kind of, I do not want to use the word forced, but had to make a priority call on PSP because your capacity access was limited? If that is the case, do you think that you are now putting in enough plans to build capacity so you never have to do a PSP-like program again?
I can't see us implementing a program like PSP again. Now, that doesn't mean that we couldn't have non-cancelable orders. The challenge with it was it was really on a first-come, first-served basis. What could happen is you could have customers that jumped on it really quickly and over-ordered and consumed the capacity on a given part and took it away from a customer that hypothetically was buying 50,000 units a month for the last four years, right? They got stuck with a lead time that went from eight weeks to 60 weeks, right? What we are doing is we are using AI and ML to analyze all of our customer data with direct customers, with distributors to look at it.
We will reserve capacity for those customers that are buying on a consistent basis and not allow that capacity to be consumed by others that are speculating and spot-buying on things that are not true in demand. We are not in that situation today. We have a massive amount of inventory. Inventory, we are projecting it is going to come down by $350 million plus this year as we get inventory days more in line with our model. We are building systems so we can better react to the situation again. I have been at Microchip for almost 30 years, and I had never seen lead time stretch out. They did in the last cycle. I hope in my career I never see it again. We need to be prepared to deal with it better than we did in the last upcycle.
I see. Makes sense. On the pricing issue, Eric, we have heard a range of views, right, from you and your peers, which is some saying pricing this year is down low single. Some are saying mid-single. I think you guys have said mid-single, right, digit, right, or so. Would you call this kind of return back to normalcy, or is this customers trying to take advantage of all the excess inventory in the system and extracting bigger? Again, is this kind of a return to normal, or do you think this is being done in a very defensive way? At what point does there is no kind of elasticity, right, in the system? Even if you cut prices, the demand is not there. It's just not. How would you describe the pricing environment?
Yeah. So we've been saying for the better part of a year now that we have been being quite aggressive at the point of design, right? So pricing on product that is designed in shipping today, that is very stable for us. We want to make sure that we are leading with customers with our newest, most cost-effective products and doing that in a way as the cost structure for those products improve over time. We're still driving reasonable margins for the company. You saw that back in March, we updated our long-term business model. We took our gross margin target down from 68% down to 65%. There is a number of things that factor into that.
One is, I think Steve Sanghi wanted to set a model that we are very confident in achieving, that we can bounce around and do better than when things are great in the industry, when utilization is really high, and then fall below it, but not by much during a down cycle. We have to do a better job of managing inventory levels to do that so we do not have to cut capacity as much as we did at the bottom of this cycle. That is one thing. Pricing does factor into it to some degree. We think that probably we are mid-single-digit decline in this fiscal year, and then it kind of returns to more normal, very small declines because these are proprietary products you get designed in, and you can hold that price for the life of the product line.
There are certain cases where we are leading for a new design. To win that design, it might be with an existing customer that you need to give a little bit of a concession on your existing business. We are willing to do that if the economics of the overall deal make sense.
I guess that's where I want to draw the distinction: is this pricing concession being given to products that they are already buying, or is this pricing concession being given as a way to attract them to the next generation of the product?
Yeah. It's typically the latter. As I say, you can have a situation in winning a design that you have to give a little bit on something that's already shipping. That's the exception to the process.
I see. How would you contrast this to, let's say, whichever was the normal period pre-COVID? Is this very different? Is it similar?
I think it is a little bit different than it was pre-COVID. We were not in a significant inventory and excess capacity position then. Obviously, our competition has excess capacity too. They are being aggressive. We need to match our competition in terms of how they are coming to the market because we want to come out of this as a share gainer. In normal conditions, I would expect pricing to be very stable for our products.
I see. This is not being done because your end customers might be getting pressured by tariffs or other situations, right? It's not because of that. This is normal industry competition.
It's normal for kind of where we're at in terms of the inventory, where lead times are, where capacity is in the industry today.
Got it. Makes sense. Now, during the downturn, Microchip had to go through a lot of headcount reductions and so forth. Where are you in that process? Because now OPEX is starting to grow. Do you think that OPEX reduction also impacted customer relationships in some way?
Okay. We had a reduction in force last quarter. We announced that in early March. All that is implemented at this point in time and is reflected in our OPEX guidance for the current quarter. OPEX this quarter is down from where it was in the previous quarter. I think investors should really look at the about $356 million that we are guiding to in OPEX to be kind of the low point. That has no variable compensation in it. We do have employees back to full salary. That was effective for everybody. That is reflected in the current quarter numbers. We need to bring back the bonus programs over time as the P&L can afford it. We will be cautious with that. We are very focused on getting to our long-term model on OPEX, which is 25% of revenue.
We're a long ways away from that today. We know that as revenue grows, OPEX as a percentage of sales needs to come down pretty dramatically.
I see. So this $356 is what, almost like $100 million-ish, right, off the peak?
Yep.
Is it possible that you get to that prior peak OPEX level even if sales are not at the prior peak?
It is not. It is not. We will be very cautious in how we add back OPEX. At the peak of the cycle that had $100 million more of OPEX, we were paying bonuses at 200%, 250%, I think 300% of target in a given quarter. That does not need to come back. That was the result of the extreme upcycle that we went through. Obviously, the financial results that we were posting at that time of 48% operating margins, which is well above our target.
Got it. On the gross margin, the drivers that you mentioned, is there a way to simplify the analysis and say that if sales are, I do not know, $1.3 billion, $1.4 billion-ish, then gross margins can be 60%? Is there a way to do a simple methodology like that? It is hard for us to track what is coming in and out of inventory and then try to correlate that, right, to gross margins.
Yeah. I talked about this a little bit in kind of the introductory that I did. The two biggest things that are impacting gross margin today are our inventory reserve charges, which is an accounting charge for write-offs. We have these underutilization charges. Last quarter, over $90 million was written off in terms of these inventory obsolescence charges. That is the first thing that's going to go away. We look at that. We take a snapshot of inventory at the end of a quarter. We look at 12 months of trailing demand on a revenue basis and then turn that and multiply it by one and a half. If inventory is over that in any point in the chain, it's a write-off candidate. Those write-offs have been extremely high. What's going to happen here is obviously inventory is coming down, right?
I said $350 million plus this fiscal year. There is going to be a lower amount of inventory that is subject to a potential write-off. We are going to turn the corner where trailing 12-month revenue starts to increase. As that happens, and we are way underproducing what we are shipping today, as those things all come together, that $90 million quarterly charge drops dramatically. It never goes to zero, but it will get much lower. Could it be $15 million-$20 million in a quarter? That is significant, right? That will happen. Our prediction is that will happen during this fiscal year. That is going to start being a very positive tailwind to gross margin. We also will start to need to increase activities in our factories.
That could happen as early as the month of September that we start to have to increase output in the factories. That would have a very modest impact on September quarter results, but could have more materially impact the December and the March quarters as we have to start adding people, running more wafers and activities through our factories. The other thing that is a tailwind to gross margin is we've written off huge amounts of inventory over the last five or six quarters. Again, this product will likely eventually sell. When it does, it will sell at a much higher gross margin. In many cases, that inventory has been written off when it's gone through the wafer fab. It is sitting in dye bank.
We get an order in, we'd have to add the assembly and test costs to get it to a finished good. It is still selling at a much higher than corporate average margin.
Got it. Makes sense. Please, Jeffrey.
I think I'll be further to put, as you just walked through that, is your unit inventory significantly different than the dollar inventory? And does that matter?
I think your question might be getting at the point that the cost per unit of inventory that we're producing today is higher, right? We do have higher cost inventory that's sitting in inventory that we'll need to sell through before we get the benefits of increasing activities in the factory. The period charges for underutilization charges will come down as we start increasing activities and then these inventory reserve charges. With taking out Fab 2 through a restructuring, our overall costs are coming down, our fixed costs associated with manufacturing. I think we can grow back into our capacity very cost-effectively. You're right, there is a time delay in your cost of inventory that's being sold to customers representing itself in the P&L. That's a good question.
Okay. Thank you, Jeffery. Is there a simple way, Eric, to think about what are the implications of this whole Section 232 that many investors are concerned about, as in if you are getting semiconductor inputs from outside, right, they will be exposed to more tariff? Does that give a competitive advantage to some of your peers who are manufacturing more in the U.S.?
Yeah. So we've looked at it. We've gone through every product that we produce and have made this information available to our customers in terms of where it's fabbed, where it's assembled, where it's tested. There's about 8% of our product that is manufactured in the U.S. that ends up shipping to China. There's another 5% or so that are manufactured not in the U.S. or China and then come back to the U.S.. There's a very small percentage, I think it's like 0.3%, that is manufactured in China and comes to the U.S.. You add all that together, and it's roughly 14% of the revenue that is subject to kind of U.S.-China tariffs. Where all that falls out, I don't know.
Now, that does not consider is there going to be tariffs on U.S. product going to the EU or EU coming to the U.S.. We are going to have to see where all these things flush out. We have some flexibility, right? There are some of our products that are manufactured in our fabs that we could move to a foundry in Taiwan. Those could be shipped into China as an example. Today, there is no tariff on Taiwan into China. We have certain of our foundry partners that have facilities in the U.S. where the product is manufactured today, and they could move that. We have some flexibility in our supply chain.
I think you could look at just depending on where tariffs fall, where your manufacturing footprint is going to be impacted, and then you'll have to make decisions on does it have an impact or not. I think the impact of tariffs directly on a company like Microchip is going to be quite small, but it ends up being what is going to be the indirect impact on our end customers and the products that they are selling and how that impacts them. That's the kind of the larger thing that is not understood at this point in time.
Got it. One or two quick ones, Eric. One is, as we are going through the cyclical rebound, does seasonality mean anything, or can you continue to have multiple strong quarters eventually?
Yeah. I do not think at this point in time seasonality has any impact, right? I mean, December quarter is typically our weakest quarter of the year, and the holidays are still going to happen. If inventory replenishment is happening, that could override that. Again, we are not giving guidance at this point that far out in time. I think this inventory replenishment cycle has multiple quarters of legs behind it.
Got it. The final one is, if, let's say, you express some confidence in kind of exiting this fiscal year at a 6-plus handle in gross margins, do you think that can somehow correlate to the ability to generate enough cash flow that you kind of get back to stock buybacks?
On capital returns, obviously, we are very committed to the dividend at the level that it is at, which is $0.455 per quarter. We did a financing transaction last quarter where we issued a mandatory convertible preferred, which is essentially an equity instrument. Common stock will be issued for that in three years when it matures. We did that to secure our investment-grade rating and make sure that our dividend is absolutely secure. Now, from a buyback perspective, we had set a net leverage target way back at our analyst day in November of 2021 of one and a half times. We are not going to be at one and a half times for an extended period of time. We have had a few quarters where we have had to borrow to pay the dividend. We get a couple more quarters out in time.
That's not going to be an issue for us. The board will have a decision to make once we get back to kind of paying our debt down for the borrowing that we've done to pay the dividend, which is about $300 million. Once we get to that point, what's the strategy from there? Obviously, the dividend is sacred and won't be cut. At that point, would they say, "Hey, you should take debt down further," or, "You should try to get back to that one and a half times levered or some other number"? Do they set kind of a total debt level that they're comfortable with? When do you start buyback again? I think buyback is further out in time is my impression.
More than the fiscal year?
It's out in time, right? I mean, we're fully committed to the dividend at the levels that it's at. We know we need to reduce leverage. That's the focus. It's not a decision that we need to make in the short term.
Right. Makes sense.
At that time, we'll take feedback from our shareholders and see what they would prefer.
Makes sense. Terrific. Thank you so much, Eric. Really appreciate your insights. Thanks, everyone, for joining us.