Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Greetings, and thank you for attending JPMorgan's 48th Annual Technology, Media and Communications conference. My name is Harlan Sur, I'm the semiconductor and semiconductor capital equipment analyst for the firm. Very pleased to have the team from Microchip here today, Steve Sanghi, Chief Executive Officer; and Eric Bjornholt, Chief Financial Officer. The Microchip team, as most of you know, is one of the few teams that actually provides the market with a nice snapshot of business trends a few times each quarter, more so in times of uncertainty. And we certainly appreciate understanding how the demand profile is changing, especially in these important times. The team did report March quarter results last week. It was a busy week, so some of you may have missed it. So what I've asked Steve to do is to start us off with some summary of the March quarter results and June quarter outlook, and then we'll go ahead and kick off the Q&A. If you have any questions, there is a Q&A button on the screen. So with that, gentlemen, thank you for joining us. And Steve, let me go ahead and turn it over to you.

Thank you, Harlan, and welcome, everyone. This is a unique event. Now my first virtual conference, investor conference like this. And hopefully, you all get something good out of that. So let me review our March quarter results. Our March quarter turned out to be quite good. You may recall that we originally, at the time of the earnings call, we guided the March quarter to be up 5.5% at the midpoint. Then when the COVID-19 situation hit, then on March 3, we revised that guidance down to 0 to be flat. And we ended up, up 3%. So compared to that, I think the results were phenomenal, and the results were also very good compared to many of our competitors that had a significantly down quarter. Now some of the difference between guiding 0 in early March and then ending up at 3%, had to do with both demand and supply. We did a little better on the supply side, and we're able to work around some of the factory closures where we had a number of our production labor employees in Thailand, not sorry, Thailand -- in Philippines, actually working in the plant and living in the plant day and night because they couldn't go out and come back in, and we were able to produce more than we had originally thought. And the demand was also strong, largely driven by a lot of concerns on the part of the customers regarding supply disruptions and all that. So we ended up, up 3%. Gross margin was within a smidgen upper all-time high, 62% gross margin, non-GAAP was extremely good, and operating margin up 36.6% was also very, very respectable. March quarter, book-to-bill was 1.17, uncharacteristically strong. I would say, in the starting backlog on April 1 was about 9% higher than starting backlog on January 1. Now don't misunderstand that. It was never going to lead to a 9% growth because of the uncertainty and longer lead times, people placed the orders earlier. Ordinarily, they would have placed those orders in April and May, they placed those orders in March, resulting into a stronger book-to-bill and stronger starting backlog. Now some of that backlog has been getting pushed out and canceled primarily from the automotive market, but some other places also as the factories closed down and the people no longer needed that product. So putting all that into picture at the earnings call, we said that backlog, which began up 9%, was now up less. It was still positive. We didn't quantify the number, but some of the strength had gone away. And the bookings had slowed down, we were getting push-outs and cancellations, which have continued. Based on all that, we guided June quarter to be down 2% to 10% with a midpoint of minus 6%. So with that, I think I'll pass it on back to Harlan.

Thanks for the summary, Steve. So despite the June quarter backlog still being up quarter-over-quarter, you talked about push-outs and cancellations, which continue to drive deterioration of the backlog. But at the same time, you also have supply constraints given some of the movement controls and restrictions in your Philippines and Malaysia assembly and test operations. Given your extrapolation of current demand trends for the rest of the quarter and your projected ramp of assembly and test, how much of the weakness here in June, you think, is due to order backlog cancellations versus what portion is due to supply constraints?

Well, I cannot numerically break it down between those 2 factors, how much is because of supply and how much is because of the demand weakness. But qualitatively, I can talk about, those are 2 totally different reasons. The demand weakness is purely driven by end markets. So data center is the strongest, communication and networking, 5G-related is the next. The military and defense business seems kind of flat, reasonable. The aerospace part of that is lower because planes and are not getting built. And on the weakness side, the weakest one is automotive. And also industrial is a little bit weak. The medical as Microchip rolls up into industrial, but medical is very, very strong with all the medical issues of ventilators and thermometers and others. So the demand picture is really much more by end market. The supply disruption is purely by plant. And whatever is built in that plant. So -- and supply disruption was largely out of the Philippines and Malaysia, the 2 countries where the implementation of lockdown was so strict that you couldn't get the workers into the factory. So many of them have been living in the factory for the last 7 weeks in Philippines, and that's how we're producing the product. Those restrictions have just been lifted in the last few days. And both in Malaysia and Philippines, now people would be going back to work, but we are starting way behind, we have lost significant production in the last 6, 7 weeks, and we'll take the whole quarter to catch up, we will still be delinquent going out of this quarter, and we'll be shipping that delinquency in the September quarter. So those are 2 totally different reasons. If you have a data center product built in Malaysia or Philippines, then that is suffering both from -- I'm sorry, if you have an automotive product that is being built in Philippines, then that is suffering both from demand and supply. But the reasons are totally different. One is plant shutdown related in automotive. Other is the lockdown in our factories in Philippines and subcontractor factories in Malaysia.

Okay. That's great. That gives us a lot of color in terms of how you guys are thinking about it and how you're seeing it. The other dynamic on the demand side of the equation is turns business, right? Obviously, you have a certain percentage of turns business every quarter, whether it's looking at it quarter-on-quarter or maybe normal turns percentage in the June quarter looking back historically, like what are you seeing in the turns business thus far here in the June quarter?

Let me have Eric comment on the turns part of it.

Okay. So we still have turns to take to meet our revenue expectations for the quarter. And so even though we are seeing push-outs and cancellations at a higher rate than what would be normal, we are still getting turns in and the backlog in total is growing for the June quarter. But some of that is being offset by these pushouts and cancellations that Steve talked about before.

Okay. So that's a relatively good proxy that -- I mean, demand isn't falling off a cliff. You're still seeing some life out there because you are getting a fair -- a certain amount of turns business as we progress here through the quarter.

I don't think we have seen any weak where the total June backlog, what you have shipped plus what's on the backlog, has not declined in any week. So the new bookings and new turns have always been higher than what got pushed out and canceled. But because of the higher pushouts and cancellations, the slope of that backlog growth has slowed down.

Makes sense. The team was pretty proactive in pulling back on the OpEx instituting salary cuts that took effect in mid-April, putting all discretionary spending, preparing for, in your words, Steve, called it downside scenario. The actions you took resulted in a step down in OpEx here in the June quarter, another step down in the September quarter, and a flattish outlook into the December quarter. Can you just maybe give us some insights into your downside scenario as it relates to the profile of second half revenues? Or do we just assume that the OpEx trajectory down in June, down in September, flat in December is probably similar to how you're thinking about the demand and revenue trajectory for the team?

So we have not guided anything for the September and December quarter. And I think assuming that our OpEx trajectory would be -- would determine -- really the revenue trajectory would just be purely a guess and probably a wrong one. Our style is when we see stress coming on the revenue, we don't layoff thousands of people right away. What we implement is a flexible structure that can move in either direction. So when all these people start getting laid off and COVID-19 hit, we knew that it was going to have some demand stress. So we wanted to position the company for a downside scenario that could also very quickly be reversed at a moment's notice. So what we did in our factories is if you take our 2 large manufacturing plants in U.S. our factory in Gresham, Oregon and Tempe, Arizona, two fabrication plants. They're now both working. Each operator is working only 11 weeks in a quarter. So they're taking 2 weeks off, 1 week is unpaid, second week, they can use their vacation. And if somebody runs out of vacation, then 2 weeks are unpaid, but the unpaid week gets paid by the government. So they're not losing money, but we have brought down the production so that the inventory does not grow. And that's a very flexible place to be. From there, if the demand gets worse, we can go from 2 weeks down to 3 weeks down. If the demand does better, we can go from 2 weeks down to 1 week down to working full time. So on the other end, when the demand recovers, we can rapidly go back to full production, that we will not be able to go if we let people go. When you let people go, they move out of town, they move with their [ parents, ] they leave town very hard to reassemble the workforce, plus they have to be recertified, retrained, some of the processes have changed by the time they come back to work. So that is very flexible. And on the OpEx side, 99% of our worldwide employees, nonmanufacturing employees are currently on a 10% pay cut. It started on April '20. And what we have told them is that they're on the pay cut between now and end of the year, December 31. And since the pay cut was instituted mid-quarter on April '20 and some of the international geographies actually didn't begin, I think, till May 1. By definition, you have more OpEx reduction in September quarter than you have in March quarter. And then you have in June quarter. That doesn't mean we're forecasting revenue in the September quarter to be lower than June quarter. We just don't know. We have no idea what September quarter will do, how the factories will go back to work, what the demand would be. We will know that as we get closer. But that is the OpEx trajectory. Now if the September quarter is very good, actually, September quarter grows, let's say, like some people are forecasting, then we can terminate OpEx reduction early. We can give it back to them in a bonus and same thing for December quarter. So it's a very flexible place to be where we could easily give the good news back to the employees if the revenue is good, but if the revenue is not good, we are placed in a good place with an OpEx reduction.

Yes, makes a lot of sense. If I look at your auto business, it's the weakest segment for the team. It seems to be following the trajectory of the auto production data. Auto production is targeted to be down here in the June quarter, 40%, 45% sequentially. But the outlook, given early China recovery, Japan, Europe and U.S., all kind of starting back up their factories. The outlook is for auto production to kind of grow 50%, 60% sequentially in the September quarter. Now obviously, that's off of a low base, but nevertheless, strong growth sequentially. If this were to be the case and given your lean inventories and pretty short lead times, when would you start to, you think, see the order inflection here for auto? Would that be more towards the second half of this quarter or towards the end of this quarter, if that would be the case?

So I think number one, over any longer period of time, like if you look at a year, the auto business will track the auto SARS reasonably well. But it doesn't ensure churn. And the reason it doesn't do it is because we don't really ship anything to the car manufacturer. We ship it to the guys in between the Contis, the Delphis, the Visteon, the Aptivs, those guys. And they have their own lead time, 6 weeks production, let's say. So when auto SARS moves around, our revenue is not in lockstep with that on a monthly and quarterly basis. It's paced in either direction. So having said that, I think the auto business is down so much this quarter, as you mentioned, too, that from that low base, it's very likely that as the factories are coming back, the auto business grows next quarter. But what I'm not sure is that most people are making that call by saying factories go back to production, so the business should do very well. But factories go back to production, but will the output of the factories, which is cars, will they be bought by the consumers when over 100 million people around the world have been laid off. That I'm not sure. That will depend on the rate people go back to work, at the rate people feel comfortable with buying cars again and really having that kind of spending and not have concerns with the COVID. So you might see that the factories go back to work producing, but then they find that the demand for their product has weakened and will take longer time to recover. I just don't know that.

Okay. In answer to my question on the earnings call, you did say that China was strong in March and April, off of a weak January and February, but not a surprise because China was the first to go into COVID-19, the first to come out of COVID-19 and open up their economy. But you also cautioned that there might be some level of pull forward business just because of the concerns around supply constraints that your customers may have. And so far, though, quarter-to-date, has China continued to stay strong? Or are you actually starting to see signs that you're starting to normalize their order rates?

Do you want to take that one, Eric?

So coming off of the extended Chinese New Year and the COVID situation, there was obviously some pent-up demand, and we saw -- part of the reason for performance above what our early March update was on revenue is that China came back pretty strong. And that continued, you know that we measure ourselves based on sell-through. The sell-through activity from China distribution was also pretty healthy for the first part of April, which was good. But we have seen some slowing in that. And although we continue to get bookings in, we think that there was some pent-up demand that got caught up in March and April, and we're not really expecting it to continue at that level for the remainder of the quarter based on our guidance range.

Got it. Okay. Well, let's appreciate all of the commentary in the near to midterm. And you guys are executing well in a very tough environment. Let's focus on the technology and product leadership. So 2019 was another strong year for market share gains for the team. As you pointed out on the call, Steve, team gained share in all 3 of its microcontroller segments, 8-bit, 16-bit, 32-bit, strong #3 market share leader in the overall microcontroller market. You actually have a top 10 share position in the analog markets. You gained share there in 2019. #3 position in FPGA, you gained share there again also in 2019. And this was after a 2018 where you had fairly strong gains in all 3 of those categories. So on a go-forward basis, how should we think of the trajectory of future share gains in all 3 of these segments?

So Harlan, what I would say is that share gains are anything but unpredictable. Share gains are a result of 2 factors: one, how the end markets will perform. So if, let's say, competitors have a lot better exposure to, let's say, NXP and Renesas, they have a much higher exposure to the automotive end market than we do. And if the automotive market becomes very, very strong, then they have an advantage. In the last couple of years, automotive market wasn't as strong, so we have the advantage. And the second factor is a design win. So -- and design win has nothing to do with what we did in 2019. It has to do with what we did in 2017 and '18. So design wins in the prior years, plus the end market in the current year are the 2 things that drive the share gains at schedule for us. And as you said, we've done very well in the last couple of years. So from a design perspective, we're very confident that our funnel is very strong, and we're gaining share, and we have very, very good print position in design that will go to production this year and next year. The second issue is the end market. If the end market moves, then that we have no control over. And if the automotive end market was very weak, and let's say, this year and next year was very strong, then there would be a headwind for us relative to those 2 names, but I don't think automotive market is headed for being that strong, so we will gain share.

So you brought up a good point, which is a big part of share gains is designs that you won a number of years ago and future share gains are going to be based off of the design wins that you are winning today. And to your point, the team does have its total system solutions, or TSS strategy, which is certainly could be a driver of future share gains. And the idea here being that over time with the portfolio, system focus that the Microchip team is going to drive more dollar content or attach per project engagement. Last you updated us, you were kind of driving double digits percentage year-over-year increases in the number of Microchip products per customer program. You have a plethora of analog, interface, connectivity, memory solutions to offer alongside your MCUs and FPGAs and processors. Are you still driving that attach metric at a double digits type of year-over-year kind of growth rate?

So we have a very high focus on TSS. All of our sales force, application units, marketing guys, business managers are all very well-trained on looking at the entire customer board and populate it with as many Microchip products as we are. So qualitatively, we can say we're much further ahead on TSS today. And with the acquisitions, we have also acquired a lot of products that are more easily attachable to our other products who are gaining on TSS. What we do not have is really a numerical measurement. There's just been so much disruption with COVID-19 and all that. There's so many other things we are tracking to be able to give guidance and all that. And I think that TSS measurement probably took a back seat, and we need to correct that next quarter.

Okay. That's fair. We'll be anxiously awaiting the update there. Maybe at our CES conference in January, you can give us an update on that. Data center and 5G are bright spots in the near term, as you mentioned. But even before COVID-19, the outlook for cloud and data center CapEx spending and 5G networking spending had a pretty strong growth profile, in fact, not only this year but for the next several years, and the team has a strong position in both of these areas. And in particular, in your data center business, strong SSD controller position, strong position in networking with your PCIe switch portfolio. You also have a broad portfolio of Ethernet, microcontroller solutions. In 5G, Microsemi actually had a relatively strong portfolio, front haul, backhaul, RF, FPGAs and so on. Question is, how big is your data center and 5G service provider infrastructure exposure as sort of a percent of total revenues. And of the 6 megatrends that you are leveraged to, are these segments probably the fastest growing segments, you think?

So let me start with restating what our end market mix was 2 years ago. We generally don't manage our business really by end markets because lots of our products are very broad-based general purpose products. And we go-to-market horizontally then by end markets. So back to 2018 was the last time we updated our end market mix after we acquired Microsemi. And at that time, it was 28% industrial, 18% automotive, 16% consumer, 15% data center and computing, 13% in communication and 10% in aerospace and defense. So now 2 years have gone by, I can qualitatively tell you which way it has moved -- I don't have the quantitative numbers, again, something we'll try to get there in the next quarter or so. So #1 industrial at 28%, automotive was 18%. I believe that automotive has moved down because of all the problems we saw on the emission issues in Europe and our COVID-19 issues in China and Europe and U.S., strongest market is data center and computing. We cannot build enough of the product for data center and computing and will go substantially delinquent. By the June quarter, we went delinquent out of March quarter because the demand is that strong. And I would say it is not unconceivable that data center and computing have moved into the #2 spot behind industrial. There was only 3% difference. Automotive was 18%, data center was 15%. And automotive has come down, data center has gone up, and they might have switched. And communication probably has grown a little bit. Rest is probably about the same.

Good. That's great. Thanks for the insights there. And we'll be looking forward to the update on that as well. Let's focus on the core microcontroller business. 32-bit microcontroller business unit hit a record all-time high in terms of sell-through in the March quarter, it represented 47% of your total MCU business. Question is, what applications are driving this product segment? And second question is, do you see 32-bit crossing over your 8-bit business over the next kind of 12 months?

Well, that one is easy. You'll be pleased to know that our 32-bit MCU business was already larger than our 8-bit MCU business in the March quarter. So I doesn't have to cross, it already did. In terms of application, we must have 1,000 applications, end market by end market, we are in every air conditioner, every ventilator, every data center, every communication gear, every router, every refrigerator, every washer dryers, every consumer products, I mean, just everywhere. So there is really no single application in all end markets. Our 32-bit MCUs as well as our 8 and 16-bit MCUs are everywhere. There's not a single application that is driving the growth. That's what's good about our business. It's very, very broad-based.

Okay. Then maybe turning to the financials. I know it's difficult to call the bottom here, given where we are. But if I look at the last cyclical trough, which was last year, I mean, the team delivered gross margins that were 350 basis points better -- 350 basis points better. Operating margins that were 570 basis points better and 2x more earnings per share versus the trough in the prior cycle, right, the '15, '16 cycle. So if I assume another quarter-on-quarter decline here in the September quarter that we're going to go into very soon. And given our economists and global view of GDP recovery starting in the second half, then I estimate that gross margins would still be 250 to 300 basis points better and op margins, 400 to 500 basis points better than the prior cycle trough. It's not really apples-to-apples comparison, but the point is that structurally, the business is financially stronger, more scale, more breadth of the portfolio, leadership in key markets, I assume are a big part of that, but wanted to get your view, Steve.

Do you want to take that, Eric?

Sure. So I think we are extremely well positioned from a financial perspective. We have continued over the course of our acquisition history to be less dependent on internal manufacturing. So if you look at today versus 2008, 2009 financial crisis back then, we probably did 85% plus of our wafer fab in-house. Today, we do a little less than 40% in-house and our assembly and test percentages have come down also. And so even though we are experiencing underutilization charges today, they are not to the same degree as they would otherwise be if we were more dependent on our internal manufacturing. So that's one thing. You've brought up the diversification of the business. And actually, we've got a lot larger scale today, and we've done a good job of being as efficient as we can be throughout the various pieces of the P&L. And I think our OpEx is extremely well-managed at this point in time. So we're really proud of the fact that even in the down quarter that we're seeing right now, at the midpoint of our guidance, we're guiding to 36% operating margins. It's really pretty incredible, how good the margin has stayed together. So I think it's a combination of the acquisitions that we've done, our continued focus on being diversified and are little bit less dependent on internal manufacturing is how I would sum up the difference.

What I would add to that is, Harlan, I think in this cycle -- and even the cycle we went through with the U.S.-China trade sanction where the business came down. You haven't seen as much deterioration in gross margin. And therefore, you shouldn't expect the same 400, 500 basis points of recovery because you didn't see the disruption. So the business has become more stable. It is a better mix. And it's a better business than maybe what we had in 2009 when we saw a deeper trough and a strong recovery on the gross margin and operating margin. This time, you're not seeing as much disruption in gross and operating margin. So when it comes out of that, you will see higher gross and operating margin but not by 500 basis points.

Yes. Makes a lot of sense. And then to that end, though, on the longer-term gross margin improvement plans, the team continues to move more assembly and test in-house where the cost profile is obviously lower. On the Microsemi side, you're still consolidating a number of network of large -- actually, I should say, a network of a number of very small fabs that Microsemi had into your 2 large U.S.-based fabs. What's the time line for executing these initiatives? And when all is said and done, how much does this benefit sort of through-cycle gross margin profile for the team?

So there are 2 separate activities. One of that is bringing more of the assembly and test inside. We haven't really completed the Atmel one and then came the Microsemi. So a large amount of product is still done outside. And the capital is very constrained because we're focused on paying debt. So that's a very slow process. Every quarter, we bring a little bit inside, a few bps of gross margin win on the back, but there is no big chunks moving. I think it's a slow process. It's a multiyear process. And when you look at the -- all the small fabs and labs that came from Microsemi, there are lots of them, they are like 5 or 6 on the East Coast, 1 in Ireland, there's 1 in U.K., there's 1 in Santa Clara. So a number of those, we're consolidating into a high-value 8-inch fabs. But that is a -- there are thousands of products in those, hundreds of processes, product going into deep defense, military, with long qualification cycle. So that's really a 3-year process also.

Got it, right.

And we had talked about back in November, kind of the rightsizing and changing our fab in Colorado at a more specialized factory. Those processes are still in progress, and we're moving some of the high-volume manufacturing for some of these historical Atmel products into our Tempe and Oregon factories, and there's benefits coming with that, along with the things that Steve talked about in terms of some of the consolidation that will happen over time from some of these smaller factories. So those initiatives are still on track and moving forward.

Great...

So the bulk of that -- that came -- that is coming from transferring of some of the Atmel processes and products from 6-inch to 8-inch that is currently underway. And you see the benefit from that in about a year. And you would have seen faster than that if COVID-19 wasn't there because the demand would have been stronger, but you see that within a year, the ones that are coming from Microsemi, it gets completed in about 3 years, when we closed down a small fab in Bend, Oregon last year. So that benefit is already there. We're closing down a small fab in Santa Clara at the end of this year, you can see that benefit. So it's really lots of different factories, and it's coming at various times, but the endpoint is in about 3 years, and there are steps along the way.

Great. Well, we're just about out of time. So Steve, Eric, as always, an insightful discussion, keep up the great execution, and thank you for participating at our conference.

Thank you, Harlan.

Thanks, Harlan.