Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

Good afternoon, everyone. This is Vivek Arya, senior Semiconductor and Semi-cap Equipment analyst at Bank of America Securities. Really honored and delighted to have Steve Sanghi, Chief Executive Officer, and Eric Bjornholt, Senior Vice President and Chief Financial Officer of Microchip, join us this afternoon to share their views. And as everyone saw, Microchip positively preannounced yesterday, which definitely helped the semiconductor stocks today. So maybe what we will do is we will go through Q&A with Steve and Eric. But if you have any questions, please feel free to e-mail them to me on the side through your console. But with that, I really want to welcome both Steve and Eric.

And as a start, maybe, Steve, if you could just maybe summarize the positive news from yesterday, what drove that? And importantly, how should we think about the state of the union right now in terms of the demand and the supply environment versus what you thought when you gave the original guidance?

So thank you, Vivek, for inviting us to the conference. So we issued the press release yesterday. And what we said is that our business has performed better than we anticipated when we had an earnings conference call on May 7. There are 2 reasons for it. One is on the supply side and other is on the demand side. The supply chain disruptions were mainly in Philippines and Malaysia, where the local governments and municipalities locked down the cities and wouldn't let the factories open or have the workers go to work. About a couple of weeks ago, they started to ease those restrictions. And prior to that, we had a large number of our employees working in the facility and living in the facility in makeshift tents to continue to help Microchip, and our thanks go out to those people who didn't see their families for 8 weeks. But with all that effort, we have begun to make up for lost production and we expect that we will continue to gain ground on the production side for the rest of the quarter. On the demand side, China factories are fully back to work and demand in China has returned to normal. But the demand in U.S. and Europe, while it is still weak, customer factories in U.S. and Europe are starting to open, and we believe that, that demand will start mending also. Vivek?

Sorry, I was on mute. Sorry. Is there a way, Steve, to break out where the upside came from? So basically, you -- the original outlook was for down 6% sequential at the midpoint, the new outlook is for down 3%. Is the upside coming more from the recovery on the supply chain side or improvement on the demand side?

I would think more of the recovery will be on the demand side and the supply side. But we have 25 business units in 3 geographies and 75 different items with 3% change, it's like slicing too small a pie. But overall, taking it in big brackets, I think our business has been strong in the data center and computing side, strong on the medical side, and very weak on the automotive side and some on the industrial side. So those are kind of the big brackets. And China has been good. Europe has been the worst. America has been in between.

Got it. One question on the supply side that I just received from an investor was that, is your supply situation completely back to normal? Or do you think the social distancing restrictions are still kind of keeping the capacity below 100% in some of your factories?

So it is not quite 100%, it's a bit lower than that factory by factory, largely driven by some of the people who are still at home, either quarantining themselves because of various exposures. These people are not sick, but they tested -- they haven't been tested and they had some sort of exposure. So it's not 100%, but it's higher than what we had modeled. So therefore, we are gaining ground rather than losing ground further.

Got it. Now when I go back to the last earnings call, Steve, and I look across the industry, we had some of your peers that were a little hesitant, I think, to call for a hard bottom in the June quarter, right? Then we had others, for example, NXP said other than automotive, every other segment was increasing sequentially and they hope that every segment would grow in Q3, essentially making Q2 the bottom. So given where we are today, do you think it's fair to assume June can be the bottom for your industry?

Well, Vivek, I look at it this way. When we announced our earnings on May 7, 5 million Americans were filing for first-time unemployment on a weekly basis. And over the last several weeks, there are now 40 million Americans that have filed for unemployment. Globally, that number is over 100 million people. In that environment and with path of COVID-19 not known, what I mean is we don't have a vaccine yet. And as the people go back to work, will there be a re-flare-up or what will happen? That not being known, I really do not know how anyone can forecast the June quarter to be the bottom. I mean it can be. But as the factories open, will the factories find demand for their products? Will these 100 million people be employed rapidly? Will they be buying large-ticket items like automobiles and brown goods and white goods and others? I do not know that. And I have made bottom calls many times before in our history, but I did that when I had the data. Today, I don't have the data based on the COVID-19 path. And the ones who are making that call, I don't trust it. I don't know how they could do that. We are seeing the same thing that NXP is seeing, where the factories were shut down and now returning to work, but do those factories know their underlying demand? I don't think they do. When we talk to them regarding how does the second half look, they tell me the same thing that I'm telling you. Well, it depends on COVID 19. It depends on whether people are willing to spend money on big-ticket items. So either sequentially or year-over-year, NXP performance into our -- Microchip's June quarter at the midpoint of the guidance is down 2.7% from a year ago, NXP's June quarter at the midpoint of our guidance is down 18.8% from a year ago. That is not even in the same ZIP code. So I think that makes any comments from them about the June quarter bottom a little bit irrelevant in my mind.

Understand. Steve, one more on the preannouncement. Was there any market that incrementally kind of stood out to you in terms of driving the upside? Because I think some of the data center and medical type end markets, I think they were strong earlier also. So was there a specific end market that picked up? Or just the markets that were strong, they just became stronger?

So data center was strong and the automotive was weak, but data was -- data center was not stronger than what we thought and automotive wasn't weaker than what we thought. The change is kind of all over the place. Like I said, we have 25 business units contributing into multiple end markets and trying to break out a 3% difference would be relatively difficult. So the upside didn't come from a single segment or a single business unit, it's pretty broad-based.

Got it. One other topic, that's kind of top of mind of -- for investors right now is the ongoing U.S.-China trade tensions. Curiously, what's your take on, first, any specific Huawei exposure and how that impacts your business, both in the near and the longer term, that is -- if let's say, the restrictions become even more onerous than, right, even stricter than they have been in the past? And just beyond that, do you see any other restrictions coming up in terms of shipping to Chinese customers outside of Huawei that can impact Microchip's business?

So the only restrictions we have right now are really shipping to customers that are making products for Chinese military, and we really don't have that. We're not shipping anything to Chinese military. Microchip's exposure to Huawei is in the 2% range. And the new restrictions from Department of Commerce does not have any effect on that. What changed in the latest guidance from Department of Commerce, really was if you make custom product for Huawei or you build a product-based on Huawei specs, then that was restricted. We don't sell anything to Huawei. That's not a standard product. So as far as now, we really don't have any impact. Having said that, it's been well rumored that Huawei has been building inventory for quite some time. Concerned about these restrictions, they have been building inventory. And at some point in time, Huawei will start building that inventory and there would be an inventory correction for Huawei's business, but it's only 2% of our business. So the impact, really, would be small. Other than Huawei, we really haven't seen a broad base of customers building inventory. We also work with the long tail of customers who don't often have the balance sheet strength to build inventory. So I think other than this issue on Huawei, where they might be building some inventory, we don't have any other issues with Huawei.

Got it. Steve, on the specific topic of inventory build. These trade tensions are not new, right? They have been there in the press for quite some time. So if I'm a customer, right, not Huawei, but other Chinese customers, why won't I have -- try to protect myself by building more inventory or maybe even finished goods? Because when I've asked this question about inventory or double ordering or pull-ins from other management teams, the answer is consistent that we have not seen that. But don't you think that by itself is surprising that if I'm a customer in China, I would have -- I would try and protect myself in case these restrictions get worse.

Well, so the restrictions from Department of Commerce have not been random. ZTE was shipping to Iran and so is Huawei and all those. So they were, based on some data and accusations, a normal run-of-the-mill customer that's building washers and dryers and thermostats and blender, iron, toasters and garage door openers is not concerned about getting restrictions on them. Plus they're buying standard products that can always be bought from distribution. They can buy them from outside the country, import them themselves. So it's really -- an average customer isn't really worried about it. Huawei knows what Huawei does, and Huawei is concerned because they do a lot of business with Chinese military. They do business with other entities that are prohibited like HiSilicon and all that. So I haven't seen a general panic with customers that they will get injunction. And secondly, since we do business with a long tail, they don't have the balance sheet strength to build inventory anyway and they do not really see a concern about their ability to buy product from Microchip.

Got it. The last one on just this topic of trade tensions. I'd just kind of lead this to this broader topic of decoupling that comes up. Do you think that your competitors in Europe or maybe even some up-and-coming kind of Chinese domestic suppliers can replicate your products or your technology over time?

Well, I mean, 98% of what we build is proprietary and many of them is protected by architectures or technology apparents and others, we haven't really seen any attempt for anybody to want to do that. But we have normal level of competition with STMicro and NXP and others, and some of them basically have been claiming that China would prefer them versus America. That's certainly possible. But it has been said for 1.5 years now and we haven't seen impact. If you look at the -- our performance in microcontroller and analog relative to them, we're doing much better than them. So that doesn't justify that they're gaining because of that. So I think I would largely say it's difficult to duplicate our products. We have a lot of proprietary content. There are a lot of sole-sourced products. And there are a lot of products with architectures and support systems and development tools and reference designs and unique firmware support. And I mean, we have a good business, that's why we have the gross margin that we do.

Got it. It's interesting, you brought up this point about the midpoint of the new outlook, down only 2.7% year-on-year. That is, as you mentioned, sort of on a different planet when it comes to some of your other competitors in the microcontroller market. Is it that proprietary nature or aspect of your business? How else can we explain why Microchip's business has held up much better despite these very severe macro headwinds that we are seeing?

So I think if we have to differentiate. I wouldn't say our business has done better, I would say their business has done worse. In the June quarter, we are still down 3% sequentially and 2.7% year-over-year. That's nothing to write home about. June is usually a strong quarter for us. We should be up 3% rather than being down 3%. So rather than saying our business has done better, I, as a CEO, am not satisfied with our business, our business should do even much better. I would say their business has done a lot worse because they're down double digit. And you should be asking them really, what happened? Why is their business down so much?

Got it. If you set aside Huawei, so excluding Huawei, is there a different dynamic to year-on-year trends? Or it doesn't matter whether we include or exclude Huawei because it's been a smaller part of your business all along?

It's a very small part of our business. It wouldn't make any difference. It's about 2% of our business.

Got it. And it was the same a year ago also, I imagine. That part has not really changed.

Yes, that hasn't changed.

Got it. Next thing is the complete platform or the solution approach that I think you started talking about a few years ago, complementing your microcontrollers with the analog products. When I talk with some of your competitors who only do microcontrollers, they say, well, customers make these decisions independently. And when I talk with some other competitors who have the microcontroller and analog products, even they think that, look, it's not as big a deal, right? Because customers, again, make these decisions independently. What has been your experience that -- what has been kind of the revenue synergy between your microcontroller and analog business? What's been the success in crafting solutions that combine both of these capabilities?

So I think when you look at across 125,000 customers that we serve, you'll find every answer. And the ones who are telling you that they buy independently, they're not wrong, but they are wrong if they're saying absolutely that is the case in 100% of the customers. A decade ago, when we went to a customer, we mainly had a microcontroller to sell and maybe a few analog parts, which were disconnected from a coordinated sales approach. Today, when we go to a customer and we take a reference design to them, whatever it is, let's say, it's a motherboard for an appliance like a washer dryer or it's something for a router, or it's something for a 5G system or whatever would be, that board has a central chip from a Microchip controller, a microprocessor or an SPGA. And depending on the complexity, it could be 8, 16, 32-bit MCU or an SPGA when -- but then in our reference design, that chip is surrounded with analog parts like sensors, amplifiers, reference devices, A to D converters, power management; connectivity products like USB, Ethernet, WiFi, Bluetooth; memory products, both static RAM and flash memory; security products, D to A converters, power drivers; discrete products like sets, diodes and transistors. And the entire board is built up of our products. And only thing that is not ours, capacitors and battery. And when you show that to the customer, you win a lot more than just the central device. We never say we win 100% of that. And customers have this preference of buying certain things from TI, ADI or whatever, but we continuously eat into it. And total system solutions is adding to it positively. We always win more than we won the prior day or prior year. So that part is positive. I think what -- but what is required, and it has taken us many years to perfect, it requires a much more coordinated sales approach to get that total system solution when other people tell you, well, customers make independent decisions because the market is like independent. One thing is from one business unit, other thing is from other business unit. And one guy goes -- tells his story and doesn't care about the other guy. At Microchip, when a microcontroller guy presents to me, then we -- that they won a design, there would be chastised if they did not show at the same time what else did we win other than the microcontroller. And same way when an FPGA guy is presenting that we won this design, what else did you win with it? And it's training, it's repetitive, it's constantly pounding and then people paying attention and presenting such to the customer. But it's a much more honed process that we have perfected, and we were struggling with it, too, for years over the -- but we've gotten really good at it in the last couple of years.

Got it. When I, again, go back to this topic of Microchip's outperformance versus the industry or just the resilience versus the declines we saw at your competitors, how much of that is due to your end market mix? Like is it that over the last year or 2 since Microsemi also joined Microchip, that the combined franchise has had a much more -- much better exposure to the data center and the communication market. So that has perhaps helped become more resilient in this downturn, where parts of the market such as data center actually held up better. So how much of the resilience is because of the end market mix for Microchip?

So some of it is that, but not all of it. Some of it is that, but you got to give the company credit for diversifying their product mix and all that. So some of it is that. A large portion of it is the total system solution. And if you take the data center mix and all that out, our business will still be very, very good and will not be down like 21% that TI is down or 19% that NXP is down. [ To sum off, ] in any market there's just lots of moving parts.

Got it. In terms of the focus that you have had or just the diversity, I should say, you have had in terms of 8 and 16 and 32, but people have been writing off the 8-bit capability for quite some time, but you have always managed to grow that business. So how much of a factor is having -- so first, how much more life is left in the 8-bit market? And then secondly, just the fact that you have this diversity in terms of solution, is that a part of the resilience also?

So it's all of the above. I mean I was hearing back in 1994 from the analysts that the 8-bit MC was dead. A lot of competitors bought into the story and they shifted their investments, but Microchip did not and we gained a dominant share of that market. Prior to the combined effect of U.S.-China trade and COVID-19, so if you go back 1.5 years ago, our 8-bit business was performing at a record revenue level. And because of the combined effects of U.S.-China trade and COVID-19, we're down some. But we believe in the future, when these things are resolved, we will make a new record in the future. So I think what is unappreciated about the 8-bit market is that there are boundless set of new applications which require adding intelligence to lots and lots of dumb devices, devices like light bulbs, door locks, pregnancy testers, digital thermometers, occupancy sensors and on and on and on. And more often than not, the 8-bit microcontroller is the easiest and simplest and most cost-effective way to achieve this. And the people who don't have that, for them, it's a hammer and everything is a nail. And you try to solve a problem with a 32-bit MCU, and it's either too expensive or too cumbersome to deal with or burns more power than they want burning or have a larger footprint. And 8-bit is the simplest way to solve that problem. By us having 8, 16, 32 and into independent business units, each having its road map and each having new products, and each doing cost reductions and performance improvement, and then we have a common marketing growth, pricing growth that takes those products to the market to ensure there's not a price competition at crossover points between 8-bit, 16-bit and 32-bit, we get best of all worlds. We have an ideal product for the customer, where it should be in 8-bit, and we're not pushing a 32-bit solution but cost-effective, and we win almost every time in that. And we don't give up price because we have a common pricing group that doesn't allow price competition between the microcontroller groups of Microchip. So as a result, we have done very, very well. And have we benefited from this total system solution also on 8-bit? We have. But we are also more benefited on the 32-bit than 8-bit because 32-bit is surrounded by more parts than often 8-bit is. So it's not only 8-bit. While our 8-bit business is doing very well, we're not sitting back on 32-bit either. Last quarter, first time, our 32-bit business was larger than 8-bit. And 32-bit business is now about $1.3 billion. For whatever reason, Gartner reported that to be $400 million smaller that we don't understand.

Steve, one thing you mentioned in the preannouncement yesterday was automotive, this has been, other than smartphones, I think, one of the weakest end markets and some industry data suggests, units down almost 40%, 45% year-on-year in Q2. So how should we view your -- the recovery in the automotive market? We are hearing of more stimulus measures in Europe, Cash for Clunkers, et cetera, right? We are hearing about some EV pickup in China. How real is this recovery in the automotive market?

So I think it is recovering from the bottom, but the absolute number is not good enough. So there were 2 different reports yesterday, one was about China and one was about U.S., and let me talk a little bit about both. The China news was good in saying that automotive sales in China have returned back to approximately 25 million units sold annualized in the last month, which is very good news because China -- we have been seeing, China has returned back to normal. So that news was good. The news from U.S. was that U.S. auto sales in U.S. were up 1.1 million cars in May from April. Now that would be good. But they were still down 33% from a year ago, and that is not good. So while it's recovering from the bottom, the absolute number in U.S. and Europe is still quite a bit down.

Got it. Do you look at the automotive market as being as attractive a growth market in the future for Microchip than it was before we got into these COVID headwinds?

So nobody talked about that being a bad market last year. And until January, COVID-19 had to be temporary. After COVID-19 is resolved, we expect it will continue to be a good market with significant growth in EV and self-driving and Ethernet in the car and general gadgets and all that in the car, yes.

Got it. Now moving to just factory utilization, and I think some of the temporary measures you had placed before. Now that you saw somewhat better trends in Q2, how are you managing your factory utilization? And I think the last time you had given a specific gross margin outlook, that was for it being significantly down in Q2. So how should we think about how you're managing utilization and hence, the impact on gross margins?

Let's have Eric take that one.

Okay. So from a utilization perspective, we're -- we've implemented rotating time off in our 2 large factories in the U.S., in Arizona and Oregon. We -- so we haven't really changed anything from a wafer fab perspective. We are running more products through our back-end factories. And with the guidance update that we gave yesterday, we had just a small uptick of 0.1% on the low and high end of the guidance range. So getting a little bit of benefit there. I think I'd like to highlight there, our margins have really held up extraordinarily well. Now at almost 61% gross margin at the midpoint of guidance for this quarter, we're only 2% away from our long-term model and that's with significant underutilization charges. So as the top line returns to higher levels and those underutilization charges peel off, we have a lot of confidence in getting to our long-term model. In addition to some of the self-help programs that we have in place to bring more outsourced assembly and test in-house, back in November, we talked about turning our Colorado 6-inch fab into a more specialized factory and moving some of that volume production to our more efficient 8-inch factories in Arizona and Oregon. So we've got a lot of confidence in getting to that long-term model on gross margin, but we're still significantly underutilizing our factories today. Probably the other thing to point out there is we only do just under 40% of our wafer fab in-house today compared to what it was back in the 2008, 2009 downturn or it was probably closer to 90% that we were doing in-house. So that also has helped us moderate the decline in gross margin in this cycle.

Got it. Then, Steve, you implemented some kind of very lean OpEx, right, throughout the company, including, right, kind of salary constraints on managers and so forth. From what you are seeing in the market, do you think it is time to step back from those OpEx constraints? Or you're not at that point yet?

Well, we're not at that point yet. What we did was we -- on a very proactive basis, we took steps to lower the operating expense in expectation of a major Category 6 hurricane, if you want to think about it that way. You board down your windows and you batten down the hatches before the storm hits. Because once the storm hits, it's much harder. So that's what we did. And we expected with -- when we're making these decisions, we were losing 5 million jobs a week in the U.S. and we expected that this storm will hit us. It hasn't hit us that badly yet. I can't say it hasn't hit at all because we should be up in June quarter and we are down 3%. So there is certainly some impact, but it hasn't hit us in a big way. So the storm kind of stalled out in the ocean. Question is, is it moving slowly and is it going to yet still hit us next quarter? I don't know the answer to that. When the people go back to work, will they be -- will all of them be rehired? Will they be willing to spend money on large-ticket items? Will COVID-19 return when people start congregating again? The answer to those questions are not known. So for the time being, we're going to stay on the salary cut. But if the business does better, you can always give money back to the people as a bonus or change the salary cut from 10% to something less than that. That is very easy to do, and it's good news. It's much harder going the other way. Especially overseas outside of U.S., where you have to get a consent employee by employee. We have 12,000 employees overseas outside of U.S., 6,000 roughly in U.S. and 98% of all employees consented, and that's a major undertaking manager by manager, leaders of various organizations, country by country, working with many works council. 100% of the works councils in France and Germany supported the salary cut. That is unprecedented. You never see that. That's what we achieved. So we don't want to be now haphazard in changing it in the other direction, you want to have the storm completely pass.

Got it. And Steve, just in the last 10, 20 seconds that we have left. I got one question from the audience which is, how do you think channel inventory will trend in Q2? And is any of this upside inventory fill in? Or is it really end market demand recovery?

Eric, you should take that.

So with a month still to go in the quarter, we don't have a public view in terms of what distribution inventory is going to do. I would say that we're in a situation where distribution inventory has been quite low. It was really at about a 15-year low just coming out of the December quarter, and then it moved by like a day in the March quarter. So it's still real low. Over the course of time, we do think that distributors will replenish inventory, but it's very hard for us to predict how that will trend in any particular quarter. So we'll watch it. And we'll give you metrics on that when we get to the end of June and report our results in early August. But for now, we don't have a public view on it.

Got it. And maybe one last quick one since I just got it, which is technology road map in 40-nanometer and below microcontrollers because one of your Japanese competitors seemed to be seeing good traction in that part of the market.

So we don't comment on use of specific technology on a specific product line, that's just too sensitive. We are using technologies all the way much lower than that, but we can't assign the technology to a product line that going to give it all. So anybody who's doing something in Japan, God love them, they're still losing market share to us. But we're not going to share what we are doing.

Understand. Well, with that, I think we are at the end of our allotted time. Thank you so much, Steve and Eric. Really appreciate the preannouncement from yesterday. Really appreciate you coming on the call and giving us your insights.

Thank you, Vivek.

Thanks very much. Thanks, everyone, for joining.

Thanks very much, Vivek.

Okay. Bye.