Microchip Technology Incorporated (MCHP) Earnings Call Transcript & Summary

So it is -- it's somewhat dependent on what the product mix is when the revenue returns to a higher level. So it was $16 million last quarter. We're showing improvement at the midpoint of guidance this quarter on our gross margin. And some of that is because we're having lower underutilization charges out of our back-end factories, assembly and test factories, this quarter. On the wafer fab side, we are not increasing starts at this point in time. We're kind of keeping that steady. So those charges will be with us until the point we start feeling we can ramp the factories and our inventory position is in a little better position. We're carrying higher inventory on the Microchip balance sheet today and particularly for the products produced in our own factories where we haven't wanted to jerk the factories up and down.

You mentioned doing new acquisitions has been one of the things that's taking gross margins down historically. Right now, the focus is on debt reduction with cash. But Steve, you made the point a few months ago that by the time the company is back at its target leverage ratio, there may not be attractive M&A assets this time around given how much consolidation has already occurred. Some semiconductor companies have ventured into the software space to do M&A. And I'm curious if that's something that would be interesting to Microchip.

We're not a hog.

Yes, we'd move on. That's -- I got a lot of questions. So maybe we could talk about market share in microcontrollers. The company has been dominant in 8-bit microcontrollers, close to 30% market share, and I actually think still taking more share in that space, 32-bit, I think by the numbers you'd shared about $1 billion of revenue and roughly 10% market share. How should we think about your market share in MCUs as we move forward?

So as we are analyzing the year ending December data, which SIA released recently, it clearly shows that we gained market share pretty much everywhere: in microcontrollers, 8-bit, 16, 32 as well as analog. So the small amount of market share gains really just almost are constant year after year after year, how we've gained market share over the years. I think beyond that, as we look forward, we have identified 6 large megatrends, which we see in the industry. Those are 5G, IoT, electric vehicles, advanced driver assist, called ADAS, data center growth and machine learning and artificial intelligence. Now you pick various companies in semiconductors, and you'll find that they have very deep exposure to one area. Somebody could be 5G. Somebody else could be totally dependent on data centers. I don't know if you will find a company that has significant exposure to all those 6 megatrends as Microchip has. Maybe you can find another one, but I think we are the only one. So that is exciting to me because 1 or the 2 megatrends in the past, IoT has been going on for many years, they haven't been able to change the growth rate of the industry, and the growth rate of the industry has been very low in the last 20 years. But as I look ahead, the 6 megatrends working together, I feel that -- and with a very, very low distribution inventory, I personally feel that we are better positioned than we have been in 2 decades.

Maybe one thing I'll add, as we typically share information based on the Gartner data that will come out in the March, April time frame, we'll share that on our next conference call and this kind of specifics that we see calendar '19 compared to calendar '18 in terms of market share in the various segments that you mentioned.

Yes. A couple of directions we can take this. But since you brought up the megatrends, maybe I can follow up on that topic. Of those 6 megatrends that you pointed out, is there 1 or 2 that the company's seen the most momentum with at the moment?

When it's only 1 or 2, we don't like it as much. So I think our focus is really distributed across all 6. Many times, they're very similar products. They're analog products from oscillator to converters to power management products. We are seeing them in data centers. We're seeing them in 5G applications. We are seeing them in IoT applications similarly with our connectivity products. Similarly, I think if you take some of the products in the networking area: USB, Ethernet and other products, they're going in multiple markets. So each of those products are going in multiple markets. So it's really not -- I can't say that all of our growth is dependent on 5G or IoT, really none of the above, I think it's all of the above.

Okay. And then on the microcontroller product line, one of the areas that the company acquired from the Atmel acquisition was a 32-bit ARM core. And Microchip has been pretty clear, it's not the core that drives the sale, it's about all the peripherals and the full solution that you're selling. But can you just talk about how helpful it's been for Microchip to have 32-bit ARM and how the growth rate with ARM has compared to the MIPS space core products?

So since we bought Atmel, our growth rate for ARM products and MIPS products have been about the same, which further justifies what we said before that we don't really believe core matters. Now some customers like MIPS, some customers like ARM and some customers don't care. So there are people in each can. What has happened by having ARM is we now have a solution for those customers who do care about ARM. There were other customers who don't care and other customers who hate ARM and they like MIPS. And I think we are the only ones who have all 2, both of them. And I think that has been beneficial. But the growth rate has not been really better on one or the other.

One of the topics that's garnered more discussion in the investment community in the automotive landscape is the potential consolidation in the number of engine control units in cars, they're getting very sophisticated with all these capabilities, and there's a movement at some of the auto Tier 1s that talked about trying to simplify the number of ECUs and have fewer -- a few of the ECU units but have more powerful compute in each one. What potential ramifications, if any, are there for an MCU supplier like Microchip?

So if we only made low-end 8-bit microcontrollers, that would be negative. Today, we make everything. We make very high-performance microcontrollers for pretty much any application. So if the trend goes in that direction, we are equally positioned like everybody else is. However, I am -- I don't think that trend would be as successful. And let me give you a historical example. There were other large-sized companies that I wouldn't name who believed 30 years ago that the cars will have one powerful engine control, 1 or 2, and they'll run copper wires to everything to really run those microcontrollers. I disagreed with that thought that long ago, and the reason being that car manufacturers do not build cars, car manufacturers assemble cars. And every single component in the car, the door module, the cockpit module, the seats, the taillights, everything comes from first-tier manufacturers who had a microcontroller with software, fully functional to each of those modules. If you think of a concept that all these are dumb, nothing works, you put it on a car together and then you expect that everything will work and you test them, that's a very, very difficult concept. That concept didn't even work in a PC, forget about a car. A PC has multiple microcontrollers, one in a keyboard, one in a mouse, one in a camera, one in a fax card, other card, sound chips and others. So there are multiple microcontrollers in a personal computer system, which is very connected, has a very powerful microprocessor, and one could think why does it need anything else. It didn't even work there. Why? Because PC manufacturers don't build PC either. PC manufacturers assemble PCs. Logitech builds a mouse. Somebody else builds something else. Somebody else builds a disk drive, and they are all put together. So when you go to car where the distances are much larger, I believe the concept of a few powerful ECUs driving the entire car is not a viable concept and I don't think would happen. But if it happens, we are equally prepared to deal with it.

That's interesting. I wanted to talk about the FPGA business. It's a segment that's had a lot of nice growth recently. And I think it's one of the underappreciated assets that Microchip has, there's just not that many FPGA providers that are available. Historically, FPGAs for your business were more tied to the aerospace and defense business, but I think there's been some focus about expanding that into new markets. And I was hoping you could elaborate a little bit on the opportunity to take your FPGA line into a broader set of the industry.

So we got the FPGA line through the acquisition of Microsemi. Our FPGAs are very nicely positioned between a very low end of Lattice and a very high end of Xilinx and Altera, Altera now being Intel. We largely do not have much competition in the middle of our space. We do have competition on the edges, both on the low end with Lattice and high end with Xilinx and Intel. But in the middle space, we largely do not have competition. I think we're putting a significant focus on that market, I would like to believe more than Microsemi was doing but -- although I don't really know what the real plans would have been. And we have recently introduced a few chips that really were developed on our clock, and we're developing the next-generation architecture that will actually take it to higher end and will compete more effectively with Xilinx and Intel. Our expertise happens to be very low power. We are the only company where the FPGAs are Flash-based,where both Intel and Xilinx chips based on SRAMs, which take much more power because you have to constantly refresh it. So there's a unique position we occupy. And with that, we're trying to take them into various different markets other than just aerospace and defense.

Steve, you spent some time at analyst days, on different investor conference calls, talking about the importance of culture. And I know it's something you've put a lot of thought into. What do you think it is about Microchip's culture that makes the company successful and unique?

Well, I wrote a book on it a long time ago. Title of the book is Driving Excellence, how we took a failing company and made it a market leader. So if anybody want to pick up that book, I think you will get the answer to that question. But a smaller version of the answer would be we run our company based on 11 guiding values, which begin with: quality comes first, customers are our focus -- I'm losing my voice. The third one is continuous improvement is essential. Fourth one is employees are our greatest strength. Fifth is growth and profits provide for everything we do, and it continues with communication and professional ethics and business partners and others. And the thinking is that we design the enterprise where all parts of the enterprise: our products, our strategies, our factories, our various initiatives all work in unison to achieve our mission. There's a culture where the company is not disjointed, and we have bought plenty of companies and acquisitions where we have found that the left hand doesn't know what the right hand was doing, and they were underperforming, spending a lot of money, having lower efficiency of product output, half of the products they will put into R&D would be canceled before they go to market. Those were signs of poor decision-making, poor coordination and R&D and decision-making across the company. When you look at that at Microchip, it is extremely efficient. 90% of the products we produce go to market successfully. And that's why you're seeing a company where our OpEx and operating profit and gross margin has been much higher than the companies that spend that much, much more money in R&D. And I think that is a result of not that we have better engineers but that's a result of we have a better culture. There is a better system. There's a better methodology, and people work together with the lowest cost to achieve our mission. If I take one example of it, and this comes from some of the companies we have bought in the past. These companies have multiple design centers around the world. And after we buy the company, we compare all the metrics to Microchip, how much CAD spending per $1 million of revenue, what is the revenue per employee, profits per employee and all those kind of metrics, and then we compare those metrics to Microchip metrics. And wherever there is a substantial divergence, we put a circle around it and assign an executive to analyze and see why is that happening and get that to a Microchip metric. And then you have learning from the prior acquisitions, you look for those things in the new acquisition, which increases the speed of how we get the synergy. So one thing we saw in some of these acquisitions were, on the CAD spending, essentially, what we found was these design centers in the various geographic regions of the world were individually getting licenses for CAD and they were all local licenses, they will use it for 8 hours a day. And then for the rest of the time, the idle will -- it will sit idle. And Germany will get their own, and Norway will get their own, and U.S. will get their own and Shanghai will get their own. And we found that the CAD spending was 2x or more higher than what Microchip was per design engineer or per any kind of metric they wanted to use. So we converted it to our system. And our system is we -- global worldwide design centers cooperate. And then with the combined purchasing power, we go and negotiate the CAD licenses, and the licenses are then available on a wide area network. So India comes to work first as the sun rises, and they use it for 8 hours. Then the sun rises in Germany and France and Norway, and they use the same license. Then it arrives in New York and Austin, and they use the same license. Then Phoenix, then San Jose, then Japan and Shanghai and around the world. So our licenses are used 24 hours a day. And so we cut down the number of license that we needed substantially. And from some of these acquisitions, we eliminated 2/3 of their spending. And that's a sign of really efficiency at its best. And that can't happen by somebody giving an order or a CEO doing it all the time. To do that, you have to build a culture. You have to build a culture of various people, company, design engineers, business units working together, pulling together in the same direction to get the maximum speed and efficiency of making improvements. That's what Microchip culture at its best.

Thank you very much for all those details. Maybe we have time for one last question. Just ramifications long-term from the trade war, one of the things investors discussed is the potential for it to become harder for U.S. companies to do business in China. Curious if Microchip has seen anything along those lines.

So we have seen a lot more talk than action. No question that during the last year with the entire U.S. rhetoric, customers will talk about it, but it didn't impact our business. What we are seeing is that Huaweis of the world and their semiconductor arms are trying to build SoCs. They will replace high-end Xilinx FPGAs and others. I think they're trying to design products that can move the needle in revenue. We make 8,000 different microcontrollers, more than 12,000 different analog parts, and no single product makes even 1% of our business. I don't know if any product makes even 0.1% of our business. So you have to accumulate hundreds of thousands of products to really move the needle. So we're not seeing that kind of effort with the larger companies, like Huaweis and ZTEs, to replace these small chips. They're working on a much larger thing. But we are seeing a bunch of start-ups making noise that they will build this, they will build that with very little success because, again, one, you can't move the needle. Secondly, I think this coronavirus is going to kill them. They just can't survive with -- they were not strong enough. They didn't have enough traction to survive a swan event like that. I think coming out of this, my sense is the Americas will be loved more again.

Understood. Well, unfortunately, we are out of time. I want to thank both Steve and Eric for being with us today.

Thank you.

Thank you.