Applied Materials, Inc. (NASDAQ:AMAT) Citi 2019 Global Technology Conference Call September 4, 2019 1:25 PM ET
Dan Durn - CFO
Mike Sullivan - VP, IR
Conference Call Participants
Atif Malik - Citi
Good afternoon, everyone. My name is Atif Malik. I cover equipment and semiconductor stocks here at Citi. It’s my pleasure to welcome Dan Durn, CFO, Applied Materials; also Mike Sullivan, VP, Investor Relations. I’m going to kick it off with my questions first and then open the floor to the audience for their questions.
Dan, it’s been almost two years since you joined Applied Materials team. Would love to get your perspective on how you think Applied Materials is doing financially as well as strategically.
I guess, what I’d say is, we are better positioned now than I think this Company has ever been. And I look at it on multiple levels. First, from a market opportunity standpoint, the next wave of compute is beginning to take hold, gain traction, and we are in the early innings. It’s going to take our industry structurally larger. When I look at Moore’s law hitting a wall and the way I see it, the simultaneous benefits of power, performance, area and cost, as we shrink across the 2D dimension, we’ve got to make trade-offs now. And you’ve got to choose which of those variables you’re going to optimize around. You don’t get them all simultaneously. And the new playbook that’s going to drive the power performance roadmap for our industry is going to have five elements and Applied is going to be a key enabler of each and every one of those five elements. So, we feel good about the market opportunity.
Then, on a Company-specific basis, we’ve got a very-strategic visionary leader who is a great product guy. And then, when I look at the opportunity to innovate in a market that’s fundamentally inflecting, from the outside looking in, Applied has always gotten high marks for being highly innovative and deep technical talent. But, what you can appreciate from the outside is really how deep those roadmaps are, how deep that technical talent base is. So, when I look at it from a market opportunity standpoint, Moore’s law hitting a wall, the roadmap going forward, having multiple elements where we can be a key enabler, visionary leadership and deep, deep technical talent, I think, this Company is better positioned than we’ve ever been.
Great. Then, you have kept investors in suspense on when to call the bottom. I feel like, you’re becoming tangy [ph] for capital equipment, which makes sense given ever-changing macro on China trade. On the July quarter call, you sounded incrementally positive and called for U-shaped recovery next year with sustained foundry logic investments and memory up led by NAND. Can you walk us through why you see this cycle different versus past cycles?
Sure. Let me share a little bit of what we see. First, it feels like our business has stabilized, and we’re really encouraged by that. When memory recovers, this is a company that’s going to return to strong sequential growth. So, we feel good about our how we’re positioned. But, taking a step back, we’re a more diverse company than we’ve ever been. Foundry logic is a strong market for us right now, and we see that strength continuing as we go forward as customers continue to make investments in their leading-edge roadmaps as well as trailing those geometries. And when we think about memory and we think about a memory recovery, we think about NAND leading, followed by DRAM.
So, in this recovery, we don’t -- it doesn’t look like everything’s going to fire in sync as we see the recovery. But, I think the foundry followed by NAND, followed by DRAM bodes well for the long-term sustainability of the recovery and I’m encouraged by it. Then, when I look at the diversification of our business and you think about display and services being a larger component of the revenues. I just think this is a company that is really well-positioned in the current environment. You mentioned geopolitical, macro risks. I just think, in environment like this, it’s just prudent to set expectations in a modest way. And we’re going to position our business with those assumptions, but we’re also going to be very nimble. When, we see signs of recovery, this is a company that’s going to be able to respond very quickly to those signs of recovery.
And like I said, we will return to strong sequential growth when we see those markets influx. So, we feel good about where we’re at. But, we think it makes sense to be prudent and set expectations in a modest way.
Got it. Now, you announced a couple of new products at SEMICON West targeting emerging memories and cloud computing, and company spearheading the AI conversation in the industry you had, Lisa Su, Victor Peng, all on the stage at SEMICON, and you’ve hired a Wall Street level talent in terms of AI. Can you share with us the Company’s or Gary’s vision on AI and why Company is spending so much time and effort on this theme?
I think, you’re onto something here. And again, let me share with you little bit about what we’re saying. We talked about the next wave of compute beginning to take hold. And we’re in the early innings of the data economy taking hold. And as you look at the history of our industry, this is the fourth wave of compute that we’ve seen take hold. First, there was mainframe and departmental compute; second wave was defined by PC and the internet; third, mobile social media; and now, we’re in the fourth wave of compute, and it’s just taking hold. What we see as common elements of each of those successive waves, each wave has brought more demand for data, data generation, networking, faster and higher capacity processing, more storage. And so, there is so many common elements to each of those waves. What’s different this time though is this is the first time we’ve had a new wave of compute come in right at the time when Moore’s law is beginning to hit a wall, and the simultaneous benefits of power, performance, area cost no longer get realized as you shrink across the 2D dimension.
So, the industry is back to innovating. And the path forward is going to be increasingly defined by what Gary calls, the new playbook, and there is five elements. 2D shrinks are an important part of it, but it’s one of five. There is going to be new materials, new architectures, new forms of packaging, and new types of devices that this industry is going to innovate and bring to market. And so, when we think about the types of things the industry needs. What better way to connect with the ecosystem than to hire a different skill set, a different set of talent that gives us that read-through capability to the ecosystem. And that’s what you see in our actions. And we think the R&D is targeted to the right opportunities, the messaging we’ve created around this industry and where it’s going, we see increasingly echoed by other participants in the industry, whether it’s the software industry or the systems companies or the design companies or the manufacturing companies, I think all elements of the ecosystem are now delivering a similar message. And so, it tells us that we’re on the right track in terms of delivering the capabilities that the industry wants.
And you talked about the product introduction at SEMICON West. I think it’s illustrative and really underscores the point we’re making. MRAM has been around as a device type for a long time. The reason the industry hasn’t talked about it -- given the benefits that that device type brings to market, the reason we haven’t talked about it that much is this industry hasn’t been able to manufacture it in high volume and in industrial scale. So, billions of devices on a single wafer, we haven’t been able to do that before as an industry. And the product we announced is probably the most complex product in the history of the Company where we’re doing multiple steps of processing all under vacuum getting 50 film sacks working together and honing the interfaces with onboard metrology to measure the height of a stack to 0.01-nanometer. That capability hasn’t existed in the industry before. It’s a significant groundbreaking set of capabilities that has the potential to introduce a game-changing memory architecture as it plays out over the long run. So, we feel good about the investments and we feel good about the insights that this talent base is bringing to the company and connecting us with a broader cross-section of the ecosystem.
Good. Dan, you provided an early view into January quarter with gross margins and operating expenses as a near-term headwind. So, the number one question I got from investors was why is OpEx going higher and January quarter EPS expected to be below October quarter, assuming, there is no major memory recovery. Can you talk about your office dynamics?
Sure. So, there is a couple elements to the OpEx story as we look forward to the January quarter. First is, the January quarter is the first quarter of our new fiscal year. So, there’s a mechanical transition that happens each and every fiscal year. In an environment that we’re in right now defined by a memory downturn, people that sit in seats that like you sit in, probably would not expect our Company to pay a full bonus, and we’re not going to pay a full bonus in 2019. But, when we start the new fiscal year, the accrual rate on bonuses needs to step up until we get a better read-through of how the year progresses. So, there’s a mechanical aspect of transitioning to a new fiscal year that causes a step up in that accrual rate.
Second thing, start of a new fiscal year, there’s merit increases that happen every year. And then, the third thing is, this new playbook that we’re talking about and customer-specific opportunities that we’re working on hand-in-hand with customers that have a potential for long-term significant value-creation at the Company are things we’re going to continue to invest in. We’re not going to let a temporary downturn, over the course of two, three quarters, get our Company out of position in the way in which we’re investing in fundamental enabling technologies to make our customers successful as the power performance roadmap begins to diversify and have multiple elements, all of which we can play a key enabling role. So, there’s multiple aspects to that story. We have a strong point of view of where this industry is going and a lot of conviction about our ability to lead. And we think we can make a real difference in enabling our customers to be successful. And that’s what we’re investing in.
Got it. Just staying on the financial model. We all know Gary’s track record, he’s a product guy, he will get it right on the new products. He’s spending R&D. When you show your financial model, fiscal ‘20 financial model, last time at $45 billion WFE you assume 17.5% OpEx to sales ratio. Does this level of spending change the long-term target model?
I don’t think, it changes the level of the long-term model. If you talk to Gary, and Gary will share with you openly, that he doesn’t manage the Company to spend ratios. What he manages the Company to is industry inflections and delivering innovation to the market at a time those inflections materialize to grow our share -- profitably grow our share, profitably grow our Company. He’s been doing that now for about three decades. And so, that’s at the core of his philosophical approach.
Now, that said, taking a step back and looking at the track record since he’s been here. When he stepped in to the Company, OpEx as a percent of sales at the Company was over 28%. As we’ve grown the Company, that’s gone down to, call it mid-teens. And in the middle of this memory downturn, it’s now crept up to high-teens, maybe touching 20%. Soon as we see the memory recovery take hold and the Company get back to strong sequential growth, I think you’ll see the operating leverage at the Company kick in again. And we’ll be incredibly disciplined about how we manage the Company.
I guess, the last thing I’d say too is, if you take a look at the proportion of OpEx towards fuel for growing our Company, the innovative engine of the Company, the things that ultimately drive that growth and market share gain, and then what it takes to run the Company, I think you’ll see on a dollar basis, things like G&A roughly flat over a five, six-year period as we’ve more than doubled the size of the Company. And R&D as a percent of OpEx today is higher than it’s ever been an all-time high of over 70%. So, we’re going to be disciplined about how we operate the Company, we will be unemotional in the decision-making. But, we do see a fundamental opportunity to drive leadership in this industry for the next decade or two. And we will invest to do that.
China, China domestic spend has been a nice anchor in second half as broader memory market recovers. Do you think domestic China spend holds next year at these levels, given the trade friction?
We do. We think China market is strong this year. We see that strength continuing. If we take a step back and look at how the markets developed. It’s been slow and steady progress of spending, building ecosystems, developing technology roadmaps, and we haven’t seen a hockey stick of spending or hockey stick of capacity adds either foundry logic or memory. And so, we feel good about the disciplined way that ecosystem is coming together and developing. And if we look at -- if we look forward, again, we don’t see that hockey stick of capacity on the horizon, or we don’t see the hockey stick of spend today, domestic China. The majority of the spend is still foundry logic and it’s still trailing no geometries, things that support image sensors and embedded processors. You are seeing an uptick, incremental uptick in memory spend this year. But, it’s still an overall -- very small percentage of the overall memory market.
And so, we think that the investment profile going forward mirrors what we’ve seen historically out of China, which is slow and steady development of the ecosystem as they progress and look to be a meaningful participant in this global market. You mentioned trade, multinational geopolitical situations. We think and are strong believers that a global interdependent complex ecosystem, like semiconductors, free and fair trade at the basis of how that ecosystem operates, creates much more opportunity for people long term. And I do think that economics will dictate behavior.
So, our strong belief is as we get to a constructive resolution of the issues and free and fair trade becomes the basis by which this ecosystem continues to operate. And if we look at specific actions, to date there’s been one action, it’s one company on an Entity List around an IP issue, there has been nothing in active that has in any way prohibited what we sell to our customers, how we service our customers, how we support our customers. And so, we feel good about the China market and the stability of that market going forward.
Great. Switching to services. Your services business has been super strong, growing 15% plus in last couple of years, but looks flattish this year because of underutilization at some of your memory customers. You have done a nice job in explaining the contract with the energy part versus the spare part business. What’s the split like this year and what makes it grow next year?
So, the services business has been a really good grower for the Company. If we take a look at what drives that growth for us, we’ve got the industry’s largest install base of over 40,000 systems. And when we look at chambers, many of our systems have multiple chambers. We have over 140,000 chambers in the field. That forms the industry’s largest install base and the platform for us to monetize in an aftermarket way for our services business.
When Gary came into the Company a handful of years ago, he changed the strategy. We were primarily a transactional spares and parts business when Gary came in. And he reoriented the focus towards more long-term service agreements and subscription like revenue. And we hired a new management team, team has executed well against the strategy and the vision for the business. And in 2018 we’ve grown the long-term service agreement business to where it was 50% of our overall services segment. So, 50% long-term service agreements, 50% transactional and spares. We talked about this year, 2019 being a memory downturn, industry utilization levels are down, we’re still seeing nice growth out of our long-term service agreement business, mid-teens growth. And this is against the backdrop of a market that is down 15%- 20%. And this year, the service agreement business will probably be 55%, maybe slightly higher, depending on where the transactional spares ultimately come out.
The setup, as we look into next year, looks really favorable for us. We continue to see strength in foundry logic; we see 2020 as a recovery year for memory; and industry utilization is coming back up. And when we think about the strong performance of our service agreement business, combined now with a transactional business that recovers with industry utilization, we see an opportunity to drive this business back to double-digit growth. So, we feel good about the strategy, feel good about the execution and the way the team is continuing to grow a more stable part of the portfolio, and something that we think is going to serve this company really well going forward.
And moving on to display, you have been a bit more conservative on the display recovery next year or single-digit type growth versus some of your peers, Tokyo Electron talking about the business getting back to 2018 levels. What are you seeing in this market and what drives the growth next year?
We think it’s prudent, I think it’s prudent to be conservative on where we set expectations. I think that makes sense. The drivers of the business, two parts of the market, TVs and handset, OLED screens, they’re both undergoing a technology inflection, which is driving near-term activity. In the TV side of the market, we’re going to larger substrates. I draw the analogy going from in the semiconductor business 200 to 300 millimeter wafers. What happened when the industry made that transition, you got 2.25 times the number of chips per wafer, which created a cost advantage for those that -- those customers that made the investment and went to 300 millimeter technology.
In display, you see the same equation playing out. On generation 8.5 technology, you can do 3 65-inch TVs per substrate. On the new generation factories, you can do 8 65-inch TVs per substrate. So, the substrates are much larger and you get more screens per substrate. You get 2.7 times the output per substrate. So, it creates a distinct cost advantage for those customers that want to make the investment to the new generation technology and we see that playing out in TVs.
In the handsets side, we’re about 30% penetrated with OLED technology. Right now, there is one source of supply of that screen technology to the handset market. I would say, there is another 7, 8 customers that are in various stages of their roadmap, and the next 12, 18, 24 months are likely going to be in high volume manufacturing. And so, we see penetration rates proliferating once the economic entitlement of that technology is realized and more handset manufacturers adopt the technology. Taking a step back though from the near-term dynamic of what’s driving activity and take a little bit longer view on the roadmap, we see technology inflections down the road. OLED technology is going to come to TVs. The handset will go towards foldable displays. Both of those trends are going to be incredibly good for Applied Materials and how we’re positioned in those markets.
And then, on a company-specific basis, I guess, the third thing I would point to is, today we’re exposed to about 15% of the spend in the display market. And we’ve got a product pipeline that potentially opens up that opportunity to 3x what our footprint is today. So, almost tripling our addressable market over time. We’ll start slow from a share penetration standpoint, but over time hope to build a share position in those new markets similar to what our core business is today. So, we like the display business. We think we’re well-positioned, and we like the technology inflections that are on the horizon and roadmaps of our customers, because we think it’s going to accrue significant benefits for the Company when that materializes.
Q - Atif Malik
Great. And since we have a standing-room audience, I’ll open up to questions for the audience. Questions?
Okay. I’ll keep going. The investors understand that WFE market share is all about your product set and end market spending mix. Memory investments are going down this year and WFE share is expected to be flattish. How should we think about your WFE share in the next two to three years?
So, not surprisingly, the share position you have is a function of the markets you’re exposed to and the markets you’re not exposed to and the differential growth rates. If we take a step back and think about the last handful of years, Gary came into the Company and this is -- Applied was a company that had just over 15% overall WFE market share. And then, in the last handful of years, we’ve gone from just over a 15% to almost 22% in a consistent drumbeat of innovating and delivering that innovation to market.
Last year was a unique set of circumstances. There were three areas of the market that outgrew the overall WFE market. And those are three areas where we did not have an exposure to, things like lithography, dielectric etch and batch processing. As we think about the setup this year, next year and in the near to mid-term, those headwinds that we saw in 2018 begin to fall away for us. We expect this year to be a better year. We expect next year to be a batter year. And the track record that’s being created since Gary’s here of incrementally growing share over time, we expect to get back on trend line. So, we feel good about where we’re at, we feel great about the product pipeline, the innovation we’re going to delivery to market, and our ability to outgrow this market over the long run. So, 2018 is the first year in seven years that we didn’t hold or gained share. We think that those headwinds dissipate in ‘19, ‘20 and we like how we are positioned.
Great. And Dan, on the -- can you just talk about EUV dynamics on the logic or the foundry side? There was a bit more reuse at 7-nanometer. So, your opportunity came down. How should we think about your incremental revenue opportunity as the foundries move to 5-nanometer?
So, as we think about the leading-edge roadmap, we think our opportunity grows, node over node. You talked about the transition from 10 nanometers to 7 nanometers, there was more reuse in that equation because fewer and fewer customers were finding a long-term home, so to speak, on 10 nanometers. And you saw a lot of that install capacity get repurposed to 7 nanometers. If you listen to what our customers are saying, they are beginning to ramp capacity statements around multiple nodes, multiple customers. It’s 7 nanometers, it’s 5 nanometers, it’s 6 nanometers later in 2020, and it’s the foundry -- or it’s the logic equivalent of those nodes that are beginning to ramp.
And so, we like how we’re positioned. Our opportunity goes up node over node. It’s going to give us an opportunity to bring in things like new films and new metals as those technology roadmaps progress. And when we think about EUV coming in, on our customers’ roadmaps, it only makes economic sense if there are three or four steps that get replaced by the new technology. The interesting thing about that is, is these are steps where we’ve had virtually 0% market share. And so, the steps that get eliminated as EUV comes in, are not steps that the Company currently participates in today, but gives us an opportunity to further penetrate the market with some great new technology we’ve developed in the etch market. So, the combination of being a key enabler to help EUV come in, replace steps that we don’t currently own today as well as an opportunity to do the things like new films and new metals and new planarization steps around the leading edge roadmap, we like how we’re positioned, and we think it’s going to create a great opportunity for us as these nodes get built out over time.
And from a reuse standpoint, our customers are saying 7 nanometers and 5 nanometers are going to be big substantial nodes over the long run. So, the reuse we saw going from 10 to 7 is unlikely to repeat as 7 and 5 are long, large, lasting nodes in our industry.
And just looking out longer term, if I look at this industry, historically NAND has been the driver for leading and bleeding edge because they are the most cost sensitive and commoditized parts of market. And then it’s followed by DRAM and then comes logic. So, given that the NAND devices already are on some kind of 3D stack and then DRAM is starting to face issues in terms of capacitor scaling, what your customers telling you about the roadmap beyond 3-nanometer, what kind of logic devices, will the logic also move into a 3 dimensional type horizontal architecture?
Yes. There is a lot of discussion about it. But, it’s really early days. So, I think at this point, the FinFET architecture introduced by one customer at 22 nanometers followed by all the other ones, 16, and 14, it’s still with us at 7, it will be the technology for 5, the tweener node called 6 is also FinFET. And then, there’s discussion that some people may carry that forward even to 3. What is being discussed on the logic side is something that’s called nanosheets. And that is a new technology where you can create a horizontal nanowire. And that will require a lot of deposition and etch to create new material stacks. So, that could be a successor to FinFET. But, it isn’t exactly -- we don’t know exactly when this is going to happen. It’s been in the domain of research papers and not yet in production. If and when it happens, it looks good for us, particularly. So, we think that’ll happen. DRAM continues to shrink. It struggles but it’s still doing it. And in the industry at large what we’re seeing is a lot of difficulty trying to get the performance of a power to scale, to get the power reductions. Right?
So, people are looking at DRAM which requires constant power, SRAM requires constant power. So, with the new incremental workloads that are AI-related, what people are looking at is, can there be a shift in the computing architecture move away from the annoyment where you’re shuttling data back and forth all over the place and consuming a lot of power on the chip off the chip. Why not try to create new architectures where you store the data and processes it in the same location. And that’s where people are looking at MRAM as a candidate and particularly ReRAM, which is still a few years out as a candidate. And that is one of the reasons why we’ve been bringing in new people to the Company, people that are system architects, people that are EDA experts to try to help figure out what those schemes will look like and begin to produce systems that can make some of these technologies. And that’s why at SEMICON West we introduced two new systems, two of them specifically, one for MRAM one for ReRAM. So, we’re trying to get ahead of that, build that knowledge and get that out there. But, I think what you will see is these will be very helpful, not going back -- backwards to existing applications but where the new workloads are the cloud, at the edge for machine learning. That’s where that will get used.
That’s good, Mike. A question, I’ve been getting recently, there is a little bit of optimism that the NAND pricing probably hit a trough. Western Digital is calling a trough in June quarter. And my question to you is, historically speaking, when the memory price start to go up, when do the memory makers engage with you in terms of better order funnel or in terms of bill plans and all that stuff?
So, we have constant dialogues with our customers. And the nature of the conversations, again, we talked about signs green shoots in the industry, we’re encouraged by the conversations we have. Given the fact that this year, factory output in both NAND and DRAM are below where they were last year, significantly under shipping true end market demand as inventories get worked down, I think, you will see, two lead indicators. You’ll continue to see the spot price stabilization in the market, followed by rising utilizations. And then, unclear at this point, whether we’re going to have a snapback as customers rush to bring supply back on line, consistent with true end market demand, or as utilizations come up, will they start to feather in more capacity and have a smoother transition back to true end market demand? Unclear right now, which of those two scenarios are going to play out. But again, we continue to be encouraged by some of the fundamentals that we see in the memory market.
And then, question on the capital allocation. You’ve done nice buybacks last few quarters, announcing acquisition. How are you thinking about capital allocation going forward?
So, I think, the framework we put forward several years ago still holds. By the framework, I mean there is three priorities for capital. First is to grow the Company, whether that’s organic or inorganic, and Kokusai is a great example of an inorganic action we can take. We want to maintain a strong and flexible balance sheet and then excess cash goes back to shareholders. If you think about our long-term track record, if you go back over the last two decades, we’ve given back 100% of free cash flow to shareholders over that two-decade long period.
If we look at a narrower slice of time, last five years, we’ve repurchased about 29% of the Company at a price below $33 a share. So, we’ve got a strong point of view about the inflection we find ourselves in, early innings of the data economy, the roadmap beginning to diversify away from just to 2D shrinks to define power performance and a disconnect between the opportunity and the value embedded in our Company. And as long as we see that disconnect persist, we are going to continue to do what we have been doing, which is taking a very strong point of view on our share price and repurchasing our own stock. We think, it’s a great investment at these levels.
When we come back to growing the Company, the bar is going to be high for inorganic actions. The standalone organic growth opportunities given what we see, given the deep technology portfolio we have, the organic growth opportunities are going to be substantial. We don’t want to become defocused on that. But, when we’ve got an opportunity to enhance the leadership position, accelerated strategy or drive value for shareholders, we’re absolutely going to devote management bandwidth to it. And Kokusai is a perfect example of that. While the bar is high, it’s great technology at a great price to pay 7.3 times synergized EBITDA in the middle of a memory downturn for high-quality set of capabilities, both team and technology. That’s a great deal for us and for shareholders. So, we’ll continue to do that, but the bar will be high.
We have few minutes left. Are there any questions in the audience?
Can you talk a little bit about the approval expectation for Kokusai? There have been challenges for mergers in the capital equipment space. Can you reassure us on the probability that you’re able to get this done?
It’s a great question. So, this is a deal that we feel really, really good about. We had a number of conversations, deep conversations with customers before we announced. We announced right before the 4th of July. We set targeting about 12 months for the regulatory review process and we filed it in all of the jurisdictions that we needed to file in. So, we continue to feel good about it. The conversations with customers pre and post announcement hasn’t changed at all. We’re seeing broad level of support across the customer base, which we’re really encouraged by. So, we feel good about where things stand and the ability to get this deal approved in six different jurisdictions.
Okay. We’re almost out of time. Thank you, Dan and Mike, for coming to the Citi Conference.