Lumentum Holdings Inc. (LITE) Earnings Call Transcript & Summary

Let's get started here. I'm Meta Marshall, I cover the networking sector here at Morgan Stanley. We're delighted to have Lumentum, Alan Lowe, CEO; Chris Coldren, SVP of Strategy and Corporate Development with us here today. I'm going to read a very brief disclosure, and then we will kind of jump into questions. So please note that all important disclosures, including personal holding disclosures and Morgan Stanley disclosures appear on the Morgan Stanley public website at morganstanley.com/researchdisclosures or at the registration desk.

So we'll start with the topic du jour and kind of how we've kicked off all of these sessions. So just -- what impacts coronavirus is having to production and overall kind of supply and demand that you're seeing right now?

Sure. So I can talk about both. I think from a supply chain standpoint, we've been working with our suppliers that are based -- manufacturing based in China. Most, if not all of them, have some level of production capability today with expectations that they'll be back to full force by early April. So our focus has been on, how do we get an unfair share of that 50% to 70% output that they have today? And how do we mitigate any risk with respect to future issues, should there be some. So I think from that standpoint, our suppliers, I think, were pretty well under control. From the standpoint of our manufacturing in China, we have a facility in Shenzhen that does manufacture most of our coherent components that then go into modules like ACO and DCO as well as high-speed components that go into our customers as components when they build their own. Because demand was so strong going into the Lunar New Year, we incentivized about half the employees to stay and manufacture through Chinese New Year. And so from that perspective, we had a base of employees that after Chinese New Year, they continued to work since then. And today, we have about 80% of our Chinese employees back to work, with some of them coming back or having come back that we're asking to spend 14 days in quarantine before they come out to the manufacturing floor. So we think that by the early part of April, we'll be back to full force. Additionally, on Monday, the Chinese government gave us approval and authority to start hiring new employees. So we're starting that in earnest this week. We'll bring them on board. We'll quarantine them for 14 days, then we'll start training them on the floor. So that should help us increment our capacity in our Shenzhen facility by the end of April. So I think from the standpoint of -- we're pretty pleased with respect to where we are, and we are impacting our ability to meet our customers' demand. So I think we're going to do what we can to continue to improve the output. From a demand standpoint, as I said, we were already short of our ability to meet demand on a lot of the products, high-end ROADMs, coherent components and modules as well as datacom chips. And so from that perspective, even if there were a slowdown in demand, we still wouldn't be impacted from that as we've been able to reallocate that capacity to elsewhere because we haven't been satisfying the overall demand.

Got it. So with that out of the way, you guys had a very successful past couple of years on the back of 3D sensing and then also the successful integration of Oclaro. But maybe ROADMs is still an opportunity at early days. And so how do you see that opportunity developing? And how do you guys maintain your lead, particularly kind of in the high-end ROADM?

Yes. So I think, as you said, we're in the early stages, especially with respect to China deployment. And we've had a partnership with a leading Chinese customer that helped us develop a leading-edge product in the NxN, and so we've been deploying that in China today. And we're seeing our other customers start finalizing system designs to incorporate those high-end ROADMs. And so we're going to see continued growth of high-end ROADMs, not only in China, but also in the rest of the world. So our focus has really been on making sure we have capacity in place to be able to satisfy those customers who are betting their future systems on our ability to provide them products because we do have a multiyear advantage on those high-end ROADMs. And so we need to make sure that we have the capacity because they're betting on us, and we need to make sure we come through for them.

Got it. A lot of what ROADM deployments will take place is dependent on the 5G rollouts in China. We went through kind of a highly predictable phase of China rollouts when it was mostly the back [indiscernible] be built out kind of up and through the '16 period, and then more uncertain environments over the past couple of years. Maybe ex coronavirus impact, do you have a good sense of kind of the phasing of builds? Or is the ROADM opportunity so big, you're just not going to be able to meet it in the coming years?

Yes. I would say that the phasing is typically tied to technology cycles, if you will. And so the time period where you said was a little more predictable was when 100-gig was initially deployed in China, if you will. And similarly, when certain -- which is now not our most leading edge, but was 5 years ago, ROADMs, there was also a rollout of that technology. So I think what I think is important to look at is, as we look at the next 2 years, there's a lot of new technologies that are coming down the pipe that our customers will be incorporating in their systems, whether that be higher speed, 400, 600, 800 gig, whether that be NxN or at least much higher port count ROADM architectures, all of that will drive lower cost to drive a certain bandwidth through a network. And that, in a sense, will create demand and drive a new product cycle at the customer's customer level, if you will. And the timing of those products seems like we're certainly in the phase where we have products either available or about to be available, and our customers are probably 6 months behind from a standpoint of really starting to get their design wins and ramp up revenues. So really, the second half of this calendar year and into the next calendar year, I think there's a lot of technology-driven tailwinds in the telecom industry. That as you started the question, 5G is just sort of, I think, of at least as the stick that's at the end of the network that are opening up increasingly and therefore, driving the need to be able to pump a lot more data through the core and edge of the network, and that's really where the ROADM technology is coming to bear.

So we should think of it much more as optical upgrades necessarily than kind of 5G, 4G build out?

Yes. I mean I think, for us, certainly, 5G directly has a benefit in terms of selling certain semiconductor laser into 5G front-haul opportunities, but at least the coherent products that we sell or the ROADMs are sort of more derivative, if you will, in that they're supplying the data that's going out over the access portion of the network.

Got it. So I mean, one of the major customers that you're talking about in China is clearly Huawei. Given all of the restrictions and kind of trade discourse over the past couple of years, how do you view the in-sourcing risk in China versus the opportunity over the next couple of years?

So I would say that certainly, the -- since last May and probably even before that, there's certainly been a concerted effort to, we think, on our -- on the largest Chinese customer to figure out ways to become more vertically integrated or find alternate supplies within the U.S. suppliers. But -- and that pace has probably picked up, very logical as they feel like they're in a self-preservation kind of mode of operation. Our response to that, and frankly, you turn the clock back 5 years ago or even 10 years ago, it's not like they wanted to buy expensive optical components from folks like us. They've always had a bias of trying to do things themselves and squeeze out that extra margin dollar or be able to lower cost. So the way we tackle that problem continuously is to always be focused on being completely indispensable with a product or a technology that they can't get anywhere else. And as we look to what drives a lot of our -- and frankly, today, almost all of our revenue with that customer are ROADM capabilities, where either they can't get it anywhere else or the alternative could be a another U.S.-based supplier, pump lasers, that's another essential element of any optical network. And the only alternative is really another U.S.-based supplier, and similarly, datacom chips and some certain tunable transceiver products. And so we feel at least for quite a while, we have a lead where there's no ability to really vertically integrate or get those products elsewhere. And certainly, we're not standing still and continuing to innovate. And on a global scale as whether it be Huawei or any other customer goes and tries to win a tender at a network operator, they can't show up without the best technology and the best capabilities. And as long as we've got that then we feel like we're in a pretty good position to get a piece of whoever is going to win that business.

Got it. Maybe moving on to Oclaro. It's been incredibly successful, both in terms of kind of transforming what you've delivered in the datacom business, but just the sale of pluggables in general maybe starting -- were you wary of becoming -- moving more towards components and away from modules? Or did you just always realize that, that was going to be the successful path forward in that market?

Well, I think Oclaro had already started down that path of selling components and chips to other transceiver companies and so -- saw how lucrative that could be. And I think that the competition at the transceiver level was very fierce. And so when we completed the acquisition, we started working on synergies, and we started working on how do we drive our gross margins to new levels because we made commitments as we did the -- as we announced the acquisition to get to different levels that we have achieved. So we looked at products that were a drag on the margins and transceivers were one of them. Some other low-margin telecom things, such as lithium niobate modulators, where Lumentum had this and Oclaro had it, and both of them were pretty tough businesses. And so we've -- we're in the wind down process of that. And I think one of the dynamics of divesting the transceiver business was, we were no longer a competitor to the other transceiver companies that we were trying to sell chips to. So they all of a sudden came storming towards us to drive the demand for our laser chips, which are best-in-breed to ramp their production of datacom, 100-gig datacom transceivers for hyperscale. We're 400-gig now since it's now starting to ramp as well as for -- as Chris mentioned, the 5G wireless. And so we've seen a real growth in demand for that. And then put on top of that, Broadcom was the other major independent supplier of datacom chips. They brought the transceiver business back in-house. And so now they're a competitor to the customers that they were selling datacom chips to. So we see now a redoubling of demand in that area as a result of that. So I think the model that we're working on as a component supplier in some places where it makes sense, and a solution supplier or a system supplier where it makes sense, I think we're pretty happy with where we are now.

Got it. Either turning to the pluggables or kind of the ACO, DCO business of Oclaro. ACO has been extremely successful for Oclaro largely on the back of kind of multiyear supply agreements that they had signed. But DCO, there was a competitor in Acacia. They're enabling newer DCO. There's rumors of -- or kind of -- or press releases about other DCOs potentially coming into the market. Is there an opportunity to differentiate there? Or is it just the rising tide is going to lift all boats in that market?

No. I think -- I mean, you mentioned that the success of the ACO was based on multiyear agreements. I'd say the success of ACO was because Oclaro was the only one that could make it. And I think from that perspective, we're seeing new levels of demand for ACOs that, quite frankly, surprised us a little bit because we were always expecting DCOs to come on board and to really cannibalize it. But what we're finding is that as we start ramping up our DCO, which we're doing today, it's a different system. It's a different market, and it's not cannibalizing the ACO. And we think that there's multiyear demand for ACOs. In fact, we're adding capacity to meet the demand that we're seeing in the second half of this calendar year and into next year on ACO. So I think, while there's a lot of press releases, there were a lot of people, including Lumentum, that tried to make ACO successful and we weren't successful at it. So I think there's a lot to be said about making a demonstration unit versus making tens of thousands of these things and having the capability, the fundamental underlying technology that's really at the chip level that I think the Oclaro acquisition brought into Lumentum. And so I think from that perspective, DCO is off and running at 100 and 200 gig. We've got our first 400-gig DSP in-house now, and we're developing a 400-gig DCO as we speak. So I think we're in pretty good position. And I think, again, we have the fundamental coherent components and technology that enables us to be successful there as well as to sell those to other customers as components incorporate into their 400, 600 and 800 gig solutions today.

Got it. I mean do you -- you mentioned that maybe you saw DCOs cannibalizing ACOs, and you haven't really seen that. Just -- is this a move where we should just expect that they're all -- the telco market is moving to pluggable form factors? Where are you seeing that make sense and not make sense versus kind of traditional line cards?

Yes. I would say that there's a diversity of what customers need and want that, typically, our customers will introduce some very high-performance systems. Generally, at least historically, it's been around much higher speeds then that isn't necessarily what's required. And they're often what you're supplying is a more novel component capability. For example, you can imagine 96 gigabaud components that supply into 800-gig systems. The first foray into that is not going to be in modules, whereas, the modules are -- or tend to be where technology is more capably being reduced to a smaller form factor, lower power consumption, and maybe more pay as you grow in the case of a pluggable form factor. And that's probably always going to be a general trend. But you're not going to sacrifice that ultimate in performance for certain kinds of long haul or high performance, under sea or other applications. So we see both continuing on in parallel. I think the only trend that is probably a little more nuanced of late is certainly there's some systems where, particularly cloud operators are trying to drive either vendor interoperability or some level of standardization. So that maybe what network equipment manufacturers had been previously ripped open their box, there probably is a bunch of modules, but they have fibers coming out in a pizza box form factor, they're really wanting to say, "Hey, can you make those pluggable modules and have that -- have some level of interoperability?" But that's not too surprising, that request. But I don't think it really changes the landscape a whole lot.

Okay. All right. So before the crowd throws tomatoes at me, I'll move on to 3D sensing. So it's been a great business for you guys in terms of share and profitability. So how do you guys, with more and more competitors kind of coming into this market, how do you think about protecting your opportunity and expanding your opportunity in this market?

Yes. Well, I'd say that there's actually fewer competitors than there were a couple of years ago. And you look at the landscape of competition, Finisar and II-VI, now they're one company, and OSRAM and ams, now they're one -- well, soon-to-be one company, I think from that perspective. So I think the landscape is getting better. I think our focus is really on how do we make sure we continue our quality, reliability and delivery performance to our customers so that they don't need or want to have second sources on these things. And so we have a pipeline of new products that we're planning on introducing over the next 2 and 3 years. And so the technology that we're working with our customers on really has that kind of a development life cycle, such that it's going to be very, very hard for people that are not going through the learning curves of the first 3 years to catch up in year 4, 5 and 6. And so we're continuing to push the technology. We're introducing new products. We're coming up with some novel ideas to be able to enable our customers to come up with new solutions and new applications that, frankly, we don't know exactly what they're going to do with them, but we're responding to their needs. And I think from that perspective, we're going to see the Android market start to adopt in a more meaningful way, I think, later this year and into 2021. That will really diversify our customer base and grow the overall market quite substantially.

Got it. And just because with your lead customer, there has been some kind of supply chain disruption for them. Has there been any changes to kind of order behaviors that you've seen kind of today?

I don't think so. I mean certainly, there's a backdrop of what's going on in the world today that is concerning. But at the same time, at least as we look to how we provide guidance and think about our business, you tend to think about when you got a limited set of customers that are driving a lot of your volumes, you need to think about that a little differently than when you've got many more customers. And so I also believe that a lot of our customers, whether it's 3D sensing or outside of 3D sensing, have an expectation that they will fix their lows, if you will, at some point and manufacture maybe more at a certain period of time to offset some of the lost time here. And given our, either capacity limitations or our cycle times for manufacturing and 3D sensing, that can be a quarter long, for example, that they don't want us to slow down quite as much as you would think because then that's got a 3-month lag time for them to correct and try to ramp back up.

Got it. And you mentioned the kind of the Android opportunity ramping up. Do you think this is the year where we start to see more -- I know we started to see a couple last year, but more high-volume Android phones kind of incorporating 3D sensing technology?

I think we will. But probably the challenge has been that the customer that we had that was on the steepest trajectory to -- basically, at least the way we think about it is, they fill out the high-end portfolio, which in Android side is not that high a volume. It's once you get down to the more mid-tier phones where the volumes really kick in. The leader was the prominent Chinese customer that's having troubles with the U.S. government, and that's really slowed their ability to slow or to sell those high-end phones. But nonetheless, we're seeing, particularly around the world, facing capability around, at a minimum, solely at a minimum using it to augment photography, let alone, keeping the capability for augmented and virtual reality at a later point in time, but the photographic implications have really proliferated across an awful wide customer base. And so we think that as that fills out over the next 12 months, that's the time period where things will penetrate down into the mid-range.

Okay. Got it. And then you've shown incredible margin leverage over the past couple of years, both on gross margin and kind of on the operating margin as well. A lot of that has been Oclaro synergies. A lot of that has been mix. But just how do you think about the sustainability of the margins that you've been able to sustain and just what potential levers you have for any additional leverage?

Yes. I think what we believe is that we still have another 150 basis points of synergies to go, and that will come over the next 3 quarters or so. And so from that perspective, lifting the non 3D sensing gross margins up makes the swings between good quarters and bad quarters a whole lot smaller. And I think the divestiture that we really complete next quarter of the datacom transceivers and low-margin lithium niobate also helps because those are basically single-digit kind of gross margins. And so we have another $15 million of that to go away will help continue to drive further improvements in both gross and operating margins. I think that said, our plan is to probably grow our R&D dollars over the coming years, not as much as we're going to grow our gross margin, but certainly, some of that to maintain that leadership that we've worked so hard to get in the product lines that we have the leadership in. So I think from that perspective, we're on the right track and we've accomplished what we said we were going to do with the Oclaro acquisition when we announced it. In fact, about 2 years ago, it's back to the date. So now we're going to take it to the next level.

Yes. And I'd add to that, that I think we -- since that time frame between what we've done, other M&A that's been announced as well as companies individually, organically succeeding or not succeeding on product lines, we're ending up as we look to the product line that we're focused on. Industry structure has really changed quite significantly over the past several years. So that what would typically happen in a product cycle where, even if you get out in the lead and start growing and having great margins, you have somebody, 6 months later, come in and steal away a bunch of share and harpoon pricing. That's just not happening. In fact, quite the opposite because there isn't competitors coming along as quickly perhaps as some of our customers would hope. They tend to have to turn to us and say as opposed to bidding off their business every 6 or 12 months and more frequently coming to us and saying, can we enter into a multiyear supply agreement so that I have both security of supply and security of cost, if you will, as they bid on multiyear deals with their customers. And what that drives is a much more healthy dynamic in an industry where we and our customer now know what price is going to be on something, and we can work to make sure our cost matches what's needed to maintain margin. But then most of the dialogue with the customer is around what can we do next together as opposed to pummeling each over price in the near term. And I think we both win in that because then we're talking about ways that they can differentiate themselves as well as us helping them to do that.

Got it. And then maybe just last question for me. The optical space had been waiting for consolidation for a lot of years. We finally kind of saw it all in one rush over the last year or year to 24 months. Do you think we're done with consolidation? And where will it make sense to kind of continue to see some consolidation in this space?

Yes. I don't think we're done. I think we've taken a good chunk out of the competitive landscape and the consolidation has helped us all. So whether we participate in further consolidation or someone else does, I think it makes the industry even more healthy. So I don't think we're done. There's not a lot of left. So I think from that perspective, and maybe Chris can chime in, we raised a bunch of money in December when we didn't need it. And I think it gives us the firepower to go do acquisitions, should we come across one that makes sense for us, whether that be for vertical integration, for broadening our product portfolio or simply consolidation for consolidation's sake. So it has to be strategic. It has to make sense. It has to have a good return to our shareholders. But I think we'll see.

Yes. I'd add too that whether it's us driving it or others. But you now have 2 substantive players emerge in the overall optical component space with both technology capabilities and R&D spend that makes it very challenging for a smaller player to either enter the industry or perhaps even continue to persist in the industry. And I think you'll -- however, it happens, the smaller players either need to choose a different strategy, join forces with a larger company or something's got to give, but it's going to be very challenging composed -- as compared to 5 or 10 years ago where maybe there was diversity of size of companies, but when you looked at an individual product line, companies tend to be a little more equal, right? Just companies had more, not less product lines. So I think that's going to change.

Got it. Any last questions from the audience? Just shout it out, and we'll repeat it.

What's the -- in 3D sensing, what's the pricing curve going to look like? Or what does it need to look like in order to sustain your market share in the premium tier? And are you confident that you can bring down costs at a commensurate rate such that you can achieve or sustain acceptable margins over the next couple of years?

Yes. So I think pricing has come down. Since we started the ramp of 3D sensing in 2017, I'd say that our gross margins are higher now than they were in 2017 as a result of continuing to drive cost improvements through yield improvement, design improvement and things like that, that allow us to have maintained the share that we have today. I think our expectations and our plans are that prices continue to come down, but that we drive our productivity and cost structure such that we keep our gross margin where it's at. Our customers expect prices to come down as we go up the volume curve, and so we're going to continue to drive the prices down. But again, I think we're able to do that, maintain our market share and maintain the kind of gross margins that we've had in the past.

[ Biggest lever ]...

Biggest lever in cost is yields. And so I think we have produced over 600 million units for 3D sensing VCSELs and that amount of data allows you to tweak and refine and tighten specifications, whether that be at the base epi level or through the back-end fab to really optimize the recipes to drive yields and get more good die per wafer. It's just like a semiconductor company. So we've gotten to the kinds of yields that have, quite frankly, surprised our customers and have allowed us to meet their price objectives while maintaining the kind of margins we need.

Great. All right. Alan, Chris, thank you so much for being here today.

Great. Thanks.

Thank you.