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

Good afternoon, everybody, and thanks, again, for joining us here on Day 2 of Stifel's Cross Sector Insight Conference. We're very fortunate today to have Lumentum presenting with us this afternoon. And with us today is Alan Lowe, President and CEO of Lumentum; as well as Chris Coldren, who runs business development for the company.

Alan, the place I'd probably like to start first, if that's okay, is just a little bit of a rehash on the 3D sensing side. A lot of, I think, questions coming out of your most recent results call about how much of this is sort of industry demand in terms of the outlook that you gave versus maybe some share shifts, some pricing, things of that nature. So maybe if you could just spend a couple of minutes there talking a little bit about that, sort of that 3DS demand environment for us.

Sure. Thanks, John, and thanks for having us at your conference today. What we said on the earnings call was that our belief, assuming no unit growth in devices, our market for 3D sensing would be down in our fiscal '22, which starts next month, over our fiscal '21 by 20% to 25% and talking about the markets. So nothing to do with share shift or anything like that. We were just trying to make sure that everybody understood our perspective on the available market out there. And there's a couple of things that led us to that conclusion, one of which was normal price reductions on existing chips that we've been in production with for a year or 2 or 3. Those come down in price over time by some percentage, and that's going to continue until the end of life of those products. New products replacing some of the old products on the front of -- front-facing cameras are coming out with smaller chips, and smaller chips mean more die for 6-inch wafer, and more die means lower cost per die. And in working with our customer, it became very clear that they understood that smaller chips mean lower cost. And in our variable cost model, where we pay a fixed cost for a wafer, whether we're making 100 wafers a week or 1,000 wafers a week, our cost comes down proportional to the size of the chip. And so we thought it fair with our customer to price accordingly so that we'd have the same margin percent but believe that, overall, that makes the unit dollar content per device for these new phones to be less than the prior generation, so the combination of price coming down on existing devices, new devices being smaller in size. And then the third thing that impacted us is we were very close to a final design-in with another major Android customer, and right at the last minute, they decided not to incorporate it in their next generation of phones. So that we were expecting to offset some of these other negatives from a standpoint of market share or market size. And therefore, we came out and said these things factored in, we wanted to set expectations for our fiscal '22.

Got it. And then a couple of things, if I can, then, Alan, just as a follow-up there. First, on the Android side, you mentioned sort of late in the game getting that. In your opinion, I guess, why has it taken Android so long to potentially adopt this technology? Is it lack of a killer app in their sense? Is it just reluctance to add that cost to the BOM because they don't think they can get that back on the pricing? So I'm just -- it's a little bit surprising, I guess, given how much Apple has chosen to incorporate this technology that Android ecosystem still seems to be that skeptical, I guess, about the overall technology at this point.

Yes. I think it's all of the above. And I think if you go back in time, we were actually shipping pretty good volume to one Chinese manufacturer that U.S. government got in the way of their high-end phones being able to be produced. That was probably as much as 5% to 10% of our 3D sensing given in any given quarter. That stopped. I think we have had small successes, although the devices haven't been very successful in Androids in the past as well. And then on our earnings call, we also announced that we had won a non-Chinese Asian-based phone that, while the volume of that phone device is not very large, the relevance to the ecosystem, if you will, to your point, of having a camera component manufacturer who also makes phones come up with more or less a reference design for the Android ecosystem to be able to incorporate. So I'm hopeful that as we look to 2022 and 2023, we'll start getting more adoption, especially as AR/VR becomes more prevalent as an application. This computational photography is here. It is shown to be very, very successful. And I think there will be some adoption on that front. But I think today in the environment, and one of the reasons we saw our Korean manufacturer not incorporate 3D sensing, was just as you said, John, no super killer app and they don't want that extra BOM cost when they're competing for share. So I do think it's just a matter of when, not a matter of yet.

Got it. And then looking out maybe longer term, As you start to see adoption beyond the handset world, whether it's automotive or LiDAR -- things of that nature, when we start to look at applications like that over time, is it easier, I guess, to differentiate when you start working with those types of customers rather than just a handset? I think there's a perception here despite all the work and effort that goes into these new chipsets, that they're relatively easy within a fairly short period of time to maybe replicate or then bring out a cost competitor and compete more on volume or share that way. But as you start to look at some of the longer-term applications for this that broaden out from handsets, should we think of it as maybe a more defensible market or one where it's easier for you to differentiate on technology?

Well, maybe I'll take the comment about the handset differentiation and let Chris talk about the new applications. I'd say that as technology is introduced, we believe we're the leader, and we'll lead the way, like we did for the last 4 years. It took 3 years for our competitor to catch up. We're introducing a new set of chips as we speak right now. And I think they're a little bit more difficult, higher-density lasers -- hundreds of lasers in a smaller chip, and we believe we're going to be in the leadership position there. I don't think it's going to be for 3 years, but it's certainly, eventually, for some period of time. So I do think that we're going to continue to add functionality and feature sets to these chips for mobile handsets that will make it more difficult to replicate and more challenging to make into commodity-type market. So I think that's on the handsets. And Chris, maybe you could talk about what we're doing beyond that.

Yes. Certainly, as we look to other applications beyond, right, you mentioned automotive, LiDAR, there's industrial kind of machine vision applications, we also have other customers interested in security and access control, adding sort of facial recognition or object recognition to video cameras, if you will, and almost an IoT like market where you're adding 3D sensing to devices so that they can sense the environment, right. You can imagine although that's appliances looking at not just people but your kitchen pantry looking to see what's in it and ordering Instacart food to be delivered as opposed to you having to be in the middle of that. And I think your point is I think there's a wide range of like -- and particularly automotive, not a lot of folks that have the underlying semiconductor laser capability. It's a very long-term market, high performance, high reliability, and long product cycles, right? I think you highlighted the handset side every year, right? There's something new out. We may not have a new product every year per se, but at least conceptually, our customer does. So there's more turnover versus an auto where you're designed in and it's a 10-year run kind of thing. So I think there is the ability, given the wide range of requirements, to differentiate on something like high performance if you're talking a long-range LiDAR or even, in some cases, LiDAR, all solid state LiDAR that involves very, very large arrays of VCSELs, if you will. So I think what's great about the situation we have is we've got a range of laser technologies, scale, if you will, which brings R&D scale, if you will, so we're able to develop products and -- that are targeted towards these different markets and very close partnership with the customers. And while some of these may not have 1 billion units per year, maybe they're measured in 100 million units per year. But for example, in automotive, you're not talking dollars per unit, which is kind of where you're looking at a handset. You're talking tens of dollars so the market can be comparable-sized, if you will, to the handset market or maybe even larger, should we, at some point, choose to not just sell the laser itself, but maybe move a little further upwards in the technology stack. Handset, that's a little harder, too, right? You're trying to use a laser module because the customer wants to control a lot of that for cost reasons. You go to automotive or industrial markets, in many cases, they just want the whole solution from somebody like us. So that also creates a bigger opportunity.

Can you talk maybe a little bit, Chris -- and I don't want to spend too much time here, but you broached an interesting topic there. I mean, one, when, I guess, would you start to see -- expect to see some of that filter in, whether it's '23 or '24, that kind of time frame? But also, particularly within automotive, right? You've made a couple of announcements on the partnership side. Is that something where you need to work directly with OEMs? Do you work with other system -- subsystems vendors in automotive? So where, I guess, should we be looking for maybe some continued milestones, I guess, is the best word, within those markets?

Yes. So we work at that sort of multiple levels through the supply chain because the folks between us and the end users, there's various -- there's different flavors of these guys, right? Some are a little more vertically integrated. Some have no optical capability and want a full module from us. But I think in terms of timing, that you hit the nail on the head with some of the dates you throw out, if you think of the traditional automotive market, I mean it's going to be a while, right? You're talking 2024, 2025 time frame before you see any kind of material inflection. That said, you need to be working today and getting those design wins and qualified today because that's just -- takes the customer that long to incorporate and design the automobile. But I think there's opportunities that are a lot nearer term. They're -- they have their risk level. And what I mean by that is you think of things like delivery vehicles, last-mile delivery vehicles to your home, trucks or warehouse robots or vehicles for moving logistics, if you will. I mean you're talking large online or even sort of traditional retailers that would be driving that. So they've got the ability to, in a shorter time frame, spend a lot of money and deploy a -- it doesn't take 7 years to develop a fleet of vehicles. They may do it in a year. And so some of our customers are targeting second half of calendar '22. That's still a ways off. So there could be some movement in that, but I think it's clearly a few years ahead of the traditional automotive market. And then as I said, the IoT and security and access control, that market for us is really about developing the fundamental with partners, right, reference designs and ecosystems, so that the thousands and thousands of companies that might incorporate this somehow somewhere in their unique capabilities. We're talking about drones or appliance. Okay, we're not going to go directly to those guys. We're going to develop or reference design that then they can go buy the parts and put it together themselves. I think that's much more of a continuous kind of growth, if you will, over time. So kind of 3 time scales: the continuous growth of the sort of non-mobile, non-auto market; and then the auto market having those 2 sort of legs of growth, the near-term delivery vehicles and then the longer-term passenger vehicles.

Got it. Okay. That's helpful. I appreciate that perspective, Chris. Maybe shifting gears a little bit to the Telecom side of the business. There was, I think, a little bit of a perception that you maybe we're a little bit more muted than some had expected on the second half of the year, but certainly bullish, I think, about the intermediate term, '22 time frame and things like that. Can you talk a little bit about maybe particularly on the ROADM side, you're starting to see shipments to the Western customers for MxN. You've got some new capacity coming out. How much of this is about your ability to meet what -- some of the demand that's out there? How much of it is still sort of COVID recovery related things of that nature are on the Telecom side?

Sure. I mean, I think we have an obligation to our customers to have the capacity when they bet their next-generation architecture on a sole source product like we have, right, on the end demand. So we're adding capacity ahead of the demand. But at the same time, demand is picking up as we expected because of China continuing to build out with MxNs as well as the western customers that we talked about on our last earnings call going from field trials to now deployments. And so that's an exciting time for us, and we're going to make sure we continue to have not only chipsets and other components that go into building MxNs but the capacity in our factories to be able to manufacture those whenever that's needed, at the right time they need it. We don't want to be the gate to our customers' ability to meet their customers' needs. So I think we're going to continue to see Telecom growth, especially in transport, both in the September quarter and December quarter, as we talked about on the call. And that's usually a pretty good median indicator that transmission [ is assumed ] as it transition from 100 and 200 gig that we've seen go through the pandemic really has turned to 400, 600 and 800 gig components. We're ramping those now, and we think that the transport as they get new highways out there deployed, they're going to have to light them up with next-generation transmission step. So that's why calendar '22 looks very, very solid for growth across the board. I'd say that the headwind to all that is what we talked about also on the call, which is our -- we call them Datacom chips, but they're basically the Datacom chips that go into 5G front haul, which some characterized as Telecom. So that's kind of a headwind, and we talked about that. We actually deferred some revenue in the March quarter as a result of the slowdown in China. But we expect that to pick up in the second half of the year. And then I think we're on the top of first inning in 5G deployment in the rest of the world. So we've got a long way to go in deploying the 5G Telecom data from [ chips to the world ].

Right. Right. And maybe just following up just because you mentioned that Telecom/Datacom split even within that Datacom chipset business, can you just talk a little bit about the difference or maybe size some of that in terms of the DMLs, which are obviously more Telecom oriented and EMLs, which seem to be doing very well on the cloud or web scale side?

Yes. I'd say our DML chips are, as you said, 5G as well as 100G 5G radio base stations stuff and 100G transceivers in the data center. I think there's a transition happening in the data center, to your point, 100G to a very, very compelling 200 and 400G using EMLs. And the economics for the hyperscalers is super compelling. And that's why one of the things we talked about on the call was 8 EML bookings during the quarter of $90 million, that's going to take us several quarters to fill just to last quarter's backlog that built. And that's because that it's so compelling from an economic standpoint in the hyperscalers. And so that transition is going to happen as fast as we can hopefully make it happen. And I think that as we talked about 9 months ago or so, we're adding capacity. It was primarily for EML capacity. Equipment is in the fab now. It's being qualified, and then we're going to start running it. And that's why by the end of the calendar year and into calendar '22, we're going to see strong growth in EML capacity to try to keep up with this transition in the hyperscalers in the U.S. and then I think a few months later or a few quarters later in China as they make that compelling transition from 100G in the data center to 400G.

Got it. And maybe while we've talked about China here with 5G, you just mentioned China on the web scale side or the cloud side. How do you view, I guess, China overall as a market for you? And obviously, they were the first to kind of come to you for the MxNs, That has started to expand a little bit globally, doing a fair bit there in other areas as well. With everything that's going on trade war-wise and with Huawei and things like that, how do you look at that China market, I guess, just from a risk and reward perspective, I guess, is the best way to put it?

Well, I mean, I'll let Chris comment as well, but it's hard to avoid 1/3 of the market, right? And so we are -- China is -- certainly was a huge part of our business. It's less so now with the restrictions you talked about with Huawei, but it's still -- and we've seen it strengthen from the other 2 main network manufacturers in China as a result of that. So it's going to be critically important to us in the future. And -- but I will say that as the headwind for Huawei in particular causes people to shift away from Huawei, our share of wallet with the other network with the manufacturers is much larger and then, of course, should be a tailwind to us as those deployments of new networks that have been traditionally Huawei networks turning to be either other Chinese NEMs or Western NEMs, whether it be in Europe or Middle East and Africa, for instance. So I think from that perspective, it's a good thing.

Yes. And I think the key point is, there's nothing unique about China's networks that require their products or technologies to be radically different. So there's not a separate R&D investment. So our focus is to make sure our technologies are indispensable to how future networks are built. And then that will naturally draw us into customers, obviously, in the West, where we've got bigger share of wallet. And -- but as Alan alluded to, there's substantial spending for domestic deployments in China. It would be a shame if we didn't ensure we got as much of that as we could possibly get. And given the differentiation we have in a lot of key products, we're very much needed.

Got it. Okay. Okay. Maybe just shifting gears a little bit down to the laser business. That obviously was one of the ones that I think suffered fairly significantly here through COVID, looks to be showing some signs of recovering from that. Just curious, as you look at that business, Alan, both sort of in the short term, if we should certainly expect that recovery to continue to pick up some speed? But more in the longer term, in light of what's going on now with Coherent and things like that, how you think about that business, both as strategically important to you and how maybe investors think about what you're thinking about doing there over the longer term?

Yes. I mean, I think, to your point, the pandemic put a hurt to our commercial lasers business, but we believe we saw the bottom at the end of last year. We saw an uptick in March continuing to believe that June throughout the balance of the year, we'll continue to have growth in our existing laser business. We are also, at the same time, organically investing in new products and increasing our R&D spend for that business so that we can capture more share, capture more market in our new markets and new applications. And so that's our organic plan. Now you mentioned Coherent, I would have loved to have gotten Coherent at the right price. And when we announced it on January 19, that was a good deal for us, a good deal for their shareholders. And as you saw play out in the public market, what happened to the point where we said, hey, enough is enough. We wish them well. That said, we're not giving up on M&A, even though we announced a share back -- a large share buyback. And before the Coherent year, we had a funnel of things we were looking at. And after the Coherent year, we have a funnel of things we're looking at. So from my perspective, whether that's in the laser space or strengthening other areas of markets where we participate today or getting an adjacent market where we can capitalize on our technology and products to a new market, I think those are all things that we're looking at.

Got it. Got it. I do want to spend a moment on the model very briefly. You've done a really nice job here over the last 18 months or 24 months or so really, I think, changing the profitability profile of the company. Margins are significantly stronger than where they were. You're now generating a fair bit of cash. Your EBITDA is growing at a very good clip. How much of that is some consolidation, rationalization, obviously, what you've done with Oclaro, what some others have done with some others. But also there's this perception, I think, that as technology continues to cycle faster and faster, particularly with some of those cloud customers, it's getting harder and harder for others to keep pace. And even for those customer or competitors that are left, scale matters and things like that, that are tougher and tougher to achieve. So I guess, how should investors think about this business from a profitability, from a free cash flow perspective as they're starting to look at this in maybe a different lens than they did 3 years ago?

Well, it is a different company than it was 3 years ago. And the acquisition of Oclaro helped catalyze some of the changes that we made with respect to lower-margin products, divesting the transceiver business where the value was in the chip and the margin was in the chip, and then you wrap a bunch of other stuff around it. And we much prefer to be able to arm our customers with that value in the chip and let them wrap stuff around our chip that has the vast majority of the value in that business. We've divested our lithium-niobate modulators, which was declining over time with volume and making it hard to make money. And so as a result of those 2 major divestitures and our focus on chip-based business, where a lot of our businesses become chip-based, where we're getting the kind of margins you would expect from the semiconductor chip business. That said, on the non-chip business, those margins are going up as well, but really due to the fact that innovation and indispensability to our customers and providing them with value that they're willing to pay for has changed the profile of that part of our business. So I think overall, getting to a 50% gross margin and 30% operating margin is something that we're very proud of, and we're going to continue to drive to either maintain or grow from that.

And you mentioned, obviously, the recently announced buyback. Should investors look at that as a step back from M&A? You mentioned you still -- you had a funnel before, you still have a funnel now. Is it just a function of, look, there was a narrative obviously around Coherent that was somehow a knee-jerk reaction? And obviously, you mentioned the price, that's not something you write a check for overnight. I think more work went into that than some they'd like to believe. But as we think about M&A looking into the future, do you expect that that's still a piece of the growth strategy going forward?

Yes, absolutely. I think it's part of our overall strategy. And we have an organic growth plan as well as an inorganic growth plan that fits into the strategy. So it's critically important than any M&A that we do, do strengthens our strategy or aligns with our strategy. That said, $700 million is what we grew our balance sheet in the last 12 months. We've got a gift from our friends at Coherent, a $218 million gift. So we say thank you very much for that. And combined with the cash from operations that we generated, about $500 million, we generated that last year. We planned that -- we announced it over a 2-year period of time. We're going to generate another $400 million or $500 million a year during that period of time. So we're not shrinking our cash balance or our firepower as a result of the buyback. And we thought at the time the best investment we can make is in our own company and putting our money where our mouth is. And so that's why we did it.

Great. Well, unfortunately, that puts us right up against the time limit here. So gentlemen, thank you both very much for spending some time with us, and appreciate the update. And we look forward to catching up soon.

Great. Thank you, John.

Thank you, John.

Thank you.